Are General Managers Myopic?

I continue my discussion of hot stove myths by focusing on the competence of general managers. Today, I explain why selling high and buying low are not reliable strategies.

GMs can buy low and sell high — So, let me get this straight: you think you know when a player is playing above or below his true ability—usually due to a small sample or by using a SGT-approved metric instead of a mainstream statistic—but guys who make a living in baseball completely miss it. For this to work, the GM on the other team has to be a colossal moron. GMs have made mistakes in the past and will make mistakes again, but they’re not dumb enough to act on a meaningless hot/cold streak. You can’t sell high or buy low and profit financially because all GMs understand these things. You don’t have to wait for a guy to get hot to sell him, nor dump him before he gets cold. In addition, the key knowledge of when the peak or trough is doesn’t exist, except in the mind of message board posters. Fluctuations in performance create uncertainty, which affect the price that GMs are willing to pay.

Several responses include historical examples of bad decisions by general managers as evidence that this is not a myth. My contention is not that GMs are infallible, just that we should not expect them predictably to make mistakes that are obvious to everyone but one sucker GM. Suckers are born every minute, and some GMs are better than others; but, sucker GMs are rare. Several years ago, I asked Oakland A’s Director of Baseball Operations Farhan Zaidi his thoughts on the issue.

Are there any GMs on other teams that are just plain suckers?

Absolutely not. Working in baseball has given me a newfound respect for GM’s in baseball. It takes a lot to rise through the ranks of the industry to one of those 30 positions. Fans and media like to deride some GM’s as being clueless, but from what I’ve seen, being a clueless GM is an oxymoron of the highest order.

Zaidi is no lifetime insider. I conducted this interview a few months after the A’s hired him out of Cal’s economics doctoral program, and his advisor was a world-renowned behavioral economist (Matthew Rabin). In the interview, he did acknowledge that GMs were subject to some types of biases that affect all humans, but that’s not the type of mistake we’re talking about here. Remember, message board posters are getting this right, while GMs make mistakes.

The type of statement I’m rebutting can be seen right now among Braves fans discussing Kelly Johnson. Before last season, Johnson had put up respectable numbers and looked to be growing into an above-average second basemen. His 2009 was awful, and Martin Prado rose to the occasion and took his starting job. Now the Braves are in search of an outfield bat, and trading Johnson as part of a package to get one is being discussed. This is when I see a statement like the following, “We can’t trade Johnson now, that would be selling low. He’s a better player than last year, the Braves should hold onto him and move him when his numbers rebound.” Or, “We should move Prado instead. There’s no way he’s an .800 OPS players. Dump him now while his value is high and get more in return than we could for Kelly.”

The assumption that underlies these statements is that the commenter can see the true Johnson and Prado, but a GM and his many close advisors can’t. I don’t care how much you liked Moneyball (I liked it, too), but no GM is this stupid. When scouting departments project players, they understand that performance is volitile. They see Johnson’s batting average was well below his historical norm, and understand that on-base percentage and power are more stable than average—in fact, I’m probably being too simplistic here. This will allow the Braves to get more for him than the typical hitter coming off a sub-.700 OPS season. There is no need to play him to get his value up.

Still, his value will be affected by his poor performance, because it raised some uncertainty about what Johnson is. The Braves might know more than other GMs because of their access to inside information about work habits and injuries; however, that he might not be that good is an uncertainty that the Braves face as well. What if they gamble on him getting better and it turns out that he’s just lost it (see Marcus Giles)?

Humans make mistakes, and general managers are human; therefore, general managers make mistakes. But, I don’t think such obvious mistakes can be so easily exploited. Teams are constantly looking for errors and do take advantage of them when they find them; however, selling/buying high/low isn’t one of those areas where inefficiencies persist unless GMs are colossal morons. I don’t think they are, and I think the proper incentives are in place from allowing incompetent individuals to rise to the GM position and stay there.

This doesn’t mean that GMs are above criticism, either. GMs can and should be criticized when we see bad moves, but I think they should always be given then benefit of the doubt.

12 Responses “Are General Managers Myopic?”

  1. Brad says:

    It’s nice to get a little perspective on this sort of thing. Today on the Hardball Times site, Pizza Cutter (one of the writers over there) observed how OBP and free agent salaries correlated (especially after Moneyball in 2002) more than the average fan would suspect — and more than batting average. Of course, considering I don’t hang out with GMs (instead I hang out on baseball message boards and blogs), I’m more likely to suspect a GM of idiocy because from our perspective (the WAY outside), it’s easier to do so — because of Occam’s Razor and because the GM can’t possibly hurt me from this distance.

    So, I appreciate a little perspective on this — if for nothing else, it at least helps me calm down and forgive possibly dumb mistakes.

  2. dan says:

    I am thinking that baseball is the main stage where individuals get to see the “Winner’s Curse” principle work itself out.

    In this case, I don’t think every manager is stupid, just that whenever a manager signs a sought after free agent, they are very likely to have overpaid for their services, which leads fans savvy in sabermetrics yet poor in game theory to believe their own judgment to be superior.

  3. Ken Houghton says:

    This is the sports equivalent of EMH: since we can’t know exactly when GMs will act stupidly, they don’t.

