Goin’ All Hanley Ramirez
Marlins‘ shortstop Hanley Ramirez is probably the most famous example of a breakout in recent history. Considered a great hitting prospect while in the Red Sox minor leagues, Ramirez consistently underperformed expectations, and before the 2006 season the frustrated Sox traded him to Florida. After posting Major League Equivalent wOBAs (Tom Tango’s weighted on-base average) of .311, .329 and .295 the three seasons prior to the trade, (MLB average for shortstop is .318) the 22 year old Ramirez won the NL Rookie of the Year Award for the Marlins, putting up an 292/353/480 BA/OB/SA line, good for a .360 wOBA. If making draft list in 2007, I had to ask “Who is the real Hanley Ramirez?” Someone who hits like Jack Wilson (without the glove), or the best offense shortstop in the league? Two and a half years later, we know that Ramirez has become even better, posting wOBAs since his rookie season of .404, .401 and .409 so far in 2009. He has clearly established a new level as one of the premier offensive players at any position. But after only one breakout season, can we tell if the improvement is real, or will the player fall back to his old level of performance?
Don’t Trust Anyone Over 29
Even a single season is not a large enough sample size to get an accurate profile of a player, so projection systems such as PECOTA take several years of a player’s past performance, weighting the most recent seasons more heavily than older ones, to get an estimate of what his true talent level is at that point in time. But like everyone tells you when you are playing the stock market, the past is not a guarantee of future performance. Roughly 15% of the time a single season hitting line exceeds the players projection by more than 35 points of wOBA, and likewise 15% of the time will under perform the projection by more than 35 points. I made a list of 175 recent seasons where the batter had a ‘breakout’ performance based on his increase in wOBA over what was projected and what his wOBA was in the season following the breakout. The second season was classified three ways, 1) the batter meets or exceeds the production of the breakout season 2) the production is below the breakout season, but still more than 35 points above the prior projection or 3) the batter produces at or below the level of the prior projection, losing his entire breakout. In cases 1 and 2, the player has shown a real improvement, which in geek speak is two consecutive seasons more than one standard error above the original projection.
The following tables show by age how many players kept all or some or lost their breakout in the following season. SDT (BABIP), XBH (extra base hits), HR, BB and SO are the improvements in those components during the breakout season.
Breakouts Age Kept Some Lost SDT XBH HR BB SO 35+ 6 .40 0 .00 9 .60 .016 .019 .019 .015 .011 30-34 9 .21 10 .23 24 .56 .023 .006 .016 .013 .005 25-29 30 .35 25 .29 30 .35 .020 .022 .020 .013 .003 20-24 11 .33 10 .31 12 .36 .025 .018 .019 .008 .004 30s 15 .26 10 .17 33 .57 .021 .010 .017 .013 .007 20s 41 .35 35 .30 42 .36 .021 .021 .020 .012 .003
64% of the players in their 20’s continued their improvement into the second season, to 43% of the players in their 30’s. The older players relied more on increased walks and lowers strikeouts for their breakouts, while the younger players showed more reliance on power numbers and BABIP. Breakdowns, where the batter produces a wOBA more than 35 points less than his projection, show the same age patterns.
Breakdowns Age Down Middle Back Up SDT XBH HR BB SO 35+ 5 .63 2 .25 1 .12 -.040 -.007 -.016 -.014 -.001 30-34 19 .56 8 .24 7 .20 -.036 -.019 -.015 -.011 -.006 25-29 23 .53 7 .16 13 .31 -.033 -.017 -.015 -.016 .008 20-24 11 .50 2 .09 9 .41 -.029 -.016 -.015 -.013 -.005 30s 24 .57 10 .24 8 .19 -.037 -.017 -.015 -.012 .005 20s 34 .52 9 .14 22 .34 -.031 -.016 -.015 -.015 .004
41% of players under 25 and 34% of those under 30 gain back all their lost productivity in the second season. When older players breakdown they show more of a loss of base hits and extra base hits compared to the breakdowns of younger players.
Players in their 20’s, especially when the breakdown was the result of an injury and they are now healthy, are a good bet to return to their previous levels. These include Jason Bay, Carlos Beltran, Paul Konerko, Pat Burrell, and Adrian Beltre.
