Calling timeouts to stop runs?

[colts18]

Is there any research on the topic of road teams calling timeouts to stop runs. ex: 10-0 run and the road team calls a timeout in an attempt to stop it. If the home team goes on a run, is it more likely to score the next basket than the road team (ie Momentum)?

[AcrossTheCourt]

The problem with this: due to regression to the mean, at any point during a great run the team will likely decrease its ORtg anyway. So I guess you'd have to compare it to similar runs where there are no timeouts.

[Mike G]

I think you'd compare the differential of the ensuing possession(s) to the normal differential during the season between the 2 teams?
You hope to 'stop a run' by scoring. But 14-2 and 21-6 spells are also called 'runs', yes?
If a team calls time, sets up a play and scores -- and then gives up the next 8 points -- they haven't really stopped the run.

My intuitive perception is that it's more a matter of time: A good run sort of expires after a while. If halftime comes, the team on the ropes is 'saved by the bell'.
A timeout, then, just eats some time out of the run.

[mtamada]

The problem with this: due to regression to the mean, at any point during a great run the team will likely decrease its ORtg anyway. So I guess you'd have to compare it to similar runs where there are no timeouts.

Correct. And even then there might be conflating factors, e.g. maybe the reason they didn't call a time out was because there was a play stoppage and maybe they even subbed LeBron James in during that stoppage. So less reason to call a TO. Or maybe the quarter ended. I.e. it's hard to get an apples-to-apples comparison of the call-timeout strategy vs. the don't-call-TO strategy.

It's almost like what do you do when you have severe, incapacitating back pain? You can have back surgery. Or you can rest and hope your back gets better. I forget the statistics, IIRC something like half of the back surgery patients have good results. And the ones who rest? About half of them have good results.

But would the ones who had surgery still have had a 50% chance of getting good results if they hadn't had the surgery? Basically we'll never know.

We're in a better position with NBA statistics, because most patients can only be observed once and they either get the surgery or they don't, and we thus all we have is one outcome with nothing to compare it to. NBA teams repeatedly find themselves victimized by runs and have to decide if they want to call a timeout, so there are multiple observations that can be looked at and compared. But making sure the comparisons are apples-to-apples is not easy.