Is Sabermetrics Helping to Ruin Baseball?

Caution: A lot of opinion to follow. And please understand I love baseball and I obviously am a believer in analytics.

Imagine a world in which baseball teams had no idea how good or bad a player could be expected to perform from year to year. All players would be total blank-slate mysteries. In this world, it wouldn't matter at all how high or low a team’s payroll is. A team that spends $1M on payroll would be just as likely to win a championship as a team that spent $200M. The correlation between payroll and wins would be zero.

Now imagine the opposite scenario in which teams had perfect crystal-ball foresight about exactly how well a player could be expected to perform. Now the team with the highest payroll would always have the most wins because there would be a 1:1 direct linkage between performance and pay. This would make the correlation between payroll and wins a perfect 1. We wouldn't even need to play out the season. Just mail the trophy to the team with the highest payroll and save everyone the trouble.

In reality we’re about halfway between the first case and the second case. We neither have no idea about expected performance, nor do we have perfect predictive knowledge.  MLB's correlation between team payroll and wins has been trending at about r = 0.4. It varies from season to season, but its recent central tendency appears to be right round 0.4.

The problem that analytics causes is this: The better the sabermetrics (or the more uniformly adopted it becomes), the better the crystal ball becomes and the closer we move toward a correlation of 1...which, by the way, is a bad thing. Of course, we'd never see a correlation of 1 because of injuries and simple randomness, but it could get much higher as it did in earlier eras. The correlation has been higher and lower in years past, varying for reasons that have nothing to do with analytics. But the point is that sabermetrics, to the degree it is effective, makes the correlation higher than it otherwise would be, and this gives an ever greater advantage to big-payroll teams.


I sense there is a widespread misunderstanding about sabermetrics and analytics, probably attributable to Moneyball. It doesn't even the playing field as many think. Sabermetrics allowed a small-market team like the A's to compete with the big boys only because it was the sole team using the right analytic tools and could exploit a market inefficiency. Once all the teams jumped on the sabermetric bandwagon, the advantage was not only lost, but turned on its head as the correlation between pay and wins improved.

Additionally, sabermetrics has helped make MLB depressingly predictable, despite some rare notable surprises. Last night, the team with the 4th most all-time championships beat the team with the 2nd most all-time championships. Sure, the Red Sox went from worst to first in their division, which to the naive observer probably seems amazing. But if you know your PECOTA, you knew last spring they would be strong contenders and among the favorites to win the World Series.

The really depressing thing is that MLB apparently likes things this way.

The short-term interest of both MLB and their broadcast partners is to see large-market teams dominate the post-season, because gives a temporary bump to broadcast ratings. But the lack of churn in the stratum of top teams comes at the expense of developing a long-term national following, where every fan has hope in April, at least every few years. The result of MLB's shortsightedness is that more sports fans will now watch a meaningless week 7 NFL game featuring teams with the worst combined winning percentage of any Monday night game in history than Game 1 of the World Series. I don't say that to gloat as a football guy; I am lamenting the national irrelevance of a sport I love.

MLB seems to be in a self-devouring spiral in which interest is dwindling nationally but strengthening locally, particularly in large markets. This is partly due to the lack of churn among top teams and a high level of predictability. Every season Brewers fans know deep down they have virtually zero chance to win a championship, but Packers fans always feel they have a decent shot. We know before the first pitch of the season how teams are probably going to finish. There is relatively little drama in the regular season, and if there is, the chances your team will be a part of the conversation are slim…except in July when your team is trading its best player to a big-market contender for cash and a player to be named later. Compare this to the NFL where guessing 8-8 for every team's record is imperceptibly less accurate than the best predictive analysis.

Defenders of MLB's status quo will quickly point to the fact that raw attendance numbers have been climbing to record levels. But as a proportion of the population, attendance has been flat since the the late 1990s when MLB added a 29th and 30th team. So I think MLB needs less predictability, and unfortunately sabermetrics makes things more predictable.

Don't get me wrong. Of course I'm a numbers guy and I understand the power and limitations of statistics. I'm a fan of analytics in general. Sabermetrics isn't a bad thing. It's that widely-adopted sabermetrics combined with a lack of a salary cap can't help but create a greater concentration of talent on large-market teams.

The NFL's correlation between salary and team wins is much lower, near 0.15 the last time I measured it. The difference is due to a lot of things, but primarily due to the short season and relatively narrow range of team salaries. The NFL has both a cap and a floor for team payroll, which keeps things less predictable. And I think a hard salary cap and floor are what MLB needs if it wants to regain the national attention it once had.

