Nate Silver is an author of Baseball Prospectus.
Nate Silver: Lots of questions folks so let's get started. Thanks for coming.
Cris E (St Paul, MN): Does PECOTA factor in injury information at all or does it simply rely on playing time as a proxy for that? You guys have been building that injury database for a couple years and I was just wondering if it was at all ready to fold into your projection tools.
Nate Silver: At some point I'm going to get together with Will Carroll and try and get the injury database integrated into PECOTA. For the time being, it only uses information based on playing time; sometimes the results are surprisingly intuitive, and at other times surprisingly naive.
Jason A (Chicago): Started reading the book - great as always. As a White Sox fan, I have been disappointed by both management and ownership ever since Bill Veeck left. We have lost Jeremy Reed, Carlos Lee and Magglio Ordonez while obtaining players like Scott "Flea" Podsednik and Jermaine Dye. Can this team ever compete with the current ownership / management in place?
Nate Silver: Kenny Williams is just bizarre. I absolutely hated the Carlos Lee deal, as well as the complacent attitude that has deterred the White Sox from going after premium free agents. On the other hand, I thought the signings of Hernandez and Iguchi were pretty sharp. At the very least, it's safe to say that he looks at things in a different way than most other GMs do, which leads to some erractic results on both sides of the ledger.
jdbranno (Boston): Can you run a PECOTA projection for Smoltz that assumes he'll be used as a starter?
Nate Silver: PECOTA is *not* adjusting for the fact that Smoltz has changed roles. And that's probably a mistake. I'd guess off-hand that the ERA forecast is around 50 points too optimistic. On the other hand, that won't matter much if Leo uses him carefully and he manages to rack up 170 innings.
Bill (Las Vegas): Why is Delmon Young rated so far ahead of Daric Barton when:
A) They are the same age
B) Barton's EqA was higher than Young's
C) Barton's humongous edge in plate discipline probably trumps Young's slight edge in ISO
D) Barton played pro ball at age 17, and played very well
E) Pecota likes Barton a lot more
Nate Silver: Rany makes a lot of effort to integrate PECOTA's input into the prospect rating process, but he doesn't rely on it exclusively, nor should he. Young is a better athlete, has a somewhat better baseball-type body, and is better liked by scouts. Those things matter. Two other things:
1) Isolated power tends to hold up a little bit more robustly than plate discipline as a player moves up the ladder.
2) Very young minor league catchers have something of a bad track record, Joe Mauer excluded.
They're both fantastic prospects.
They're both fantastic
Evan (Vancouver): But if Smoltz goes down with an injury after only 70 innings this season, you'll look like a genius.
Nate Silver: Well, PECOTA is not predicting that Smoltz is only going to pitch 70 innings this year, because it isn't taking the context of the position change into account. To reiterate, PECOTA's approach is to rely solely on empirical information, rather than human input - it doesn't know how to account for the fact that he'll be a starter. And frankly, even if it did, I don't know that it would come up with an especially reliable projection for Smotlz. There just aren't any data points for a pitcher making the transition when: a) he's as good as Smoltz, and b) he's as old as Smoltz.
Personally, I'm expecting the transition to be an unfettered success, just based on the fact that Smoltz is such a competitive mofo. But that isn't a scientific opinion.
jtrichey (muncie, in): Hi Nate, have been enjoying the book this week. Noticed the trend to looking at projections for up to 5 years. Don't you think the very nature of baseball, and life in general, makes 5 year projections almost worthless?
Nate Silver: I don't see how the 5-year projections are worthless. You need to hedge more significantly when looking five years into the future, but that in itself can be a revealing exercise.
Jazz (Boston): The PECOTA cards list Jeromy Burnitz with a weighted mean projection of .253/.328/.480 in 393 AB, while the Player Forecast Manager has him at .276/.349/.536 in 542 AB. Why the difference?
Nate Silver: It looks like the Player Forecast Manager version has Burnitz still in Colorado. We missed a few team changes in the last iteration of PFM but will get that fixed before the next iteration, which is due out on Tuesday.
Cardinal991 (New York City): How much better, if at all, do you think Hank Blalock will be than Dallas McPherson this year? His baseline is more favorable, but McPherson's improve/breakout numbers are huge. Hank's BP2005 comment seemed pretty bearish, considering his record and projection.
Nate Silver: I think you've captured the difference, which is that McPherson is higher risk and higher reward. There are a lot of guys like Jim Thome high up on McPherson's comparables list - on the other hand, there are also a lot of guys like Franklin Stubbs.
Cardinal991 (New York City): Do you really think Adam Dunn is more likely than not to break out from his already monstrous "baseline" projection? And over 2-1 favorite to improve on it?
I'm a Dunn believer, and I own him in a keeper league, but it's hard to consider anyone (not named Bonds) a near mortal lock for a 4-digit OPS, as he seems to be.
