> “We found that when people broke the rules, teams were less likely to win games. Rule breaking hurts teams, despite the fact that people in positions of power, or coaches, might look at the rule breakers as people who are facilitating a better team,” Wakeman said. “The big caveat is that this is correlational, not causational.”

This is a really surprising piece of commentary considering the finding in the immediately prior paragraph:

> Different situations had different effects on coaches’ assessments of penalized players. Their generally favorable views [were] absent during winning streaks.

So the thought process here is, first we observe that coaches like fouls when the team is losing, and don't like them when the team is winning. And then we say that the coaches must be misguided (unless there's some kind of bias in the sample, but come on, look at the data) because teams committing a lot of fouls are doing worse than teams that aren't.