Tempo and Turnovers

by Jeff Fogle 7. January 2011 00:54

What's the relationship between tempo and turnovers? Does speeding up invariably lead to more turnovers because teams get out of control? Or, does a faster pace just mean more dunks and layups...which are pretty hard to screw up?

A couple of days ago, our article on the Lakers inspired some discussion across cyberspace related to tempo. It was noted that, during the recent Los Angeles mini-slump, the Lakers had slowed down their pace but increased turnovers? Some (including me) thought that odd because slowing down should decrease turnovers. Others were under the impression that slowing down creates more turnovers...possibly because it forces you attack a five-man defense rather than get your team easier shots in transition.

What's the reality? I have a personal bias towards "faster" meaning "sloppier." We saw that with Charlotte last week when Paul Silas tried a short-lived up-tempo approach with a group that immediately went out and committed 20 turnovers. I've logged the stats from many slow playoff games over the years where each team treasured every possession because so much was at stake.

Yet...I think we've all seen games where a change in tempo took a team out of its comfort zone in either direction. If you force a fast team to slow down against a good defense, they can have a lot of miscues in the halfcourt game. If a slow team gets caught up in a track meet, they're prone to throw the ball away. 

So, what is it? What's the relationship between tempo and turnovers?

I did a quick and dirty study focusing on the fastest and slowest teams in the NBA the past three seasons. The numbers were tabulated Wednesday afternoon, so they may have changed a smidge by now. I wanted to compare the turnover rates (turnovers committed per 100 possesssions) of the fast teams to the slow teams. If you just look at raw volume, there's a potential that faster teams create more turnovers just because there are more possessions. The increased tempo doesn't make them "sloppier." It's just the same level of sloppiness in a game with a few more possessions. Looking at turnover rate will give us a better sense of the true impact.

6 fastest teams: 24.7
6 slowest teams: 23.8

That's about one extra turnover per 100 possessions if you round up. To this point in the 2010-11 season, faster has meant a little sloppier.

6 fastest teams: 23.7
6 slowest teams: 22.9

Same story last season, which is a larger sample size obviously. Faster meant sloppier by about one turnover per game.

6 fastest teams 23.4
6 slowest teams 22.5

Consistent reads, with the faster teams once again committing about one more turnover per 100 possessions. Can't ask for a better indicator than that. I won't suggest this is anything close to a definitive study. But, it's a quick and dirty study that passes the eyeball test. Slower teams take care of the ball better than up tempo teams. Fast teams hope to get enough dunks, layups, and kick out treys to make up for the extra miscue (at least when you're looking at the extremes over the last two and a half seasons!).

So, until someone has time to do a more comprehensive study (and, please feel free to comment with a link to one if it's already been done), we can assume a slight general tendency for faster to mean sloppier...with the understanding that moving out of your comfort zone in either direction can lead to problems protecting the ball.

In terms of the Lakers situation, the high turnover numbers in their slower games may have represented being taken out of their comfort zone, or may have been a red flag that the team lost focus for a stretch. The increase in turnovers was at least going against the grain of what we normally see with slower tempos.

And...in terms of the Lakers...that tendency immediately stopped! You get the sense that players said to themselves "This is ridiculous!" after watching recent game films.

6 turnovers in 91 possessions vs. Detroit
10 turnovers in 89 possessions at Phoenix

They cleaned up their turnover mess in a hurry. Interesting that they slowed THAT far down from previous norms in the two-game set. Smart move in a back-to-back. It's tough to slow Phoenix down to just 89 possessions. That tied for Phoenix's third slowest game this season.

Lakers fans had to be happy with the improved level of play. Though, the sweep still came against losing teams. It's amazing how weak LA's schedule has been so far this year!

I promised last time to take a quick look at the Clippers/Nuggets game from late Wednesday.

Efficiency: Denver 96.9, Clippers 110.4
Two-Point %: Denver 36%, Clippers 45%
Rebounds: Denver 43, Clippers 57

You can see the big inside edges for the Clips. Blake Griffin had another double double. By halftime! Baron Davis continued his recent impactful play with 16 points and 8 assists in 31 minutes. Eric Gordon added 28 points. For Denver, Carmelo Anthony and Chauncey Billups combined for 40 shots and just 5 assists.

