I have a post over at the Hardball Times today about a way to improve replay review. Check it out!
Summary: I look at averages for hitters and pitchers in the postseason to see how their quality (relative to league average) has changed over time. Unsurprisingly, the gap between postseason and regular season average pitchers is larger than the comparable gap for hitters. The trend over time for pitchers is expected, with a decrease in quality relative to league average from the 1900s to mid-1970s and a slight increase since then that appears to be linked with the increased usage of relievers. The trend for hitters is more confusing, with a dip from 1950 to approximately 1985 and an increase since then. Overall, however, the average quality of both batters and pitchers in the postseason relative to league average is as high as it has been in the expansion era.
Quality of play in the postseason is a common trope of baseball discussion. Between concerns about optics (you want casual fans to watch high quality baseball) and rewarding the best teams, there was a certain amount of handwringing about the number of teams with comparatively poor records into the playoffs (e.g., the Giants and Royals made up the only pair of World Series teams ever without a 90 game winner). This prompted me to wonder about the quality of the average players in the postseason and how that’s changed over time with the many changes in the game—increased competitive balance, different workloads for pitchers, changes in the run environment, etc.
For pitchers, I looked at weighted league-adjusted RA9, which I computed as follows:
- For each pitcher in the postseason, compute their Runs Allowed per 9 IP during the regular season. Lower is better, obviously.
- Take the average for each pitcher, weighted by the number of batters faced.
- Divide that average by the major league average RA9 that year.
You can think of this as the expected result you would get if you chose a random plate appearance during the playoffs and looked at the pitcher’s RA9. Four caveats here:
- By using RA9, this is a combined pitching/defense metric that really measures how much the average playoff team is suppressing runs relative to league average.
- This doesn’t adjust for park factors, largely because I thought that adjustment was more trouble than it was worth. I’m pretty sure the only effect that this has on aggregate is injecting some noise, though I’m not positive.
- I considered using projected RA9 instead of actual RA9, but after playing around with the historical Marcel projections at Baseball Heat Maps, I didn’t see any meaningful differences on aggregate.
- For simplicity’s sake, I used major league average rather than individual league average, which could influence some of the numbers in the pre-interleague play era.
When I plot that number over time, I get the following graph. The black dots are observed values, and the ugly blue line is a smoothed rolling estimate (using LOESS). (The gray is the confidence interval for the LOESS estimate.)
While I wouldn’t put too much weight in the LOESS estimate (these numbers should be subject to a large bit of randomness), it’s pretty easy to come up with a basic explanation of why the curve looks the way it does. For the first seventy years of that chart, the top pitchers pitched ever smaller shares of the overall innings (except for an uptick in the 1960s), ceding those innings to lesser starters and dropping the average quality. However, starting in the 1970s, relievers have covered larger portions of innings (covered in this FiveThirtyEight piece), and since relievers are typically more effective on a rate basis than starters, that’s a reasonable explanation for the shape of the overall pitcher trend.
What about hitters? I did the same calculations for them, using wOBA instead of RA9 and excluding pitchers from both postseason and league average calculations. (Specifically, I used the static version of wOBA that doesn’t have different coefficients each year. The coefficients used are the ones in The Book.) Again, this includes no park adjustments and rolls the two leagues together for the league average calculation. Here’s what the chart looks like:
Now, for this one I have no good explanation for the trend curve. There’s a dip in batter quality starting around integration and a recovery starting around 1985. If you have ideas about why this might be happening, leave them in the comments or Twitter. (It’s also quite possible that the LOESS estimate is picking up something that isn’t really there.)
What’s the upshot of all of this? This is an exploratory post, so there’s no major underlying point, but from the plots I’m inclined to conclude that, relative to average, the quality of the typical player (both batter and pitcher) in the playoffs is as good as it’s been since expansion. (To be clear, this mostly refers to the 8 team playoff era of 1995–2011; the last few years aren’t enough to conclude anything about letting two more wild cards in for a single game.) I suspect a reason for that is that, while the looser postseason restrictions have made it easier for flawed teams to make it in the playoffs, they’ve also made it harder for very good teams to be excluded because of bad luck, which lifts the overall quality, a point raised in this recent Baseball Prospectus article by Sam Miller.
