CSS Button No Image Css3Menu.com
New! Search comments:
(NOTE: Relevance, Author, and Article are not applicable for comment searches)
Regarding steroids, until we know better who used steroids and other PEDs, as well as when they did and didn't use them, our concepts of career trajectories are sadly underresearched. You can't just make part of your science go away. It's easier, for example, to postulate frictionless surfaces when modeling in physics, but you compromise the accuracy of your results when you do that. The same applies in sabermetrics.
Right now PECOTA includes the career years of Jose Canseco, Alex Rodriguez, and Manny Ramirez. Should it? It also includes the career of Tom House, who has openly admitted steroid use. Should it? It also includes the career of Dave Johnson, who joined pioneer steroid user Tom House and became the next Jose Bautista before Jose Bautista was even born. Should it? There's absolutely no proof or suggestion, or even a hint except for sharing a clubhouse with House, that Johnson used steroids, but his going from an average of nine home runs per season to a one-time mark of 43 home runs looks...well, unusual.
If the sabermetric community wants to be taken seriously, the sabermetric community needs to be able to assess the effects of PEDs on baseball quantitatively. We cannot wish that away.
David, this is my second favorite BP interview ever. You've got some great descriptions of players' reactions at being elected to the HOF, reactions that don't make the press. You've also got the candid admission that being a beat writer is much harder than it was a decade or two ago, with the effect of reducing the number of ten-year members significantly.
Regarding players who get just one or two inexplicable votes for HOF, I think that it's a normal and appropriate situation. I know as much about stats and JAWS as most BP readers, but I can think of two players far short of normal eligibility standards I would've voted for regardless of their stats falling short: Jim Bouton and Jim Abbott. Other people would find reasons to support other players. Merely making the ballot is a significant accomplishment.
Jay, several players have no minor league playing time. Al Kaline and Winfield had none; John Olerud had one MiLB home run in the last year of his playing career. An "n to zero" ratio isn't unusual. One could call Olerud's career ratio "unusual," but it's essentially an outlier of "no minor league playing time." Bagwell played in MiLB for two full seasons with three teams and amassed only six home runs: that's unusual for a HOF-candidate power hitter, as is the ratio.
Is there another player who played enough to be HOF eligible with such an unusual ratio of MLB home runs to MiLB home runs as Jeff Bagwell? Bagwell hit 449 MLB home runs and 6 MiLB home runs, a ratio of almost 75:1.
Bagwell was a fourth-round draft choice, picked 110th overall, and was never higher than #32 on Baseball America's top prospects list. Few players go from such modest college and minor league careers to contend for the Hall of Fame, especially with a comparatively short MLB career. Whatever the circumstances were that allowed Bagwell to do that, they were unusual.
John, BP has seen a lot of changes over the past two seasons. Your regular "On The Beat" columns are already the greatest positive change at BP, and I thank you for your work.
I'd like to see a two-part feature at the end of each "On The Beat," titled "What I See." You see things about baseball that most fans don't, simply because of your experience and wisdom. I'd like the first part of "What I See" simply to be a short candid observation about some small part of the game that stands out in your mind: a catcher who suddenly looks much better throwing out baserunners, a team whose chemistry looks better in April, or a pitcher who just looks lost on the mound. The second part would be a different view, also in very few words, from a different member of your staff who is well-suited to see the same thing from a different perspective: Matt Swartz could look at obscure stats, while David Laurila could comment on team chemistry, for example.
John, you're an excellent journalist, and journalists are objective. I don't want to lose that. I would like you to add a little bit of an editorial section to your work, complemented by the perspectives of others on your staff in their areas of talent.
So if I understand what you're saying, Dan Fox joins the Pirates, and three years later we realize that during his tenure with the team they possess the highest home field advantage in MLB?
It's probably coincidence, but it's also exactly the sort of thing I'd expect Dan to be able to affect.
While I can see the Red Sox trading both Scutaro and Ellsbury, I don't see Scutaro for a second-tier relief pitcher as a trade that Boston would be likely to accept. Boston wants to concentrate talent by trading depth for star-quality talent, so it's more likely that they would trade both Scutaro and Ellsbury, and maybe a prospect or two, for a single, better player.
If a reader were to make a post alleging that a 32-year-old player who had played a combined half of a season at a more demanding defensive position two to three years ago was fine to move back full time, disregarding that he'd played first base all of the last season and the majority of both of those two (and disregarding that he'd finished the previous year on the DL with a serious thumb injury), just because he'd been average at that harder position in part-time play in his twenties, the reader would be laughed off the site.
Kevin Youkilis has never been a regular third baseman in MLB, never even playing a simple majority of his team's innings at the position. His career UZR at 3B is 8.3, with 7.0 of that value earned in 2004 at age 25 and a net 1.3 earned at ages 26-31 the last six years. He played 15 innings at third base last year.
Kevin Youkilis might be able to play third base. He might very well do badly were he asked to play there, though. Red Sox fans are not kind to infielders who struggle. Were I Boston's GM, I might not pencil in Youk at third, just as I might not expect the Padres to surrender their ultra-popular franchise player first baseman for a "fair price."
First, Colin, excellent work. Thank you.
The trend line seems very sensitive to the first year used in the study. If one takes out the three oldest data points, the slope of the trend line might be remarkably more positive. There's a noticeable dip in replacement level after the 1961 expansion. The talent pool seemingly gets worst right around 1968, the Year of the Pitcher, suggesting that the rule changes to correct pitching dominance might have made it possible for more individuals to compete at an MLB level.
Looking at this I'd be tempted to do one of two things:
1) Trying to research further back, so that all post-World War Two seasons were considered; or
2) Starting the generation of the trend line at a natural break in the level of competition, probably in an expansion year.
Again, though, thank you for a great article.
I'm glad that you made the trip, John.
Wouldn't it make sense for the Pirates to serve as a "get well" center for free agents wanting exactly one more chance, just one year, to put up decent numbers before hitting free agency again? It sure worked for Adrian Beltre and the Red Sox last year. Beltre will probably get at least $20 million more on his last big multi-year contract than he would've last year, and Boston will get two excellent draft picks if he leaves as expected. Pittsburgh could try offering similar deals to a comparatively large number of free agents, with the expectation of using all of the successful ones for deadline trades, building for that 2014 window of opportunity. Pittsburgh has some money to spend, and they need to spend it in way that ends the current ridicule while building for three years from now. Why not bring in a comparatively large number of established players, especially pitchers, and see which ones thrive in the weak NL Central?
OK, now we're in there. This is pain: this is pain that, merely by description, many can tell must exceed their worst life pain, exceeding their own "10."
I didn't post my worst pain above because this was a "scar contest." I posted it as a true extreme because I disagree with this key passage from the article:
"The fact is that the worst is the worst. One man's hangnail "10" is the same as a compound fracture "10." While individual responses to pain are very different, the disabling nature of pain is the same. That doesn't make it "right" or "wrong"—it just is."
My point is that there are injuries and procedures that we understand at a gut level are really, really painful. Many of those were listed above. There are those that are a step worse. Gangrene of internal organs is worse, not because gangrene hurts but because the neighboring tissue hurts. I think that we all get that. What you're describing, serviceoutrage, is worse than a hangnail (a lesser neighboring tissue hurt)...for anybody.
What Will Carroll is writing that I've quoted I've seen written by other professionals in the field. I don't dispute that it's the current accepted wisdom. I disagree with it: I think that compound fractures trump hangnails, and that the difference person-to-person isn't as great as the difference injury-to-injury for extremely different injuries.
Ellsbury had cracked ribs, not compound-fractured ribs. I can't find any indication that such injuries ever take more than ten weeks to heal. I know that UTK expected Ellsbury back in the lineup months ago. We're more than a month past expected healing time for an injury that affected performance only through pain, and Ellsbury claims that he's still not able to play. At some point, when other players on the team are playing through cracked bones and following the team on crutches, and when Ellsbury looks fine in rehab but defers at being reassigned to Boston, it goes from a personal-injury-pain-perception issue to a team focus issue.
Worst pain ever: second-degree and third-degree burns over two-thirds of my body, before they used induced comas to relieve the pain. My doctor didn't want to risk narcotics addiction or anesthesia use, so I had debridement without narcotics and grafting surgery without anesthesia.
I win scar contests.
Matt, excellent article, as usual.
I'd offer the possibility that we all, as a community, overestimate the weight given to sophisticated statistical analysis in evaluating trades. I think that the decision makers still rely significantly upon players' reputations, scouting reports, and obsolescent-but-popular stats (such as ERA and RBI). That does nothing to counter your observation that "The Winners Curse" may apply here. It merely adds my perception that the people really making the calls (at the GM and ownershp levels) are stupider than we might choose to believe.
David, this is one of the best interviews I've ever read regarding any aspect of MLB. Great work, and thank you.
Yes, the Yankees' infield is making more money this year than the Brewers' 25-man roster...but what really strikes me is that the combined salary liabilities on the contracts of the four Yankees infielders is somewhere around the entire fair market value of the Brewers.
And that's with Jeter in a contract year.
Clay, I haven't yet posted how happy I am that you've made a greater time commitment to BP. Thank you.
Is Eckstein's job really safe?
Nomar's UZR/150 was -37.2 at SS in his games with Boston that season. You're an articulate fan and you know your stuff, krissbeth: I'm confident that you know how far below acceptable a UZR/150 of -37.2 is.
Nomar had UZR/150 of +6.1 at SS with the Cubs in late 2004. Something changed.
Want another stat? Nomar had a WPA of -0.09 with Boston in 2004. Yes, he was getting a high batting average, but his outs in clutch situations were costing his teammates victories more than his hits were earning wins. With the Cubs Nomar was +1.78 WPA.
Well, that's a gauge. The Wald stat is, IIRC, a squared ratio, but the square root of 2,000 is still significantly larger than the square root of 25.
On August 1, 2004, Joe Sheehan wrote at BP:
"I think they made this trade not because it makes them better, but because they didn't have it in them to stand up to Garciaparra, who by most accounts had been a jackass since the Alex Rodriguez trade fell through. I rarely—perhaps never—factor non-performance issues in to my analysis, because they tend to be filtered through the press and tailored to create the best story. In this case, I'm convinced that this trade happened because Garciaparra wasn't going to come out of his full pout until he was dealt or filed for free agency."
zzbillfitz isn't writing anything that BP authors weren't writing at the time. I regard all BP authors--even Joe Sheehan--as both credible journalists and serious analysts. I see his words as a fair comment. I followed much of the 2004 Red Sox season game-by-game, and I agree fully that Nomar's attitude was a problem.
Marc, good article. As an older fan, I remember where I was when I heard of the trade, too, and I remember reacting much the way that your father did. It's not easy to write in a way that stirs memories that way: well done.
"As Baseball Prospectus’s own Joe Sheehan stated, “I don’t think that the Sox are a better team today than they were Friday (before the deal)… I think they made the trade not because it makes them better, but because they didn’t have it in them to stand up to Garciaparra…"
Joe Sheehan was a Yankees fan first and a BP writer second. It shows here.
Attitudes aside, both Nomar and Millar were horrible defensive liabilities in 2004, and Theo plugged both gaps in one trade, giving up just one free agent-to-be and one prospect who never made it. As a rational Red Sox fan in 2004 (a role verging on oxymoron status), I thought that the trade made sense the next day, as a calculated risk if not as a sure thing.
An excellent question, and one that becomes more significant when the author chooses to ignore it.
What is the entire regression equation, and what are the associated statistics?
Remember, Theo Epstein did not make the Hanley Ramirez plus three prospects for Josh Beckett/Mike Lowell trade. That trade was made during his hiatus. He's been very careful never to say whether or not he would've made that one.
What would've happened? Well, the Red Sox released Carlos Pena to keep room for Lowell, Youk and Big Papi, so without Lowell Boston probably would've had Carlos Pena, an All Star first baseman, after the midpoint of the 2006 season. (Jeff Bailey or a different replacement-level player would've started 2006 at first base for Boston.) AJ Burnett signed as a free agent for less than the salary of Beckett and Lowell combined, so Burnett probably would've gone to Boston instead of Beckett. That still leaves salary room for one more free agent signing for Boston in the 2006-2008 timeframe, so I'd suggest that Boston would've contended in 2006 as well as 2007 and 2008 with Hanley Ramirez, Carlos Pena, and AJ Burnett on the team instead of Lowell and Beckett. Over those three years, I'd expect Boston to have had about three chances in eight to win a World Series. That's less than the guaranteed win in 2007 (given hindsight), but it's more than the zero percent chance of winning they enjoyed (with hindsight) in 2008 and, especially, in 2006. Going forward
Third-Order Marginal Effectiveness (TOME)
If you want to know how well a Front Office is doing, it's all in the TOME.
You're right: I pulled up the wrong column of data. Thank you!!
Yes, but MLB as a group exceeded the .009 difference in BABIP shown by your right-handed slugger sample in several years. I'd grabbed 2007 as a middling year for Howard's career, and MLB as a whole hit for .011 more BABIP points with baserunners that season. Looking further, the BABIP difference varies more year-to-year than I'd guessed, though, and it's often below .009, the difference for the dozen RHB you studied.
The BABIP differences with runners on base from your sample were as follows:
Name (BABIP without baserunners, BABIP with baserunners, difference)
Jason Giambi (.327, .277, +.051)
Barry Bonds (.308, .267, +.041)
David Ortiz (.324, .283, +.041)
Todd Helton (.346, .327, +.019)
Ken Griffey (.297, .285, +.012)
Jim Thome (.327, .317, +.010)
Carlos Delgado (.303, .303, .000)
Larry Walker (.331, .332, -.001)
I don't see it as noise, given the distribution of the differences; I see it as a clear bimodal distribution of two sets, one with a median of .010 and the other with a median of .041, not as a single set with a mean of .022. Frankly, given the frequency of extreme shifts for these various players, a bimodal distribution might make your point regarding Howard better: I think that Bonds, Giambi and Ortiz have had extreme infield shifts played against them more often than the other five players you mentioned.
But I'd expect, then, for Howard to show a decline in overall BABIP on ground balls concurrent with the fairly swift introduction of the shift in 2007, making a higher overall BABIP in clutch circumstances possible. Here are the numbers:
Year / Howard's BABIP on GB
I don't see any change coinciding with the introduction of the shift.
Eric Seidman attributed Howard's high BABIP with runners on base as indicative of his disproportionate ability to make pitchers pay for mistakes, and the correlation between pitchers' mistakes and runners already being on base. There may be merit to that concept, too.
You know, Will, given the information that came out below this in the discussion, I'm now unsure that it was unfair of me to expect a personal anecdote about McCutchen from Brian this week. Yes, expecting it in 24 hours might've been a long shot, but if BP is aiding Idol writers in gaining access to MLB teams, then the authors have an obligation to make best use of that access.
A quick check of Ryan Howard over his career shows (if I've run the math and referenced BR correctly) a difference of only .016, not .073, for his career BABIP with men on base, a BABIP of .316 with no men on and .332 with men on. I believe that the BABIP for a man on first and nobody on second is .341, and that the BABIP for all other on base scenarios is .325, just nine points higher than his BABIP with no baserunners.
Certainly your linked work regarding the success of eight left-handed power hitters was interesting, but Howard apparently differs far more from the mean for the group of LHH you studied than he does from the mean of all MLB players. (As an aside, almost the entire effect you discovered was the result of just three players in the sample set: Bonds, Ortiz, and Giambi.) Also, much of the effect for Howard is the effect of the first baseman's holding a runner at first with second base open, the effect I regarded as previously well-known.
Regarding steroids, it's unusual that a given player would have quite so many of his PECOTA comparables either directly linked to PED use or as hitters who peaked in the "Steroid Era," regardless of personal usage. I'd suggest that careers seemed longer for power hitters in particular during those years, and that changes in the game--either drug testing or otherwise, such as possible changes in the baseball itself--may mitigate against Howard's managing "to lead the league in homeruns a few more times over the next several years." That's just opinion, of course, but not many players have quite so great a concentration of comparable players at that time. Notably, though, in 2009 the current PECOTA shows Howard more comparable to several players from before the Steroid Era, including Mike Epstein, Jim Gentile, Willie Aikens, Boog Powell, John Mayberry, and Don Mincher. As a group, these players did not excel once past their twenties, as Howard will be next year when he turns 30 this November.
Thanks for the perspective, Will. Perhaps I'm spoiled by David Laurila's success at getting unique tidbits from those with whom he does have the chance to speak.
I'd agree that Howard is underperforming at home, especially given that there's a home ballpark advantage across MLB regardless of what ballpark players call home. Thus far in 2009 the average NL player is hitting .262/.341/.414 at home and just .253/.324/.395 on the road, a difference of 36 points of OPS, even though the ballpark factor should even out. Players just play better when they're at home--but Howard just plays better on the road. My question is why, and it's just not answered here, even in the comments section. Obviously the trend was obvious enough that Howard has been given his off days at home, not on the road, given the raw numbers. It's a big factor, but Matt overlooks it.
As an aside, given how long the article seemed, I did a word count, and it came to 1,998 excluding words in charts and tables. The trouble I have is that words in charts and tables, if anything, slow down the reader more than words in the paragraphs forming the body of the work. Counting words in charts and tables put this piece over 2,100 words, and it seemed even longer to me as a reader.
The issue with "clutch hitting" because of higher BABIP with runners on base isn't new. BABIP jumps from .295 to .313 this year in the NL going from the "no baserunners" state to the "runner on first" state. That's normal for all hitters, even if it's a bit more pronounced for Howard.
Finally, the expectations regarding Howard's future are based on some very small-sample trends, as well as some spotty data: how reliable are listed weights for large MLB sluggers? Frankly, given the names included as Howard's comparables and the era in which many of them played, one is almost compelled to wonder if overlooking steroid use, effects, and aftereffects is enough to make gauging the future of Howard by this list of comparables an act of guesswork.
Many readers have voiced their strong support of this piece. Rereading it a day later, though, doesn't change my assessment: others did significantly better on this topic. I'm sure that Matt will be back next week, and I look forward to a strong article, but this one didn't earn a thumbs up from me for the reasons cited here.
Your article on the downside of chaining MLEs was, for me, one of the two best in this competition and one of the best I've read regarding any aspect of sabermetrics this year. You probably didn't have the words to show what you just posted in your article this week, but it's good stuff.
Brian, if I'd somehow have been gifted enough as a writer and an analyst to be sitting on the article that you wrote with 24 hours until the deadline, I would've tried to find a way to talk to McCutchen's mother and to squeeze in something personal about him for either the second or third paragraph, as well as something for a concluding paragraph. That .709 batting average with power in high school, coupled with his declining Isolated Power and Isolated Discipline as he approaches MLB, is ripe for a couple of lines about the frustration he might feel having once been a feared power hitter and his now being, at best, a guy with a "Ron Gant swing." I don't know if that would've been possible, but that's what I feel this article needed.
