APBRmetrics

True story - I started doing preliminary research today on writing an article about the myth of good coaching and time out play. After my first google search, I see this published yesterday: https://sports.vice.com/en_us/article/t ... ay-calling

Fantastic article seth!

I'm more meh. The first part of the article, about coaches calling known practiced plays rather than drawing up new plays, has been known for over 35 years. In the last 1970s CBS started putting a camera and more importantly a microphone into one of the teams' huddles during timeouts. TV viewers thought they saw Lenny Wilkens drawing up a new play and asked isn't that a risky strategy? And the NBA or the announcers or whoever (maybe it was Wilkens himself) eventually explained that Wilkens wasn't drawing up a new play, he was merely reviewing an existing play to make sure the players knew what to do.

As for the lower scoring after timeouts compared to other halfcourt plays, I'd want to see how they define halfcourt plays. They seem to be including more than just deadball situations; a better comparison would be to compare the after-timeout plays only to true deadball situations where both teams can set up as they wish, e.g. at the beginning of the 2nd, 3rd, and 4th quarters. (That might over-correct, because those teams have to travel the entire length of the court whereas teams after timeouts usually half to travel only half the court.)

The other item that they appear to be ignoring is that timeouts are more likely to be called when a team has just experienced a scoring streak by the opponent and a cold streak of its own. If there's any autocorrelation to cold runs by teams (or good defense by their opponent), then unless the timeout breaks up that autocorrelation the team is likely to be colder than average after the timeout.

Finally, late game timeouts will sometimes be called because the team needs a 3-pointer to tie the game. The other team knows this and can defend appropriately, resulting in desperation 25-footers with a defender close by and 0.7 seconds on the clock. We would certainly expect lower offensive efficiency in those post-timeout situations. This also applies to less extreme situations where the team is only down by one or two points and can choose to attempt a 2-pt FG. But if there's only 7 seconds left in the game, their ability to generate a good play after the timeout is heavily reduced.

tl, dr: the article doesn't seem to take into account the ways that after-timeout plays can differ from each other, and from average play situations.