Archive for May, 2006

Sabernomics in the WSJ

Thanks to Matt Phillips for his flattering write-up of Sabernomics in his May 15 Blog Watch column in The Wall Street Journal (RR).

University of the South economics professor — and Atlanta Braves partisan — John Charles Bradbury’s Sabernomics blog offers a good glimpse into “sabermetrics,” the mathematical and statistical analysis of baseball.

Mr. Bradbury posts items on sports, business and economics. But the site also includes links to fascinating, if wonky, studies on baseball. For instance, did you know that from the 1970s to the early 1990s, the rate of batters hit by pitches was almost 20% higher on average in the American League than in the National League? Interestingly, the higher plunking rate coincides with the introduction of the designated hitter in the American League — leading some to theorize that pitchers are more likely to throw at batters when the pitchers don’t have to bat themselves and face possible retaliation.

Other studies dive into the numbers to help determine how a beer-guzzling, hot-dog-gobbling slugger like Babe Ruth might have fared in today’s game, or to figure out whether there is a rational basis for the conventional baseball wisdom that bars lefthanders from a spot behind the plate as catcher.

Thanks to John McA and Uncle Andy for pointing this out to me.

Why Don’t More GMs Do This?

Josh Byrnes, the GM of the Arizona Diamondbacks, just signed Chad Tracy to a 4-year $13.25 million deal (guaranteed) with an option for a fifth year at an additional $6 million ($7 million minus the $1 million buyout that is guaranteed). Mark Shapiro, of the Cleveland Indians, did something similar with several of his young players earlier this season, and it’s something his predecessor, John Hart, did as well.

As a general rule, individuals tend to be risk averse while firms are risk neutral. Baseball teams can diversify away risk—by signing many players—while players have their financial nest-egg in one basket. For a reserved player who has the potential to be a star, he’s likely to forgo some of his expected future earnings for a guarantee of stability. This represents a huge opportunity for teams to free up financial resources. As David Pinto puts it:

Tracy is a solid offensive player who is now locked up through his peak years. Not too long that he can cost the team a lot of money with a career ending injury, and they don’t need to worry about arbitration.

The league minimum may seem like a lot of money, but it’s not over the course of a lifetime. I expect, as must the D-backs, that Tracy is going to be worth more than $3.5 million/year for the next four years—I estimate that Tracy was worth $8.15 million in 2005—and Tracy is happy to trade some of that potential income for stability.

I also wonder how players behave after signing deals like this. You might think that a player who gets some guaranteed money over the next few years would shirk, but I think the exact opposite might happen. There is still the potential for a bigger payday down the road, and I’m sure Tracy wants to win. Maybe he’ll be more likely to do some little things that might injure his chances of getting a long run payoff—like dive for balls or get into collisions, or do strenuous weight-training. He might also be willing to share knowledge with younger teammates who could compete for his job. And let’s not forget generating good will with players.

March 7, 2007

Frequently, people ask me when my book will be coming out, but I haven’t had more than a vague time-frame. However, I talked to my editor yesterday, and he tells me that the official publication date is March 7, 2007. We’re close to a title, but I don’t want to reveal it yet, as these things can change. I’ll keep you posted as I learn more.

Age Cut-Offs and Month of Birth in Baseball

Steven Levitt writes this interesting post on the impact of age cut-offs and the typical birth-month of players in the NHL. This follows up an article with Stephen Dubner in the NY Times Magazine on the phenomenon in soccer. I’m swamped with grading so I won’t write much about the theory. I’ll just provide you with a quote from the article and two bits of information.

Since youth sports are organized by age bracket, teams inevitably have a cutoff birth date. In the European youth soccer leagues, the cutoff date is Dec. 31. So when a coach is assessing two players in the same age bracket, one who happened to have been born in January and the other in December, the player born in January is likely to be bigger, stronger, more mature. Guess which player the coach is more likely to pick? He may be mistaking maturity for ability, but he is making his selection nonetheless. And once chosen, those January-born players are the ones who, year after year, receive the training, the deliberate practice and the feedback — to say nothing of the accompanying self-esteem — that will turn them into elites.