Two miscellaneous tidbits from the preparation of this article:
- I used data from the Lahman database and Fangraphs for this article, which means there may be slight inconsistencies. For instance, there’s apparently an error in Lahman’s accounting for HBP in postseason games the last 5 years or so, which should have a negligible but non-zero effect on the results.
- I mentioned that the share of batters faced in the postseason by the top pitchers has decreased steadily over time. I assessed that using the Herfindahl-Hirschman index (which I also used in an old post about pitchers’ repertoires.) The chart of the HHI for batters faced is included below. I cut the chart off at 1968 to exclude the divisional play era, which by doubling the number of teams decreased the level of concentration substantially.
I wrote an article for The Hardball Times about using subterfuge to obtain the platoon advantage. Check it out here.
Summary: Based on a suggestion heard at SaberSeminar, I use a few different means to examine how rookie umpires call the strike zone. Those seven umpires appear to consistently call more low strikes than the league as a whole, but some simple statistics suggest it’s unlikely they are actually moving the needle.
Red Sox manager John Farrell was one of the speakers at Saberseminar, which I attended last weekend. As I mentioned in my recap, he was asked about the reasons offense is down a hair this year (4.10 runs per team per game as I type this, down from 4.20 through this date (4.17 overall) in 2013). He mentioned a few things, but one that struck me was his suggestion that rookie umpires calling a larger “AAA strike zone” might have something to do with it.
Of course, that’s something we can examine using some empirical evidence. Using this Hardball Talk article as a guide, I identified the seven new umpires this year. (Note that they are new to being full-fledged umps, but had worked a number of games as substitutes over the last several years.) I then pulled umpire strike zone maps from the highly useful Baseball Heat Maps, which I’ve put below. Each map shows the comparison between the umpire* and league average, with yellow marking areas more likely to be called strikes and blue areas less likely to be called strikes by the umpire.
* I used the site’s settings to add in 20 pitches of regression toward the mean, meaning that the values displayed in the charts are suppressed a bit.
The common thread, to me, is that almost all of them call more pitches for strikes at the bottom of the zone, and most of them take away outside strikes for some batters. Unfortunately, these maps don’t adjust for the number of pitches thrown in each area, so it’s hard to get aggregate figures for how many strikes below or above average the umpires are generating. The two charts below, from Baseball Savant, are a little more informative; red dots are the bars corresponding to rookie umps. (Labeling was done by hand in MS Paint, so there may be some error involved.)
The picture is now a bit murkier; just based on visual inspection, it looks like rookie umps call a few strikes more than average on pitches outside the zone, and maybe call a few extra balls on pitches in the zone, so we’d read that as nearly a wash, but maybe a bit on the strike side.
So, we’ve now looked at their strike zones adjusted for league average but not the number of pitches thrown and their strike zones adjusted for the relative frequencies of pitches but not seriously adjusted for league average. One more comparison, since I wasn’t able to find a net strikes leaderboard, is to use aggregate ball/strike data, which has accurate numbers but is unadjusted for a bunch of other stuff. Taking that information from Baseball Prospectus and subtracting balls in play from their strikes numbers, I find that rookie umps have witnessed in total about 20 strikes more than league average would suggest, though that’s not accounting for swinging vs. called or the location that pitches were thrown. (Those are substantial things to consider, and I wouldn’t necessarily expect them to even out in 30 or so games.)
At 0.12 runs per strike (a figure quoted by Baseball Info Solutions at the conference) that’s about 2.4 runs, which is about 0.4% of the gap between this year’s scoring and last year’s. (For what it’s worth, BIS showed the umpires who’d suppressed the most offense with their strike zones, and if I remember correctly, taking the max value and applying it to each rookie would be 50–60 total runs, which is still way less than the total change in offense.)
A different way of thinking about it is that the rookie umps have worked 155 games, so they’ve given up an extra strike every 8 or so games, or every 16 or so team-games. If the change in offense is 0.07 runs per team-game, that’s about one strike per game. So these calculations, heavily unadjusted, suggest that rookie umpires are unlikely to account for much of the decrease in scoring.