No mention of park factor for a guy whose primary skill is hitting home runs and whose ballpark is Citizens Bank Park?
Granted, over his career Howard has hit almost as well on the road as at home--with more home runs by raw count--but Citizens Bank Park was such a hitter's park early in Howard's career that I'd've expected some passing mention of how it affected Howard's stats...especially given that word count seemed not to be an issue.
I didn't have trouble following Oliver projections or normalized projections. The "Oliver" is similar to "Marcel" with a longer memory, and normalized projections adjust for league and ballpark, per the article. Brian's point, I believe, was that McCutcheon had outperformed expectations every single year of his career save 2007, and that regression might be more likely for such a player than it would be for one whose range of likely outcomes were better defined by his MiLB career. Furthermore, neither reasonable projections nor after-the-fact normalized stats showed McCutcheon as the peer of McLouth.
Will, you wrote, "This is by far Brian's weakest article of the competition in a week where I was really expecting him to shine." I don't know why you would have expected that. Brian's strength is synthesis of data we've all know to be available in a way that offers us knowledge we didn't have before. If the week's assignment had been a league profile instead of a player profile, Brian would've been hard to beat. Writing about the history and future of a single soul on the mound or in the batter's box is, to my mind, not Brian's strength.
Frankly, many of this week's entries lacked the style that's present when current BP authors write a profile. That's good. I want the skill of the current staff to show, just as I want the Idol writers, at their best, to do better in a paragraph or an article than the established marks of excellence for the BP Staff.
Kevin and Christina, you both wrote of "expectations." It's hard for all of us to separate expectations from round-specific standards, especially for those contestants whose work we know. But I feel that this article missed two things: in the first half, it failed to capture Andrew McCutcheon as a living human being distinct from his stats; in the second half, it failed to bring together all of the excellent statistical anecdotes into a comprehensive picture of what we should expect McCutcheon to become. There was a whole lot of great information, but it came in impersonal bits and pieces, not as a single coherent theme.
In a week where I gave just two thumbs up, this article didn't make the cut.
Brian, you surprised me. This topic challenged many of the remaining Idol finalists, and you wrote a piece that flowed smoothly from various statistical metrics through game-by-game descriptions to a final discussion of "genetic maximum" and how an MLB team could best use a player such as Cecil.
This was, by far, your best work contrasted to the performance of your peers on that same week's topic. Thumbs up.
"Kila Ka'aihue (pronounced KEY-luh Kuh-eye-HOO-a)?" Normally I have to learn how to pronounce the acronym for the flashy new stat, but in your case I have to learn how to pronounce the player's name.
Well done: this was my favorite of the week. I like the way you seamlessly integrated so many different sources while discussing a player whose mix of skills may reasonably leave him in either AAA or MLB.
I suspect that Boston would find it challenging to move either Lugo or Ortiz without picking up most or all of their respective contracts. To a lesser extent, that might go for Drew, too, but J.D. Drew is still a good right fielder with streaks of both elusive greatness and nagging injury. David Ortiz is a pure DH struggling to stay atop the Mendoza Line, and Lugo is a singles-hitting shortstop trying to keep his UZR/150 above -40.
This was a pretty good idea, IMHO.
I'd possibly hire all of the remaining writers, but by the time we get to the final five we'll certainly being voting away writers with the capability to contribute significantly. Whether BP retains seven, five, three or one of the remaining contestants, I'll be happy to see my subscriber dollars going to pay for those articles or Unfiltered pieces.
As an aside, I'd pay an extra ten dollars a year to see the remaining seven all brought on board...and, if you could, maybe that subscriber who used to post such long article comments under the pseudonym "Oleoay," too.
It's almost more interesting reading the comments than reading the (excellent) article.
The Law of Large Numbers suggests when sample size is big enough that a normal distribution will be "close enough" to binomial distribution, but just because np >= 10 it doesn't mean that the distributions will be identical. Eric makes that point in the comments. Furthermore, the discussion of Feliz where the number of trials was just four demanded use of binomial theorem instead of normal approximation.
While the point that plate appearances are not truly independent trials has merit and needs to be (and was) stated, almost every baseball statistic has, at its root, an assumption of independent trials, and it's an assumption that we usually accept. We might question "Wins" and "W-L Pct" as a fair stat for pitchers, but few of us dig into the available tables regarding strength of pitchers faced by batters when evaluating hitters by their OBP. Judging the likelihood of four consecutive walks isn't so great a stretch from one's acceptance of OBP as a statistic.
Once the decision to model plate appearances as independent trials is accepted, using binomial theorem is almost always a better modeling system than normal approximation for exactly the reasons Eric stated in the article. It may be very little better with large enough samples, but it's still better. With smaller samples, it's important to use binomial theorem if it applies.
The absence of mention of David Ortiz and Julio Lugo from this article demonstrates how the Boston Red Sox always get less media attention than teams such as the Rockies, the Royals, the Blue Jays, and the A's.
"The data may not necessarily be normally distributed, however, making the binomial distribution the more accurate measure."
I find myself wondering if this article was inspired by the discussion regarding one of this week's BP Idol submissions. In any case, great article. Thanks!
Matt, my understanding is that normal distribution is an underlying assumption for a t-test, and that binomial distribution approximates roughly, but does not truly model, normal distribution.
If you know statistics better than I do, which is certainly possible because my fields are engineering and business, not statistics, refute my points. This is a competition, I've nothing at stake, and I am cheering for you: if my understanding is imperfect, please educate me.
"How does a number like 517 convert to success in minor or major league baseball? That is an answer I can't honestly give. Conversion statistics are based on actual conversions."
When you had to write this, you lost my thumbs up. If you're going to write this article, you have to offer a rationale for an approximation to conversion.
I've thought for a long time that there are three roles at which the very best women might possibly compete in professional North American baseball:
2) Knuckleball pitcher
3) Second base
I was dying for you to come closer than a reference to Chad Bradford on how a particular female might break the gender barrier. This was a good topic. I wanted your article to succeed...for me it did not.
Had I seen the terms "MLE" or "MORP" (or "summation of MORP over duration of obligated service adjusted for salary considerations") I might've given this a thumbs up.
Brittany, if you're not the ultimate victor in this competition, try writing about what you know better. I can't help but feel that I know more about your topics than you've offered in your articles, but that doesn't mean that you're any less than excellent as a writer.
This article marks the first time that I felt critical of the BP writers' opinions. Assessing the success of a league's All-Star Team, regardless of the details of the selection process, is an excellent way of judging the talent of its best players. The disproportionate success of hitters over pitchers is important information. I'd have guessed that to be true, but I didn't know it 'til now.
Big thumbs up, Matthew.
Tim, thumbs up, and a great article.
A caution: the transition between the opening paragraph's style and the rest of the article almost lost me. The opening was good, and the body of the article was great, but I felt that the two did not fit together.
I see this as one of the best articles thus far. I'm impressed with the flexibility and wit Ken demonstrated.
I strongly concur with your criticisms. For me, it was enough that I didn't give the article a thumbs up.
If this were Baseball America Idol, thumbs up. BP is different, and, while there's overlap in readership, I don't subscribe to get to read this sort of article.
+1, molnar. Yours is a non-trivial point.
Thumbs up, Matt. Your article wasn't as good as last week's, but your next-to-last paragraph rocked. Great job.
Better, but not much better.
Brian, you did what you had to do: you demonstrated that you could carry an article without heavy statistical analysis. At some points your style reminded me of Joe Sheehan; at others you reminded me a bit of my childhood memories of Curt Gowdy.
Still, I see you as Yaz stuck in the 1972 season, and I see this article marking the second half of June. Everybody knows your past accomplishments and your talent, but there's a lot of good young talent in the league, and suddenly this year your own productivity hasn't met your reputation. Yaz had a tough start to the 1972 season, including an injury costing him a month of playing time. On June 15, Yaz ended the day hitting .216/.315/.243, a dismal small sample size start to his season. In the second half of June, 1972, Yaz did everything but hit a home run, and all of New England tuned to the Red Sox Radio Network each night on AM radio, cheering all the singles that gave Yaz a .404 batting average the last half of the month, but wistfully saddened because the home run never came.
Good article. Thumbs up. I'm waiting for the home run...from you, I'm waiting for the three home run game.
Pardon: Wakefield is 115 wins from 300 as I type, "hypothetically" putting him in range at age 50. I overlooked momentarily his 14 wins with the Pirates.
This reminds me -- Tim Wakefield is only 129 wins from 300. At the pace of 13.5 wins per year he's established over the past two years (a pace he's exceeding thus far in 2009) Wakefield would be within reach of 300 wins in his age 51 season in 2018.
Tim Wakefield "seems to have no chance at all right now," but, upon reflection, we don't know how long modern sports medicine will enable Wakefield to throw a 68 mph knuckleball. My money would be on Felix Hernandez, who will probably be racking up 22-win seasons with the Yankees in a few years, to be the next 300-win pitcher, but Wakefield might have as good a chance as a few of the names mentioned by Jay in this article. Barry Zito or Tim Wakefield on total career wins? I'll take Wakefield.
Doesn't Felix Hernandez have a JABO of 344, higher than that of anybody you've listed?
ofMontreal, too kind, especially given my earlier typo, but thanks.
Regarding gauging replacement level over time, the best metric with which I'm familiar is tracking pitchers' hitting. The article here at BP I liked best was written by Dan Fox. Here's the key graph, which pretty much speaks for itself:
And here's the whole article:
There's some evidence of recent improvements in replacement level, but I think that changes are coming at a lesser rate than they were in and around the first half of the 20th Century.
I caution, however, that some recent BP articles fit this format. It's not enough to change my vote, but it's enough that I wonder if Tyler was misled.
Jeff, this wasn't what I was hoping to read from you this week.
How do these potential trades affect fantasy owners? Can we quantify through analysis which players are most likely to be traded? Can use of the Retrosheet trade database show anything useful?
I hate to write this, but I perceived yours as an article without value this week. Your previous work has been better, but votes go week-by-week.
I'll bone up on Pivot Tables and VLOOKUPS and give you a thumbs up. You're aiming for the high end of the BP community with regard to statistical analysis, and, even if I don't normally use those Excel functions, I want a writer capable of writing at this level to win.
Brittany, I've yet to see an article of yours that I liked. Indisputably, you're an excellent writer. I'm just not sure that you're best suited for BP, where subscribers such as me value quantitative insight over the finer points of writing. You know baseball; you know stats; it's just that your vision of how to provide new insight through statistics may be less than I want as a BP Idol voter.
Ken, your topic was far from expectations but within the topic, you offered rational analysis, and, most importantly, you made me smile.
Well done. Thumbs up.
Christina nailed it. This article is too limited in its value without further analysis.
My weakest thumbs up of the week. I don't mind when Brian gets into challenging analysis. What I mind here is that the analysis is good, not great, but that the writing still isn't as good as that of his best competitors.
I enjoy reading articles from those who know what stats do tell and don't tell. I liked the article. More importantly, I now know that Matt is smart.
Thumbs up, with credibility in future rounds.
That's "thumbs-up" with a "b' both silent and hidden.
Kevin Goldstein hit it: "Good-not-great"
This is the best article to which I didn't give a thums-up.
Wow, Rob, tough crowd yesterday.
I thought that you raised a mildly interesting point, and I thought that you wrote well enough that I saw to need to analyze the quality of the prose...at least until I hit the comments. Whoa.
But your core point was, as I read it, this: baseball is a sport challenging enough and varied enough that skills coached as "fundamentals" can be momentarily forgotten by even the top 750 players* on the North American continent. Fans need to understand that; retired veterans paid to banter about the games need to realize that. The frequency of criticism of fundamentals suggests a mindset that modern players are too lazy to work, and that would seem to be far from the truth: if there's been any dip in the level of talent at MLB, be it median level or replacement level, I've not seen the analysis showing that dip.
I'm reminded of Ty Cobb having had the gall to claim that Joe DiMaggio was one of the few players of the 1930's and 1940's who could have made it as an MLB player when Cobb was playing. That's absurd: a simple check of pitchers' batting stats suggests that the replacement level was much lower before World War One than it was on either side of World War Two. While changes in replacement level are probably less extreme in recent decades, there's still nothing to suggest that modern players are worse because of inattention to fundamentals...while there's more than a bit of evidence that Ty Cobb might not have posted such gaudy batting averages if he'd been forced to face, say, the same pitchers Ted Williams had to face.
On a personal basis, there are players whose fundamentals I criticize, but I do it a whole lot less ever since Manny Ramirez went to LA and decided to try (trying so hard that he was suspended for his efforts this spring). I don't think that I've voiced a fundamentals criticism this season, and I'll be less likely to do so now.
Good article. Thanks!
* Now that Matt Wieters, too, is on an MLB 25-man roster.
I'd like to be able to do so easily because BP made available to Premium Subscribers the "career" search function.
Maybe both your source and my source are correct by the algorithms each choose. That's why I try to cite the external sources I use, and that's why I try to use BP stats when posting here: park factors can vary by definitions such as actual runs scored or runs created, or runs per game or runs per out, as examples.
Yes, the "600 Club" and a new press box were added in 1988.
From 1967-1987, Batter Park Factors* at Fenway ranged from 118 (1977) to 99 (1986/87) with a median of 107.
From 1988-2008, Batter Park Factors* at Fenway ranged from 111 (2007) to 97 (1997) with a median of 105.
There's a lot of talk about how the change affected air currents, and I guess that there's maybe a little change, but Fenway is still a hitter's park most years. In terms of the article, though, the HRpf hadn't changed: it was 1.02 both seasons cited.
* source Baseball Reference
"To use an Idol metaphor, Brian doesn't have that low range that we were looking for, but he's got a voice."
Brian Cartwright has one heck of a voice!
Every Week One article was good, but Brian tackled an issue currently consuming the baseball writers' attention and he offered unique value to the readers of BP. Perhaps he wrote at a level that could have been misunderstood by some baseball fans, but he wrote at a level easily understood by the readers of BP and he offered that for which we seek, insight missing from sites such as ESPN.com, MLB.com, or our local Sunday newspapers.
I gave just one thumbs up this week. Congratulations, Brian: you got my vote. Superb analysis for working on so short a deadline!
Thanks for reading and considering my thoughts on the issue, Richard!
Richard, I see two key points you're trying to make: first, that Ibanez had no incentive to use PEDs in KC that he didn't have in Seattle; and second, that it's not unusual that four players coming together mid-career would all have late career peaks if one considers role changes. Let's look at those one at a time.
First, there's an obvious reason that Ibanez might not have risked PED use in Seattle and then chosen to consider it in KC. The Basic Agreement holds players to near MLB minimum salary for three years, but allows for players to go to arbitration and, eventually, free agency after that. Ibanez's rise as a hitter neatly coincides with his coming into his later years of arb eligibility. Setting aside the possible career concern a player would have in being traded away from the contending Mariners to the struggling Royals, suggesting that his skills possibly were no longer adequate for a contending team's roster, Ibanez had an opportunity for greater compensation through acceptance of risk in KC that simply wasn't there in Seattle.
Second, you made me look up Keith Woolner's 1997 study regarding hitters' peak ages. Given his relationship with BP and my attribution of source, I'm hoping that he won't mind my reproducing his key table here:
AGE PEAK %Peak Cumul%
18 1 0.1% 0.06%
19 5 0.3% 0.38%
20 15 0.9% 1.33%
21 29 1.8% 3.17%
22 55 3.5% 6.65%
23 87 5.5% 12.16%
24 90 5.7% 17.86%
25 148 9.4% 27.23%
26 160 10.1% 37.37%
27 163 10.3% 47.69%
28 155 9.8% 57.50%
29 138 8.7% 66.24%
30 141 8.9% 75.17%
31 107 6.8% 81.95%
32 91 5.8% 87.71%
33 68 4.3% 92.02%
34 46 2.9% 94.93%
35 25 1.6% 96.52%
36 22 1.4% 97.91%
37 13 0.8% 98.73%
38 9 0.6% 99.30%
39 6 0.4% 99.68%
40 1 0.1% 99.75%
41 1 0.1% 99.81%
42 2 0.1% 99.94%
43 1 0.1% 100.00%
As I wrote above, the four players I cited had their peaks at ages 31, 34, 34, and 37. Only a quarter of hitters' peaks come after age 30; only eight percent come at age 34 or later. That counts all role changes.
But let's look at something else: let's look at how other KC players in the same age range as these four players in 2001, 29-33, fared with respect to peak age by WARP3.
Player Peak Age (WARP3)
Joe Randa: 29 (7.7)
David McCarty 30 (2.1)
Hector Ortiz 30 (1.0)
Cory Bailey 30 (1.4)
Rey Sanchez: 31 (4.0)
Brent Mayne: 31 (3.8)*
Of ten players on the KC Royals in the just-past-peak age range of 29-33**, all ten had their peaks at age 29 or later. At first approximation, a back-of-the-envelope estimate, the chances of that, using Woolner's chart, look to be around (1-0.5750)^10, or roughly a one in 5,200 chance. There are a number of reasons that this could be an underestimate of the odds: pitchers weren't considered by Woolner, there's selection bias in the sample because all of the players were still active at age 29, and selecting players from the Royals might lead to enough changes in roles to affect the odds. Conversely, it's not as if these players all peaked at age 29. Three of them, including Ibanez, peaked much, much later, and the median peak age in the cohort was 31. That would make these late peaks more unusual than the back-of-the-envelope estimate suggests.
When I see rough estimates of something being a less than a 0.1% possibility, I'm not easily swayed by suggestions that it's attributable to a change in role, especially given that truly improved performance can precipitate a player looking better in their new role, possibly making change in role a correlation but not a causative factor. If you can show me anything that backs up your suggestion, though, Richard, perhaps with many late peaks from 29-33 year old players on the same-era Pirates or the Devil Rays, or other teams that made part-time players starters, I'm eager to read it. But I've done my utmost to explain the numbers behind my posts, Richard, and the case is pretty robust.
In summary, I see salary structure of the Basic Agreement as giving Ibanez motivation to use PEDs in KC but not in Seattle, whether he used them or not. I see a very unusual group of late peaks by players aged 29-33 on the 2001 KC Royals. Accordingly, I continue to see Ibanez's late peak as less than unusual for players from the 2001 Royals, for whatever cause, and I discount choice of role as the probable reason for the late peaks.
This doesn't prove PED use. I still find the numbers interesting.
Caveat: While I have endeavored to be thorough with all research and math, I am human, and humans are fallible, and I've not had access to an editor. Apologies if there are any typographical errors, especially any that might have influenced my conclusions. Also, apologies for any math errors: I last took statistics before some current BP authors were born, and geezers may occasionally err.