1) Here is a graph of of the birth-month of every major league baseball player through the 2005 season (16,067 players), thanks to the Lahman Baseball Archive.

Birth-month of MLB players
(note: the horizontal line at 8.33% is the mean)

2) The cut-off for participation in Little League TM baseball is July 31.

This seems to fit with the theory. Of course, it could be that baseball fans—who are most likely to raise baseball players—have a lower opportunity cost for procreation after the World Series.

Addendum: I re-read the Little League page, and there is a proposal to move the cut-off to April 30. If it passes, in a decade or so, someone remember to check on this. 🙂

Further Addendum: Here is a revised graph of the percent born by birth-month for US-born MLB players. There are four samples for players who made there debuts after 1970, 1980, 1990, 2000. If anyone wishes to play around with the data, it’s freely available at

Birth-month of  US-born MLB players

Then pattern still generally holds; although, the rise in March births is interesting.

Hating Adam

I’m curious why Adam LaRoche is so hated by Braves fans. And look, I have a realistic view of Adam.

— He is NOT a good defensive player. Those touting him for future gold gloves are either deluded or flat lying. Yeah, he’s not Giambi, but, he’s not good by any defensive metric out there. And my own perceptions of his play confirm what the stats show.

— He is slow, very slow. Because he is a lefty, he will never be able to play any position other than first.

— He doesn’t have the team-leader “intangibles.” By all accounts he is quiet, and his nickname “three-second delay” indicates he’s not the guy teammates rally around.

— He is NOT the type of first baseman you build around. The Braves should have moved Chipper to first and held on to Andy Marte.

Given all of this, Adam LaRoche is a quality MLB ballplayer who will have a decent career. And it’s possible that he will gain some more power that might make him even more valuable. I suspect he will play for a lot of teams in a first base platoon and pinch hit. The Braves have very few options at first now that Marte is gone, so I don’t see what the deal is. In fact, I’m quite happy the Braves still have him around.

A player like LaRoche is hard to rally around. His swing is ugly, and he’s generally an unexciting player. His .220 batting average this year and 33 strikeouts are easy targets for criticism. But his 18 walks and .250 Isolated Power are beautiful. He has an OBP of .336 and a SLG of .471 for a healthy OPS of .807.

Compare this with Jeff Francoeur, whom everyone is cheering for and says, “but if you take out his slow start he’s hitting nearly hitting a lot better.” Blah, blah, blah. Really? If you take out the bad stuff, of course you’re going to look pretty good. Jeff has been stinking up the joint in a way that was totally expected from his minor league stats, and he’s not as good a hitter as LaRoche. Adam has produced 81 outs for the Braves this year compared to 105 by Francoeur—to lead the team. That’s 24 more more outs, nearly an entire game’s worth. My point isn’t to disparage Francoeur, but to point out that Adam is facing a different standard. He’s been a lot more valuable than Francoeur this year, but getting much harsher criticism.

Checking on My Predictions

On March 20, I made several predictions relevant to the Braves that I want to check up on.

Prediction: Oscar Villarreal will pitch less than 30 innings.
So far: 14 and 1/3 IP. Although his line of 7 Ks/ 8 BBs/ 2 HBPs/ 1 HR isn’t instilling me with any confidence.

Prediction: Horacio Ramirez is going to have his worst season yet.
So far: Injured after 3 IPs, although still no home runs allowed.

Prediction: Lance Cormier will be the most-reliable middle-reliever in the pen.
So far: Well, before his injury he was Cox’s favorite middle-reliever and his ERA is good (3.00). However, his peripherals are pretty smelly (7 Ks /9BBs/ 1 HR in 15 IP) and not in the spirit of my prediction.

Prediction: Chris Reitsma will be a solid set-up man or closer.
So Far: He’s been awful. 5Ks/ 4 BBs /3HRs in 12 and 2/3 IP. His K-rate continues to fall, and he’s already given up as many home runs as he allowed last year. This isn’t our hit-unlucky Reitsma. If anyone might be missing Leo, it’s him. Although, this is pure conjecture on my part.

Prediction: Andruw Jones will make 2005 look mediocre.
So Far: He’s hitting about the same as he did last year.

