Dan walks through some extremely clear-cut sports examples where people continued making decisions that were
wrong costing them wins in ways that were easily calculable. They did eventually change, but even that took years, sometimes decades. Given that this was happening in something as easy to measure as sports, he suggests we should expect bad decision making in corporations or governments to take even longer to correct.
My major concern with this post is that the goal of pro sports is a mix of winning and making money via attention from fans. This first means that the feedback loop isn't as tight as it would be otherwise. Additionally, many of the optimizations that lead to more wins make games more boring, which ultimately costs the entire league money. Using these optimizations is in some ways defecting in a prisoner's dilemma: you win more, which your fans like and other fans don't, but if everyone does it the game becomes more boring, which drives away everyone's fans. Under this light, base stealing attempts that cost you counterfactual wins could be viewed as a prosocial sacrifice to the entire league's benefit.