Turnovers and Rebounds: Significant Trend or Random Streak?
I was perusing KU’s Game Plan page this week and something jumped out at me: the Hawks’ sudden drop in turnovers forced. In non-conference games, opponents averaged turnovers on about 28% of their possessions (this isn’t exact, but it’s a good enough estimate). The lowest turnover rate of any opponent had been Boston College at 21.8%. But as soon as Big 12 play started, the turnovers stopped.
@Nebraska: 18.7%
Oklahoma: 16.8%
@Missouri: 12.8%
Iowa State: 15.3%
Nebraska: 21.6%
Now, some of the drop is simply because these teams are surprisingly good at not turning the ball over (the worst among them is Nebraska, which ranks 101st in T0%). But these rates are even below their season averages; not a good sign for a defense that has previously been ranked near the top in forcing opponents to cough it up, right?
Well, not so fast. Along with the drop in turnovers, there was a simultaneous drop in opponents’ offensive rebounding percentage (OR%) and effective field goal percentage (eFG%). Their previous season averages had been approximately 31% and 45%, respectively:
TEAM: OR%, eFG%
@Nebraska: 21.1%, 47.0%
Oklahoma: 29.6%, 35.9%
@Missouri: 25.2%, 41.5%
Iowa State: 21.1%, 36.1%
Nebraska: 25.0%, 32.5%
So while the turnovers decreased, they’ve done a better job of crashing the defensive boards, and they’ve made it even tougher for opponents to get open looks. Could this be the result of a change in defensive strategy - gambling for turnovers less and playing “straight up;” putting more priority on staying between their man and the basket, and keeping a hand in their face? I didn’t get to watch any of these games closely, so maybe those of you that did can answer that question.
Whatever the reason, I wanted to know how often other teams show similar changes, and what it meant for those teams going forward. Unfortunately, Pomeroy’s site only has Game Plan data for 2007 and 2008, so I could only use last year as a guide. I searched the data to find all major conference teams with a stretch of 5 or more games in which their TO% dropped 10+ points from its previous level. I found 5 teams: Arizona State, Duke, Florida, Georgia, and West Virginia. But the Mountaineers were the only team whose streak wasn’t accompanied by a simultaneous general worsening of their defensive efficiency (which hasn’t happened in KU’s case).
So what happened to WVU after the streak? Their next game was on the road against a top-20 (according to Pomeroy) conference opponent with a talented freshman forward (Notre Dame and Luc Harangody). Their turnover and rebound rates dropped even more, their defense softened a little, and they lost a close game, 61-58. Going forward, their turnover rate never recovered to its previous level, but their opponents’ offensive rebounding DID, not a good combination.
By no means does this imply the same thing will happen to KU, but it is something to look out for tonight. K-State will have success on the offensive boards, as they have all year. If they also manage to hold onto the ball as well as other conference opponents have, tonight’s game may be closer to a toss-up than Pomeroy’s 76-68 prediction would indicate.
* Have the Phog Blog come to you! * Subscribe to Comments
5 Responses to “Turnovers and Rebounds: Significant Trend or Random Streak?”
Leave a Reply
Comment Policy:Be polite.You must be logged in to post a comment.
January 30th, 2008 at 3:57 pm
David, I can’t fathom why (My recipe for beating Kansas would be to 1) shoot lots of threes, 2) attack the offensive glass, and 3) commit as many fouls as reasonably possible.) but it appears to me that the conference opponents have a made a conscious decision not to try to rebound their own missed shots.
If the other team is going to send four guys back on defense after every shot attempt, the increased reward for creating a (live-ball) turnover rather than forcing a missed shot diminishes.
As long as this is a conscious decision by the Jayhawks to force bad shots and take the defensive rebound given, I’m not worried that when an opponent (and I suspect K-State will be such an opponent) plays offense in a manner which increases the value of a steal the team will be able to adjust.
January 30th, 2008 at 4:47 pm
Hoopinion, thanks for the analysis. I got to watch about one half of the Iowa State game, and I had noticed the Cyclones were doing a fantastic job of preventing KU fast breaks. I failed to catch on to the fact that they were getting back on D before Kansas had even secured the ball.
I can understand why a coach would choose to do this. The knock on Kansas for a couple years has been that too much of their offensive output is dependent on getting easy transition points, and that their half-court offense was one of their weaknesses. That’s why, watching the ISU game, I was thrilled with how well they did offensively. It put to rest that fear for me. So, while I don’t think it’s really the best idea, I can at least see why teams might choose that strategy.
And I’m pretty sure you’re right about KSU not playing that way tonight. Let’s hope you’re right about Kansas adjusting.
January 31st, 2008 at 10:45 am
Well, to follow up, the turnovers stayed down (18.4%), the rebounds went back up (44.4%, actually their worst game of the season by that measure), and KU had by FAR its worst defensive showing of the year.
By the way, was anybody else frustrated at the guards’ insistence on going for layups in the last 4 minutes? After KU’s last time out at around 4:30 left, they only shot TWO 3-pt shots, despite being down by a pretty significant margin. Sure, if they miss a few it’s over. But they’re not gonna catch up trading 2’s.
January 31st, 2008 at 10:51 am
No one can acccuse Phog Bloggers of being homers. Well, we are, but not in our analyses. David H stated before the game that it would be closer to a tossup than Pomeroy’s 78-68 prediction and that k-state would have some success on the offensive boards. The REAL Standings projected k-state as the winner. Try to get that kind of straight talk from a politician.
February 1st, 2008 at 9:22 am
That game was certainly a data point in favor of my recipe for beating Kansas.
K-State shot about 30% more three-pointers (as a percentage of their field goal attempts) than normal and made 46% of those attempts as opposed to their season average of 32.7%. That’s not going to happen a majority of the time but you only need it to work once to beat Kansas. And what are the chances you beat them playing your normal game anyway.
I’m sure all of us would have taken 20+ three-point attempts from Walker, Pullen, Young, Sutton, Stewart, and Kent before the game. The only things I’m unhappy about are Russell Robinson taking 8 shots (3 of them threes) and Darell Arthur’s sustained mediocrity as a defensive rebounder.