Prediction: Chipper Jones will continue his excellent play. If he stays healthy, he is an MVP candidate.
So Far: He’s already been on the DL. He’s hitting for average but is Iso-Power is down about 100 points. He’s only hit 2 homers so far.

Prediction: Marcus Giles will have an OPS of less than .800.
So Far: Well, unfortunately I was really right on this one. To be fair, Giles has been injured. But his power has been non-existent. At least he’s walking.

Prediction: Adam LaRoche will have an OPS greater than .800 (does not apply if he isn’t platooned).
So Far: Surprisingly, I’m right on this one and he hasn’t been platooned. Currently, LaRoche’s OPS is at .809. Despite a .220 AVG, he’s getting on base and hitting for power. Jeff Francoeur could learn a lot from Adam about the difference between productive and empty batting averages. Speaking of Jeff…

Prediction: Kelly Johnson will out-hit Jeff Francoeur (assuming KJ stays healthy and isn’t sent to AAA).
So Far: KJ looks to be heading for Tommy John.

Prediction: Andy Marte will out-hit Jeff Francoeur (the Indians will find room for him).
So Far: The Indians haven’t found room for him, and Marte is strangely homerless in Buffalo with a .350 OBP.

Prediction: Brian McCann will out-hit Jeff Francoeur.
So Far: Brian McCann has been fantastic. Again, here’s someone Jeff should seek to learn from. Unlike Francoeur, McCann learned how to take a walk in the minors and it has paid off with a .337/.379/.522 /.901 line. He’ll probably drop off some, but he’s in no danger of seeing the minor leagues for any reason other than a rehab assignment. He’s really been overshadowed by Francoeur and Marte, but he’s the real shining star of the Braves farm system.

Prediction: Johnny Estrada will have an OPS greater than .750.
So Far: Johnny is off to an excellent start (.298/.326/.488/.814). He has to hit for a high average to be an excellent-hitting catcher, because he doesn’t walk much. But, he’s shown he can do that. I expect he’ll regress some, but I predict he’ll end the season over .750, which is good for a catcher.

Prediction: The O’s ERA will drop and the Braves ERA will rise. Their ERAs will be within 0.25 runs.
So Far: Wrong, right, and wrong. The O’s pitching hasn’t looked good with the second worst ERA in the AL (5.71). However, the Braves aren’t doing so hot either with a 4.29 ERA. And the Braves have actually been a little bit lucky, with a FIP of 4.77. Atlanta had a 3.98 ERA in 2005 with a very similar pitching staff. The O’s had an ERA of 4.57 with some significant turnover from last year.

Open For Business

As regular readers have probably noticed, Sabernomics is no longer advertisement-free. I’ve decided to start hosting ads to help cover the cost of moving to a service provider that doesn’t crash every 15 minutes. If you’re interested in supporting Sabernomics, please visit the Blogads link.


I ran into two interesting posts this morning.

Jeff Merron lists the top-20 greatest TV moments in sports. I have to say that I get chills when I think about some of those events, at least the ones that I was old enough to have actually witnessed, and it just reminds me of why we like sports. The uncertainty and drama are key elements in all of these events. And you can help Jeff rank the events after reading the article. Although, I can think of one event that was left off, which would be my number one: Sid Bream’s slide to win the 1992 NL Championship for the Braves.

Also, Mac Thomason has a series of reviews of John Schuerholz’s Built to Win. I think it’s a very interesting take, and it goes to show how influential Moneyball was for baseball. It seems JS can’t shake the bitter taste or admit that he does some things the same way Billy Beane does.

Pictures and Steroids Again

I would post this in the comments to the previous post, but I don’t think I can put images in the comments. There are lots of good comments there, too. A few posters implied—or maybe I’m misinterpreting—that if the best hitters dominated the worst pitchers, shouldn’t the best pitchers dominate the worst hitters too? Well, in fact, pitchers also improved in striking out batters just as they were giving up more home runs and hitting more batters. But, the effect of dispersion has been greater on pitchers than hitters, which should lead to more great feats by hitters.

Strikeout Rate

Also, Skip notes that “HR rates vs. pitchers should diverge in the 90s: the best pitchers dish out fewer, and the worst dish out more.” So, I looked at the divergence of home runs allowed across pitchers, using the coefficient of variation as the measure of dispersion.

