Related to (Eliezer Yudkowsky): Inadequacy and Modesty
Epistemic Status: Confident. No sports knowledge required.
In 2005, Willie Randolph became manager of the New York Mets.
In his first five games as manager, all of which he lost, Willie made more decisions wrong than I thought possible. If he needed to change pitchers, he waited. Other times he changed pitchers for no reason. Starting lineups made zero sense. Position players bunted. And so on. He cost us at least one of those games. My friend Seth and I called for Willie’s head.
He would go on to an excellent 97 win season in 2006, come in second in manager-of-the-year voting, get a contract extension, and only get fired after wearing out our starting pitchers so much that we experienced one of the most epic late season collapses in baseball history in 2007, followed by a horrible 2008.
Willie’s in-game decisions did not improve. If anything, they got worse.
Despite this, we came to understand why Willie got and kept his job.
Willie Randolph was a leader of men.
Players liked Willie. They wanted to play for him, work hard for him, be the best they could be. They put the team first. He created a positive clubhouse atmosphere. He inspired good performances, spotted ways players could improve.
That is what counts in baseball.
Do bad in-game decisions cost games? Absolutely. But not that many games. Maybe they lose you 4 a year out of 162.
If the lineup makes your players unhappy, that costs a lot more. If your pitchers lose motivation or have their rhythms disrupted, that matters more than getting high leverage for your best reliever. Maybe bunting inspires team unity. The reason we hate bunting so much isn’t because it’s a huge mistake. It’s an obvious mistake. A pure mistake. An arithmetic error.
Plenty of people could get those technical decisions right. I could do it.
What most of us can’t do is lead men. Leading men is what counts. That’s the real job, but it comes with these other tasks.
Sometimes these other tasks land in good hands, at other times they land in terrible hands. Those who do the little things right do succeed more, but you can still win championships without them. If you can lead men.
Other sports follow the pattern. Why do football teams employ Andy Reid, who could not manage a two-minute drill if his life depended on it? Why do highly successful basketball players refuse to cooperate with their teammates or practice key skills?
Because those are small mistakes. The things those people do right matter more.
Could Willie Randolph hire someone to micromanage the game? Could Andy Reid hire someone to manage his two minute drills?
No. The people who are capable of that, are not leaders of men, and how they make those decisions is part of how they lead men.
Even if you could do that, fixing such penny-ante problems is too disruptive. You want their eyes on the prize.
If a position calls for a leader of men, you often find a leader of men. If it instead requires super high levels of another skill, whether it’s coding, raising money, lifting weights, intricate chemistry or proving theorems, you’ll find that. However, if you need rare levels of such skills compared to what you can offer, you won’t select for anything else. You can’t demand ordinary competence in insufficiently important areas. There aren’t enough qualified applicants. Plus it wouldn’t be worth the distraction.
This helps explain why people in unique positions are often uniquely terrible. They’re not replaceable. Some incompetence and shenanigans are acceptable, so long as they deliver the goods.
The same goes for other groups, organizations, religions, software and most anything else.
If a system has unique big advantages, they’re not effectively competing on less big things. They might be optimizing small things, but they don’t have to, so you can’t assume such things are optimized at all. Even when a system does not have unique advantages, anything insufficiently central is likely not optimized because it’s not worthy of attention.
It is much, much easier to pick out a way in which a system is sub-optimal, than it is to implement or run that system at anything like its current level of optimization.
Thus I generally believe the following two things:
It is relatively easy to find ways in which almost anything could be improved on the margin, were one able to implement isolated changes. Well thought-out such ideas are often correct.
The person making such a correct suggestion would likely be hopelessly lost trying to implement this change let alone running the relevant systems.