* Peak achieved not with the KC Royals, but with San Francisco Giants in 1999. All other players listed peaked during or after their time with KC.
** Sal Fasano, who had one PA with the Royals as he passed through between trades and AAA time, not included.
What sort of turnaround would be considered "remarkable?" You didn't define that, and it makes it challenging to defend my point against your criticism. I'd offer, though, as a starting point, that MLB players are usually more productive in their twenties than in their thirties, so I'd suggest that a career where a player had his peak season at or after age 30 and achieved more aggregate value at and after age 30 than before that age would be remarkable. Let's use WARP3 as a metric, and let's actually check stats. We'll check peak seasons and aggregate WARP3 before 30 and after 29.
Peak overall: 4.5, age 37 (2009; his previous peak was a 4.3 in 2008)
Peak before 30: 0.7, age 29
Before 30: 0.8
After 29: 25.8
Ibanez had a remarkable turnaround, but we agreed on that.
Peak overall: 7.6, age 31
Peak before 30: 2.9, age 27
Before 30: 2.8
After 29: 15.1
Byrd had a higher difference between his pre-30 and his post-29 peaks than Ibanez did, and he was over five times more valuable in his thirties than he was in his twenties.
Peak overall: 2.8, age 34
Peak before 30: 1.5, ages 22 and 26
Before 30: 0.7
After 29: 7.6
Grimsley had a far lower peak than Byrd, but his best season after 29 was almost twice as valuable as his best season before 30, and he was over ten times more valuable in his thirties than he was before.
Peak overall: 2.8, age 34
Peak before 30: 1.7, age 26
Before 30: 4.9
After 29: 12.3
Zaun's was the least striking of the turnarounds I cited, but it was still remarkable: he peaked at age 34, and his value in his thirties was two and a half times his value in his twenties, despite his reaching MLB at age 24 and acquiring over five years' service time before age 30. That very rarely happens.
As an aside, Zaun's EqA went from .200 and .229 his last two seasons before coming to KC to .273 and .294 his two seasons with KC. You suggest that Zaun's improvement was due to his quitting drinking before joining the Blue Jays, but his EqA with Toronto ranged from .264 to .285, slightly below the range he established with KC. Looking at numbers, I don't see your point regarding the allegations of drinking.
Frankly, looking at numbers, I don't see why you don't consider these career turnarounds to be remarkable, either.
I'm wondering if it's just a perception, but it seems that more sluggers are having trouble this year than usual.
Does anybody have a handy benchmark of how many established power hitters, by some metric of "established power hitter," are usually slugging poorly a quarter of the way through most new seasons?
"PEDs are not the reason for his rebound."
Hmmm...that's a statement of fact. Are you sure?
You know, a few people dropped hints here, and, yes, Grimsley, Byrd, and Zaun all either admitted PED use, were suspended for PED use, or were named in the Mitchell Report. They also played alongside Ibanez at the moment all four turned their careers around.
Ibanez might have had injury issues with the Mariners, but in his last two full AAA seasons, ages 24 and 25, he only posted OPS of .783 and .847 in the PCL, and given the amount one discounts PCL hitters' stats it didn't look as if he'd ever make it as an MLB corner outfielder. He then posted a .241/.295/.383 career line in 518 plate appearances through his age 28 season, as Geoff Young pointed out. That's five years of replacement-level hitting, give or take, right in what should've been his prime. At age 29 he went to KC, and there, alongside Byrd, Zaun, and Grimsley, he started a very odd career trajectory.
Here's why I posted my original comment: yes, there are odd career trajectories, but it's very, very rare to see four of them take off simultaneously. If there's one chance in ten that such an event would happen on any team in a given year, it's a one-in-10,000 shot that four players would independently start remarkable second career halves on one team. That's what happened, though, on the 2001 Royals.
I fully agree that odds and coincidence prove nothing. I would, however, want to make sure that it's known that Ibanez turning around his career wasn't an odd event for 2001 in Kansas City, whatever the cause.
What's really amazing is that Ibanez made his turnaround after he went to the Royals, and that the team had four players who had the same odd career pattern of playing their best years, by far, in their 30's: Ibanez, Paul Byrd, Jason Grimsley, and Gregg Zaun. All four of those players joined the Royals in either 2000 or 2001, and all four improved significantly with the Royals.
I keep thinking that there's a common link between Byrd, Grimsley, and Zaun that might explain their career turnarounds, but I can't quite put my finger on it. In any case, if there's no common cause it's an absolutely mind-boggling circumstance that four such late bloomers played together on one team in 2001.
You know, the more hours that pass after my reading this, the more I understand how important this article is. I understood the first time what you'd proven, Brian, but I'm getting a better grasp with reflection of why your findings are important.
Thanks, Brian. I'm realizing that this is not just a great contest entry, but that it's also going to be one of those articles that might influence thought in the community for some months to come. It took me a while to appreciate what you'd written, but it was well worth the time.
Bryan Price has interviewed for MLB managerial positions in Seattle and Florida, and he may have felt slighted at being passed over for the job upon Melvin's departure. CBS called Price a "Melvin loyalist;" that may also play a part in his decision to quit.
Were I in an MLB Front Office, I'd be very reluctant to hire Price now. Had he quit without comment, it would have made his point. Publicly criticizing the credibility of the man his boss's boss selected to be his manager insults the Front Office and ownership team more than it insults Hinch, and one doesn't risk hiring individuals who lack professional demeanor for jobs in the public view unless their credentials clearly outweigh the risk and cost of a media fiasco.
In any case, I liked the article, and I hope that Hinch makes the fans of the Diamondbacks happy with his management. Best wishes to Bob Melvin, who got far more from his team than Pythagoras could have dreamed in the 2007 season.
My concurrence. All of these articles are excellent, just as I'm sure Oleoay's was. I suspect that it was a difficult call.
As an aside, if this is ever repeated, either offering a little more time for creating the qualifying article or disqualifying previously published work would be helpful.
Brian, this is an excellent article but a rigorous article. It's tough communicating the results of so much research in so short a piece. Furthermore, while I enjoyed your research and your results the most, I'm not sure that BP is aiming for articles that demand quite so much concentration and study to comprehend. (Well, articles that require so much concentration and study for ME to comprehend...)
In any case, thanks for sharing the research, and best of luck: I know your work, and I want the winner of this contest to be a writer capable of bringing to BP the sort of insight and analysis that one always finds in your articles.
Is the organizational skill of selecting good bench talent, back end starting pitching, and relief pitching something that varies directly with FO talent? Can it be repeated year after year? If one takes away payroll and its influence, is there any trend attributable to FO talent remaining?
Tim, you've written a good article, one clearly in the top half of the selected submissions. I'm just not sure of your conclusion: I don't know if you've discovered decision-making talent, luck, or payroll.
Actually Wang made it into the fourth inning his first start of 2009, but your point still stands.
So does Christina's point, though: disregarding Wang's transition innings and their 24 runs, the Yankees are still the fourth-best AL East team at transition inning runs allowed. If one deducted three bad transition innings from each other AL East team, as well as removing Wang's transition innings, I'd expect that the Yankees would return to fifth place.
We're overlooking the comparable pitcher who offers the most hope: Jack Quinn, who also started six or seven games (accounts differ) at age 46 for a Philadelphia team, the 1930 Philadelphia Athletics. Jack Quinn was shifted to the bullpen to become a Depression-Era equivalent to a closer, finishing 16 games and earning six saves, had they been counted. The next year, at age 47, he played for the Brooklyn Dodgers and earned enough MVP votes to finish 17th, not a bad showing for a guy two decades older than the average MLB pitcher of his day. Here are his WARP3 totals by age in 1930-32:
Jack Quinn pitched effectively through his age 48 season, but he posted an ERA over 4.00 at age 49 and left after 14 games that year.
Seriously, though, Jamie Moyer was doing adequately until he gave up 14 runs, six on home runs, in 6.2 IP his last two starts. In both of those games there was a strong wind blowing out to CF. Any pitcher is likely to see his ERA inflated in such conditions. Moyer had an ERA of 5.65 entering those two games, and a 5.65 ERA, while sub-par, is probably better than replacement-level this season.
I see two other reasons for hope:
1) Moyer's HR/FB rate over the whole season thus far is 22.9%, roughly double what one would expect. It's unlikely to stay that high.
2) Moyer has actually lowered his LD% allowed this year relative to 2008, but he's being killed because his fielders are allowing opposing hitters a .353 BABIP on ground balls. Maybe the ground balls are being hit harder than usual, but .353 is around a hundred points higher than MLB norms, and over his career Moyer has allowed only a .212 BABIP on ground balls. Furthermore, in seven 2009 games featuring plenty of opposition baserunners, his defense has turned only two double plays to support him. I'd expect a big change in hits allowed once Moyer's past the current period of flukishly bad infield defense.
An ERA of 8.15 after 35.1 IP is bad, but Chien-Ming Wang will probably have an ERA at least that high if and when he reaches 35.1 IP, so Moyer's not the only well-regarded MLB pitcher struggling that badly this season. Yes, one could say that Wang has better comparables, but it's possibly more accurate to say that Moyer has NO comparables: Niekro, Wilhelm, Paige and Ryan wouldn't even count if there were any reasonable number of guys who'd at least tried to start in MLB at age 46.
I expect Jamie Moyer to be better than replacement level from this point forward. I don't know that he'll be enough better to pull his overall season stats back within normal levels for MLB starting pitchers, and I don't know how long MLB GMs will give a guy Moyer's age a chance to pitch if he struggles. Still, I think I remember reading that MLB fans of every team are supposed to be entitled to hope and faith, so I'm not yet giving up hope on Moyer.
While purists demand three years' data, not nine games' data, before assessing fielding, the "Scout's perspective" that Lugo isn't looking good is strongly supported by UZR. Over his career, despite a tendency to errors, Julio Lugo has had a UZR/150 games of +3.8 runs at shortstop, suggesting that's he's been just barely better than average. His UZR/150 for 2009 is at -48.8 runs through his first 67 innings at shortstop. If that's not a fluke - and his FRAA of -2 and his DFT Fielding Rate of 73, both calculated very differently than UZR, suggest that it's not a fluke of the UZR system - until Lugo's knees are better, the Red Sox are better off with Nick Green, or maybe even Gil Velazquez or Dustin Pedroia, playing shortstop.
Clay, thank you. It's good to hear that you're better.
Best wishes on continued great health from here onward.
Thanks for a good analysis, and pardon my weighing in a day late.
1) David Ortiz has been having trouble hitting at night this year. Over his career his OPS has been only six points by day, but this year, despite a higher BABIP at night, his OPS is 87 points higher in the daytime. If that were to continue, I'd suspect vision problems.
2) Over his career, Big Papi's SLG and OPS have been highest, by far, when he's pulled the ball. This year a majority of his balls in play have been up the middle, where he's slugging just .300 (remember, this doesn't count strikeouts or foul outs, so that's even worse than it sounds), and ten of his other fifteen hits have been to the opposite field. Almost a quarter of the way through the 2009 season, Ortiz has only five hits pulled to right field. Two years ago fans would have expected twice that many home runs to right field by now, as well as a dozen or more doubles into the corner and a few assorted singles through the shift. Big Papi ha a lifetime SLG of .958 on balls pulled to right field. This year he's slugging just .471 on balls pulled to right.
My suspicion as a fan watching games on TV is that normal aging and wrist injury challenges are being exacerbated by vision trouble. I'm not a pro scout, nor am I a BP-caliber analyst, and I've heard neither group mention vision issues, but my suspicions remain.
I believe that the MLBPA was asking/advising/directing its members not to correspond with the Mitchell investigation, and one could cite that as the reason that Sosa chose not to cooperate.
Notably, though, of McGwire, Bonds, Palmeiro, Sheffield, and Sosa, the five players cited in my quote above, all but Sosa have been linked to PED use.
That said, back again to The Onion, which is indisputably more fun and potentially more informative than whatever I might post regarding PED use in MLB.
In John Perrotto's defense, here's an excerpt from the Mitchell Report:
"...Canseco, who repeated the allegations from his memoir, said he had knowledge of McGwire’s alleged use of steroids. Through his personal lawyer, I asked McGwire to meet with me for an interview about these issues, but he declined to do so. I then sent his lawyer a list of specific questions about whether McGwire had ever used steroids or other performance enhancing substances without a prescription during his major league career, in the hope that McGwire would be willing to provide a response outside of the context of an interview. Neither McGwire nor his lawyer responded to that letter. (I sent similar letters with specific questions to lawyers for Barry Bonds, Rafael Palmeiro, Sammy Sosa, and Gary Sheffield, none of whom provided answers to my questions either.)"
Some other sources, such as LIFE magazine, considered Sosa's refusal to deny steroid use as involving himself in the issue. Most news sources agree with you, cams68, that there's no evidence, and I know of nothing except this refusal to answer George Mitchell that would qualify as evidence, but John's not alone in his position.
Of course, The Onion Sports Network pictured Sammy Sosa returning to Spring Training in 2007 carrying his man-bag full of steroids, but there are occasional claims that, despite The Onion's popularity, not everything written there is entirely accurate.
After several Red Sox perceived that Joe Girardi was threatening players, Joba Chamberlain drilled Jason Bay without repercussion. Earlier this year Josh Beckett might have, but didn't, hit Bobby Abreu after he called time while Beckett was winding up, and it resulted in a suspension for Beckett. From the perspective of the Red Sox, that must be frustrating.
David, good interview, as always. Thank you.
Listening to Blyleven speak, I've always had the perception that he almost hides a very sharp intellect by using simple words and brief, to-the-point sentences. I feel that way reading this Q&A, too. You certainly captured Blyleven's "voice" in this interview.
Eric, thank you. The number-crunching for this article was considerable, but you've demonstrated some points regarding LOOGYs that many of us have claimed as likely fact without possessing the exact research to back it up.
Now that you've demonstrated the importance of Spec% and REqA vs. same-side hitters, where will we be able to find them in the BP sortable statistics? I'd like it as an automatic part of the Relievers Expected Wins Added Report, but I'm certainly happy to have the stats available anywhere in the menu.
"Can't prove it?"
One cannot prove that stat changes with age are the result of age, but it's generally accepted in the community. One cannot prove that stat changes with change of home ballpark are the result of home ballpark, but, again, it's generally accepted. One cannot prove that personal joy or sorrow changes stats, but it's generally accepted in casual discussion of baseball players, just as the joys and tragedies of life are accepted to affect all of us.
Jason Varitek batted .155/.247/.233 in 39 games between his return to Boston when his wife Karen moved out and his filing for divorce on July 28 last year. He was hitting .269/.349/.462 before she left him, and he recovered to .225/.333/.362 after that, normal decline for a starting catcher over the course of a season. One cannot prove that his decline from a quality MLB catcher to an inadequate backup catcher in June and July was the result of personal issues, but the coincidence is more striking than the norm for changes in either age or ballpark.
Kevin Youkilis used to have a reputation among some fans for enjoying the nightlife and staying out late, and, perhaps, tiring over the course of the season as a result of his social life. Here is the difference between his first-half OPS and his second-half OPS for each season of his MLB career, 2004-2008:
2005: -.045 (only 22 PA in second half)
While one "can't prove it," something changed, and the key difference that I can cite is that Youk proposed to Enza Sambataro between the 2007 and 2008 seasons. If you see another key factor that would increase Youk's OPS by 115 points and increase his second-half OPS by 251 points between the two years, I'm always eager to learn. It's not trivial, and it's not a tiny sample size, and I've heard of no other factor that changed that should have made such a difference.
If you choose to maintain that I "can't prove it"...well, Oleoay, you make many posts here, and it strikes me that you have, on several occasions, made accurate observations that meet reasonable standards of readers' interest without meeting rigorous standards of proof. If you'd choose that I challenge all of your posts of that ilk with correlations to the rise and fall of politicians and silly potential consequences, I could do that, but I'd rather that you reconsider the possibility that, on occasion, significant changes in lifestyle might herald significant changes in performance, and that Kevin Youkilis might, possibly, be a player for whom such a change might apply.
Steven, thanks for a great analysis.
As a Red Sox fan, I've noticed a correlation that wouldn't be as visible to a student of the game reading the national media. Kevin Youkilis became a better power hitter and showed less of a late-season decline when he started dating Enza Sambataro, who is now Mrs. Kevin Youkilis. Enza is a beautiful, intelligent woman who handles herself gracefully with the media, and I would expect that it's possible that Kevin Youkilis is simply "a better man" for being with her, "better" being more tangibly measured in VORP and MLVr for an MLB player than it is for most lucky husbands.
Our workplace successes and failures are functions of both our professional skills and our personal ability to focus and to succeed. Marital relations factor into that. One can look at Jason Varitek's collapse in June and July last season as evidence of what marital challenges can cause: his horrific slump correlated directly with the span between his wife moving out on him and his eventual choice to file for divorce. I'm sure that there are other cases like Tek's. I feel that Kevin Youkilis is an example of the converse of this; I feel that Enza Sambataro's influence on his life has made him a better ballplayer. I know that his run production has improved while he's been seeing her, because this article documents that change perfectly.
The loss of Kevin Youkilis does hurt the Red Sox badly. Youk holds together the infield defense by being both a solid first baseman and the most-capable backup to Mike Lowell at third base.
Still, I don't think that Boston would even consider using David Ortiz at first base in anything but an interleague game without a DH. Jeff Bailey is an adequate first baseman, and Terry Francona seems to like giving him playing time, so I'd expect to see Jeff Bailey getting starts at first base. If Youk were put on the DL, I'd expect to see Chris Carter back with the team for a platoon arrangement at first base. Carter has a bad defensive reputation, but he's done much better defensively since leaving the Diamondbacks organization, and I'd expect him to be, if not a great defender, better than David Ortiz with his glove work.
I can't wait either. Let's do it!
Disco Night: BP Idol contestants link their article of the week somehow to disco. They can write on lessons of the 1973-1979 era, especially if they can find something from then that was suddenly and totally rejected in the early 1980's, only to find a few new practitioners of the skill in the last decade. They can write on "Thank God It's Friday," teams and players who seem to excel at weekend play. They can write on "Saturday Night Fever," either players who are exceptionally good in Saturday night games or the ten Italian players who have just looked the best playing the game. They can write on "The Hustle," the successes and failures of trick plays at the MLB level. They can write on "Stayin' Alive," the teams on the verge of elimination in September who somehow stayed in contention for far longer than anybody expected, or players like Tony Fossas whose careers extended far beyond what PECOTA might have imagined. They can't write on "Rock the Boat," a review of umpire objectivity and accuracy using detailed splits of Pitch f/x data, because there are still some taboos in the sabermetric baseball media. But if anybody really wanted controversy, writing on "Got To Be Real," a study regarding exactly how effective PEDs really were in the Steroid Era, would be far more ground-breaking than a study on sliding home versus running through the catcher titled "Get Down Tonight."