So, we have three different imperfect calculations, plus a hearsay back of the envelope plausibility analysis using BIS’s estimates, that each point to a very small effect from rookie umps. Moreover, rookie umps have worked 8.3% of all games and 8.7% of Red Sox games, so it seems like an odd thing for Farrell to pick up on. It’s possible that a more thorough analysis would reveal something big, but based on the data easily available I don’t think it’s true that rookie umpires are affecting offense with their strike zones.
I was fortunate enough to be at SaberSeminar this past weekend, held at Boston University and organized by (among others) Dan Brooks, the titular Brooks of PitchF/X site BrooksBaseball. I took some notes throughout the weekend, and I’ve typed them up below, broken into smaller observations. (All of the reflection was done on the bus home, so any mistakes are my own due to typing with a fried brain.)
One other thing to tell is that I presented some research I did on shifts in the strike zone (it actually came out of this article about high socks), and I’m going to be writing that up as an article soon, though it may appear at a different site. All in all, quite an enjoyable weekend even after factoring in the scattered criticisms below.
It was my first time at a baseball conference and the first time I’d been at any conference in quite some time, and it actually struck me as pretty similar to an indie music festival. The crowd wasn’t huge (a couple hundred people), and they all knew most of the speakers, who mostly stuck to greatest hits sort of things. (Most of what I saw presented wasn’t novel, especially by the more prominent folk.) That’s not to take away from the sessions—it was still interesting to meet and hear people that I had only read, and it was still great to be around a group where everyone was interested in the same sort of stuff and you could bring up things like SIERA and wOBA without much risk of confusion.
They had a panel discussion featuring three of the Red Sox baseball operations interns, and I was reminded of how skeezy some aspects of that system are. The moderator talked about how there were fewer MLB intern slots than there used to be because the feds cracked down on illegal internships, which he framed as a bad thing. I found that a bit horrifying; it seems odd to me that a team with a payroll of hundreds of millions would cut staff rather than pay a semi-reasonable wage to their junior people. (Even if it makes economic sense, it seems like a bad way to treat people.)
The interns, for their part, didn’t offer a whole lot of insight into things. (Not that I blame them; it’s hard to be insightful during a panel discussion.) They twice dodged the question of how many hours they work, only saying “a lot.” (It’s possible I’m being too harsh and they actually don’t know; because I’m billed out by the hour at work, I have to keep very accurate time logs, which means I know how much I’ve worked every week since I’ve started, but I may be an outlier in that regard.) One of the failings of the panel was that it didn’t include anyone who had been an intern and washed out (either quit or wasn’t offered a job), which would have been more informative and more interesting. (I understand not wanting to irk any of the teams, but I don’t think this is too inflammatory.) This is the same problem I ran into in college a lot, where most of the advice I got about whether to get a Ph.D. came from people who had not only loved grad school but also met with astounding success afterward.
I also thought about the fact that they are hiring recent grads of extremely expensive schools (Columbia, Yale, and Georgetown, in this case) to be extremely underpaid interns; I wonder how much of their labor pool is indirectly disqualified simply due to a lack of connections or a need to make money for family or student loan reasons. It’s very puzzling to me that teams, despite being flush with cash, hire people similarly to high-prestige, low-money companies like magazines rather than high-prestige, rich firms like banks and tech companies. I’d love to see more discussion from people who know more about the industry than I do.
One mostly unstated theme that kept occurring to me throughout the weekend was the issue of class, opportunity, and privilege, which popped up in a number of different ways:
- Baseball Prospectus’s Russell Carleton discussed how teams can (or should) help their players develop into better adults by focusing on their financial, practical, nutritional, and mental well-being. While he focused mostly on the positive effects it would have on a player’s career and thus a team’s investment (fewer distractions and a better makeup will help talent win out), it seems to me that it’s a good thing in its own right to improve the life skills of the wash outs—especially the ones that skipped college and/or come from poorer backgrounds. Fewer guys in society who behave like Dirk Hayhurst’s teammates is probably a good thing.
- Tom Tippett, a senior analyst for the Red Sox, talked about how he always appreciates players that are diamonds in the rough, i.e. guys who went undrafted, played in the independent leagues, etc. The thing is, though, that equality of access doesn’t exist, and it’s an interesting thought experiment to think of how many guys get cut before they figure this out. In particular, I wonder how many guys are able to get to a good college or (building on Carleton’s point) hang around the minors longer (thus increasing the opportunity that they make the big leagues) because they have better “makeup” that really comes from growing up with a few more advantages.