Nate Silver: It's important to remember that the breakout/collapse numbers are based on comparing a player's expected performance to the *weighted average* of his performance in the past three seasons, and not just his performance in 2004. PECOTA thinks Dunn's subpar 2003 was pretty much a colossal fluke, so a lot of the high breakout number is in comparison to that.
TThunder (Omaha, NE): Hey Nate, why is Kenny Rogers's VORP so high? Looks like he'll be a league leading starter, no? Then again, those peripherals are awful. Is the discrepancy a matter of pitching in Texas, or does he just lack comps at his age?
Nate Silver: Well, Rogers has good command, stays healthy, and does a reasonably good job of keeping the ball in the yard, which is an good thing to do in Texas. Those things go a long way toward compensating for a lowish strikeout rate. PECOTA compares him to Warren Spahn (!), which I think is overstating the case, but he's an underrated pitcher at this point of his career.
rotosavant (DC): I did a comparison of BP and BHQ projections to identify big divergences (e.g., Upton, Phelps, Bellhorn). You've got BJ Upton approaching 475 ABs (while Baseball HQ has him at 200 ABs). Do you see him sticking out of spring training or getting a near immediate call up?
Nate Silver: I think it would be utterly silly for the Devil Rays to do anything other than play B.J. Upton every day. Our relatively optimistic forecast for him in the depth charts is based on the notion that that sort of common sense is going to prevail. The only question in my mind is whether that playing time should come at shortstop, DH, or somewhere else.
jdg995 (Hilo, HI): PECOTA seems to be jumping on board with everyone else who says that Pedro is going to be back to Pedro this year. Why should we believe that last year's 3.90 was an aberration, not a trend?
Nate Silver: Some of the optimistic-seeming forecast for Pedro is simply because he's switching leagues and switching ballparks - those things alone knock about 75 points of his projected ERA. That said, it seems to me that the sort of stathead conventional wisdom on Pedro has been too pessimistic, rather than the other way around. History teaches is that it's worthwhile to look at more than one year's worth of historical data when making projections, and if you give some weight to Pedro's 2002 and 2003 performance, you get a more favorable projection for him. I don't know why we should be ignoring that data and treating his 2004 as his baseline going foreward - Pedro has been a fantastic pitcher for a long, long time, and now he's got Rick Peterson to work with.
Arathorn (Chicago-area, IL): What's the next big thing for PECOTA? Is it adding human input on playing time? Or something else entirely, like using opponent-quality-neutral numbers for similarity matching?
Nate Silver: We incorporate human input on playing time into the depth chart and fantasy manager versions of PECOTA, so that wouldn't be anything new.
It's a system that feeds off of data, so the question is really what sorts of data we could incorporate into the system that it isn't evaluating now. The two particular things that I'd like to get implemented within the next year or two:
i) Detailed information on player injury histories.
ii) Groundball-flyball data for hitters.
Bill Johnson (New Mexico): I was disappointed that Ron Santo got stiffed yet again by the HoF Veterans process. Any insights as to why? Is the bar for HoF 3rd basemen unrealistically high for some reason? Do you think that implies anything for the top 3Bs of our day like Scott Rolen? I look at Rolen and see "Future Hall of Famer" written all over him, but then, I'm a Cards fan ...
Nate Silver: Yeah, I can't believe that Santo has been shut out. I'm convinced that there's some sort of psychological phenomenon in which bad decisions can be sort of self-reinforcing, because the person(s) making the bad decisions need to defend them publicly, and become more entrenched in their beliefs in the process. This is the same phenomenon, I think, that led to Ed Wade retaining Larry Bowa for as long as he did.
Rolen, I think, had been something of an underrated player for a long time, though last year did a lot to correct that. What the mood of the BBWAA will be in twelve years or so is something it's hard to predict, but I see him as a Hall of Famer.
collins (greenville nc): Nate, Thanks for doing the chat. Why is PECOTA so down on Joe Nathan? His 50th percentile ERA is over 4. Seems awfully high.
Nate Silver: Nathan's PECOTA forecast still includes some input from his 2002 season, when he was a starter in Fresno and pitching just awfully and his arm wasn't really healthy. As I said in the Pedro comment above, you'll be far better off in the long run when you incorporate three years' worth of data, but there are probably some exceptions, and Nathan might be one of them.
That said, one thing to watch for is that Nathan has been a strong flyball pitcher who as also had very low home run rates. One of those things has to give, and it's usually the home run rate.
Rob in WI (Milwaukee): Nate,
Follow up on Santo: is the the best elgible player not in the HoF?
Nate Silver: He's #2 with a bullet after Blyleven.
akrech (Chicago): Nate:
Have you come across any evidence to suggest that certain teams have lower variation in average runs per game over the course of a season than other teams? It seems to me that there may be certain benefits (especially for a team with good, consistent pitching) to scoring an average number of runs very consistently, rather than alternating blowouts with shutouts. (and yes, I am thinking of the cubs here)
Nate Silver: I've looked at this, and in fact come to the opposite conclusion: there isn't really any meaningful predictor of variation in team run scoring over the course of a season.