Will, thanks for the note. I'm looking forward to reading the competing articles.
Given that I'm getting negative feedback on the simple question, I'll explain further.
Joe Mauer is a good defensive catcher. Using FRAA adjusted for all time, he's at +40 after four full seasons, about +10 per year in his uninjured youth.
Most of the catchers I mentioned were about that good in their primes. Bill Dickey earned +50 FRAA in his best four years. Mickey Cochrane was +55 in his best four years. Thurman Munson was +54 over his best four years. Mike Scioscia was +50 over his best four years. Because these players didn't all have their peaks in their youth, there's a potential bias, but there's no guarantee that Mauer will ever get better than he is now, either.
Some of the other players I mentioned, including Terry Kennedy, Ray Fosse, and, especially, Shanty Hogan, were objectively worse than Mauer by the metric of FRAA. Still, I feel that the overall defensive performance of the catchers I mentioned, measured by FRAA, puts them in the same range as Mauer.
FRAA isn't a perfect metric, though, especially for catchers. We don't have all of the contemporary fielding metrics available for players from different eras, so it's tough to assert that Cochrane or Dickey were indisputably better in their peak four years than Mauer has been in his first four full years. Conversely, though, it's tough to claim that Dickey and Cochrane were necessarily worse than Mauer, too...unless there's a reliable metric for assessing historical catchers' skills of which I'm unaware.
Thanks for the feedback. By what metric do you assess Mauer's skills as excellent?
This is a great ESPN Insider article. I like it less as a BP article.
For a BP article, rather than only giving anecdotal stories about various catchers, both HOF-caliber and others not so gifted and/or durable, one could grab the twenty best PECOTA comparables for Jason Kendall at three points in his career: age 25 (right before his ankle injury), age 31 (at the point he was traded away from Pittsburgh), and this year. At each of these points we could look at the key PECOTA graphs we know and love--the Stars to Scrubs chart, the EQA Distribution, the Seven-Year Charts--and we could immediately see what the possible futures held for Kendall at each of those turning points. Beyond possible parallels, we'd have quick visuals of the whole range of likely outcomes. As BP readers, we'd understand that. Most ESPN Insider readers might be puzzled, though, so those graphs have to be left out of an article intended for that audience.
I can't pull down PECOTA comparables by myself at home for any desired player-year. I can, however, pull the less-useful but still enlightening Baseball Reference comparable player lists. In Kendall's case, for the three career points I listed above, here they are. For the first group, I've used the format player name, followed by (ages during peak years, age of last full-time season), asterisked if the player was not primarily a catcher:
Mickey Cochrane (24-30, 32)
Shanty Hogan (22-25, 26)
Heinie Zimmerman (24-26, 32)*
Bill Dickey (29-32, 32)
Joe Mauer (22-25, active at age 26)
Jose Vidro (25-28, 32, semi-active)*
Charlie Gehringer (25-36, 38)*
Glenn Wright (23-29, 31)*
Tony Fernandez (24-28, 37)*
Rod Carew (23-32, 39)*
Of the nine comparable players other than Mauer, only three were catchers. They had between four and six prime years, and none played full-time after age 32. Of the six position players, their primes lasted three to twelve years, and half of the group played full time until ages 37-39.
Here are his best comparables at age 30:
And here are his best after last season (age 34):
Well, Craig Biggio started as a catcher, but he caught only one game after age 25. The other 19 "best comparables" were never primarily catchers. Jason Kendall hits so much like an aging middle infielder that his best comparables aren't catchers. Accordingly, while we can see that his career path has veered away from that of Dickey or Cochrane, we can't necessarily draw inferences of Kendall's future from these comparables.
We do, however, have a current PECOTA forecast for Jason Kendall. Most, but not all, of his best comparables are catchers. Astoundingly, 13 of the 20 best comparables showed a 20% or greater drop from their baseline, documented by red down arrows on the PECOTA Card.
But the original article was more about Joe Mauer than it was about Jason Kendall. What can we learn from this regarding Mauer's future?
Mauer's PECOTA comparables include many players who weren't catchers or who moved to another position while still in their prime years. Those who kept catching, including Scioscia, Fosse, Munson, Kennedy, and Kendall, were all in significant decline by their early thirties. (Munson is a special case, but his value was way down from his peak by ages 31-32.) The outlier is Ted Simmons, who had some time at other positions, but who put in two good years as a catcher with Milwaukee in 1982-83 at ages 32 and 33 before his career went downhill. One outlier does not a trend make: looking more closely at comparables seems to reinforce the theme of the article with respect to Mauer's future.
While I'm as big a fan of small sample size analysis as any well-informed stats-minded observer can be, I'm not sure that I'd consider the Mariners 38% likely to reach the ALDS. By third-order wins, the Mariners are last in the AL West and ahead of only the Twins and the Orioles in the entire AL. Their current odds are escalated by the irony that early-season injuries to other AL West teams affected both the early Pct3 and the PECOTA Depth Chart, while the PECOTA projections already included an estimate for total time lost to injury as an element of the forecast. When Clay Davenport adjusts the depth chart to account for an injury that's already factored into Nate Silver's projections, there's a possibility (not a certainty) of an overadjustment against teams facing injury challenges as they leave spring training. Both the A's and the Angels had ace pitchers down at the start of the regular season, and that might have caused their projected wins to be deflated.
Still, the Mariners have a lead in what looks to be a tight race. I might quibble with 38%, but I don't disagree that we now should see the Mariners as surprise contenders. If we never accept a small sample size at face value, or at least run the numbers to attempt to gauge the relevance of the small sample, we'll miss every turning point. I'm happy to see one of BP's gifted analysts and writers taking a stand in favor of a small sample for a change.
+1...whoops, can't give kudos to a staff member.
Eric, my strongest congratulations on your excellent article. I've seen no better description of the system for users at the level of BP readers. I've used Pitch f/x quite a bit since 2007, and I still found your article a great guide--I sure could've used it at the inception of Pitch f/x. Thank you.
Maybe "last-place" seems flawed, but it accurately reflects the actual performance of three of the five teams the Yankees have faced. The Orioles have also struggled, especially from a third-order runs scored and allowed perspective. In terms of actual performance, only the Royals have fared well in 2009. The combined record of the five teams the Yankees have faced is 31-42. While any single team's performance might be a small sample, a .425 winning percentage in 73 games is fairly indicative of non-contending talent. Using projections instead of W-L record after 15 games is reasonable for any single team, to some extent. Using projections instead of performance across a 73-game sample is starting to stretch things.
One could reasonably remove the 15 games these teams have played against the Yankees, reducing the aggregate W-L record to 25-33. That's a .431 record in 58 games. That's still a full eight games below .500 in a bit more than a third of an MLB season. That's still a pretty significant bad record.
Any of the five teams the Yankees have faced might get better: I'd expect the Indians and Rays, in particular, to do that. Any of them might get worse, too: most folks consider the Royals and the Orioles to have overperformed significantly thus far in 2009. I see it as a tossup. Others' mileage may vary.
I guess that one could challenge the premise of considering any team to be good or bad at this point of the season. Certainly some others as well as you have challenged my doing so. I'd suggest, though, that actual W-L performance over a given stretch of games may be a more accurate metric of team quality--during that particular stretch--than pre-season predictions for 162 games would be. Many well-respected pre-season forecasters miss teams' ultimate performances by eight or more wins on average, an average error equating to the difference between a contender and an also-ran. We can do better: we can use historical records to determine strength of competition, even if we then use that as a tool in estimating a team's true talent, in this case the Yankees' true talent.
If you check the Quality of Pitchers Faced and Quality of Batters Faced, it looks as if Sabathia and, especially, Pettitte have faced easy opposition, while Burnett has faced tough hitters. It looks as if most Yankees position players have faced substandard pitching.
But if it's too early, in your estimation, to consider opposing teams' talent as truly known under any metric, I'll only add that the article about which I was commenting used as a premise that the strength of schedule could be used to judge the relevance of a given team's record, and that I was doing exactly that.
Joe, good points regarding schedule strengths for teams in the early going.
Another team that might be considered to have overperformed thus far due to schedule structure would be the New York Yankees. I haven't seen it mentioned much, but here's what the Bronx Bombers have faced in 2009:
The Baltimore Orioles
The Kansas City Royals
The last-place Tampa Bay Rays
The last-place Cleveland Indians
The last-place Oakland Athletics
Despite this easy schedule, they're only tied for second place in the AL East by luck. They're 2.6 wins over Pythag, and at a +1.9 win difference with respect to third-order runs scored and allowed, in only 15 games played. Their pitching has allowed 6.47 runs per game, 28th-best of the 30 MLB teams. By actual, second-order, or third-order runs scored and allowed, they'd be fourth-best of the five AL East teams.
Certainly the double-digit win totals of the Marlins and the Blue Jays were good examples to choose for teams that have capitalized on a weak schedule. Neither were projected to do well in 2009. I'd offer that the Yankees, with weak opponents and the second-highest win total in the AL, might be almost as good an example, even if many forecasters and fans had them bound for the playoffs, not for the cellar.
Correction: J.J. Hardy is hitting .125, but I transcribed 50 PA instead of 48 AB.
That changes the 99% confidence level to .277 if I've run the numbers right. Pardon the typo, especially as it influenced the computation.
You're right that it's easier to get a higher confidence with a larger sample size. It's also easier with more extreme performance, and what I'm trying to demonstrate is how bad Wang already is at the 99% confidence level.
You cite J.J. Hardy. I show Hardy as 6-50, a .125 BA. I'm 99% positive that his true talent level with respect to BA is .267 or less. J.J. Hardy is a .266 lifetime hitter: I'm not convinced anything is wrong.
I posted above that binomial theorem suggests that Wang's true current talent level is a .445 BAA or worse. See the difference? A BA of .267 is well within MLB norms, and the sample size is too small to expect that the talent level is less than that, even if the current stat line is much lower. Wang's actual BAA is .622, suggesting that we should be 99% confident that the current talent level would be a BAA of .445 or worse. While a BA of .267 is within MLB norms, a BAA of .445 certainly isn't. That's why I think that we might pay attention to this particular small sample.
Patrick, there are certainly shortcomings using binomial theorem as a modeling tool here, and there's a 1% chance that 99% confidence would be misleading, too. I understand your skepticism. Still, I'd urge you to consider that this particular small sample appears to be more significant, by far, than most.
Yes, but extreme cases are significant without a large sample. If an MLB pitcher were to throw three consecutive perfect games, none of us would hesitate to disregard sample size and recognize that we were watching history. We might not expect 30 more perfect games, but we'd know that any pitcher who could prevent 81 consecutive batters from reaching base was extraordinary.
Wang's work thus far is so far from MLB norms that it is extraordinary in its own way. Few pitchers have ever been this bad. A .622 BAA is almost what one would expect were every batter to hit a line drive. Consider an MLB batter who hit 42 line drives in 45 PA. We wouldn't be quibbling sample size, we'd be heralding the arrival of something special. That's how bad Wang's pitching is--it's something especially bad.
Let's look at it differently. Wang has a -18.5 VORP in just 6.0 IP. Let's look at previous seasons for pitchers with a VORP below -18.5 for the year, and let's record, for each season, the one who had the fewest IP with a VORP that low or lower:
2008 Luis Mendoza 63.3 IP
2007 Scott Elarton 37.0 IP
2006 Hayden Penn 19.7 IP
2005 Dewon Brazelton 71.0 IP
It usually takes three times (or more) as many IP as Wang has accumulated to be as bad as he's been.
We're too accustomed to discounting small sample sizes if we're serious about baseball stats, and we tend to do it without checking the significance. Doing that, we'll miss every turning point. I'd suggest that the evidence against Wang is, after just three games, enough to indicate that something is very wrong--this is not reasonably a small sample size issue.
I understand the small sample size issue, but let's use binomial theorem to check the hypothetical batting average allowed (BAA) of which we're 99% sure the pitcher in question was worse during the period in question. Mussina was a .327 BAA or worse pitcher during his three games. That's bad, but not inappropriate for an MLB guy with some problem for a small set of games. Wang was a .445 BAA pitcher or worse in his three game sample. That's horrific. We see starting pitchers having single games that bad, but few, if any, have three consecutive season-starting games that bad.
It's hard to imagine how an MLB pitcher could be so bad. At the core of the problem is that Wang is allowing a third of his opposing batters to hit line drives. That's roughly double the MLB average in most years. The other factor is that his BABIP allowed on ground balls is .417, while his career norm is .206. That's too big to be luck: the ground balls are being hit harder by batters. The strongest and fastest sluggers in the game might approach such a BABIP on ground balls; Wang is making every opposing hitter a superstar.
I agree that this won't last. Where I disagree is with the expectation going forward. Wang just lost his old Yankee Stadium, and away from there he looked pretty mediocre. Given that his current issue is so much worse than that faced by Mussina (or Sabathia), and given his lack of established excellence in neutral ballparks in previous seasons, I see little reason to expect a return to the form that Wang showed in home games earlier in his career. Without those home games, he's not anywhere near the caliber of Mussina or Sabathia.
"...the difference between a .472 BA against and a .622 one is pretty insignificant..."
No, it's not. Batting average differences of 150 points define the difference between roughly league-worst and roughly league-best. It's the difference between the worst shortstop and the best first baseman. It's the difference between the Cy Young starter and the defining replacement-level long relief pitcher.
There are other differences between Wang and Mussina and Sabathia, most notably that the latter two look to be roughly HOF-caliber pitchers and that Wang, away from his old ballpark, has looked pretty bad. His career "Away" ERA is 4.61 as I type. Few great pitchers, if any, posted an ERA over 4.00 on the road.
I'd have faith that Wang will bring his batting average allowed to within roughly 150 points of what Mussina and Sabathia allowed over their respective careers, given that he's currently 150 points worse than they were at their worst. Were I a Yankees fan--or a Yankees GM--that wouldn't comfort me right now.
By the way, tossing in a couple of other names they acquired, guys who weren't among the eventual seven best:
Jorge Julio $2,525,000
Roberto Hernandez $2,750,000 (shared with Pittsburgh)
There are a few other Mets bullpen pitchers from that year...The Mets invested on the order of $20 million in their MLB-best bullpen.
Given the number of free agents and the extraordinarily high total bullpen salary, I'd call it buying a bullpen. Others' standards and mileage may, of course, vary.
"No team has ever bought a great bullpen."
I'm not sure of that.
Not going too far back, the 2006 New York Mets led MLB with a 17.787 WXRL. They bought almost their entire bullpen. Here are their top seven relief pitchers, along with how they were acquired:
1. Billy Wagner 5.953 WXRL Free Agent
2. Aaron Heilman 3.282 WXRL Drafted*
3. Duaner Sanchez 2.795 WXRL Trade
4. Chad Bradford 1.834 WXRL Free Agent
5. Darren Oliver 1.334 WXRL Free Agent
6. Pedro Feliciano 1.263 WXRL Free Agent
7. Guillermo Mota 0.820 WXRL Free Agent
* Heilman was drafted by the Yankees and the Twins before being drafted by the Mets. He didn't choose to sign for the money offered by those teams. He signed with the Mets for a $1,508,705 bonus, pretty reasonable money for an 18th pick in 2001. While he wasn't acquired as a free agent, money was, to some degree, a factor.
I'd suggest that a team can possibly buy a bullpen, but I'd acknowledge that it's not easy.
"Lastly, regarding your comment that only a "rich little boy or girl" would be the ones attending the Yankees games---you may well be right. And you know what? That's OK."
Setting aside the rest of your post, I'd differ with these three sentences.
If the only children attending MLB games are rich--and I mean rich to the extent that their families can reasonably afford to go to see the Yankees a couple of times a year, which would require disposable income of at least several hundred dollars each game for a family of four--then MLB will die out over the course of the next two generations. "Die out" may be an exaggeration: certainly polo is not dead; it's just that it's only followed by the rich. But if MLB becomes a venue of only the rich, then it won't be the nation's pastime that it was when middle-aged individuals such as myself were growing up.
Elsewhere at BP today, John Perotto quoted New York Jets cornerback Darelle Revis as saying about baseball, "It looks like fun, and I thought about trying it, but it's a white man's game. It's not for black kids." That's the attitude engendered by a position that it's OK if only rich kids get to go the games. There are deserving kids who aren't rich who will grow up to support the teams we love if we only give them a chance--African-American kids, Hispanic kids, white kids, Asian kids, all from poverty-stricken families where baseball means not only cheering, but a chance to see, in person, other poverty-stricken youths who escaped the trap of destitution. For some kids, baseball is hope.
For many of us who love the game, at some level, baseball is hope. Yes, we know the stats by rote, but we watch because the struggles of each player mean something to us. If those to whom the battles of each pitch appearance mean the most are excluded from the ballparks because they cannot afford a single game ticket, with coming years the sport as we know it will be only a memory.
Yes, Yankees fans...excuse me, supporters of Jeter's defense...have some challenging explaining to do.
Wang's propensity to induce ground balls starting in 2005--and the propensity of Yankees groundskeepers to groom (well, soak) the infield for his games, a grooming contributing to Wang's having a .205 BABIP Allowed on ALL ground balls and a BABIP allowed a whopping .058 better in Yankee Stadium than on the road over a 2,644 PA career--strongly tilts everything in favor of Jeter posting exceptional fielding stats from 2005 onward.
Given that the stats of Jeter's given in the article above range from bad to mediocre, just as more advanced fielding metrics do, the extraordinarily positive effect of home ballpark and one of his five starting pitchers just make the analysis done here more damning.
"There are no miracles in baseball...no team that has ever deserved to win 90 has lost 90"
Hmmm...the 1982 Oakland A's come to mind.
The team's starting rotation remained "intact" but Oakland's starters increased their ERA from 3.16 to 4.67, a 48% increase in runs allowed with no turnover in the five-man starting staff. The pitchers' ERA+ went from 105 to 85. The hitters' OPS+ dropped from 103 to 90, too, even though the only change to starters was replacing Shooty Babbitt with Davey Lopes at second base, hardly a move expected to lower team run production.
The 1981 Oakland A's were 64-45, a 95-win pace had it been a full season. The 1982 A's were 68-94, fifth in the seven-team AL West.
I remember the season. I was surprised. A few others were, too.
I wouldn't expect any of this year's projected 90-game losers to reach the playoffs, but "never" is a very long time, and a whole lot can happen over the course of 162 games.
Wang has consistently put up an ERA almost a run better at home and a BABIP allowed fifty points lower at home because they groomed (well, soaked) the infield to work well with his extreme ground ball tendencies.
Wang finished second in Cy Young voting in 2006. He's an excellent pitcher. On the road, he's posted a 4.21 career ERA, largely because the AL East is tough.
