Do Timeouts Stop Scoring Runs? (A Thrilling Statistical Analysis)

Donate To GopherHole
Unregistered User

Unregistered User

Wild animal with a keyboard
Joined
Jan 17, 2010
Messages
14,747
Reaction score
4,919
Points
113
Not by me. This was posted on Reddit, and the analysis may have gotten the OP a job. Refer to the "Original Post" link at the end to read about the potential job offer. I did not devour the entire article, but thought some might be interested in taking a very deep dive.

I've done a bit of statistical analysis using ESPN's play-by-play database of college basketball games and found that calling timeouts doesn't stop teams on scoring runs and may even have a negative effect.
Read more here.
If you're interested in the code, the Jupyter notebook that this analysis was performed in can be found here.
Any thoughts or feedback would be greatly appreciated!

I did not embed the link from the "can be found here" spot, but the link below is the direct link to the analysis reference by "Read more here".

Link

Original Post
 

stop runs? of course not. You usually arent bringing back all your starters after you call a timeout when your opponent goes on a run, and your opponent will likely come out with the same group that went on the run, so they will still likely have the better 5 on the court at that moment...you slow the game down though with the timeout. Easy strategy in all competitions...speed the game up when you are winning, slow it down when you are losing.
 

I don't think it's possible to statistically determine this. Too many variables. How much better is the other team than you are? How'd they go on the run...them pressing you, rash of unforced turnovers, while you were resting your only player, you are controlling the ball by strategy but played too uptempo for a stretch, were you pressing and got burned for baskets repeatedly, how does your team respond out of timeouts (not all teams are the same), did they hit 3 threes in a row? Million variables...timeouts can make absolutely no difference or they can turn the game around. Good coaches know when it has the potential to make a difference and when it doesn't... if you don't take the timeout your players feel you have given up on them or your fans do. Timeouts to stop runs can absolutely work...you can make adjustments to tempo, to your type of defense, to your personnel, to your mindset...a million things but they aren't guaranteed or magic. And sometimes the timeouts are just for show because you are dead on arrival. 100% games can be won with timeouts.
 

60's Guy said:
I don't think it's possible to statistically determine this. Too many variables. How much better is the other team than you are? How'd they go on the run...them pressing you, rash of unforced turnovers, while you were resting your only player, you are controlling the ball by strategy but played too uptempo for a stretch, were you pressing and got burned for baskets repeatedly, how does your team respond out of timeouts (not all teams are the same), did they hit 3 threes in a row? Million variables...timeouts can make absolutely no difference or they can turn the game around. Good coaches know when it has the potential to make a difference and when it doesn't... if you don't take the timeout your players feel you have given up on them or your fans do. Timeouts to stop runs can absolutely work...you can make adjustments to tempo, to your type of defense, to your personnel, to your mindset...a million things but they aren't guaranteed or magic. And sometimes the timeouts are just for show because you are dead on arrival. 100% games can be won with timeouts.
Click to expand...

You more or less nailed it here. Interesting project by the OP, but certainly does not tell the entire story. Momentum in sports is undeniable for those who have played and lived it. A timeout can't cure all of your ailments, but it can certainly be a coaching opportunity, and can at the very least alter the pace of play.
 

Seems to me the biggest variable is the coaching done during the timeout. If you make the right adjustments, I'm seen the timeout strategy work.

I remember in High School we were getting killed by a kid against our 2-3 zone, we came out of the timeout in a box and one, half the team thought we were in a zone and half thought we were in a man to man. Their run stopped cold and we won the game.
 


Otis said:
Seems to me the biggest variable is the coaching done during the timeout. If you make the right adjustments, I'm seen the timeout strategy work.

I remember in High School we were getting killed by a kid against our 2-3 zone, we came out of the timeout in a box and one, half the team thought we were in a zone and half thought we were in a man to man. Their run stopped cold and we won the game.
Click to expand...

Absolutely! Coaching can win or lose games 100%. In your case, the other coach didn't recognize or couldn't get his kids to adjust to the change in strategy by your coach. It's fun of coaching...having adjustments to the other guys moves and seeing how he responds to yours.
Last year versus Iowa. Iowa half court trap, little spurt, Pitino timeout and made adjustments. May have even taken two timeouts? And Dupre was no longer the guy with the ball attacking the press, plus how and where the Gophers attacked across the 10 second line. Iowa countered but it wasn't effective for them. Gophers win. No timeout, 100% we lose.
 