CoV of HRA

Now some of this has to do with the increased specialization of pitchers, which might also be a function of a depleted talent pool, so I also looked at the dispersion of home run rates. The outcome is not as pronounced, and the change in dispersion since the 1993 expansion has been quadratic—decreasing then increasing. And. in a historical context, the dispersion in home runs allowed rates is low.

CoV of HRA rate

Pictures and Steroids

When people talk about steroids, they tend to talk about pictures like those below. Maybe they don’t use an actual picture, but the point is clear: the rate at which an extreme achievement—like home runs—has changed recently. And that change coincides with a possible increase in steroid use by players.

Rate per year

Because the shift in the 1990s was so dramatic, it’s hard not to wonder what happened. There is no denying that home run balls are flying out of the park more than they used to, which makes it hard to know why steroid skeptics, such as myself, do not acknowledge steroids to be the cause. Yes, Juiced and Game of Shadows tell us that steroids were rampant in the locker rooms of MLB teams during the home run boom. I really can’t argue with the fact that some players may have taken steroids. As to how much and what effect they have had on hitting power I still don’t know. It’s going to take more than accusations from Ken Caminiti, Jose Conseco, and anonymous sources to convince me of player guilt. Regardless of whether or not players used or not, I’m still not convinced that steroids, or any other PEDs, have much to do with the increase in offense. Here are some reasons not to blame the surge on steroids.

Steroids ought to impact both sides of the ball. If there are benefits, both pitchers and hitters will reap them. Recent testing has busted both pitchers and hitters. It is possible that batters benefit more than pitchers, though. So, this reason is not all that powerful, although I would like to see better arguments for the asymmetric effects of steroids on hitters and pitchers.

Also, the introduction of homer-friendly ballparks may have facilitated the home run surge, but it’s not enough. The balls are still flying out of the big places too.

Expansion is the explanation that gets little attention, yet I believe it is the most important factor for the power surge. The rise in home runs coincides with the expansion of the leagues in 1993 and 1998, which has increased the league by more than 100 players a year—a minimum of four 25-man rosters.

The late evolutionary biologist Stephen Jay Gould put forth a novel theory of extreme achievements competitive environments. His idea was that as the variance of the quality of participants shrinks, opportunities for great performances diminish. Being an avid baseball fan, Gould used baseball as his example. He argued that because MLB was being populated by better baseball players, great achievements, such as hitting .400, were decreasing. You see, very good players dominate very bad players, and as the very bad disappear from the game, the very good can no longer rack up good performances against them. Expansion has the effect of letting in the riff-raff for baseball’s elite to exploit. And we can see that as evidence of the inferior pitching entered the leagues that pitchers began to hit more batters. As the graph above shows—did I forget to mention it was a digram of the hit batter rate, not home run rate, over time?—the current era of baseball could just as easily be referred to as the hit batter era rather than the home run era. And most certainly, no one believes steroids cause pitchers to hit more batters—well, maybe it’s roid rage, but I don’t buy that.

Below is a graph of the hit batter and home run rates since 1960. They track together somewhat, but more importantly, they both shoot up in the 1990s, which is what Gould’s theory predicts.

HBP (HR) Rate per year

Furthermore, we can examine Gould’s hypothesis by looking at the talent distribution of hitters and pitchers over time. The graph below maps the coefficient of variation (standard deviation/mean) of batter OPS and pitcher ERA by decade relative to the 1920-2005 average.

CoV of Baseball Talent

The higher the bar, the greater the distribution of talent and the greater the opportunity for great players to dominate poor players. It’s clear that the distribution of pitchers has been rising over the span that both home runs and hit batters have increased. The direction of this effect is exactly what we would predict, and the change is as dramatic as the change in home runs and hit batters.

Most certainly, this does not mean that steroids have not influenced the game. I just think it’s incorrect that there are no other plausible explanations for the changes in performance. Personally, I don’t think steroids have had that much of an effect, even if players have been taking them, but that is just my opinion. Furthermore, I believe Art De Vany is right.

Addendum: David Pinto noticed the effect back in 1994.

Further Addendum: My follow-up post.