I'm not so sure that Sabathia is headed for an ERA in the mid-threes. I know that some other pitchers coming from the NL to the AL East in recent years, particularly Josh Beckett, had some challenges acclimating to their new division. If Wang is a 4.21 ERA pitcher away from a home ballpark groomed for him, and if Beckett posted an ERA over 5.00 his first season in the AL East, maybe Sabathia is in for some more rough outings, even if he hasn't declined at all from his contract year's health and talent.
We'll see. I wish no player ill; I hope that, six months from now, Sabathia is one of many AL East players whose superlative work has made it a season to remember in three (or more) AL East cities.
Perhaps CC Sabathia will be just fine, but his great improvement last year was largely due to the opposition getting easier. MLB pitchers with 100+ IP faced hitters with OPS from .770 (Livan Hernandez) to .707 (Randy Johnson) in 2008. CC Sabathia faced hitters with a mere .736 OPS in the AL and a paltry .710 OPS in the easier NL. Both figures were near easiest in their leagues.
Furthermore, Sabathia's marked improvement in the last 14 of his 18 games with the Indians might have been related to quality of opposition: of those 14 games, three were against weak-hitting NL teams (the Padres, the Reds away from home, and the Dodgers before Manny's arrival), three were against the Royals; and two were against the A's and the anemic first-half Blue Jays. The remaining six games were against stronger teams, and CC Sabathia lost five of them. His win against the Twins was their sixth loss in a row, suggesting that timing might have played a role.
Interestingly, CC Sabathia faced the easiest opposition of any AL pitcher with 100+ IP in 2007, a .738 OPS. In 2006 his opponents' OPS was .767, just six points higher than that of the batters faced by Liriano, who got the AL's easiest hitters (again, minimum 100 IP).
Sabathia is an excellent starting pitcher. Still, he's unproven in the very competitive AL East. Given his advantages with weak competition in the past three seasons, and remembering that his ERA+ was barely over 100 in four of the five seasons before that, and tossing in the possibility of a decline now that his salary is guaranteed for several years, I'm not yet convinced that his Opening Day was "utterly meaningless." It could be meaningless; it could be a harbinger of a CC Sabathia who finds that the 2009 AL East is harder than the other places and times that he's played.
Ben (and Bil Burke), kudos.
I look at the first graph, (D_OPS+)-(F_OPS+) vs. Consecutive Days Caught "Catcher Fatigue Curve," and I wonder if the Red Sox knew the shape of this graph over half a decade back when they decided to give Tim Wakefield a personal catcher. While one can look at it and see a linear decline from two to eight days' catching, with an outlier at six days, I see a discontinuity between five and six days. Catchers decline little by little over five days, but they lose a lot between the fifth and sixth day. Given that few other MLB catchers approached Jason Varitek's hitting ability until 2008, Boston would've wanted to use him most of the time, but giving Tim Wakefield a personal catcher offered Varitek an excuse to avoid the most challenging catching assignment endured by Boston catchers, as well as offering him "guaranteed" rest opportunities often enough to avoid serious fatigue-related effects. The "Era and Catcher Fatigue" graph makes the decline after day five seem even more pronounced.
Limiting catchers to no more than four consecutive games, as well as the traditional rules regarding catchers working neither both ends of a doubleheader nor day games following night games, seem to be good ways of limiting lost effectiveness due to fatigue. One doesn't have to use the backup catcher as a "personal catcher" to reach that goal, but it does offer what seems to be an appropriate amount of rest.
I read Google News to see who just went on the DL.
I read Baseball Prospectus to see what Will Carroll says about Ichiro going on the DL.
I'm surprised that Rivera is projected to earn a 2.21 ERA, better than his career ERA, in his age 39 season. Yes, he had a 1.40 ERA last season, but he also had a .782 DER behind him and a flukishly good IF/FB ratio relative to his career norm.
Of Rivera's top 19 PECOTA comparables, none had a green up arrow and six had red down arrows. (His twentieth, Mike Timlin 2005, flourished when Keith Foulke lost his job due to injury.) Four of the top six struggled at age 39. Frankly, there are only five good comps to Rivera (with scores higher than 32), and three of those five struggled. Another was Hoyt Wilhelm, whose similarity to Rivera in terms of durability with age might be questioned.
I see Rivera as a potential weakness on the 2009 Yankees, especially given the pressure of pitching for the Yankees and the absence of another proven closer. Certainly others might have more faith in Rivera despite age or point out that the value of a closer is less than many perceive. I'll still point to Rivera as a potential issue.
Will, you got a great value on Jody Gerut. When I emailed you regarding Gerut, I think that I recommended a different Gerut strategy, but I've got to admit that getting him for $3 is better than what I'd suggested.
Good team. Good luck!
I'm eager to learn.
What are the first dozen of several "dozens of metrics that are a lot better than WARP(3)," indicating that Schilling's age 25 season wasn't one of his three best?
I like to think that Clay Davenport's WARP3 is a pretty good metric. It's at the core of the JAWS system used by Joe Sheehan in his article. But if there are dozens of better metrics, all season-adjusted like WARP3 to account for changes in the value of counting stats, by all means, please share.
And could you share that 2009 Curt Schilling PECOTA Card link showing Clemens as his top comparable, too? I don't see a link on his DT Card nor an entry in the January 30 PECOTA spreadsheet, which would've listed at least his top four comparables. I don't even find an active link for Schilling's PECOTA Card using his name and the format of web address used in other PECOTA links. His 2008 PECOTA Card listed Rick Reuschel, Gaylord Perry, Bert Blyleven, and Danny Darwin as his top four comparables: it's odd that it would've changed to Roger Clemens after Schilling missed an entire season.
Thanks in advance.
Lou Marson's seventh-closest PECOTA comp is George Kottaras. Somehow I don't think that Boston will give up Buchholz or Bard for Kottaras redux.
"And the pitcher most statistically similar to Old Schilling according to Baseball-Reference.com? None other than known steroid user Kevin Brown."
Baseball Reference's career similarity scores use career totals without respect to the career trajectory with respect to age. In a discussion of PEDs and aging curves, a Baseball Reference similarity score is an odd fact to, um, cherry-pick.
If you check my posts here, though, I don't believe that I accused any player of steroid or PED use. I merely pointed out that Curt Schilling hit his peak, using Clay Davenport's WARP3 as a metric, at age 25, while Roger Clemens, again using WARP3, hit his at age 34. I also commented that many do consider Clemens to have used PEDs, but I believe I didn't assert that myself.
You say that Clemens had his best year earlier, at age 23. While I respect your opinion, I'll continue to point out that WARP3 strongly indicates otherwise.
When you choose an arbitrary cutoff of before age 30 for Schilling, you're averaging at least three injury-affected seasons--1994, 1995, and 1996--with just two other seasons as a starting pitcher. Furthermore, Schilling's 1993 season appears affected by pitcher abuse: he frequently threw over 130 pitches in a start, and he allowed over five runs per nine innings that season after he threw four complete games and a median of over 120 pitches per game starting eleven of his team's first 49 games. Describing Schilling's "peak" performance by averaging a preponderance of injury-affected seasons with a few other seasons is misleading and, perhaps, disingenuous.
What you're showing, though, is that cherry-picking stats in manners of your choosing can be used to try to demonstrate PED use in misleading ways. I don't dispute that. My position, however, that a statistical method of proving (or, more accurately, suggesting) use is not "demonstrably worthless," but possibly not yet developed, isn't disproven by your intentional choice of misleading combinations of players' seasons. If such a method were developed, it would involve, I'd expect, the search for a player's true peak, using a metric constant across seasons such as WARP3, not the attempt to aggregate seasons ruined by injuries with those where the player was healthy.
"If Feinstein can do that to Clemens (and many other media members can do it to Bonds, ARod, and most other non-white players), then it's only fair to apply it to Schilling."
I disagree with the logic that the lowest standards of behavior of others defines what the standards should be with respect to any ethical choice, including accusing Curt Schilling of steroid use.
"...the statistical method of proving use is demonstrably worthless."
We don't yet know this, though: there's too little information yet available on the effects of PED use to create a statistically significant test.
For the general public, though, the late peak of Barry Bonds, who became one of the three greatest hitters in the history of the game in what would have been past-prime years for most players, defines the perception that late peaks equal PED abuse. Steroid use was admittedly prevalent in MLB in that era, and perceptions of steroid use by Bonds preceded any evidence of use: the farcical absence of drug testing agreed to by the MLBPA and the owners led to flurries of accusations of PED use, several of which now appear to have been correct. Bonds hit more home runs in one season than anybody else, as well as posting unimaginable SLG and OBP rates. Without effective testing, that resulted in allegations of steroid abuse. The allegations of Clemens' steroid use, coupled with his late peaks in Toronto and Houston, are the corresponding image for pitchers.
But in Schilling's case his peak was at age 25, not in his 30's. His second-best year was at age 30. All that he did in his mid-thirties was to avoid injury. That improved durability corresponded directly with lessened pitch counts per start. It still didn't result in individual seasons better than those he had at ages 25 and 30.
It may be that, with the hindsight of years and the eventual release of more drug testing data, somebody will discover a measure of how likely it is that a late peak suggests PED use. That model might be used to consider whether or not the great players of the Steroid Era were more or less likely to have used PEDs.
But I'll be surprised if a pitcher who hit his peak at age 25 is determined to be a likely steroid user by any model that might later be discovered.
Joe, good article.
I'm glad that you explicitly raise the point regarding steroid use, rather than dancing around it. Certainly many have pointed to, say, Roger Clemens as an example of a pitcher whose late renaissance suggested use of steroids. One could make similar assertions regarding Schilling. Let's compare the two of them.
Roger Clemens had his best season at age 34, with a staggering 12.8 WARP3. He had four other seasons above 10.4 WARP3, three of them between ages 23 and 27 and one at age 42. Perhaps his health was better in his mid-twenties, because that's when he had his best stretch of consecutive strong seasons. His peak, however, was clearly his age 34 season, where he posted a WARP3 1.9 wins higher than he did at any other point in his career. That's a late peak for a starting pitcher.
Curt Schilling had his best season at age 25, his first full season as an MLB starting pitcher, with an 8.5 WARP3. He struggled with injuries during his twenties, possibly because his average Pitcher Abuse Point (PAP) score at ages 25 and 26 was higher than that of any MLB pitcher in 2008, and his second-best season was his age 30 season, 8.1 WARP3. That year he finished fifth in PAP; the next year he posted a 7.1 WARP3 and finished fourth in PAP. The next two years he was less effective...still very good, averaging 5.6 WARP3 per year, but less effective. The next four years with Arizona and Boston he earned fewer PAP and no CAT 5 starts, and he remained healthy enough to put up his best four-year stretch from ages 34-37. None of those four years, though, were as good as his seasons at either age 25 or age 30. Schilling clearly had his peak at age 25 or age 30. He had his best-managed and healthiest years in his middle thirties.
I don't see Schilling's stats as indicative of steroid use, regardless of his "quotability, skin tone, and affability." Others' mileage may vary.
"In the meantime, we appreciate that in these economic circumstances, plopping down 40 bucks on a subscription to a baseball website may not seem like the best use of your resources."
How much is entertainment worth per hour?
Taking my wife out to a nice dinner can cost over a hundred dollars per hour. Taking a family to an MLB game, with refreshments and other associated expenses included, costs over ten dollars per person per hour, significantly more than that in many ballparks. Taking the family to a movie approaches that same cost. Enjoying MLB Extra Innings at home, MLBAM blackout rules willing, costs a couple hundred dollars a year, a cost probably on the order of magnitude of a dollar an hour for the average viewer.
If every time I check PECOTA mid-season or consult Will Carroll's latest UTK or consider Kevin Goldstein's scouting reports or laugh at Joe Sheehan is counted, at forty dollars a year BP is one of my best entertainment bargains. There's no need to apologize, even if PECOTA was late this year. You guys (and gals) are great.
I'm delighted to see that PECOTA is being automated. Will this permit mid-season updates?
Yes, but I don't care about what David Foster Wallace wrote, and I do care about what Kevin Goldstein wrote.
Eric, I'd expect differences between writers in predicting the future, but not in reporting the past. I understand your perspective that tossing knuckleballs is less demanding than hurling fastballs and hard breaking pitches, but Nate Silver has been explicit that it's irrelevant to a pitcher's longevity. You're saying something different. Your position meets the common sense test; Nate Silver has done a little research on the issue of player forecasting.
My subjective opinion based upon personal observation is that you're right with respect to "pure" knuckleball pitchers, those few craftsmen who throw knuckleballs almost to the exclusion of other pitches, and that Nate Silver is right with respect to those more common pitchers whose arsenal includes a knuckleball as well as one or two other frequently-used pitches, especially an MLB-caliber fastball. The two knuckleball pitchers you mention, Phil Niekro and Charlie Hough, were in my memory almost "pure" knuckleball pitchers by their mid-forties, so for the purpose of this article I'd subjectively agree with you that Jamie Moyer's longevity is more impressive. Nate Silver could still be correct, or close to correct, given that all knuckleball pitchers occasionally throw other pitches, and that there's nothing magical about their ability to throw other pitches at increasing age.
I'd love to see Bil Burke's take on this issue if he had time. In any case, Eric, thanks for fielding the question.
Eric, you write, "Niekro and Hough were both knuckleballers, so their ability to pitch while creeping up on AARP eligibility isn't really that impressive." Tim Wakefield's player card reads, "Knuckleballers decline rapidly with age, just like everyone else."
Which is true? I'd expect a consistent position from Baseball Prospectus writers, and these quotes are diametrically opposed.
With the recent injuries to Julio Lugo and Dustin Pedroia, it may be prudent to reverse the playing time scheduled for Lugo and Lowrie at SS, and maybe Nick Green should be added to the mix with some playing time at second base.
Shawn, good article. Thank you. I'd been a firm supporter of an International MLB Draft for years. I'll reconsider my position.
I can't help but notice that the two nations with the large recent increases in MLB Debuts, the Dominican Republic and Venezuela, are the very same nations whose young players are most associated with PED use in that same period. Possibly being included in the MLB Draft has reduced Puerto Rican players' opportunity to reach MLB, but perhaps the skyrocketing opportunity enjoyed by Dominican and Venezuelan talent is partly due to factors other than their being free agents.
The Mets led MLB in home attendance last year, but I still found an average attendance of 53,295 fans at Pedro's games and just 49,530 fans at the other 73 home games, suggesting that Pedro on the mound might have been worth an extra 3,765 fans in the ballpark. It's a small sample of eight games, but what data we have suggests that, despite your perception that he's no longer a big draw, having Pedro Martinez as a starting pitcher possibly does improve paid attendance.
Most of the rest of my expectations were based on Pedro's PECOTA projection, not my personal opinion. My gut feeling is with yours, Eric: I expect Pedro to blow out his arm. I just look back to 2004, though, where I attributed Pedro's spike in ERA as a sign that the labrum problems we knew he had were growing critical. He left Boston for the Mets with something to prove, and he looked very effective in his 2005 season, particularly before he tired in August. He even had two good months in 2006 before he showed signs of injury. I see a parallel situation here: Pedro again seems to feel he has something to prove. Despite the obvious health concerns, if the cruel, unemotional PECOTA system says that Pedro might be effective, I'm standing by PECOTA.
Eric, true, but a couple of points:
1) The Dodgers are still well below their 2008 player payroll: they should be in a position to make a move.
2) Clay Davenport's PECOTA Depth Charts have the LA Dodgers in a position to win the NL East by just one game at 89 wins. The Diamondbacks are projected for 88 wins; the Phillies, Braves and Brewers are projected for 87, 86, and 85 wins, respectively, all trailing the Mets and the Cubs significantly. It's shaping up to be another epic finish, reminiscent of 2007, in the NL West and NL Wild Card races. One win could mean a lot, particularly in Marginal Revenue.
3) This is Pedro Martinez. If he's a starting pitcher, LA fans will pay to come out and show him to their kids. He'll boost attendance for his name's marquee value. Also, if he's healthy, he could be great: his Stars and Scrubs chart shows a 10% chance that he's a superstar in 2009, and his reputation supports that possibility. He hasn't proven in the WBC that he's still an MLB-caliber pitcher, but he's done nothing to disprove it, either, so that potential upside still applies. So does the 25% chance that he might exceed his 75th percentile PECOTA forecast of a 4.08 EqERA in 123 IP, enough to make a difference of three or four wins in a tight division. Finally, and subjectively, if it's the 163rd game of the season, you're Joe Torre, and your top three starters are tired or injured, who do you want on the mound if Pedro is healthy and rested?
While I'm not sure of this year's exact MORP formula, I can't help but think that, if the specific situation the Dodgers are facing regarding postseason opportunity is considered, Pedro would be worth to LA enough guaranteed money and incentive opportunities to convince him to sign.
Interesting...nobody from the Red Sox, forecast for the most wins in MLB last time I checked the Depth Charts, even makes this amended list, let alone the top ten.
While this was a good article, those who missed this comment of yours, Jay, missed the best part of the work.
According to Baseball Reference minor league stats, Teixeira had a .917 career minor league fielding percentage at third base. The next year he had an .811 fielding percentage playing third base with Texas. I don't believe that he's played as much as an inning at third base, at any level, since 2003. Disregarding the challenges he might have regarding range in trying to move the wrong way on the defensive spectrum, it's not unreasonable to expect that he might make 30-40 errors if allowed to play half a season at third base. His 2003 third base Davenport Fielding Translation Fielding Rate of 73 suggests that he might cost the Yankees over 20 runs by playing 80-odd games at third base. These are admittedly small sample sizes, but Teixeira is admittedly six years older and no longer a third baseman.
I'd enjoy watching Mark Teixeira at third base, too, but that's only because I believe that Alex Rodriguez would receive a warmer welcome upon his return to pinstripes if Teixeira had been playing third base in his absence, and A-Rod has legitimately had a tough time with the media and the fans so far this year.
Don\'t put words in Joe Sheehan\'s mouth? Let\'s check his actual, exact words from the close of his article again:
\"Not wasting money on old relief pitchers and giving those jobs to $400,000-per-year players is one way to afford market-value commitments to free agents. The Yankee roster is increasingly bifurcated into eight-figure and six-figure players, and that\'s the proper way to assemble a baseball team these days. Concentrate your resources in your stars, then use player development, scouting, and performance analysis to fill out the roster with inexpensive players who are worth far more than they\'re being paid. The assembly of the 2009 bullpen, as much as the $400-odd million spent on free agents, is what should be scary for the rest of the AL East. That kind of quality decision-making goes a lot further than a checkbook does.\"
See, I don\'t see the assembly of a group of pitchers whom Clay Davenport\'s Depth Charts estimate to be fourth-best of five AL East teams\' bullpens this year as \"scary for the rest of the AL East.\" I also don\'t see \"not wasting money on old relief pitchers and giving those jobs to $400,000-per-year players\" as the Yankees\' secret to affording high-salary free agents; I see setting their payroll roughly $50 million higher than any other team\'s payroll and roughly 150% higher than the median MLB team\'s payroll as their secret to affording so many free agents, and I think that it wouldn\'t have made much difference overall had they invested another two percent of their payroll - maybe four million dollars - in set-up and long relief, as both Toronto and Tampa Bay have chosen to do this year. Calling the decision to rely on some lower-priced pitching talent as \"quality decision-making (that) goes a lot further than a checkbook does\" seems misleading when the difference between the Yankees\' payroll and the median MLB payroll is over 20 times greater than the difference between the Yankees\' non-closer bullpen payroll and the Orioles\' non-closer bullpen payroll, the highest such payroll in the AL East. It just seems, at least for the Yankees, that the checkbook can and does go further...much further.