Matt Wieters for MVP
In order to minimize the error for all players, projections will normally appear to split the difference between the previous year’s projection and the actual production that year. Players with true breakouts will be under projected in the following season. Matt Wieters was the Orioles 1st round pick (5th overall) in the 2007 amateur draft. He did not play professionally until 2008, when he hit a combined .355 with 27 home runs and 91 RBI in High-A and Double-A. PECOTA liked what it saw, giving a most likely ‘weighted mean’ projection of 311/395/546, with it’s 90th percentile projection a Pujols-esque 344/437/635. This week the Orioles announced they were calling Wieters up for his major league debut, but after hitting 305/387/504 in 39 games at Triple-A Norfolk, some are asking “What happened to him?”.
Based on Wieter’s stats at Georgia Tech, I projected him at 266/346/431 for a 341 wOBA, which is very good compared to the MLB average of 312 for catchers. I translated his 2008 minor league stats at 331/411/561 for a 416 wOBA, which generated a 294/373/487 projection for 2009. This year at Norfolk, one of the best pitcher’s parks in Triple-A, Wieters’ MLE is 295/364/504, splitting the difference and nailing my projection.
The Jay Bell Syndrome
Sometimes the breakdowns occur the moment the player steps onto a major league field. Jay Bell made his major league debut in 1986 at age 20 for the Cleveland Indians. In his third professional season, Bell had hit .298 and .277 in parts of two seasons at Double-A. Promoted to Buffalo of the Triple-A American Association in 1987, Bell hit .260 with 17 HRs and 60 RBI in 110 games, very good production at shortstop. In 38 games with Cleveland, he hit .216. Back a second time to Triple-A in 1988, Bell hit .276 with 7 HRs in 49 games. Promoted again to the Indians, Bell ‘hit’ .218 in 73 games. In his third trip to Triple-A in 1989, Bell was hitting .285 with 10 HRs in 86 games. Instead of recalling him, the Indians shipped him off to Pittsburgh where he would become the regular shortstop for the next 7 ½ years, plus another five years as a starter in Kansas City and Phoenix, finishing his career with a 265/343/416 line.
The modern poster boy for not performing up to expectations in half a season is Nelson Cruz. After being traded twice in three years, and into his third season in the big leagues, something finally clicked. In 76 games since his August 2008 recall by the Rangers, Cruz has hammered major league pitching for 19 HRs at a 308/385/588 clip.
Minor league statistics can give an accurate profile of a player. For whatever reason, some players do not immediately adjust to the majors. Try to have patience that some team will have the same faith as you and reward them with enough playing time to prove their value.
Honestly, other than age, I have yet to spot any statistical indicators of which player will go one to a second season at a new level of performance, and those who will go back to their previous level. Younger players will rebound from breakdowns much better than those in their 30’s. Try to be patient with the young studs who struggle in their first attempt or two at the majors, but stay away from those veterans who sign minor league contracts.
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.Subscribe now
On the negative side, there are a handful of typos and a few awkwardly phrased sentences, but that shouldn't matter too much, as under "normal" circumstances, he'd have editor to clean that up. The only substantial negative point here is that I still don't understand the tables so much. Could you please explain how you got these numbers from your table: "64% of the players in their 20's continued their improvement into the second season, to 43% of the players in their 30's"?
Also, I'd love to hear more about how you use both NCAA stats in your projection analysis and only a partial season's worth of stats (minors or otherwise).
As a reader, what I'd love to see more than anything else for the future is someone who can provide regular updates to PECOTA (or an equivalent projection system) throughout the season. It seems like Brian could do this (and so much more).
However I really enjoyed this article and felt like he did a good job from start to finish. The only part I didn't love, was exactly what Kevin touched on that college stats scare me. I'd really have to see how they are translated and there simply wasn't room for that. I think just the minor league translations are enough without the college stats.
Still, I will be giving it a thumbs up. Continue the good work.
Every week, Brian has some great things to say, and even better, it seems he's getting better and better at making his great work more readable. If this trend improvement continues, I can't wait to see what Brian's articles are going to be like in a few weeks
2) You could have taken this article to another level by helping us figure out which breakouts might be for real. As Kevin points out, merely mentioning the role that scouting plays in evaluating breakouts would have made the article stronger.
3) What's wOBA? I'm a stat geek and I've never heard of it.
If I'm giving a presentation in Paris, I'm going to speak in French. Sure, English and French accomplish the same objective, and if I used English, most of my audience would still understand what I'm saying. However, if one is attempting to explain a concept to people, it's best explained in terms that are familiar to the audience.
Minor quibble though; strong effort.
Although I do prefer wOBA, but that is a discussion for another day and thread.
The idea of figuring out breakout players is a fantastic topic, but again the numbers and graphs don't tell me anything. A whole bunch of numbers that lead to nothing. It's not that I don't comprehend them, it's that there are so many they are undistinguished.