ScotMartin (Bedford, TX):
Which player do you believe is most likely to exceed their PECOTA projections? Why do you think PECOTA has trouble with that player?
Nate Silver: The first name that comes to mind is Jeremy Bonderman. I don't think that PECOTA has adequately accounted for the steady course of improvement he's made over the past 24 months, particularly in the last portion of last season. I expect him to be a top-tier starter this year.
Jason R (Chatham County Line): The talk of Tiger Town is that Dave Dombrowski's next move *should be* trading Uggie Urbina for Mike Cameron or (insert ML centerfielder here). Part of the logic is that with luminaries as Farnsworth, Percival, Jamie Freakin Walker, etc, the pen is now a point of strength (cough).
What about sitting tight and playing Higgy in center?
Nate Silver: I'm a huge Cameron fan, but the center field solution lies in house with Curtis Granderson.
Cris E (St Paul, MN): Quicky: what does the defensive rating for a DH mean?
Nate Silver: PECOTAs defensive projections are based on a composite of the positions that a player has played over the past couple of years, but it only lists his primary position on the forecast. So you'll get a few guys who mostly played DH, but turned in a positive defensive forecast when they did take the field. They may wind up with a weird-looking rating like "DH 109 +1".
I do need to go into a lot more detail about the defensive forecasts at some point, as they were an area I worked on a lot this winter. Look for an LDL soon.
Rob in WI (Milwaukee): Wrigley field once again has bumped over the magic 100, playing as a slight hitters park as opposed to the pitcher's park it had been rated over the last few years. How much of this can be related to the Cubs lineup (and the good hitting teams such as Astros and Cards), and how much is related to the injuries to Cubs pitchers last year? And does this make the Big Z an even better fantasy pickup this year (assuming he's healthy all year)
Nate Silver: Park factor adjustments, properly done, should not be at all dependant on a team's relative strengths and weaknesses. If the Cubs had a better year at the plate last season, that should have had about the same impact on the road as it did at home.
The fluctuations in Wrigley's park factor have mostly been caused, I think, by the weather patterns that the Cubs have encountered during their homestands. I'd have to double check this, but it seemed like the Cubs were fortunate enough to have an awful lot of warm-weather ballgames last year, and days when the wind was blowing out, which makes something like a 40 percent difference in run scoring at Wrigley.
As for Zambrano - I love the guy to death but I'm concerend that the workload is going to catch up with him. So is Will Carroll, and he knows a lot more about that sort of thing than I do.
Matt (Oakland): Will the A's contend this year or are we looking at a rebuilding year for them?
Nate Silver: The A's, I think, are vastly underrated by the mainstream media. I've got them at about 92-70, and taking the AL West fairly easily.
What's funny is that a lot of the individual PECOTA projections for A's players aren't all that optimistic, especially on the pitching side, but the overall package is very good. It matters a lot over the course of a 162-game season that Oakland has folks like Bobby Kielty and Adam Melhuse and Justin Duchscherer coming off the bench, instead of say Jose Macias and Henry Blanco and Todd Wellemeyer.
prinaldi (newark): What park factor did you use in PECOTA for the Nats?
Nate Silver: Going to go about five more minutes here; thanks for all the great questions.
We used a park factor of 98 (slight pitchers' park) for the Nats. That's based on two things:
1) RFK generally played as a pitchers' park when the Sens played there, and that was at a time when there were more big ballparks than there are today;
2) RFK's overall dimensions and layout look an awful lot like Shea Stadium, and Shea has been a pretty strong pitchers' park.
prhood (St. John's): I sense that the Braves streak is coming to an end. They don't have Turner's millions to throw around and have some serious retooling to to do. Results for the last year or two seem somewhat to have been the result of "smoke and mirrors" by the coaching staff.
How do you see 2005? and what is the likelihood that we'll see some prospects (e.g. Marte, Langerhans, et.al.) geeting a significant amount of playing time?
Nate Silver: Rumors of the Braves' demise have been greatly exaggerated so often in the past that I just don't know what to say at this point. They look like about a .500 ballclub on paper but the same could have been said last year. I do think that they're probably the smartest front office in baseball, Boston and Oakland included. Whether Andy Marte is the starting third baseman by the first of July I don't know - what I do know is that the organization is highly likely to make an efficient decision in that regard.
nate (n.y,): in fantasy baseball today the lemprikans beat the unicorns 7 to 3. a homerun was hit by the easter bunny.
Nate Silver: Thanks for the update.
Cris E (St Paul, MN): Talk about MIN vs CLE. The projections seem to say that the Indians' hitters are due for some regression and the new pitching isn't going to make up the difference.
Nate Silver: On the contrary, PECOTA sees the AL Central as pretty much a dead heat, with the Twins and Indians both at 86-76. The White Sox and Tigers aren't too far back if things break right. The Indians are a big sleeper club in my opinion. Nobody much noticed, but they scored a hell of a lot of runs last year, and I expect the pitching staff to be better.
Nate Silver: That's all, folks. Thanks for all the questions, and for supporting Baseball Prospectus.