If that\'s \"sneering at Sheehan over something Joe never said or implied in the article,\" as you wrote below, well, my apologies to Mr. Sheehan. I\'d thought that I had been disagreeing with him on the exact core points of his article\'s summary, not sneering at something never written. Joe Sheehan was and is a good professional writer, and I\'m just an amateur who saw a different perspective on this one issue, researched it, and posted it.
Perhaps, though, the results of the 2009 season will differ from what seems likely from PECOTA. PECOTA\'s not perfect, and the Yankees\' bullpen could prove deep and dominating this year. Still, PECOTA\'s been as good as any system in recent years, especially with the recent improvements in how it projects minor league players, and I\'ll take PECOTA over any single expert\'s subjective ideas. But Joe knows his stuff about MLB talent way better than most professionals and almost all amateurs, and it\'s certainly possible that he sees something PECOTA doesn\'t.
But my figurative money is almost always on PECOTA in early March, except where injuries have altered the Depth Charts, and PECOTA and the Depth Charts consider the Yankees\' bullpen less strong than those of their divisional rivals, excepting the Orioles\' bullpen. If the Yankees bullpen is a comparative weakness, I\'m not sure that Joe\'s analysis is accurate, especially if the Rays and the Blue Jays did better for little more investment.
Eric, superb article. Thank you!
You ask, \"Why would more fluctuation in horizontal movement between starts cause a higher BABIP?\" I\'d hypothesize that if pitchers frequently missed while \"painting the corner,\" one of two things would happen: either the pitch would be within the heart of the strike zone, increasing LD%, or the pitch would become an obvious ball, forcing the pitcher to pitch from behind in the count and indirectly increasing LD%. In either case, the increase in LD% would increase BABIP allowed. If between-game deviation was accidental, while in-game deviation was intentional, then increases in between-game deviation would result in more missed corners, and it might possibly explain the higher BABIP.
But that\'s just one possible explanation...frankly, I\'m struck that Between-Game metrics and In-Game metrics have such different correlations to your five performance-based indicators. Yes, they\'re all lower than .35, but I think that you\'re onto something structuring the data for analysis the way you have.
I was interested by Joe Sheehan\'s concept that the Yankees might be using their farm system to achieve relief pitching excellence at bargain price, and I checked it using Clay Davenport\'s Depth Charts and salary figures from Cot\'s Contracts. (Rather than AAV I took 2009 salary, and I assigned a $400,000 2009 salary where none was provided for pre-arb players.) Here\'s what I got looking at the AL East (pardon the formatting):
AL East Bullpens
Team VORP Salary
Red Sox 108.4 12.9
Blue Jays 94.6 19.7
Rays 72.8 16.8
Yankees 65.3 22.8
Orioles 50.0 16.0
The Yankees seem to have the fourth-best bullpen at the highest salary.
Considering that Joe Sheehan seems to be discounting the issue of Mariano Rivera - certainly a very valuable closer both in 2008 and throughout his HOF-caliber career, and probably worth his salary - I looked again at the AL East bullpens, this time removing the VORP and the salary of their closers:
AL East Bullpens (No Closer)
Team VORP Salary
Red Sox 86.0 6.6
Blue Jays 83.8 9.7
Rays 68.3 12.8
Yankees 42.8 7.8
Orioles 39.3 13.3
Now the Yankees drop to fourth in salary, but they\'re also fourth in projected VORP. The Blue Jays and Rays appear to have bullpens - excluding closers - worth three or four more wins than the Yankees\' bullpen at an additional cost of only two to five million dollars\' salary. That would appear to be exceptional value, far exceeding norms of Marginal Cost per Marginal Win for a contending team.
I find two different lessons learned here than Joe Sheehan does:
1) The Red Sox seem to be doing something right. Theo Epstein is generously supplementing home-grown talent with bargain-basement free agents and judicious trades. The Boston 2009 bullpen includes Jonathan Papelbon, very different from Rivera but possibly similar in value. It also includes Japanese free agents Hideki Okajima and Takashi Saito, trade acquisitions Javier Lopez, Ramon Ramirez, Wes Littleton, and Randor Bierd, and farm products Manny Delcarmen and Justin Masterson. Perhaps considering and using all options is better than just supplementing eight-figure free agents with pre-arb talent.
2) The Yankees have certainly put together many excellent teams in recent years, but it\'s difficult to escape the point that their \"quality decision-making\" has at its core the decision to exceed the median MLB team salary by astronomical amounts. In 2008 the Yankees spent about $75 million on player salaries than the Red Sox, the second-highest spending team in their division, and almost $130 million more than the median MLB team. The median MLB team could be expected to earn about 81 wins. If the Yankees can\'t exceed that by a very significant margin by spending 250% as much money on salary as the median MLB team, then there\'s little reason to be considering their salary allocation strategies as any approach to an optimal strategy.
How does it favor the small-market teams to allow the Yankees to sell their own video rights, keep most of the money if revenue sharing applied and all of the money if the profits were buried in YES, and give a pittance, at best, to the small-market teams? How does it improve competitive balance in MLB?
If you check his linked blog, Shawn Hoffman attacked John Henry\'s support of an MLB salary cap as anti-Yankees on February 18th and defended A-Rod with respect to stats and years of admitted PED use on February 9th. Hoffman lives in New York (per Twitter), his posts occasionally (at least) support the Yankees, and this article seems to be an attempt to convince BP readers that giving more to the rich and less to the poor is somehow good for the game.
I notice that the bulk of the BP readers taking the time to post have already pointed out that Hoffman\'s ideas would favor big-market teams. I agree. I infer, though it is only a perception, that Hoffman\'s views may be influenced by his support for the New York Yankees.
I\'m happy that BP is making space for articles regarding the challenges of the absurd blackout rules; I\'m upset that this article is chosen as a viable solution to bizarre rules blacking out games of a half-dozen or more \"home teams\" at any given site. I\'d offer a simpler solution: alter blackout rules to ensure that every viewer can somehow watch every game with their chosen premium subscription, be it an MLB.tv subscription or an Extra Innings subscription. Restrict the feed to the \"home market\" RSN if necessary or desired, but don\'t create a situation where fans can\'t watch their team because of blackouts.
And keep the revenues in the MLBAM pot, not in the pockets of the rich.
I wouldn\'t have guessed that Roger Clemens would be Tim Wakefield\'s 16th-best comp and that Hoyt Wilhelm wouldn\'t make the list.
Certainly a valid counterpoint, Dr. Dave.
I\'d offer that Belle was 10-8 in stolen base attempts the two years prior to the one you cite, and that he was 0-5 the next year, his last. His career 3B/2B ratio was 21/381, 5.5%, in an era when league norms were roughly double that. His range factors and Davenport Fielding Rates were average for the corner outfield positions he played, and corner outfielders aren\'t always noted for speed.
I still consider Belle a slow-moving slugger, just as I consider Jason Varitek the same way despite his 10-3 stolen base success rate in 2004. Belle had three or four good seasons stealing bases, not just one, so your point that he could occasionally have real value as a base stealer is valid--it\'s just that I don\'t see occasional success as a base stealer offsetting all the other things, including some bad seasons trying to steal bases, that make me consider him a slow-moving slugger.
As others have commented, PECOTA seems to be less accurate projecting careers of extreme players, including players like A-Rod, so good at playing baseball that he is among the very few best ever to have played the game. It\'s tough choosing 20 comparable players to A-Rod because there aren\'t twenty other position players as good as A-Rod at his age. Picking comparables means minimizing ways that other players are inferior.
That said, some of the players listed as A-Rod\'s best comparables aren\'t reasonable choices because they\'re slow-moving sluggers prone to swift decline, not excellent shortstops who moved to third base to make room for another HOF-caliber player. Greg Luzinski is, to me, the most laughable comp: he had earned -132 FRAA over a career at left field and first base by age 29, and he played his last four years as a DH. Saying that a 32-year-old Greg Luzinski is the twelfth-best comparable to a 32-year-old A-Rod is, to my mind, absurd. Yes, Luzinski played just one more year and retired after a disappointing season at age 33. But he\'d only played 1.2 games on defense in the last four years before retiring. That\'s not comparable to what A-Rod has done at third base his past few years.
There are other among A-Rod\'s comps who were declining slow-moving sluggers. Albert Belle fits into this group: while his career ended swiftly, lots of slow-moving corner outfielders and corner infielders see their careers end swiftly in their early thirties. Jimmie Foxx, mentioned in the article, fits that mold as well.
The trouble with A-Rod is that his defensive comparables were lesser hitters. Still, there are four of twenty comparable players who were reasonably similar to A-Rod in the field through their careers: Brett, Grich, Sandberg and DeCinces. (Caminiti was clearly a lesser fielder.) Of those four, Sandberg oddly retired after a bad start in 1994 at age 34 before returning for two years at age 36, Grich and DeCinces had reasonable decline curves, and Brett was still a 7.1 WARP player at age 37 and a 19-HR hitter at age 40, when he retired. It\'s tough calling Sandberg\'s retirement choices \"comparable,\" but he, DeCinces and Grich all ultimately retired at ages 36-37 because they needed to do so. Brett, probably the hitter of this group most like A-Rod, was still an MLB-caliber player when he chose to retire at age 40, and he was hitting home runs at better than half his peak rate from his prime. This article has A-Rod hitting just eight HR at age 40: discounting the lesser fielders (Luzinski, Belle) and the lesser hitters (Sandberg, Grich, DeCinces), and maybe discounting the known juicers (Caminiti), A-Rod\'s odds look better.
I\'d caution, though, after all of this writing intended to bolster A-Rod\'s case, there\'s one thing that still troubles me. If you check A-Rod\'s best comparables at Baseball Reference, there\'s one from the top ten who\'s better than any BP comp in many ways: Rogers Hornsby. Both Hornsby and A-Rod were probably the second-best hitters of their era, each surpassed only by a corner outfielder who defied the standards of their times.
That said, Hornsby had just one more MVP-caliber season left at A-Rod\'s age, his 14.9 WARP 1929 season at age 33. He injured his foot badly in 1930, ruining the season. He came back for a 7.6 WARP 1931, but then he accrued just 2.8 WARP and just six home runs over the next seven years before retiring. His last MLB-caliber year was at age 35, just one year earlier than Grich, DeCinces and Sandberg all found themselves nearing retirement.
While I\'ve questioned the validity of PECOTA comps for a player as good as A-Rod, maybe the answer of \"probably not\" remains valid for this question all the same.
Let\'s look at Jon Lester\'s career 2005-2008, considering a consecutive stretch of his games from mid-2006 to late-2007 as cancer affected:
2006 (MLB): 2-2, 7.75 ERA
2007 (A/AA/AAA): 5-5, 3.47 ERA
2007 (MLB): 2-0, 5.67 ERA
2005 (AA): 11-6, 2.61 ERA
2006 (AAA): 3-4, 2.70 ERA
2006 (MLB): 5-0, 2.38 ERA
2007 (MLB): 2-0, 3.34 ERA
2008 (MLB): 16-6, 3.21 ERA
PECOTA can\'t see that. There\'s a big part of his last three years where he was influenced by a factor seemingly unlikely to recur, his cancer. The year before that he was great; the year after his illness he was great; at the beginning of 2006 and at the end of 2007 he was great. PECOTA looks at the past three years, takes those horrible games around Lester\'s chemo as indicative of his talent, and discounts the forecast accordingly.
You know, it\'s funny that PECOTA thinks that Lester will be an average pitcher and that Jody Gerut, who lost two years completely, has a 55% chance of being a superstar at age 31 because he hit LHP well in 68 PA. Over 320 MLB games before 2008 Gerut proved that he couldn\'t break the Mendoza Line vs. LHP. In 68 PA last year he hit .308/.338/.585 against LHP, seven hits--five or six for extra bases--more than might\'ve been expected. He hit RHP almost exactly as well as expected. PECOTA apparently doesn\'t look back to his work in 2003-2005 (although the platoon charts reflect it), and the forecast is very, very optimistic for a guy who had, well, seven more hits than expected in a small sample last year.
If PECOTA just discounted Lester\'s 2006-2007 the same way, he, too, would look much better for 2009.
Dave, thank you.
You might want to check the Chris Carter (Boston) PECOTA Card. I think that you\'ve attached the 2008 BP comments section for a different Chris Carter.
You ask, \"How do you know A-Rod hasn\'t stepped forward, been totally honest, and accepted responsibility? What would satisfy you?\"
I guess that nothing short of his providing us the entire history of his drug testing from 2004-2008 to back up his words would do, and that would suffice only if it happened that there were tests that exactly coincided with his highest performance peaks in 2005 and 2007, disproving that he was using PEDs before and during those few weeks of great hitting. That\'s an exacting standard, but that\'s what it would now take for him to prove his honesty to me.
The challenge is that it\'s not just that A-Rod was caught. It\'s that he asserted to a national audience that he never used PEDs after he\'d used them and even though he knew that he\'d been caught using them. He\'s displayed an arrogance of dishonesty that is so great that I simply can\'t trust the man at his word any more.
If the national reaction to A-Rod\'s press conference is underwhelming, perhaps it\'s because others don\'t trust him, either, not because others dislike his agent nor because of class warfare. A-Rod didn\'t just try to use PEDs without notice; he lied about whether he used them to a national audience. It\'s understandable that his words of apology only invite skepticism: the man is a proven liar.
\"To answer another question in the thread, hitter comps should definitely be there this year.\"
This year will bring comps
We expect the data soon
Subscribers can smile
Spring training arrives
A-Rod speaks while few believe
Certainly Bard might have been better with more practice, but Flaherty retired on March 7, Wakefield\'s first start was April 4, and Bard\'s worst game catching Wakefield (four passed balls) was his last game catching Wakefield on April 26. Flaherty was gone almost a month before Bard\'s first time catching Wakefield, and Bard and Wakefield practiced almost every day in that timeframe. Certainly, given seven weeks to practice, Bard should\'ve been able to allow fewer than four passed balls in 5.2 IP. Heck, we\'re only about seven weeks from Boston\'s 2009 Opening Day as I type this comment: if that wasn\'t enough time in 2006, there\'s no guarantee it\'ll be enough this year.
As an aside, the Red Sox sent Josh Bard a knuckleball catcher\'s mitt immediately after he was traded to the team in late January, and he worked out from that point onward knowing that, if he made the team, as expected, he was probably catching Wakefield.
It\'s surprising that there\'s a report that Josh Bard is slated to be Tim Wakefield\'s personal catcher. Bard was -4 FRAA in 6.0 adjusted games with Boston in 2006 for a DFT Fielding Rate of 33. Yes, it\'s a small sample size, but those of us who watched Bard try to catch Wakefield don\'t think of it as a fluke. Bard allowed 10 passed balls catching Wakefield just five times, and I believe that opposing baserunners stole nine bases without getting caught at all against the Wakefield/Bard battery.
It will be interesting to see how this works out. Bard is competing with Kottaras, and I understand that Kottaras is out of options and that Bard\'s contract isn\'t guaranteed. Kottaras is a poor defensive catcher, but he has caught Zink several times in the minors, and Kottaras is a better match for Varitek in terms of being a pure platoon partner.
I\'d second that. Kevin Goldstein is really good at describing why the scouts like this or that player, but PECOTA is almost as good and it offers a very different perspective.
Will, thanks for the clarification of how you fit together the quotes I cited with what you meant regarding the six percent figure. You\'re able to stick with your old words because you\'ve stuck with facts, coupled with diligent research, throughout your years of writing on this.
Thanks also for the bit you added to your priginal post regarding Tejada: it\'s a good link as well as a good explanation of how your views might change.
I\'m struck that you perceive that steroid use in MLB was probably higher in 1988 than it was in 2003. I\'ve thought that for a long time--I\'ve suspected many players in the 1989 World Series of steroid or precursor use for years--but I\'ve seen few writers suggest that. You carefully phrased your opinion on that as opinion, not fact, but it\'s an opinion I\'m happy to read.
Will, you yourself wrote in 2004 that the 2003 drug testing \"has been widely disparaged for spotty testing, weak penalties, and a very limited set of banned substances.\" You quoted Dr. Lewis Black in 2003 as saying, \"I\'d say maybe 15% are doing something, whether that is steroids or some other substance.\" You quoted Dr. Charles Yesalis as feeling that the number was probably higher than 15%. Gary Huckabay had an excellent article in 2001 where he wrote that current estimates of use ran as high as 50%, and he quoted a former MLB trainer as saying, \"...from what I\'ve seen, I\'d say (steroid use is) very widespread; maybe a third of the players.\" After reading all of these contemporary estimates I found just by using the BP Search function, I find it hard to look back and estimate that only six percent of MLB players were using PEDs. All we know is that only 104 were caught. We don\'t know how many were casual users who didn\'t use PEDs near the testing date, nor do we know how many adjusted their usage patterns in a way to avoid a positive test result in 2003.
That written, let\'s put my criticism in the proper perspective: I\'m commenting because you read feedback and care about it. Your writing is always good, and your words on the topic of PEDs a half-decade back were overall strikingly on target: I urge all BP readers to search \"steroids,\" author name Will Carroll, to read your articles from 2003-2004. In this one \"Unfiltered\" post, I take issue with one figure, the six percent figure, which I find misleading although I concede that it\'s true. I\'d ask, as others have, that you reconsider the validity of the six percent figure. I thank you, though, for being, in retrospect, perhaps the best national writer on this issue in MLB.
Joe, I share your frustration at the leak of supposedly anonymous testing data, and I would support the harshest penalties possible for those responsible for the leak. I don\'t feel, however, that it in any way exonerates Alex Rodriguez. He evaluated the risks, he made choices, he made literally hundreds of millions of dollars from those choices, but his reputation is now ruined for his cheating and his lies. It\'s a harsh consequence, but he knew the possible consequences of his actions, even if he didn\'t know how the information might someday be leaked.
Will, great sidebar. I just reviewed much of what you and Joe have written regarding steroids at BP over the past several years--your words certainly stand the test of time.