- There was a demo of TrackMan, which is a portable radar system that can be used to evaluate pitch speed and rotation. One of the guys I was talking to pointed out that they’d sold as many of the systems as they could to clubs and agencies and were hoping to move on to selling it to amateurs. I don’t know what one costs, but the idea of buying a portable radar system for your high school pitcher seems like a caricature of what a rich family gunning for a scholarship would do (analogous to all of the academic tutoring and test prep that a lot of people I know did).
- Relatedly, a number of guys talked about the various high school showcases that pit the best high school talent in a region against each other, and one casually mentioned how much money they bring in. Again, seems like the sort of thing that serves to extract cash from hyperzealous parents and limit the opportunities for kids of less means, but I don’t know enough about the system to comment.
- Internship opportunities for big league teams, which I discussed above.
- I’ve been of the conviction for a while that pro sports would be equally or more enjoyable and substantially less ethically problematic if the teams were run as non-profits in the same general manner as art museums and whatnot (the Green Bay Packers are something like this already), and while I won’t go into that further here, if that were the case I think it’d be easier for a lot of the explicit privilege issues to be brought up in the game. While teams are still nominally concerned with profits, it’s a lot easier for them to sidestep problems that in a better world they could help address.
There were a number of talks with a more medical and scientific focus, and they provided good examples of how hard it is to apply these things rigorously to baseball (or any other real world application). There are lots of studies with very small N (“N=4” appeared on one slide describing research that had been published), and they are presumably subject to the same sorts of issues that all public health and social science papers are. While I’m sure lots of teams (in all sports) would love to bring in scientists to help them with things like sleep and vision, I imagine there’s a lot of stuff that falls apart between the lab and the field (if it even exists at all).
I should mention that the first talk of the conference, by a UC-Riverside professor who focuses on vision, did have experimental evidence of how improved vision helps college player performance, but it’s still tiny samples and only college students, and thus to be taken with a grain of salt.
There was an interesting panel featuring Matt Swartz, Ben Baumer, and Vince Gennaro discussing the relationship between winning and teams’ making money that prompted at least a couple article ideas for the future. Gennaro said he thought that the way teams spend money might change a bit after the addition of the second wild card, as there is now much greater variety across playoff teams in terms of how valuable the postseason slot is—the first seed became more valuable and the wild card slots substantially less. I have some thoughts on that, but will leave them for future articles.
Astros’ GM Jeff Luhnow gave a pleasant enough if relatively fact-free talk, the main focus of which was the importance of convincing the uniformed personnel of the importance of the sabermetric principles that buck conventional wisdom. He used the example of the shift and how it took the Astros three years to actually get people on board with it; obviously, if the players and the manager don’t like it, it won’t work as well as it would otherwise. I honestly wouldn’t be surprised if this becomes much less of an issue in 10 or so years, when the reasoning will have permeated a bit more through the baseball establishment and managers and young players will be a lot more open to things.
Vince Gennaro gave a very similar talk, and one thing he brought up was that you need to strike a balance between sticking to general principles (about shifting, pitcher workload, etc.) and making exceptions where warranted. Given that Luhnow talked about how he had made too many exceptions about when to shift last year, the point was hammered home, though it’s a vague enough point that it’s hard to really implement. (I was also reminded of this recent article about Ruben Amaro and exceptions.)
While Red Sox GM Ben Cherington didn’t discuss anything much more novel than what Luhnow covered, he was a lot more personable and down-to-earth while doing so. I imagine some of that is personality and a lot of it is the result of being the GM of the defending World Series champs and talking in his own backyard instead of presiding over three years of horrible teams and a lot of criticism from around baseball.
The last question asked of Luhnow was a minute-long ramble that wasn’t really a question and basically turned into “haha, you screwed up the Brady Aiken situation,” and Luhnow looked pretty peeved afterward, prompting Dan Brooks to tell the crowd not to be jerks to the presenters. There were a lot of bad questions all weekend, especially to Cherington, Red Sox manager John Farrell, and Luhnow, who clearly couldn’t say anything about their teams to us that was any more interesting than what they tell the media after a game. That didn’t stop people from bugging Farrell about his bullpen, though.