  • Spread The Love
  • Digg This Post
  • Tweet This Post
  • Stumble This Post
  • Submit This Post To Delicious
  • Submit This Post To Reddit
  • Submit This Post To Mixx

35 Responses to “Is Sabermetrics Helping to Ruin Baseball?”

  1. Harper says:

    Quesiton : Do you really believe unpredictability is a substantial driving force in the national NFL popularity? I would imagine it matters but only marginally. Every little bit helps I guess, but I don't think cap/floor -> unpredictability will raise national attention.

  2. James says:

    I think unpredictability matters a ton. Everyone loves underdogs, and underdogs only have a chance if things are unpredictable. The first Patriots-Giants Super Bowl was HUGE in no small part because of how unlikely it was for the Giants to make it there, much less win it all. Same for the rematch a few years later, the Cardinals SB run, Flacco's absurd run, etc.

    After all, unpredictability is the primary factor in March Madness's appeal! Without unpredictability making brackets would be boring, and without brackets not nearly as many people watch the tourney.

  3. Xeifrank says:

    The Red Sox were 22 to 1 to win the World Series. And before you ask, smart people who know sabermetrics bet too.

  4. Anonymous says:

    1) the entire MLB system of the draft / service time is a huge component of team success, surprised you ignored that
    2) there is a luxury tax that changes teams behaviors
    3) the Red Sox had the 14th best odds to win the WS in the preseason

  5. Anonymous says:

    I agree with a lot of the points being made here, but preseason sabermetric projections had the top 6 teams each with between a 4-6% chance of winning the world series....there is still a ton of randomness and unknown involved.

  6. Dan Szymborski says:

    You're vastly overrating how much can actually be predicted and how much we can improve the correlation (and projecting players is a good chunk of what I do for a living).

    If you had absolutely perfect knowledge that every team was a .500 team and would play like a .500 team, you would *still* expect around 25% of the league to have win totals outside of the 74-win-to-88-win range. And we do not have anything approaching that absolutely perfect knowledge.

  7. Brian Burke says:

    We can debate the magnitude of the effect of widespread analytics on the correlation between salary and wins. But I think everyone here would agree that its very purpose (at the player evaluation level) is to increase that r. Otherwise, what's the point?

    Randomness is, of course, a very important factor in almost all sports. Team payroll is another very important factor.

  8. DSMok1 says:

    Baseball is still WAY more unpredictable than basketball, and the NBA's popularity is increasing while baseball's is decreasing.

    I'm sorry, but I think predictability is peripheral consideration at best in how popular sports are.

  9. Brian Burke says:

    Also, I don't think the Vegas odds in one season prove or disprove the my point. I mentioned that only in the context of the Red Sox's worst-to-first improvement and that it wasn't a surprise at all to the sabermetrically-savvy.

  10. Anonymous says:

    Vegas set the Red Sox at 83.5 wins preseason.

    If the Red Sox making the playoffs (let alone winning the World Series) was such a sure thing I guarantee you that this number would've been much higher.

  11. Doug says:

    This column reminds me a little bit of something that someone (it may have even been Dan Szymborski!) said (in jest) on rec.sport.baseball in 1999: "baseball sucks because large market teams like the Braves win every year, whereas football is great because small market teams like the Falcons have a chance to go to the Super Bowl."

    Point is, the goal posts always seem to move in these kinds of columns.

    Baseball is too predictable because the same teams win every year.

    Wait, didn't the current champion finish last a year ago?

    Yeah, but they weren't really bad. Plus they're big-market. Big-market always wins.

    Wait, isn't St. Louis a small market?

    Yeah, but St. Louis is "the team with the 2nd-most all-time championships."

    To recap: a team that finished last in its division a year ago meets a small-market team in the World Series, and this is evidence that the same big-market teams win every year.

    I guess that's what happens when you spend every ounce of your energy for two decades badmouthing your own product in hopes of getting a salary cap. Nice work, baseball owners.

  12. Mike says:

    Three of the bottom five teams in payroll (A's, Pirates, and Rays) all made the playoffs. Considering that the Astros and Marlins weren't really trying this year, you can effectively say that the bottom three teams all made playoffs. That the Red Sox and Cardinals made the World Series instead of those teams was likely more randomness/luck than any other single thing.

  13. Brian Burke says:

    Yes it's true that was a wildcard in . That's fine and I like that. But it doesn't change the fact that r=0.4. If we boil down the all the anecdotal cases and pick every last cherry, we get r=0.4.

  14. Marver says:

    Only one team, as far back as my team salary figures go, with a payroll in the bottom 50% has ever won the World Series: the '04 Marlins. Since 1990, over 50% of World Series winners were in the top third of spending.

