The Ben Howland timeout. You know the reputation — they come at random times, more often stopping a UCLA run than bringing any benefit to his team, leaving the Bruins without any timeouts for the stretch run. I’ve said it myself plenty of times, but then in poking around the issue last week, I couldn’t find any good factual analysis interpreting Howland’s timeout pattern and the results. So, I did it myself. And here are the facts, based only on this season, with details to follow: Ben Howland makes pretty darn good use of his timeouts.
Before we get into the hard numbers, here’s the method to my madness. I only looked at timeouts called by UCLA prior to the four-minute mark in the second half and I excluded any timeout in the final minute of the first half as well, figuring those timeouts are almost universally accepted as a “normal” usage of timeouts. Using that group of timeouts (there have been 77 this season over 26 games), I looked at a few things: the total score for each team’s possession following that timeout; the total run prior to the timeout; and, the total run after the timeout. Defining a run is a bit more perilous, but I limited it to consecutive possessions with points scored by one team, while allowing for one possession without a score provided that team’s opponent does not score. In other words, if both teams failed to score on the first possession following a timeout, there is no run recorded. So, two UCLA possessions and two hoops mixed in with two opponent possessions with one hoop equals a 4-2 UCLA run and three UCLA possessions with two hoops coupled with three opponent possessions with no hoops equals a 4-0 UCLA run. Also, any timeout, media or otherwise, ends a run, and any free throws interrupted by a timeout get counted with the run.
So, with the methodology out of the way, here are the numbers. In the 74 UCLA offensive possessions following a UCLA timeout, the Bruins are scoring 1.42 points per possession, while they are giving up just 0.94 points on average in the 68 opponent possessions following a timeout. Prior to a UCLA timeout, in runs over the course of 179 possessions, the Bruins are being outscored 1.82 PPP to 0.67 PPP, an amazing differential of 1.15 PPP. Conversely, in runs following Bruin timeouts, UCLA is scoring 1.55 PPP and allowing 0.85 PPP (differential = 0.70 PPP). Also notable is that Howland called nine timeouts immediately following a UCLA basket, something Bill Walton has notably pointed out as problematic in several ESPN broadcasts.
So, what does it all mean? Well, first I think you can throw out the idea that Howland’s misuse of timeouts is stopping his own team’s runs. Only 12 of those timeouts came at a time when UCLA had been on a run, and of those, three of them were one-possession/one-hoop “runs.” Clearly the data shows that Howland uses timeouts most often to effectively stop opponent runs and begin UCLA runs. Now, why exactly Howland’s teams are getting outscored by 1.15 points per possession over the stretches that precipitate those timeouts is another question entirely. But you can look through the play-by-play and find a lot of commonalities in the events prior to the timeouts. For instance, 31 of Howland’s 77 timeouts have been preceded either by an opponent’s layup, dunk or transition hoop; nineteen of those timeouts followed an opponent’s three; and eight have followed a defensive possession on which the Bruins allowed an offensive rebound. So, the idea that Howland is just randomly calling timeouts in the normal flow of the game? That one’s pretty much shot down too.
As for the idea that Howland’s rampant use of timeouts leaves his team absent any in his pocket for the stretch run? Only three times all season (vs. Oregon State, at Arizona, and at Arizona State) has Howland used as many as four timeouts prior to the final media timeout. And in the Oregon State game, which turned out to be a 10-point UCLA win, Howland’s timeouts stopped four runs totaling a 12-0 OSU benefit and helped to generate opposite runs totaling an 11-7 UCLA advantage. Likewise, in the Arizona game (an 11-point UCLA win), Howland’s timeouts stopped runs totaling 15-2 for Arizona and turned them into 4-0 UCLA advantages after the timeout. Even in the Arizona State game (which turned into a blowout loss for UCLA), UCLA had given up 13-8 runs prior to timeouts and earned a 3-0 run following them.
As for those nine times where Howland has called a timeout directly following a made UCLA basket? Well, those certainly don’t seem to be doing UCLA any good. Prior to those timeouts, UCLA had been on a combined 31-14 run. And in none of those cases had UCLA been outscored over the course of a preceding run. Moreover, following those nine timeouts, UCLA has been outscored 17-15. So, yes, in those nine cases, Howland’s apparent misuse of timeouts has robbed his team of some momentum. Of course, in six of those occasions, UCLA was already up double figures and one of those was a use-it-or-lose-it late first half scenario against San Diego State. But there are two occasions that appear most troubling. First off, on that fateful late November night when Cal Poly upset UCLA at Pauley Pavilion, Larry Drew II hit a jumper with 7:15 left to play that gave the Bruins an 11-point lead. However, with a media timeout pending at the next stoppage in action, Howland used one of his timeouts. And when the Mustangs came out of that break, Dylan Royer hit a three, the first of seven straight possessions on which Cal Poly would score (20 total points over those possessions). Over those same seven possessions on the other end, UCLA would score six points of its own. It is impossible and ridiculous to say that Howland’s timeout caused that run, but it certainly seems to have contributed to it.
The other occasion where some timeouts by Howland can be questioned was in the Utah game. It was an ugly game, but after taking an 11-point lead into the half, a Shabazz Muhammad jumper early in the second half expanded the Bruins’ lead to 12. And then Howland called a timeout with 17:11 remaining. Utah scored on five of its next nine possessions with UCLA scoring on just three and by the under-12 media timeout, the Bruins’ lead had been cut down to five. Later in the half, a Travis Wear jumper extended the UCLA lead to six points and it too was immediately followed by a Howland timeout, although this one appeared to have little effect on the flow of the game.
Ultimately, it is hard to say exactly what effect a timeout has on the momentum of a team and a game. Who’s to say that a team in the process of getting run off the court wasn’t on the verge of making a play to turn the game around? Who’s to say that a team on a run is going to stay hot? But the commonly held belief that Howland uses timeouts randomly and ineffectively does not appear to be the case, at least not this season. The data shows that Howland’s timeouts are called largely in response to defensive breakdowns not only stop opponents’ runs but to create new ones for the Bruins.