There was a bit more offensive humor in some of the talks than I would have expected—one professor made a joke about George W. Bush being brain damaged, and another managed to have a slide showing him wearing a t shirt that said “Drunk Bitches Love Me,” a slide with a cartoon captioned “I Will Fucking Cut You, Bitch,” and threw in a couple fat jokes for good measure. As anyone who knows me will attest, I have a reasonably sharp sense of humor, but throwing around jokes with misogynistic overtones at a conference that I would estimate was 90–95% men made me cringe.
More bullets from John Farrell’s talk:
- When asked (I believe by accomplished sabermetrician Mitchel Lichtman, aka MGL) about his handling of the bullpen on Friday night, he discussed how “it’s about 162 games” (specifically about pulling Koji Uehara after an easy inning). While that’s certainly important, I wonder how often it’s just used as a crutch to justify poor decisions. As Grantland’s Bill Barnwell has written before, you basically only ever see nebulous qualitative concerns like Uehara’s overwork invoked to defend conservative decisions, never more aggressive ones.
- Farrell thinks that the spree of Tommy John surgeries is more due to guys overthrowing (to impress scouts) than it is due to the sheer volume of pitches. The surgeon that spoke later in the conference disagreed.
- He mentioned that it’s easier to use less conventional strategies once opponents start doing it, because it seems more normal to the players. One obvious consequence of this is that really conspicuous tactical advances like shifting are going to be relatively short-lived advantages.
- He thinks that introducing replay and the corresponding promotion of several umpires has led to an expanded strike zone and that’s part of the continued downturn in baseball. That seems like a testable hypothesis, and something I’ll probably look at soon.
- Farrell mentioned that figuring out how guys will react when they fail as baseball players for the first time is an important part of helping players move through the system, which I guess is a big part of what Russell Carleton later talked about. It reminded me a lot of what people say about elite colleges and places like Stuyvesant High School, where a lot of people have to adjust from being in the 99th percentile of their peers at their prior institution to being median or below at their new one.
Russell Carleton gets major points for treating data as a plural noun rather than a singular one; he was the only one I noticed doing that all conference. I’d be curious to see what the usage rates are depending on background, with my guess being that people with more academic experience use “are” more than people who mostly use data in a private sector setting. (Yes, I’m a pedant about some of these things.)
Two White Sox notes from people’s presentations:
- Apparently Tyler Flowers is fifth in the bigs in saving runs by catch framing. Some of that is surely because of his workload, but it’s still a surprise.
- Not surprising, but something I’d forgotten: Erik Johnson was BP’s #1 prospect for the White Sox at the beginning of the year. Sigh.
Another one of those themes that kept popping up to me during the weekend was the idea of how teams preserve their edges, especially the ones they derive through quantitative and sabermetric means. (I’m reminded of the Red Queen hypothesis, which is an evolutionary biology idea I learned about through quiz bowl that applies pretty well to baseball analysis. Basically, you have to keep advancing in absolute terms to stay in the same place relatively, because if you are complacent people will catch up to you naturally.) A few places that this issue manifested itself:
- MGL said that, for general short term forecasting, any of the major projection public systems (Oliver, Steamer, PECOTA, and ZiPS) will do. He actually suggested that teams were probably wasting their time trying to come up with a better general forecasting system and that understanding volatility and more specific systems is probably more important. Jared Cross, developer of Steamer, disagreed a little bit; his view was that the small gap between current projection systems and perfect estimates meant that a seemingly marginal improvement would actually mean a lot because of how competitive things are.
- The sort of innovations Carleton discussed are the sort of thing that would pretty quickly spread throughout baseball, making any one team;s advantage relatively fleeting. Of course, if it spreads then it’s likely to increase the overall quality of the talent pool, which would lead to either better games or more teams, which is probably good. (It’s not necessarily good for the current players; any progress in player evaluation and development is unlikely to actually help players, given that the total number of jobs isn’t increasing; if one player does better, another loses his job.)