  15. Marver says:

    '03 Marlins, but you get the point. If you seemingly can't win the World Series without a top payroll, there's something wrong with the competitive balance of the sport. It wasn't always this way.

  16. Dan Szymborski says:

    Now I'm curious as to which Doug that is from rsbb. Obviously not Pappas. Riblet?

    Does sound like something I wrote, seems sarcastic enough. But I sometimes have trouble remembering something I wrote 14 days ago, let alone 14 years ago!

  17. David Kravitz says:

    One thing you didn't mention is how random the postseason is, especially now with the wild card. The better team may win about 60% of the time. All your points are correct about who emerges in 162 games. The long playoffs just ensure that the better team doesn't always win.

  18. Anonymous says:

    The r=0.4 is sound, however as a Seattle Mariners fan it's painfully evident that not all big market teams are using analytics at all, in fact baseball seems (at least to me) to be one of the sports most resistant to change. My guess is there is a greater correlation of success among teams that are adopting analytics (Rays, A's) vs. those that are not.

    When I saw the title I was expecting this article to take an entirely different direction. It seems to me that analytics points baseball in the direction of being a less interesting game, in terms of in-game tactics. That is, it emphasizes home runs, walks and strikeouts, and de-emphasizes bunts, baserunning and most situational strategies. Compare this with football where analytics points the game in the direction of more interesting in-game decisions e.g. going for it on 4th down more often. This is something I hope gets addressed eventually, although I'm sure it won't be soon.

  19. Joe says:

    The correlation between payroll and wins in the Premier League in England is about 0.7-0.8. And that is the most popular sport in the world. I agree with absolutely everything you say, except for the notion that popularity is linked to how random the results are.

    Joe

  20. LamKram says:

    David Kravitz hit the nail on the head: the key difference between baseball and football which explains their respective correlations between payroll and win % is the 162 game season vs the 16 game season. As any amateur statistician knows, 16 games isn't nearly a large enough sample to truly distinguish the good teams from the bad teams. So baseball is much better at truly rewarding the better (and better paid) teams with playoff berths. Which is good thing, isn't it?

    That said, I believe there are still inefficiencies in the Xs and Os strategies in baseball (if I may mix my sports metaphors) which can allow lower salaried teams to overcome teams with superior talents. For example, the (admittedly well paid) Boston Red Sox are masters of driving up starting pitcher's pitch counts to get to the opposing bull pen quicker. I'm not sure average number of pitches seen per plate appearance is a stat that all teams are looking at seriously to evaluate talent, but they probably should be. Also, no MLB team, in my opinion, uses their closer optimally - a lot of room for improvement there.

    Maybe this supports Brian's point: lots of foul balls and pitching changes don't make baseball more exciting. On the other hand, I find the evolution of baseball strategy and sabermetrics fascinating. And as a statistics and probability geek, I don't want randomness - I WANT to see the best team win the championship. Call me crazy...

  21. Anonymous says:

    Well, it was a pretty good World Series, despite MLB's flaws... The most efficient linup (Boston, sooooooo many PPA) against one of the most efficient rotations (St Louis, they get deep into the game on very few pitches).
    Boston's linup:
    http://www.baseball-reference.com/teams/BOS/2013-batting.shtml
    Very well-rounded with the RAA spread throughout the line-up.

    http://www.baseball-reference.com/teams/STL/2013-pitching.shtml
    St. Louis pitching staff. Their rotation gets into the 7th inning or later on very few pitches (~100). Even if their bullpen was terrible the Cards would be set because their starters are so efficient they could go even DEEPER into games if it were necissary. That was the principal weakness of the Tigers pitching; their power/K pitchers naturally threw a lot of pitches even when they were 'hot' and Boston's linup exacerbated this allowing the Sox to get to the Tigers weak bullpen all the more quickly. Game 1 of the ALCS was a good example where Anibal Sanchez only went 5.2 innings on ~120 pitches. Sure, he struck a lot of guys out, but the bullpen was getting overworked from the very beginning.

    Cheers,
    J

  22. Anonymous says:

    Joe--that's got to be Posnanski, right? Anyway, he makes an excellent point: drama isn't necessarily killed if the rough parameters of expectation are well understood.

  23. Anonymous says:

    As a lifelong KC Royals fan I think that even with a salary cap the Royals would still find a way to be mediocre. The last time they were in the playoffs if was 4 years old... KC seems to have an institutionalized hatred of OBP and PPA as well as an innate ability to destroy, or misidentify, promising young starting pitchers (Lord how painful it was to see the Greinke show...and let us not speak of Luke Hochevar the first overall pick in the draft where Max Scherzer and Clayton Kershaw were chosen AFTER Hochevar...). At least I can console myself with the fact that old KC talent like Carlos Beltran is one of the best post-season hitters of all time...