Clay, thank you.
What I\'m waiting for is how long it will take for the turmoil to begin when more people realize that you\'ve predicted a 79-83 record for the LA Angels. I can immediately understand why, but it\'ll be tough explaining that one to folks less familiar with what you and a few other have been doing over the last decade.
Keith, this an excellent discussion of the situation. Thank you. I hope that your schedule allows time for more articles as the case and the trial develop.
Christina, thanks. The article isn\'t particularly clear in two areas: first, it appears (at least to me) that the 216 free agent figure combines players eligible for free agency because of all Sections of Article XX, whereas the limit on Type A and Type B signings is based upon Section B free agency only; second, it says that an exception was granted, but it doesn\'t specify whether the exception was given by MLB and not challenged by the MLBPA, or if it was negotiated with MLBPA leadership and made with full concurrence of both parties.
Tomterp, the only reason I see for the MLBPA to oppose unilateral action by MLB on this issue would be to prevent the established past practice of the Commissioner\'s being able to unilaterally modify the negotiated labor agreement...but that\'s a huge reason to grieve the action. If the Commissioner were to be allowed (or has been allowed) to unilaterally waive any part of Section XX, next time the waiver could hypothetically be a waiver of player free agency rights at six years, done, for instance, to limit payroll expense for MLB during a time of sagging revenues. Normally labor unions guard the right to modify any negotiated agreement very carefully, and I\'d be surprised if this were a unilateral, unchallenged waiver by the Commissioner.
I believe that the rule on Type A and Type B free agent signing quotas by team is found on page 74 of the 2007-2011 Basic Agreement, Article XX B.(5), and the Commissioner shouldn\'t be able to waive any aspect of the league\'s agreement with the MLBPA without the union\'s concurrence. If Bud Selig has, indeed, waived this, I\'d be eager to read more about it.
I didn\'t miss the decrease of importance of every home run. It was diminished by runs per game averages, like every other stat. In 1995 the AL averaged 5.06 R/G per Baseball Reference. In 1985, the median point of the six-year span I cited with respect to Phillips\' youth, the AL averaged 4.56 R/G. WARP adjusts for the change, discounting 1995 runs by 10% with respect to 1985 home runs. But, Jay, c\'mon, Tony Phillips racked up his entire home run total of ages 23-28, what should\'ve been his prime, in one year at age 36 without moving to an extreme hitters\' park. The 10% adjustment for offense is dwarfed by the normal decline in talent with age...and we\'re still talking about a huge difference in home runs per plate appearance.
Jay, you\'re frustrated at my use of Phillips--whose career you chose to cite earlier today as an example of a left fielder better than Rice--as my primary case to counter your position. I\'ll tell you what: cite any one other player active in MLB from ages 24-28 with 1,500+ PA at those ages who matched his home run score for those five seasons in a single season anywhere from ages 34-38 where the latter season was outside the \"Steroid Era\" and the earlier seasons all were after 1919, and I\'ll withdraw my position and apologize.
Otherwise, while I apologize for frustrating you, I believe that my point stands: some players improved radically during the \"Steroid Era,\" and the increase in deviation of WARP3 scores undermines the JAWS system. I\'ve cited the arithmetic. This is the example to drive the point home.
If you\'re surprised, as I am, to find a player who played more games at second base than left field and more games at infield than outfield classified as a left fielder, perhaps the system is flawed.
As an aside, if a player spent the latter half of his career in the \"Steroid Era\" and his home run total at age 36 matched his sum of home runs for ages 23-28, the first six years of his MLB career, one might consider that, regardless of his personal use or avoidance of PEDs, maybe the conditions of the latter part of his career somehow favored offensive contribution more than those of the early part of his career. Phillips hit 27 HR in 1982-1987, just as he hit 27 HR in the single year of 1995, his age 36 season, in Anaheim, not an overwhelmingly favorable ballpark for power hitters.
Maybe the JAWS system should normalize across decades more than it does.
While others are frustrated with your position on this issue, it\'s supported overwhelmingly by the MVP votes of BBWAA members of Rice\'s era--he\'s 29th in all-time MVP Award Shares--just as it\'s supported by three-quarters of current BBWAA HOF voters.
I hesitate to call Joe Sheehan or his article stupid, though.
I point again, as I have in recent comments, to the situation where the deviation of hitters\' performance bottomed out just as Rice hit his peak. Normalizing WARP3 by run average while overlooking deviation puts Rice at a disadvantage. The possible shortcomings of WARP in evaluating left field defense in Fenway Park, a notorious issue with Zone Rating systems, could contribute to an underestimation of Rice\'s defense: I really don\'t think that a left fielder with a greater Range Factor than Carl Yastrzemski was roughly 130 runs worse over a career than his peer, George Foster, a well-known defensive liability, but that\'s what the FRAA online leads us to believe. Furthermore, with WARP we\'re forced to accept astoundingly high Park Factors limiting Rice\'s contributions without knowing the derivation of those factors--reverse engineer the BRAA to see what I mean. Yes, Rice hit better at home, but so did almost every player of his era, and Rice\'s career road stats aren\'t too different from those of George Foster or Jose Cruz, hitters far above him in JAWS scores.
Joe Sheehan is a writer. He accepts, I expect, the Davenport Translations and the JAWS system at face value. Had he had a chance to read, before he went on ESPN News today and committed himself to the excellence of Albert Belle, that WARP favors the 1992 era dramatically over the 1979 era, he might have chosen different words.
BBWAA members see the game as it is at any point in time. During the career of Jim Rice, and again 20 years after his retirement, they considered him HOF-caliber. I\'d like to consider Jim Rice a \"Perfect Storm\" of shortcomings in JAWS and WARP and continue to respect Joe Sheehan, Jay Jaffe, Clay Davenport, and the BBWAA as well, despite radical differences regarding this one player.
Your mileage may vary.
Jay, sorry that you\'re not enjoying my comments as much as I enjoy your work.
My theme is that JAWS scores vary by period, with 1979 being a relative low and 1992 being a relative high. Would you care to offer a comment in response regarding how your JAWS system overcomes that, offering fair opportunity for excellence to players across differing decades? That\'s what I\'m not seeing, but you know far more about JAWS than I do: what would a critic be missing?
Do you seriously believe that 19 position players in 1992 were better than any position players in 1978, even after adjusting for runs per game averages?
It\'s not just averages that vary with time, it\'s the ability for great players to achieve high deviation from the norm. Babe Ruth once had more home runs than any other AL team. No player could slug that well today. Are all modern players lesser sluggers than Ruth? Not necessarily, and certainly not by such a margin: it\'s just that Ruth played in an era where very high deviation from norms was possible. Home runs are an evident example of that; check the WARP3 scores by season relative to peers for Honus Wagner, Rogers Hornsby, Ruth, Lou Gehrig, and Joe Cronin for another set of examples.
The 1992 era--the era defined by Albert Belle, the poster child of the anti-Rice crowd, as well as the era defined by Mark McGwire--saw far more high WARP scores and high JAWS scores. Regardless of individual allegations of steroid use, players of that era need to be considered within a context of knowledge that there were many high WARP scores, just as players of the late 1970\'s should be considered knowing that WARP had very few outliers in that period.
You might check the previous posts I\'ve made in this debate where I smoothed the data over seven years, the duration of the JAWS peak. It still shows a very, very significant bias toward players in the Steroid Age.
As an aside, your chosen years for rebuttal weren\'t symmetrical around those I chose: were you cherry-picking yourself?
Jay, I can\'t access what you\'re using, but I can access the WARP3 data currently online. The frequency of HOF-caliber seasons for qualifying hitters--assigning 10+ WARP3 as a rule of thumb for HOF-caliber--ranges from 19 in 1992 and 2007 to zero in 1978, if my research is correct. That\'s a huge difference in players\' opportunity for accruing a HOF-caliber JAWS score.
If the new Davenport Translations have equalized opportunity across generation, I\'m eager to learn that. If not, my point that McGwire enjoyed an unusually favorable era for sluggers seems to stand.
Actually, I don\'t care about McGwire one way or the other.
What upset me was the impassioned plea by Jay Jaffe that we disregard the relevant circumstances regarding PEDs, disregarding the point that WARP3 itself isn\'t normalized properly over generations. It was easiest to rack up a high JAWS score around 1992, and it was toughest around 1979. JAWS takes seven years\' peak scores twice, so it\'s far easier for a player like Mark McGwire to rack up a high JAWS score than a player like Dwight Evans or Dave Parker. A difference in peak scores of 2.7 over a seven-year stretch could translate into a difference of fifteen points or more--18.9 points over an absolutely exact difference in peaks--in JAWS scores for identical players playing in different eras. That\'s the difference between McGwire and either Mattingly or Grace, all mentioned in this article. It\'s most or all of the difference between McGwire and Jim Rice (18.7 points), and we all know from his recent article that Jay Jaffe considers Jim Rice absolutely to fail any test for HOF consideration.
JAWS is not properly normalized across eras, unless we accept that the MLB hitters of the 1979 era were demonstably and significantly worse as a group than those of the 1992 era. If not, JAWS favors McGwire, just as it discounts the accomplishments of others who played in other decades.
From the article:
\"It remains unclear whether the electorate intends to permanently withhold election for every suspected but otherwise qualified player to hit the ballot, and if so, what the standards of proof are; certainly, they\'re lower than the existence of a positive test. If McGwire is being made into an example, will successors like Bonds, Palmeiro, Sammy Sosa, and Roger Clemens—each with vastly different cases—receive similar treatment? Does it matter that the Hall itself is filled with spitballers, sign-stealers, racists, alcoholics, drug addicts, cheaters, wife-beaters, booger-eating spazzes, and other \"role models\" whose place in baseball history is nonetheless secured for eternity? Does it matter that the electorate itself is complicit in the entire steroid narrative, abdicating journalistic responsibility in favor of preserving access to the press box and the locker room? Or that we as fans played right along, flocking to the ballparks in ever-increasing numbers to celebrate record-breaking home-run totals—and that we still haven\'t left despite BALCO, the Congressional debacle, and the Mitchell Report (of which McGwire was not a part)?
I don\'t pretend to have an answer here, but I would hate to see McGwire drop off of the ballot before we gain perspective on his career and the scale of his alleged misdeeds. Ultimately, the less emotion that\'s attached to a vote on his candidacy either way, the better.\"
While emotion may be suboptimal, an understanding of how the conditions of the game changed may put McGwire\'s candidacy in perspective. Using the current online database of 108 years of MLB, the year where the mean highest WARP3 of the best batter of the surrounding seven years was the highest was 1992, right in the middle of McGwire\'s career. That figure is located only 13 years from the minimum found in 1979. The difference between 1992\'s 14.8 and 1979\'s 12.1 was caused by some factor creating an increased ability of the very best hitters in the game to do better than they might have a mere few years earlier.
Failing to consider the changing conditions of the game that alter a player\'s ability to reach a HOF-level JAWS score would result in decisions as incorrect as those failing to consider the change in pitchers\' wins with the change from a four-man rotation to a five-man rotation. Mark McGwire enjoyed a period where sluggers enjoyed an exceptionally high deviation from mean performance. There is evidence that McGwire used performance-enhancing drugs to achieve his stats, as did many of his peers. It may be unfair to penalize him for allegedly using PEDs, given the nature of the game in that era, but it is only correct to normalize the deviation of JAWS scores across decades before crediting McGwire with performance even approaching HOF value.
Bil, thanks for the good news. Although these weren\'t my top picks, these are all good projects that will add significantly to my enjoyment of your site.
On a personal note, I\'ve noticed your name popping up as a research contributor on many of my favorite articles. Thank you for your work behind the scenes. I look forward to more Bil Burke in 2009!
UZR measures a player\'s fielding with respect to location of balls in play in zones. FRAA measures a player\'s defense as an element of his team\'s overall performance on defense.
The equivalent to UZR for hitting would be evaluating batters by nature of contact on a baseball--home run, fly ball in play, line drive, or ground ball--with respect to pitches received, with pitches broken down by exact location, velocity and movement by Pitch f/x. That would be a great talent metric: it would answer the question of the effect of great hitters without lineup protection getting \"pitched around.\" It might be \"better\" than other offensive stats in many people\'s estimation. It wouldn\'t, however, answer the question of a hitter\'s contribution to his team in a given season as well as VORP does. Great hitters don\'t get as many good pitches to hit as weak hitters do. Opportunities vary.
While I don\'t believe that the current exact formula for FRAA is publicly available, I know that the foundation for FRAA is looking at the effectiveness of the team defense and then using fielders\' stats, as well as information such as the handedness of the pitching staff, to apportion credit for the team\'s success. While many consider UZR simply \"better,\" it doesn\'t necessarily capture all of the information that might be provided by FRAA.
One example of a player with good UZR but bad FRAA is Alex Gonzalez. His career UZR/150 is +5.4, while his career DFT Fielding Rate is just 94. His UZR has been positive every year but 2005, and his FRAA has never been positive. It\'s not a small-sample aberration causing a difference, it\'s a difference in what the systems are measuring. If one watches Alex Gonzalez, one sees that he\'s sure-handed and that he has a quick release, but he doesn\'t position himself especially well, he doesn\'t range exceptionally far, and his velocity on his throws is only average, all of these things limiting his ability to make difficult plays from deep in the hole. UZR, as I understand it, doesn\'t penalize players for failure to make occasional plays outside their assigned range. FRAA, as I understand it, credits players for assists irrespective of where the player stands while making the throw. UZR seems to capture Alex Gonzalez\'s being better than average in his assigned area, and FRAA seems to capture his unusually low frequency of working beyond his assigned position.
If I had to choose just one fielding metric, I\'d probably choose career UZR/150, but I\'d rather have a variety of fielding metrics, just as we have a large variety of metrics for batters and pitchers.
The Davenport Fielding Translations and UZR measure different things, much in the same way VORP and RBI measure different things. I\'d be very upset if FRAA simply went away: it\'s not perfect, but it sometimes explains things that other popular fielding metrics don\'t.
My one vote: introduce an \"Ultra-Premium\" Subscription with none of the banner ads that now plague this site. I don\'t fault the decision to introduce them; BP contributors deserve to make good money for good work. I\'d just like the opportunity to be rid of them for a higher annual subscription price.
How does history suggest that players\' OBP was strictly the result of skill sets and not of managers\', coaches\', owners\', and fans\' expectations? Certainly Dwight Evans, mentioned often in posts following this article, changed significantly and permanently regarding his plate discipline in mid-career when he was moved to the second spot in the batting order by Ralph Houk in 1981. Jim Rice batted between third and fifth. I don\'t remember his being expected to change, and he didn\'t. I don\'t know that he couldn\'t. I just know, from memory and from research to confirm my memories, that Houk expected Rice to bat in runs, not to reach base by walking. In Houk\'s words, \"Jim Rice may have been the strongest player I ever managed. And I managed some of the great ones. He powered the offense the whole time I was there. He was one of the hardest working players you will ever see.\" I don\'t see those as the words of a manager disappointed that his past-prime player wasn\'t being more adaptive in pitch selection.
Still, if there\'s a preponderance of history--fact, not legend, as Jay Jaffe indicated was important--that I\'ve missed, I\'d be eager to have it revealed. I\'ve mentioned one well-known case where a slugger from Rice\'s team was expected to change and he changed; I don\'t know if there\'s any massive record of league-leading sluggers who were asked to change who failed to change because of inability to change. I am, however, eager to learn.
Furthermore, I didn\'t know that MVP voters disregarded defense. In 2008, I didn\'t think that Dustin Pedroia won MVP strictly because of his bat. In Rice\'s era, I didn\'t think that Thurman Munson won MVP in 1976 (without leading the AL in a single offensive category) except for his glovework. From the 1960\'s, I think that Brooks Robinson and, especially, Zoilo Versalles might have made their MVP cases primarily on the strength of their defense.
It\'s tough looking back at defense in the days before advanced metrics and being sure of what we\'re using. I know that Rice\'s reputation was bad, but I also know that the local media was hostile, and I know that Rice had a better career Range Factor in left field than either Carl Yastrzemski or, by far, than Mike Greenwell, the All Star players preceding and succeeding him in front of the Green Monster. The Davenport Fielding Translations suggest that Rice wasn\'t very good, but I could list a host of articles from other sources criticizing the DFTs. Even the DFTs credit Rice with a FRAA of +11 in 1984, a year far past his prime but a year when he shared an outfield with the sessile Tony Armas (-15 FRAA) and an uncharacteristically slow late-career Dwight Evans (-8 FRAA). Maybe Jim Rice was suddenly a defensive star in 1984 after supposedly struggling for almost a decade, most of which he spent sharing an outfield with the speedy Lynn and a younger Evans. I\'d prefer to offer that there might be a measure of doubt regarding the usually excellent DFTs in a ballpark as unique as Fenway, as another poster here has commented, and that maybe Rice wasn\'t quite as bad as we now tend to believe.
Finally, it seems that the writers who vote on MVP Awards are being disrespected by many posters here at BP in much the same way that I see \"statheads\" disrespected in posted comments at other major sports blogs. As a BP subscriber for several years, I\'m surprised to see that: usually BP writers and posters ask for more information rather than belittling others and their ideas. It seems that when the name \"Jim Rice\" is raised, though, standards change.
Well, thank you for the apology.
You seem discouraged by my use of contracts as a metric of how players\' contemporaries defined excellence. Pardon me if I\'m missing something, but I\'d thought that baseball contracts in the free agency era reflected the amount that GMs were willing to pay players to help their teams win. That\'s the point with the Joe Carter - Wade Boggs contrast: the two peers excelled in very different ways, and the one whose performance we know to be less valuable by many modern metrics was, by far, the better paid of the two. Jim Rice took the Joe Carter path, whether through choice or talent, and he was considered among the greatest hitters of his era. It seems to me that players were rewarded for swinging away at marginal pitches and maximizing hits and RBI at the expense of walks in Rice\'s era, and I can neither remember nor find via Google any article where a manager or Boston sportswriter regretted that Rice wasn\'t more selective in his pitches, walking more and slugging less.
If you check what I\'ve written, the MVP Award Share isn\'t the sole metric I\'ve used: I\'ve looked at the traditional Black Ink and Gray Ink stats, and I\'ve looked in considerable detail into how WARP3 scores varied over the years, hopefully shedding some new light into how players of the late 1970\'s may have had a harder time reaching the wins-added standards currently in vogue for assessing which players are most deserving of election to the Hall of Fame. My recurring point, though, is that MVP Award Share is an important metric, not legend nor fiction, and that when exactly two ballplayers, Rice and Parker, are so far above those players with otherwise similar career statistics in their MVP votes received, maybe their body of work deserves more credit than the JAWS system suggests. Maybe the conditions of the game, despite fairly constant aggregate statistics, were such that it was exceptionally difficult for hitters to excel. There seem to have been far fewer hitters achieving high WARP3 scores in 1978 in particular and the period of Rice\'s early career in general, and, while I\'m not about to allege that Rice was greater than Rickey Henderson, I still suggest that we may not be considering the whole picture if we discount Rice\'s career with a quick look at JAWS and WARP3.