- Tom Tippett said that if he had his druthers the public wouldn’t have access to PitchF/X data, because it allows teams who don’t have good analysts to borrow heavily from the public and thus decreases the advantage that analytically-inclined organizations hold. While I think that’s true from his self-interested perspective, I think it’s a bit short-sighted overall, and I wish he’d answered with a “good for the game” perspective. When you think about what’s best for fans, I think defending closed systems is probably harder; one of the things I like about baseball is how freely available the data are, and to the extent that becomes less true I think it’s a sad thing.
- On that note, I’m a bit amused by what people think of as being “trade secrets”; there were lots of teams bouncing around and a few presentations by companies that are using highly proprietary data analysis methods that they are trying to sell people on. Again, it’s hard for me to evaluate how meaningful that stuff is (though people love kicking around the figure that pro teams are five years ahead of the public in their understanding of things), but even if it does represent a competitive advantage it’s still pretty funny to step back and think about the secrecy that’s applied to sports.
Several different people brought up StatCast, which is the new data collection system MLB Advanced Media is going to roll out some time soon; it will provide a huge amount of data on how fast players move and how quickly they react that will allow for analysis that’s a bit more along the lines of what the SportVU cameras do in basketball. (See the videos in the above link for examples.) There’s still no sense of whether or not it will be made public (and in what form it might be made public), but people were uniformly excited about it.
The projections folk were united in the belief that it would have a huge effect on projecting defense, to the point where MGL thinks defense will go from being the hardest component of the sport to analyze and predict to the easiest. There was a bit more divergence about what it might do for pitching and batting analysis, as well as about when it would come out—one speaker quoted MLB and said it would be ready to go by the beginning of next year, whereas Dave Cameron pointed out that test data hadn’t been released to the teams yet despite what was originally promised and thus thought it was highly unlikely that the data would be ready for teams by next year, much less ready for public consumption.
Dan Brooks jokingly introduced a hitting metric he called “GIP,” for Google Images Performance. It was prompted by the fact that a Google search for “miguel cabrera hitting” or “david ortiz hitting” finds pictures of them hitting home runs, whereas a query for “jose molina hitting” gets mostly pictures of him behind the plate.
More notes from Tippett’s talk:
- He started by mentioning that the hardest part of his job is deciding when a player’s underperformance is real and not just noise; I don’t think I’ve ever seen a rigorous evaluation of how to do that using Bayes’ rule and the costs of Type I and Type II errors (booting a good player and keeping a bad one, respectively), but I would love to read one. (How projection systems react to new performance data is in principle just Bayesian reasoning regardless.)
- He non-snarkily talked about the “momentum” that the Red Sox had, and nobody asked about it. All that probably means is that when he’s talking outside the office he’s perfectly willing to let a bit of narrative creep in.
- He talked about their decision to cut Grady Sizemore and how it was related to certain incentives in his contract that would have vested. I casually wonder how cavalier teams are allowed to be in explicitly making decisions based on players’ contract incentives, given that there have been talks about how players might opt to file grievances about these things in the past. (The one that comes to mind is Brett Myers, who would have had an option vest if he finished a certain number of games for the White Sox a couple years back. The commentary I read suggested that conspicuously changing his usage pattern would have been grounds for a grievance, but I don’t know how that applies to situations like Sizemore’s.)
Quick summary: I test the ZiPS and Marcel projection systems to see if their errors are larger for players with larger platoon splits. A first check says that they are not, though a more nuanced examination of the system remains to be conducted.
First, a couple housekeeping notes:
- I will be giving a short talk at Saberseminar, which is a baseball research conference held in Boston in 10 days! If you’re there, you should go—I’ll be talking about how the strike zone changes depending on where and when games are played. Right now I’m scheduled for late Sunday afternoon.
- Sorry for the lengthy gap between updates; work obligations plus some other commitments plus working on my talk have cut into my blogging time.
After the A’s went on their trading sprees last week at the trading deadline, there was much discussion about how they were going to intelligently deploy the rest of their roster to cover for the departure of Yoenis Cespedes. This is part of a larger pattern with the A’s as they continue to be very successful with their platoons and wringing lots of value out of their depth. Obviously, when people have tried to determine the impact of this trade, they’ve been relying on projections for each of the individual players involved.