    Cheers,
    J

  24. Wizard says:

    Brian, your letter is right on. I am as much of a numbers guy as you are (without the pub :) ), and I have often thought there is something about sabermetrics that is damaging to the sport. In fact, I am an avid backgammon player, and for those of you play backgammon, you are probably aware of the software that will give you the best move. Although I use it to keep up with everybody else, I wish it never existed. Backgammon was much more 'romantic' when players debated a move, and there was a mystery about it. The mystery is now gone.

    There is somewhat of a resolution in baseball. I would think that if you had a much smaller standard deviation of player salaries, then the correlation would go down, because intangibles (injuries, managing, etc.) would be more prevalent.

  25. Anonymous says:

    It seems like what you're saying is that the lack of a salary cap is ruing baseball, and yet you've framed it as an argument that advanced analytics is ruining baseball instead.

    As a few people have already pointed out, you grossly overstate how good the proverbial crystal ball can get. The correlation will never ever ever be 1, and there's nothing wrong with it getting closer to 1 if there's a salary cap. It's not a problem with knowing too much. It's a problem with a poor salary system. Without a salary cap, teams with more money can get better players. Understanding better what makes a good player only further exposes the existing problem; it doesn't create a new problem.

    Plus, as at least one other person pointed out, you're using anecdotal evidence unrelated to your argument. "The team with the 4th most all-time championships beat the team with the 2nd most all-time championships." If your argument is that increasing the correlation between salary and wins in a sport with no salary cap is bad, not only does it not matter how many championships the teams in the championship series had, it's counter-productive to point out that a small-market team has the second most championships.

  26. Jeff Clarke says:

    Lamkram is absolutely right. The central difference is the length of the season. In each individual game, the baseball underdog has a much greater chance of winning than the football underdog. The season guarantees that the better teams will have a greater playoff chance.

    I think the longer season also plays into the lack of a national audience. Watching only your own team's games doesn't satisfy most fans. There simply aren't enough games. When your favorite team plays 162 times, there isn't as much reason to watch two random teams play.

    Obviously this has a huge impact on the economics. National TV money is basically everything in football. In baseball, local TV and the gate are much more important

  27. Guy says:

    I think the central insight here is important: the development of sports analytics will inevitably -- contra "Moneyball" -- make the strong stronger rather than boosting underdogs. Over time, the wealthy teams will match the analytic investment of any upstarts, and eventually will probably outspend them on this front as well. And this must increase the correlation between payroll and wins (although Brian may overstate the magnitude of the effect).

    Unfortunately, I think this insight will be lost on many readers because it's set within a larger argument about why baseball is (allegedly) in trouble as a sport that fails to make a lot of important distinctions, and makes some dubious assumptions. A few that come to mind:

    *There is little evidence that fans value uncertainty as much as Brian assumes. As others have noted, the NBA is the obvious counterpoint here: Brian's argument applies 100-fold to the NBA -- where we all know the top 4-5 teams with 95% certainty on opening day! -- yet the sport appears to be thriving. (Also, uncertainty at the game level -- much higher in MLB than in the NFL or the NBA -- may matter as much or more for fan interest.)

    The spread of talent and spread of outcomes are not the same thing. MLB actually has much MORE talent parity than the NFL, despite the lack of a payroll cap. To the extent MLB standings may be a somewhat better reflection of talent than NFL standings, is that really a bad thing? Outcome uncertainty has some appeal, but we sports fans also want talent to be rewarded.

    *Payrolls are not simply a function of size/wealth of metro areas. Both St.Louis and Boston have developed a strong fan base that supports larger payrolls than underlying demographics alone can explain.

    Personally, I'd be in favor of stronger measures to force payroll parity in MLB. I do believe small markets are at a structural disadvantage. But it's hard to know if such changes would improve ratings for the World Series (assuming that's an important goal). It's at least as plausible that the casual fans who would boost WS ratings are the kind of people drawn to battles between dynasties, as the NBA often offers.

  28. J.D. Krull says:

    I think statistical analysis is ruining sports a bit for me in another way. It tears away a lot of the mystique of the game when you realize that outcomes are heavily influenced by randomness. The typical naive fan believes that the game is *always* won on the field by the team that played the best that day, and they are deaf to any caveats about outliers and probability distributions, etc. Unfortunately, most fans like that have a greater capacity to enjoy watching a game than I have, because whatever the outcome of a game is, I'm never satisfied that it "settled" anything.