    Under that condition, there would be no GMs with good reputations (Brian Cashman, Theo Epstein, Pat Gillick) or bad ones (Moore, Minaya, Bowden).

    It may well be true that Warren Buffett’s old doctrine about “good” managers and “bad” companies applies here as well, but the evidence of trend-following noted above makes it clear that the market can be arbed. (Whether it is worth the trouble to do so is a separate, interesting question.)

  4. toad says:

    I think the Winner’s Curse might come into play in the case of “commodity” FA’s – outfielders within a certain range of skill, say.

    But often a particular FA is worth more to some teams than others. How much would Sabathia have been worth to Kansas City? That doesn’t mean the GM can’t overpay, but I’m not sure they consistently overpay ex ante.

    In fact, my own unsubstantiated theory is that mid-range FA’s, who will often be commodity types, are the most likely to be overpaid, and that the ones at the low and high ends are paid more accurately.

  5. steven says:

    I agree with this analysis,

    but it still doesn’t explain how players like Adrian Beltre and Gary Matthews can get monster contracts when they’ve done so little to earn them. Also, this may be deviating from the path here, but how come GM’s routinely sign pitchers such as Carlos Silva to contracts that all fans can see are not worth it? They could get similar performance from career spot-starters or long relievers for pennies on the dollar.

  6. Theo Epstein recently made a comment to the point that he and new Padres GM Jed Hoyer, after working together in Boston, value players similarly. Epstein went on to say that this will likely made trades between the two less likely, and that trades more happen most often when GMs have different styles of player evaluation.

    It stands to reason that if different styles exist among GMs, then one style must be more accurate than another.

    One might make the argument that different styles are dictated by different conditions.The most important difference being market size, which might explain differences in style more accurately than classifying them “right and wrong”.

    However, going back to Epstein’s comments, he was referring to Boston and San Diego, two very different markets. Yet according to him, each leader still preferred similar evaluation methods.

    While the different methods among GMs may all be more accurate than some fans believe, I think there is still enough of a difference to classify one as more accurate than another.

  7. JC says:

    Melvin,

    I had not read that opinion, and I think it’s interesting to get his take. But, I’m not so sure he’ll be dealing less with Hoyer. Billy Beane has made many deals with his former associates and saber-minded GMs to the point where it almost seems that he prefers dealing with them. Five years ago, I wrote up a post explaining why the saber-GMs were trading so much with each other. Here’s what Beane had to say.

    “We have similar beliefs and knowing each other allows us to make deals [quickly]. With J.P. and Theo, I can get a deal done in five minutes.”

  8. Here’s the link:

    http://www.weei.com/sports/boston/red-sox/alex-speier/2009/11/08/why-adrian-gonzalez-may-remain-limits-sox?page=full

    And partial quote:

    “When you value players the same way, sometimes it’s hard to make a deal than with an organization that emphasizes different attributes of players. If you have different evaluations, you can get a deal done quicker.”

  9. Ben says:

    Isn’t it possible that GMs can evaluate player talent and performance incredibly well, but be poor negotiators?

    What message board posters are usually arguing, as in the Johnson/Prado example you cite, is that a GM is foolish to deal someone (Johnson) because his leverage is reduced, and he has a weaker hand. Wren knows Johnson’s value, and an opposing GM may value Johnson exactly the same, but Wren is still at a disadvantage because the opposing GM can make a convincing argument that Johnson is actually worse than both GMs really think he is, and there are plenty of non-talent arguments, too (i.e. “Our fans won’t accept a deal like that;” “The media would eat me alive, and you know it,” etc.)

    While not an insider myself, I imagine there are lots of other factors involved in why trades do and do not happen other than player evaluation. For example, why has Brian Roberts not been traded the past couple off-seasons? It’s very likely (almost certain) that the Orioles could have made a trade that benefited them, but they didn’t because Roberts is the “face” of that franchise, and Angelos didn’t want to further alienate the fanbase.

    So while I agree with the point of your article, I think you’re ignoring a big part of the dynamics of negotiations.

    I don’t know much about game theory, but I imagine it is relevant here, too.

  10. Gary says:

    Point blank, Steve Phillips is a sucker GM.

    If he tells me something, I assume the exact opposite is true.

  11. JC says:

    Not “is,” but “was.”

  12. Eric M. Van says:

    I think there’s a bit of a straw man attack here. We are not talking about one moron GM, but rather the ten or dozen who do no in-house analysis. And we are absolutely not talking about merely selling high or low based on the plain vanilla numbers that all 30 GMs know, but rather about getting an edge by going into something like PrOPS. I can guarantee you that there were c. 20 GMS who were more confident about Nick Swisher bouncing back from 2008 and c. 10 who were more skeptical because they hadn’t looked into the evidence that his decline in BABIP (which is usually a measure of hardness of contact) was mostly bad luck. The Prado / Johnson message board posters you cite are being silly, but the more sophisticated posters who can do the sort of analysis that the 10-12 MLB clubs aren’t interested in doing — they really can be smarter than the anti-analytical GMs, who can often be guilty of selling low or buying high.

    I can testify that, within a saber-oriented F.O., a great deal of thought goes into trying to figure out who is undervalued by all but the smartest analysis. If all GMs were doing that, it would be pointless, and there would be no economic reason to hire sabermetricians.