What prompted my specific question is that Jonny Gomes is one of those helping to fill Cespedes’s shoes, and Gomes has very large platoon splits. (His career OPS is .874 against left-handed pitchers and .723 against righties.) The question is what proportion of Gomes’s plate appearances the projection systems assume will be against right handers; one might expect that if he is deployed more often against lefties than the system projects, he might beat the projections substantially.
Since Jonny Gomes in the second half of 2014 constitutes an extremely small sample, I decided to look at a bigger pool of players from the last few years and see if platoon splits correlated at all with a player beating (or missing) preseason projections. Specifically, I used the 2010, 2012, and 2013 ZiPS and Marcel projections (via the Baseball Projection Project, which doesn’t have 2011 ZiPS numbers).
A bit of background: ZiPS is the projection system developed by Dan Szymborski, and it’s one of the more widely used ones, if only because it’s available at FanGraphs and relatively easy to find there. Marcel is a very simple projection system developed by Tangotiger (it’s named after the monkey from Friends) that is sometimes used as a baseline for other projection systems. (More information on the two systems is available here.)
So, once I had the projections, I needed to come up with a measure of platoon tendencies. Since the available ZiPS projections only included one rate stat, batting average, I decided to use that as my measure of batting success. I computed platoon severity by taking the larger of a player’s BA against left-handers and BA against right-handers and dividing by the smaller of those two numbers. (As an example, Gomes’s BA against RHP is .222 and against LHP is .279, so his ratio is .279/.222 = 1.26.) My source for those data is FanGraphs.
I computed that severity for players with at least 500 PA against both left-handers and right-handers going into the season for which they were projected; for instance, for 2010 I would have used career data stopping at 2009. I then looked at their actual BA in the projected year, computed the deviation between that BA and the projected BA, and saw if there was any correlation between the deviation and the platoon ratio. (I actually used the absolute value of the deviation, so that magnitude was taken into account without worrying about direction.) Taking into account the availability of projections and requiring that players have at least 150 PA in the season where the deviation is measured, we have a sample size of 556 player seasons.
As it turns out, there isn’t any correlation between the two parameters. My hypothesis was that there’d be a positive correlation, but the correlation is -0.026 for Marcel projections and -0.047 for ZiPS projections, neither of which is practically or statistically significantly different from 0. The scatter plots for the two projection systems are below:
Now, there are a number of shortcomings to the approach I’ve taken:
- It only looks at two projection systems; it’s possible this problem arises for other systems.
- It only looks at batting average due to data availability issues, when wOBA, OPS, and wRC+ are better, less luck-dependent measures of offensive productivity.
- Perhaps most substantially, we would expect the projection to be wrong if the player has a large platoon split and faces a different percentage of LHP/RHP during the season in question than he has in his career previously. I didn’t filter on that (I was having issues collecting those data in an efficient format), but I intend to come back to it.
So, if you’re looking for a takeaway, it’s that large platoon-split players on the whole do not appear to be poorly projected (for BA by ZiPS and Marcel), but it’s still possible that those with a large change in circumstances might differ from their projections.
Short summary: I look at how much year-over-year persistence there is in the lists of best position players and pitchers, using Wins Above Replacement (WAR). It appears as though there is substantial volatility, with only the very best players being more likely than not to repeat on the lists. In the observed data, pitchers are slightly more likely to remain at the top of the league than position players, but the difference is not meaningful.
Last week, Sky Kalkman posted a question that I thought seemed interesting.
Sky’s Theorem: the group of aces has a half-life of two years. If you’re an ace, 50/50 you’re one in two years. Anyone want to research it?
— Sky Kalkman (@Sky_Kalkman) July 9, 2014
Obviously, this requires having a working definition of “ace,” for which Kalkman later suggests “the top dozen-ish” pitchers in baseball. That seems reasonable to me, but it raises another question: what metric to use for “top”?
I eventually wound up using an average of RA9-WAR and FIP-WAR (the latter being the standard WAR offered by FanGraphs). There are some drawbacks to using counting stats rather than a rate stat, specifically that a pitcher that misses two months due to injury might conceivably be an ace but won’t finish at the top of the leaderboard. That said, my personal opinion is that health is somewhat of a skill and dependability is part of being an ace.