    I used to be the biggest baseball fan I knew...I could recite every team's starting lineup and pitching rotation from memory. Now I don't even watch the World Series, because I see it as essentially seven coin flips.

  29. Boston Chris....now Atlanta Chris says:

    J. D. Krull....I and xkcd agree w/ you.

    http://xkcd.com/904/

    It takes attempting ignore some cognitive dissonance to enjoy watching as much as I might have in the past!

  30. Stat Guy says:

    The truth is, is that the MLB isn't predictable. Moneyball didn't use real sabermetrics. They mainly used on base percentage. That's been around for a long long time. We mainly just try and get a well balanced team. A couple guys for speed, a couple more for defense, and couple more for power.
    This year, we all thought the Blue Jays would be World Series champs, but after the first month, we knew that they couldn't turn it around.
    And with teams like the Yankees and the Dodgers who have unlimited salary cap, or so it would seem, they overpay for most of their players. Chad Billingsly was extremely smart to take his 15 million dollar/year salary because now no one even talks about him. Players generally will have one good year and have a regression the next. Chris Davis probably won't crack 35 HR's next season now that pitchers know how to pitch him.
    About the lack of churning of top teams, a couple years ago, the Yankees were sitting pretty when they beat the Phillies, but now, they hope to get attention by talking about Mariano Rivera. The NFL is more predictable because there less games in the NFL. The Red Sox may have been predicted to do well, but there are way more games for a person like Dustin Pedroia to get hurt, and if he did, they wouldn't have a shot at the playoffs.
    Don't get me wrong, I love both sports equally, but I think the NFL is more predictable.

  31. Wizard says:

    to the last two comments, I agree. I was trying to say as much in an earlier post. and while J.D. expressed his sentiments quite well, I personally think the 7 coin flips is a bit of an overstatement (I , of course, cannot speak for him). I think what statistics have done has *brought to the forefront* the randomness of it all. But, we knew about it back then. We just weren't cognizant of it. Two friends might play golf together, and know they are evenly matched, and each win 50% of the time, but they do not look at the result as a coin flip (i.e, not really worth golfing).

  32. Anonymous says:

    I have to echo everyone else's disagreements on this. Sabermetrics isn't ruining baseball by making it too predictable (DSMok1 hit the nail on the head w/ the comparison to the NBA, which is WAY more predictable than baseball but growing in national popularity every year).

    I can see sabermetrics ruining baseball simply by ushering in certain stylistic changes that make the game harder to watch as a fan... For instance, DIPS theory emphasizing pitchers who get a ton of strikeouts have reduced the variation between pitchers -- where in the past you'd see a mix of hard-throwers with crafty soft-tossers, they're almost all power arms now. And emphasizing OBP + the realization that strikeouts are no worse than any other out from a batter's perspective has led to a bunch of whiff-happy hitters who walk a ton. (This is why the ball is getting put in play far less than ever before.) And I didn't even mention the reduced emphasis on base-stealing.

    Unfortunately, baseball is at its most exciting/watchable when the ball gets put in play a lot and there are exciting baserunning plays, not when dudes force pitch counts up and games last 3 1/2 hours. Nobody wants to watch a guy either strike out or walk to 1st base in every AB. So if anything, THAT is how sabermetrics is ruining baseball relative to football -- not because MLB is more predictable than the NFL.

  33. Dennis Warner says:

    Many things have ruined baseball over the years; (1) expansion from 16 to 18 teams, (2) the lowering of the pitching mound in 69, (3) the elimination of the reserve clause, (4) local TV blqackouts, (5) The introduction of the DH, (6) lax enforcement of the PED ban (7) inter-league play (8) excessive enforcement of the PED ban. (9) continual lengthing of the season well into the fall........yet, somehow baseball survives these calamities. I doubt that a gaggle of exuberant number crunchers will cause any permanent harm.

  34. J.D. Krull says:

    Dennis, baseball "survives" in the sense that the league hasn't folded and is still generating record revenue. But surviving and thriving are two different things. As Brian pointed out, World Series TV ratings are in the tank. The modern game experience is "meh" compared to its past glory days, for various reasons. I think this is indeed a candidate for "permanent harm", even if the pathology hasn't shown up in the financial bottom line...yet.

  35. Unknown says:

    I don't think sabermetrics are ruining anything except the traditional measurements for player value - batting average, wins, ERA, etc. We still use those to give a general idea, but I think the advanced stats will help lower market teams find value in role players. Are there any advanced NFL statistics?

Leave a Reply

Note: Only a member of this blog may post a comment.