When we hit the Matt Wieters topic, I thought the breakout player section was over. That it wasn't is thoroughly confusing.
I think this author is a great statistical mind. I think he would be a much better research assistant at BP--pulling numbers and getting things ready for other people's writing--than he would be a writer for the site himself.
That being said, I won't be reading past the introduction on this one. It is a flood of errors and awkward phrases that add up quickly and make it hell to get through. Simply put, if the author can't be bothered to ensure a minimum level of readability in his very first paragraph, then I have little interest in giving him a chance. Sorry.
Writing aside, I have a bit of a philosophical problem with the fantasy analysis here... it is asked if we know whether Ramirez will repeat his numbers in 2007 or not. My response is "I don't care, he was a shortstop who stole 50 stolen bases!" Jack Wilson's best years were tied into his batting average and a modest power spike, but nowhere enough to put him in the top tier. People still draft Juan Pierre as a part-timer in NL-only leagues merely for his SB. Those 50 stolen bases are extremely valuable and how repeatable his other numbers are become superficial, to an extent. It would have been more instructive not to compare wOBAs, but to compare SB performance dropoffs from one season to another.
The age-29 thing is nothing new, and in effect, felt a bit oversimplified here. Some people have age 30 peaks, and some peak at age 28.
You write: "Players in their 20's, especially when the breakdown was the result of an injury and they are now healthy, are a good bet to return to their previous levels. These include Jason Bay, Carlos Beltran, Paul Konerko, Pat Burrell, and Adrian Beltre." Bay is 30, Beltran is 32, Konerko is 33, Burrell is 32 and Beltre is 30. Are you saying these players should be treated as injury rebounds? What previous level can Pat Burrell return to? I just don't understand this list of players since none of them are in their 20s...
Fun projection for Wieters, but I don't know a fantasy league that uses wOBA, and though some leagues use SLG and OBP, more use BA, HR, R and RBI.
"..stay away from those veterans who sign minor league contracts." - This didn't really work as a concluding sentence to me... shouldn't it be obvious in a fantasy league?
Maybe people are right in that you're responding too much to our feedback. Stay innovative, in depth, and precise. I'd prefer that to this kind of article. Other than that, I'm sorry Brian, but I'm not sure what happened...
There is real value in seeing if Hanley is his minor league self or his breakout self. 50 SB is great, but what about the rest of the picture that makes Hanley the great fantasy player he is.
He was talking about those pleyers bouncing back from previous injuries I believe when they were in their 20's and doing so well to become productive again since.
No league may use wOBA but it is a very predictive measure, and can help to find out who might do well in those other categories.
Maybe just a clash in taste similar to medium vs. medium rare but this article was in my clear cut top 2.
Assuming Rodriguez gets playing time, I don't care if he regresses if he puts up those kinds of SB numbers. Similar to how, with the recent scarcity of home runs, drafting Adam Dunn or Ryan Howard is worth it even if there are issues with batting average or health.
If he is referring to players in the past who rebounded from injuries, such as Jason Bay, etc, then the word "include" needs to be "included", putting it in the past tense. I know Jason Bay had a major knee injury but did those other players have major injury issues besides the typical 15-day DL stuff?
Maybe it is just a clash in taste and it's not like I'm some BP-fanbase authority figure or anything, it just didn't work for me especially with all the typos.
The typos, and slight grammatical errors I cannot defend, however I still felt the article carried them.
I hope you know I am not meaning to dispute your points all over the place just that I don't see eye to eye.
I don't see eye to eye with Will on almost any comment, because I think he is too quick to be afraid of any article with deep stats anaylisis, and any article not directly using BP stats he is familiar with. That doesn't mean I am right or Will is (although he is kind of an authority around here). Just that we don't see eye to eye.
That is the glory of this competition.
BTW, I will be 100% surprised if 1 of 2 people is not eliminated.
I definitely agree with you that he has grown up and filled out. Since 2006, he has become a much better player since then, adding an extra 10-15 HR and hitting at or above .300 since. Last year he lost about 10 SB and this year he might be closer to a 30 SB pace.
Does that clarify things a bit?
BTW I was 100% surprised one of three people were not eliminated last week. I have no good hunch on how this week would turn out. Shows what I know even with all my ramblings :P
If I got to see every player as much as I see Hanley (I watch about 130 Marlins games a year) I might be able to see the prgression, besides that it is difficult. He has more than made up for the drop in SB with power.
We mostly agree really. Just semantics which aren't (in my opinion) worth going any further as we have aren't really that far apart.
Yeah we don't know who will be gone I guess. Glad so many people respond with intelligent conversation.