Lmao. I can skim the article and identify you guys didn't read it and automatically dismiss the study because, you know better.
 





I skimmed it. Relevance because 100% timeouts can make a difference in stopping or starting runs. Adjustments and counter adjustments is what coaches do. Often you do it on the fly in basketball but when you need your players attention, or you need to quiet the crowd, change the momentum, again a million things you need a timeout. Relevance because 100% experience says a timeout can be a huge factor. How could any coach say it wasn't?

Do timeouts stop scoring runs? All by themselves they probably don't make much difference. But the adjustments you make can make a huge difference. How you gonna quantify all the variables? Who is coaching? What are the matchups? Who is refereeing...how is it being called? What is the foul situation? There are more variables to basketball than any other sport.

Timeouts are precious. Timeouts can absolutely change the outcome of games. I really don't care what any analysis says Winasota...on this subject, coaching 1,000 plus games and watching twice that many I know timeouts matter...depending how you use them.
 



Also, this is the first paragraph of the intro:

"Basketball fans generally believe that calling timeouts can slow or stop scoring runs. Certain coaches (famously, Roy Williams) are often criticized for not calling timeouts in an attempt to end a run. However, to the best of my knowledge, this general wisdom has never been tested nor supported by any actual analysis. I intend to use a large dataset (all of the play-by-play data available from ESPN) and some straightforward statistical analysis to show that calling timeouts midgame to end scoring runs has minimal effect on stopping scoring runs."

How is that fair analysis? Talk about looking at data through a lens. This kind of goes against the basic principles of a study. He should've stated that he was looking to find out if it had an impact, not that he intended to disprove it.
 



GopherHoopsFan said:
Also, this is the first paragraph of the intro:

"Basketball fans generally believe that calling timeouts can slow or stop scoring runs. Certain coaches (famously, Roy Williams) are often criticized for not calling timeouts in an attempt to end a run. However, to the best of my knowledge, this general wisdom has never been tested nor supported by any actual analysis. I intend to use a large dataset (all of the play-by-play data available from ESPN) and some straightforward statistical analysis to show that calling timeouts midgame to end scoring runs has minimal effect on stopping scoring runs."

How is that fair analysis? Talk about looking at data through a lens. This kind of goes against the basic principles of a study. He should've stated that he was looking to find out if it had an impact, not that he intended to disprove it.
Click to expand...

There is a reason this isn't ground breaking research on the front page of ESPN, it was a Redditor, as the first post clearly states. Original poster didn't even read the damn thing.

YET....

We have a bunch of posters grandstanding acting like they are way smarter than an obviously flawed post on a website.

Like I said,

Did you guys read the damn thing?
 

Winasota Gopher said:
There is a reason this isn't ground breaking research on the front page of ESPN, it was a Redditor, as the first post clearly states. Original poster didn't even read the damn thing.

YET....

We have a bunch of posters grandstanding acting like they are way smarter than an obviously flawed post on a website.

Like I said,

Did you guys read the damn thing?
Click to expand...

There is definitely a poster in this thread with an attitude problem. That we can agree on.
 

The article is crap. It tries to base mathematical models on a basketball game. This is ridiculous as it doesn't factor in skill of the teams and coaches. If UNC plays Little sisters of the Poor they are going to have some insane runs based solely on talent and a time-out isn't going to help one bit even if UNC runs out their bottom 5 guys out of the time-out.

Now if Minnesota is playing Northwestern (fairly evenly matched teams) and the Wildcats go on a run, Pitino can call a timeout make some adjustments by bringing in a different player or switching up the defense or offense to actually stop the run. i.e. if NW's center scores 10 consecutive points, Richard can bring in Reggie to counter the run and thus stop it. (until Reggie fouls out) Or instruct a back-up to go bully the center out of his flow.

It's situationally dependent whether a timeout can stop a run or not and is up to the genius of the coach calling the timeout on whether it is effective or not.
 

You must log in or register to reply here.



Top Bottom