Again, though, thank you for the kind words.
As an aside, you write, \"Joe Carter got more career MVP shares than Wade Boggs. That\'s pathetic.\" Perhaps true--but Joe Carter played two fewer years and still made almost half again as much money as Wade Boggs, too. It seems that Joe Carter was doing what was valued by contemporary sportswriters and GMs, even if those who read the Baseball Abstracts knew that Wade Boggs was the vastly better player.
You write, \"There was nothing about Rice\'s era that made it hard for players to stand out offensively.\" Did you check the numbers before you wrote that?
I took the liberty of checking every MLB season from 1946 to 2008 for the highest WARP3 score by a hitter and for the number of WARP3 scores of 10 or higher. The season with the lowest top WARP3 score by a hitter was 1978, Jim Rice\'s best year. It was also the only season with no hitter earning a WARP3 score of 10.0 or higher. Other seasons have had as many as 19 such players--but 1978 had none.
If you smooth the data over seven years, the number of peak years considered by JAWS, the post-war minima for both highest WARP3 and for number of players with WARP3 over 10.0 is 1979. The seven-year average centered on 1979 is a 12.1 highest WARP3, with 6.0 players each year over 10.0. For contrast, the seven-year average for 2005 was a highest WARP3 of 13.4, with 11.4 players over 10.0. Given a 10.0 WARP3 as the level expected of a Hall of Famer, using JAWS it\'s twice as easy to look like a HOF player now than it was around 1979. It was also easier for players before 1979: Rice played at the exact worst point for evaluation by the JAWS system.
You mention a list of other players who should also be Hall of Famers, in your estimation, if Rice is. My response is that exactly two players meet the 2.50 MVP Award Share criterion I\'ve mentioned who are eligible but not inducted, Rice and Parker. Coincidentally, both peaked right at that WARP3 minima period surrounding 1979. There may be a correlation. Certainly, though, players with lesser WARP3 scores who weren\'t considered valuable by writers shouldn\'t be considered, and I\'m not suggesting that. That\'s the equivalent of checking Tim Raines\' home run and RBI totals and claiming that a vote in support of Raines opens the door to every left fielder with more home runs and RBI. My position is that we shouldn\'t discount the wisdom of the writers voting on awards, especially in the case of a player whose career accomplishments appear in the Hall of Fame range by traditional metrics.
It\'s tough trying to project how players might have done in different eras, and it\'s useful having a tool such as JAWS. Before we start to discount the opinions of fans and sportswriters who were contemporaries of a HOF candidate, though, we should be sure that our own metrics offer a level playing field over the generations. When a few hours with a spreadsheet suggests that the JAWS system shows such wide variance across generations, perhaps we should keep that in mind when criticizing the few players who were most appreciated by contemporaries at the exact moment when JAWS was least generous in its assessment of the game\'s stars.
I don\'t know if you\'d noticed, but this year the Internet Baseball Awards paralleled the writer-voted ROY, Cy Young and MVP voting results exactly. Given the involvement of metric-savvy Baseball Prospectus readers in the Internet Baseball Awards, I\'d suggest that perhaps the writers aren\'t as ignorant as you suggest. One can pick exceptions such as Utley, but on the whole current writers reflect the perceptions of the current MLB fan base.
Is it really reasonable to argue that 25 years ago the writers of Rice\'s day were better equipped to vote more equitably? Absolutely: in an era where advanced metrics such as Equivalent Average, the Davenport Fielding Translations, and Win Probability Added were simply unavailable, the veterans who had seen 150 or more games every year for decades were invaluable in assessing the true value of ballplayers.
One might argue that Rice\'s skills, heavy on slugging and light on reaching base and defense, were less broad and less valuable than those of many of his peers. In response, I reiterate that Rice was expected to slug, not to walk, and that we lack the advanced metrics to know whether or not FRAA captures the significant Park Factor of Fenway Park prior to its recent reconstructions accurately. We do know that voters were very impressed with Rice, and we know that his traditional statistics meet the standards expected of Hall of Fame players. The exceptions seem to be WARP and JAWS, and I\'ve yet to see a demonstration that those metrics offer a fair playing field across generations.
Kind of you to appreciate my being around.
I did a quick check on Andre Dawson\'s salary after his 1987 MVP season. He got the fourteenth-best contract in MLB in 1988 for hitting lots of home runs and getting lots of RBI in 1987, earning a salary just $460,714 shy of the highest salary in the game at that time. He only had 32 walks in 1987, but MLB hitters didn\'t earn much money for bases on balls.
Likewise, while Rickey Henderson had a very respectable third-place finish in MVP voting in 1985, Mattingly, with a lower OBP (and defensive value) but a higher SLG (and batting average and home run tally) finished first. Mattingly went on to earn excellent money in MLB, starting with a 202% raise in 1986.
Players in Rice\'s era were rewarded for hitting home runs. Rice tried to hit home runs, not to draw walks. That may have been due to many factors, but the foremost may have been that his manager and his fan base wanted him to hit home runs, not to draw walks.
Evaluating his performance years later on other factors than those by which he was evaluated at the time may not be fair to Jim Rice.
Well, thanks for the kind words.
You mention the writing of Bill James. I remember an early Bill James Baseball Abstract pointing out that its readers should pay attention to early equivalents to Most Valuable Player votes, because the voters recorded the best conventional wisdom of their time regarding which players were best. I think that James captured an important idea there: MVP votes recorded for posterity which players were best at doing what was valued by the audiences of professional baseball of that era.
I don\'t dispute that Jay Jaffe\'s JAWS system accurately records which players excelled to the greatest degree in the respects which we now value the most. I do dispute that players from all eras had equal opportunity or equal encouragement to reach those standards, and I\'d suggest that the extraordinary respect which Jim Rice earned amongst sportswriters of his era suggests that he may have been better than a JAWS score leads us to believe.
\"Albert Belle is not a Hall of Famer.
Neither is Rice, at least according to any comparison that rests on fact as opposed to legend.\"
I\'d respectfully disagree.
Good writing includes and disproves potential counterarguments, and this well-written article has the foresight to include the fact--not the legend, the fact--that Jim Rice was among the top five in his league in MVP balloting six times. Where that objectively ranks is overlooked, though. The article links that recognition with the arbitrary issue of Rice striking \"fear\" into opposing teams, and the words used to disprove the counterargument are all dedicated to showing that Rice\'s stats shouldn\'t have caused \"fear.\" No mention is made of how noteworthy his MVP votes received were, contrasted to other players as a benchmark.
One can check Rice\'s MVP Shares, a measure of how his votes received compared to the values of unanimous first-place selection, to contrast his MVP votes received against other players\' MVP votes. Jim Rice has 3.15 MVP Shares, the 29th-highest figure in MLB history. This places him in the range of players who were all inducted into the Hall of Fame. In fact, there are only two players who received 2.50 or more MVP Shares who are eligible for Hall of Fame induction who haven\'t been granted inclusion in Cooperstown: Jim Rice and Dave Parker. Both of those players hit their peaks at exactly the same time, 1975-1986. By the old Black Ink and Gray Ink standards, Rice earned enough credit for probable Hall of Fame entry, while Parker was just borderline. By WARP, Parker emerges as the slightly better player. But the two players are exceptionally similar in their career marks, and they are the only two eligible players over the 2.50 MVP Share mark, let alone the 3.15 MVP Share mark, who aren\'t in the Hall of Fame. Why not?
One could point to the stats, explaining that we now know far more about how valuable hitters are, and that these hitters fall short of the historical standards of excellence at their respective positions. By itself that would be true, but it would leave out so much. Rice was primarily a power hitter, and it\'s easy to forget that his 46 home runs in 1978 were hit in an era where no other AL batter had accumulated as many as 40 home runs since 1970. No AL batter accumulated more than 46 home runs in a season until the hitter\'s year 1987. Rice\'s achievements were earned in a context uniquely challenging to hitters. With the enhanced hitting stats of the past two decades, a period dubbed by some \"The Steroid Era,\" it\'s easy to discount Rice\'s work.
But there\'s more than that. It seems, although I lack the stats (or the time) to prove it definitively, that the conditions of the game from 1975-1986, especially in the AL, made it difficult for either pitchers or hitters to excel. Implicitly, the JAWS system assumes that players of every era had equivalent opportunity to excel. I\'d suggest that, unless we can somehow check the number of players achieving excellence by WARP in a given season and determine that all eras were equivalent, using WARP and JAWS as the sole metrics by which to disqualify a player is unfair.
What I do know is that the MVP Awards are voted on today, as they have been for decades, by women and men who follow baseball for a living, professionals who would lose their reputations (and meal tickets) if they voted for awards in a way thought to be ludicrous. That makes MVP Award Share a viable metric based upon fact, not legend, of how a player was thought to rank with his peers by contemporary experts. By MVP Award Share, Jim Rice is certainly far above the threshold for near-automatic induction to the Hall of Fame. Perhaps there are other perspectives, but unless and until those perspectives\' metrics are demonstrated to treat players from all eras equitably, pardon if I disagree with the closing statement that \"Rice\'s admission to Cooperstown would flatter neither the institution or the process.\"
Given the repeated suggestion in this discussion thread that management, not payroll, determines success, I decided to check that. For each of the 330 team-seasons in the 11-year history of the 30-team MLB, I calculated every team\'s payroll as a percentage of the MLB total each year using the USA Today salary database, and then I checked whether or not each team had made the ALDS or the NLDS. Breaking it down by percentage of MLB payroll, here\'s what I got:
Percentage: Chance of reaching playoffs
It seems to me that payroll does matter, especially as one approaches the level where team payroll is double the MLB average of 3.33%. Nothing is guaranteed, as the 2008 Yankees proved, but money seems to make success more probable.
Conversely, for smaller-market teams, lack of money seems to make success roughly a one-in-seven chance. A team such as the Twins, with good management and luck, might do better, and a team such as the Pirates might do worse, but the group of teams with below-average payrolls--and that\'s a majority of MLB teams because the payrolls aren\'t distributed normally--can expect to see postseason baseball only once every seven years or so.
Pardon my entering this discussion so late, but there\'s an element regarding Clemens\' turnaround that\'s frequently missed, and nobody in this discussion thread has mentioned it. It\'s not that Clemens turned around his career his first year after free agency, 1997; it\'s that Clemens turned around his career in ten days in August 1996.
Roger Clemens posted a WARP1 over 9.0 for seven consecutive years from 1986-92. He then posted WARP1 of 5.6, 7.9, and 4.9 in 1993-95, and he was on track for another mediocre season in 1996. On August 1, his season stats were 4-11, 4.36 ERA, with 168 strikeouts of 719 batters faced (23.4%) and a .723 OPS allowed. Then he tore a fingernail, missed a start, and went home to Houston. After he came back on August 11, his stats through the rest of 1996 were 6-2, 2.09 ERA, 89 strikeouts of 313 batters faced (28.4%), and a .566 OPS allowed. His first four months of 1996, with the start on August 1 included, fits right into his 1993-1995 decline. His last seven weeks of 1996 establishes the performance pattern of his second, higher, career peak in the late 1990\'s.
I don\'t know what was responsible for the change, but whatever we choose to attribute has to work its magic within ten days. There\'s a paucity of reliable information regarding the speed with which one could gain such results from PEDs in MLB, but my gut feeling is that, if PEDs do work, they don\'t work that well that fast.
As a Red Sox fan since the 1960\'s I remember all of Roger Clemens\' years with the Red Sox. My perception was that he\'d gained a lot of weight in the mid-1990\'s, and that he might have been partying too much. He had trouble in day games in 1994-95, contrary to his career splits, and I wonder now if he\'d been out too late nights before scheduled starts. This is sheer speculation, but this I do know: ten days is enough time to get rested and to come back to a team determined to get enough sleep and to do one\'s utmost for the last seven weeks of one\'s contract season, and it\'s not time enough for much of anything else that could make enough of a difference to cut one\'s runs allowed by half.
While I was responding to a series of posts that had extolled the virtues of Yankees management and the shortcomings of Pirates, one could look at the performance of large-market teams and small-market teams in 2008.
Looking back at Nate Silver\'s excellent articles on market size:
The biggest markets are New York, LA, Chicago, Boston, Philadelphia, and, for media market purposes only, Atlanta and Washington DC. Six of the eight playoff teams were from the five biggest attendance markets. Of the eight teams from the five biggest markets, the only two missing the playoffs were the New York teams, and they missed postseason play by very thin margins indeed. All of the playoff teams from the major-market cities had high payrolls.
It truly seems that the Rays and the Brewers are the exceptions, well-managed smaller-market teams able to reach the playoffs--but not able this year to win a World Series--on years when things \"came together.\" But when 75% of the teams from major markets and only 9% of the teams from all other markets reach the Divisional Series, it seems that revenue plays an overwhelming role in determining which teams can compete.
It’s an interesting premise that the Pirates are bad and that the Yankees are good for reasons irrespective of payroll.
I decided to check it out. Per USA Today, the Yankees had an Opening Day payroll of $209,081,577 last year, while the Pirates had a payroll of $48,689,783, a difference of just over $160 million. That difference in salaries would roughly translate into a Yankees team without their nine top-paid players: Alex Rodriguez, Jason Giambi, Derek Jeter, Bobby Abreu, Andy Pettitte, Mariano Rivera, Jorge Posada, Johnny Damon, and Hideki Matsui. Those nine players, whose salaries were slightly less than the $160 million difference in payrolls, combined for 48.4 WARP1. The difference between the two teams last year was 22 wins, so at first glance it seems that the payroll more than accounts for the difference.
Certainly there are other factors, including Yankees prospects who might have earned more WARP1 with more playing time and the strength of the respective divisions, but 48.4 WARP1 represents a lot of wins. It’s tough to escape the point that the big Yankees payroll is a major factor in its multi-year dominance of MLB’s toughest division.
MLB is not close to a free market.
Were it a free market, I would enter it tomorrow, positioning my team in Brooklyn or Newark and asking to be placed into the NL Central. It would be a rough couple of years without any expansion draft, but with high draft picks and the massive New York metro area market, within about half a decade I\'d have a team worth a couple hundred million dollars or more, and I bet that I\'d be winning half (or more) of my games.
OK, barriers to entry preclude that--one can\'t just \"join the NL.\" I\'d like to move the Milwaukee Brewers to Newark, then. I\'d also like to move the KC Royals to Mexico City. Each team would be far more competitive with their new market.
What? Moving a team takes league approval? That\'s not a free market--it\'s a Congressionally-sanctioned exemption to anti-trust laws. And, because it\'s not a free market, unless the leagues (or the lawmakers) take action to ensure competitive balance, the current situation will continue.
\"Well, no one\'s talking about Terry Francona\'s tactical missteps any longer.\"
Joe, you clearly don\'t know me. I almost always criticize Francona, one of Chris Jaffe\'s lowest-rated managers, and I was berating Francona\'s decision to lift Tim Wakefield from the moment he communicated his thoughts through his non-verbal communication in the dugout.
Look at the Pitch f/x record of Wakefield\'s knuckleballs. They were moving: if some of them ended in the strike zone, and if a disproportionate number of them became home runs, that\'s a matter of luck, not of Wakefield\'s skill yesterday night. Wakefield could have been left in the game for several more innings without harm. Francona chose to burn bullpen arms instead.
Unless Francona is considering pitching Lester in Game Six and Wakefield in Game Seven, it didn\'t and it doesn\'t make sense.
Regarding the AL MVP vote, the top three of your top ten by WARP1 are Cliff Lee (10.4), Roy Halladay (9.8), and Dustin Pedroia (9.8). As far as I know, WARP1 doesn\'t take into consideration the actual strength of opposing players, just the strength of the league, and Cliff Lee had an unusually lucky year in drawing easy starts: of 39 AL pitchers with 162 or more IP, Lee had the 38th-easiest set of opposing batters. Discounting Lee both for that and for his not being on a contender, as was done for Johan Santana in this article, brings Halladay and Pedroia back into consideration. Dustin Pedroia was the only one of those three who DID play for a contender, and he\'s the only position player of the three...I\'d favor Pedroia for MVP, not Cliff Lee. Frankly, I\'d put Halladay and Lee behind both Pedroia and Mauer (9.6 WARP1).
Good interview. I\'m sorry that it was so short - Chris Carter is an interesting guy.
Chris Carter faced tough pitching in getting that .333 average. The average AL batting line was .268/.336/.420, and Boston, in the tougher AL East, faced pitchers who allowed around a .250/.333/.390 batting line. Chris Carter\'s opposing pitchers posted a .234/.316/.382 composite batting line allowed. It\'s a small sample size in any case, but there should be little doubt that Carter can hit MLB pitching.
That said, there might still be some doubt regarding whether Carter can, as he suggested, be a good pinch-hitter. Carter hit .571 as a LF and .133 as a PH/DH. If it were just the 18 MLB AB, that could be completely dismissed, but Chris Carter had never gotten a pinch hit in MiLB. Furthermore, his career batting average as a MiLB DH was over 30 points lower than his batting average as a 1B/LF, and he hit better as a position player than as a DH each of the last four years. He\'s a little weaker in batting average and OBP hitting in the first inning, too. Carter may not be a good fielder, but he may be a much better player when he\'s out there trying to play defense between his plate appearances.
John, you write, \"The numbers give the nod to Mauer, as does baseball common sense. Mauer leads the AL in WARP3 (11.5), is fourth in EqA (.317), and eighth in VORP (55.3); a strong showing across the board. Beyond that, he has been a steadying influence behind the plate in guiding a young starting rotation, and that counts for something in the intangibles department.\"
While that is all true, I\'d have written this instead:
\"The numbers give the nod to Pedroia, as does baseball common sense. Pedroia leads the top five candidates in both WARP1 (9.8) and VORP (62.3). No position player for a contending AL team meant more to his team than Dustin Pedroia by either of these metrics. Beyond that, he has played an above-average second base in a Boston infield riddled with injuries, playing a crucial role in keeping Boston sixth in MLB with a .706 DER. He\'s also performed at an MVP-contending level despite being one of the smallest position players in the modern era, and that counts for something in the intangibles department.\"
In his sophomore season, Dustin Pedroia led the serious contenders for AL MVP in VORP and WARP1, and his brash cockiness made him a clubhouse leader on a Red Sox team that needed a leader badly. Given Quentin\'s September injury, given Pedroia\'s edge over Mauer (and Morneau) in both VORP and WARP1, and especially given Boston\'s clubhouse dynamic, I\'d give my top MVP vote to Pedroia, not Mauer.