It is a good discussion to have, and any discussion of a Marlin I am more than willing to contribute.
"For fantasy baseball enthusiasts, some of the biggest question marks heading into draft day involve players coming off of a 'breakout' season. Which ones will sustain their level in the upcoming season? Can any kick it up another notch? Who to avoid like the H1N1 flu? Separating the repeat performers from the one year wonders can help your team rise to the top of your league."
The notion that an increase in BABIP is an actual improvement is not generally accepted by the baseball experts I know.
Complaints about the first half of this article aside, I quite liked the table he produced which gave a fine indication of the odds players will revert back to their norms given the various age slots. In this competition that is enough substance for Brian to walk to the next round.
The point of the Wieters section was unclear to me. Do I understand correctly that Brian made all his own projections based on each players previous yearâ€™s performance or were those projections averaged in with PECOTAâ€™s projections â€“ or some other combo? As with the other critics, I agree the jump to â€œnailingâ€ Wieters is suspect.
In the Jay Bell Syndrome, it is left out how Nelson Cruz performed in the minors.
Like me, Brian is obviously not a natural writer and probably needs to get what he has written out of his head before going back to proof read for mistakes (â€œoneâ€ instead of â€œonâ€) and smoothness of prose.
The conclusion is honest, but to overlook scouting reports and luck indicators such as BABIP proves you are not yet a BP level annalist. As Iâ€™ve said before, you have some catching up to do to understand what the rest of us know already. Overall, the creativity and substansiveness of your sabermetric insights in my opinion outweigh the problems in your articles to keep you in the upper half of the competition.
I did look at how much each player over or underperformed his own established level of BABIP, and reported those in the two tables.
As for scouting reports, that would probably take another article. I didn't have time to do the necessary research for this piece. We can look at scouting reports for Hanley Ramirez and Miguel Cabrera that turned out to be prophetic, but how many scouts wrote glowing reports of guys who didn't make it? If you don't have a nearly complete list of what the scouts said, you're going to be cherry picking.
Of course, you didn't have room to completely cover scouting reports with the same weight that you covered the ideas you introduced here, but it seems fair to, at least, mention it. Rereading your piece, though, I see you did qualify your conclusion with "statistical indicators".
Given the article's early use of wOBA (an excellent stat, but one readers might not see as meaningfully different from EqA in terms of measuring overall offense), it might have been useful to explain why it was being used.
Just kidding. I liked your first article, and I liked this one better. Nice job.
Upon doing so, I looked more favorably upon his work, but still could not bring myself to offer the thumbs-up. Perhaps next week.
"After posting Major League Equivalent wOBAs (Tom Tango's weighted on-base average) of .311, .329 and .295 the three seasons prior to the trade, (MLB average for shortstop is .318) the 22 year old Ramirez won the NL Rookie of the Year Award for the Marlins, putting up an 292/353/480 BA/OB/SA line, good for a .360 wOBA. If making draft list in 2007, I had to ask 'Who is the real Hanley Ramirez?'"
The first sentence is tortuous, the second is meaningless. I care about grammar, and any time I read published work with this many errors in the first couple paragraphs, I stop reading. I scanned the rest of the article, and didn't find enough to outweight my initial impression.
In each of his last three seasons at UNC, Cleveland improved his stats, coming out of college with a .345 projected wOBA, mlb average for a corner outfielder, and the only season he projected as mlb quality after factoring in past seasons and regression. I would worry if his Sr year was an outlier, but it's a positive that he improved two years in a row.
Wieters was very consistent in his last three years at Ga Tech, all projecting to a 340's wOBA. Although average for a corner of such as Cleveland, this is well above average for a mlb catcher (312), in the range of Russell Martin or Ryan Doumit.
Cleveland went to Class-A the same season he was a Sr in college, and continued to rake, showing power and high BABIP. Ever since that one year, for whatever reason, his HR% tanked (.060 to .020, mlb avg .040) and stayed low the rest of his career. Was he injured, such as Jason Kendall's thumb?
Wieters, coming out of college with the same absolute numbers as Cleveland but much higher compared to his position, spent half a season in Class-A and half in Double-A, outhitting everyone at both stops. In 2009 Wieters was promoted to Triple-A and his stats appeared disappointing compared to his rpevious year, but after adjusting for park factors (home Harbor Park in Norfolk is worst hitters prk in Triple-A) Wieter's 2009 line almost exactly matched the projections based on his college numbers combined with his 2008 minor league stats. As far as I know, PECOTA only used 2008 and relied on the outlier, giving a much higher projection.