I chose to use this blend of WAR (it’s similar to what Tangotiger sometimes says he uses) because I wanted to incorporate some aspects of Fielding Dependent Pitching into the calculations. It’s a bit arbitrary, but the analysis I’m about to present doesn’t change much if you use just FIP-WAR or plain old FIP- instead.
I also decided to use the period from 1978 to the present as my sample; 1978 was the expansion that brought the majors to 26 teams (close to the present 30), keeping the total pool of talent reasonably similarly-sized throughout the entire time period while still providing a reasonably large sample size.
So, after collecting the data, what did I actually compute? I worked with two groups—the top 12 and top 25 pitchers by WAR in a given year—and then looked at three things. I first examined the probability they would still be in their given group the next year, two years after, and so on up through 10 years following their initial ace season. (Two notes: I included players tied for 25th, and I didn’t require that the seasons be consecutive, so a pitcher who bounced in and out of the ace group will still be counted. For instance, John Smoltz counts as an ace in 1995 and 2005 in this system, but not for the years 2001–04. He’s still included in the “ace 10 years later” group.) As it turns out, the “half-life” that Kalkman postulates is less than a year: 41% of top 25 pitchers are in the top 25 the next year, with that figure dropping to 35% for top 12 pitchers who remain in the top 12.
I also looked at those probabilities by year, to see if there’s been any shift over time—basically, is the churn greater or less than it used to be? My last piece of analysis was to look at the probabilities by rank in the initial year to see how much more likely the very excellent pitchers are to stay at the top than the merely excellent pitchers. Finally, I ran all of these numbers for position players as well, to see what the differences are and provide some additional context for the numbers.
I played around and ultimately decided that some simple charts were the easiest way to convey what needed to be said. (Hopefully they’re of acceptable quality—I was having some issues with my plotting code.) We’ll start with the “half-life” graph, i.e. the “was this pitcher still an ace X years later?” chart.
As you can see, there’s a reasonable amount of volatility, in that the typical pitcher who cracks one of these lists won’t be on the same list the next year. While there’s a small difference between pitchers and position players for each group in the one year numbers, it’s not statistically significant and the lines blur together when looking at years further out, so I don’t think it’s meaningful.
Now, what if we look at things by year? (Note that from here on out we’re only looking at the year-over-year retention rate, not rates from two or more years.)
This is a very chaotic chart, so I’ll explain what I found and then show a slightly less noisy version. The top 25 groups aren’t correlated with each other, and top 25 batters isn’t correlated with time. However, top 25 pitchers is lightly positively correlated with time (r = 0.33 and p = 0.052), meaning that the top of the pitching ranks has gotten a bit more stable over the last 35 years. Perhaps more interestingly, the percentage for top 12 pitchers is more strongly positively correlated with time (r = 0.46, p < 0.01), meaning that the very top has gotten noticeably more stable over time (even compared to the less exclusive top), whereas the same number for hitters is negative (r = -0.35, p = 0.041), meaning there’s more volatility at the top of the hitting WAR leaderboard than there used to be.
These effects should be more visible (though still noisy) in the next two charts that show just the top 12 numbers (one’s smoothed using LOESS, one not). I’m reluctant to speculate as to what could be causing these effects: it could be related to run environments, injury prevention, player mobility, or a whole host of other factors, so I’ll leave it alone for now. (The fact that any explanation might have to also consider why this effect is stronger for the top 12 than the top 25 is another wrinkle.)
The upshot of these graphs, though, is that (if you buy that the time trend is real) the ace half-life has actually increased over the last couple decades, and it’s gone down for superstar position players over the same period.
Finally, here’s the chart showing how likely a player who had a given rank is to stay in the elite group for another year. This one I’ve also smoothed to make it easier to interpret:
What I take away from these charts are that, unsurprisingly, the very best players tend to persist in the elite group for multiple years, but that the bottom of the top is a lot less likely to stay in the group. Also, the gaps at the far left of the graph (corresponding to the very best players) are larger than the gaps we’ve seen between pitchers and hitters anywhere else. This says that, in spite of pitchers’ reputation as being volatile, at the very top they have been noticeably less prone to large one-year drop offs than the hitters are. That said, the sample is so small (these buckets are a bit larger than 30 each) that I wouldn’t take that as predictive so much as indicative of an odd trend.