Great showing for the Clips as they continue their long stay in the Golden State. Their next opponent will be Golden State in a Sunday afternoon tip at the Staples Center. Blake Griffin and company won't leave California until a January 20th game in Portland. Can they really make it to .500 by the end of the year? We'll know more by that Portland tip off.

The Dallas Mavericks had another disappointing offensive showing Thursday in the early TNT game. This despite a great first half that saw them score 55 points. Only 40 came after the break, in a 99-95 loss.

Offensive efficiency without Dirk Nowitzi in the lineup:
87.4 vs. Toronto
100.0 vs. San Antonio
90.6 vs. Milwaukee
113.0 vs. Cleveland
101.2 vs. Portland

They were at 106.0 for the season heading into Thursday night. Won't have Thursday's official numbers until Friday. Only the game vs. Cleveland was above their norm...and Cleveland's defense has been horrible lately.

Back late Friday night with notes from a busy 11-game schedule...

Bookmark and Share DotnetKicks dotnetshoutout


1/7/2011 1:22:03 AM #


I have a question regarding the 12 teams you used for the tempo analysis. What were the combined records for the 6 fastest vs the 6 slowest? Does the fast tempo compensate for the turnover by way of a better record, or were the records comparable?

I like your analysis. Like I said in a previous comment, I am going to be doing similar things on my blog. I am looking at what has been done already to see if I get any good ideas.

DanTheNBAFan United States

1/7/2011 1:48:02 PM #

David Hess

Are you sure you don't have the cause and effect mixed up?  It seems like turnovers themselves increase tempo.  Tempo is essentially the inverse of [time per possession].  A turnover cuts short a posession (e.g. a team might have taken the full 24 seconds, if the ball hadn't been stripped from their point gaurd with 12 seconds left on the clock).  This decreases time/poss and increases tempo.

I'm not saying I'm sure the causality goes this way, either.  Just that I don't really know, and I don't think your quick calculation sheds light on the question.

Here's an idea.  Along the same lines, offensive rebounding should lower tempo, by making a possession last longer.  Maybe you could do the same calculation for the top/bottom teams in OR% (or, offensive rebounds per possession - that would be more analagous to TO%).  I don't think tempo should have any effect on OR%, so if you see a difference between the top and bottom teams in this case, it seems reasonable to conclude that OR% is affecting tempo.  The magnitude of the observed effect (if there is one) might shed some light on the effect of TO% on tempo (and vice versa).

David Hess United States

1/7/2011 4:42:52 PM #

Jeff Fogle

Can't disagree with your points David. So much of basketball is intertwined that it can be hard to get a true read on various individual elements.

Worth noting I think that that the quick and dirty study showed just one extra turnover per 100 possessions...but the nature of extremes suggests that it's not that one extra turnover causing the extra tempo.

*The six fastest teams this year range from 97-100 possessions per game.

*The six slowest teams this year range from 91 to about 92.5 possessions per game.

The style of these teams is independent of turnovers. Coaches are choosing to run a lot, or hardly run ever. But, it's certainly possible that your concerns are relevant in terms of how it all comes out in the wash.

Would be interesting to see in a possession-by-possession study some time what turnovers were doing to tempo. Meaning...are teams typically losing the ball early in a way that also sets up a fast possession for the opponent? Or, are teams typically losing the ball late...and then the other team has to take it out of bounds (off an errant pass) or walk it up in a way that's maybe a second or two slower than they're average possession. Note that Charlotte was the 5th slowest team last season, but the very worst in offensive turnover rate. Tough to pull off if a high volume of turnovers adds significant tempo to a game.  

I agree with you that there's potential pollution here. Whatever we do at first will be baby steps toward eventually getting things figured out. For now, I'm comfortable with the thought that tempo increases sloppiness. But, I was biased toward that line of thinking coming in. Can definitely see how sloppiness could influence possession counts.

Will consider trying to do your offensive rebound study when time permits, as well as the thoughts from Dan on game-by-game results when extremes meet. Thanks very much to both of you for commenting and presenting ideas for additional study.  

Jeff Fogle United States

1/7/2011 9:20:32 PM #

David Hess

Good point about the difference only being 1 turnover.  One TO can't cause a several-possession difference.

I would love to see time of possession data for basketball.  Two teams in a game always have nearly the same number of possessions, but is one of them using way more clock than the other?  Does efficiency change for different length possessions?  What about TO% based on possession length (i.e. of all possessions that last at least X seconds, how many have a TO at time X+1, X+2, etc)?  So many fun questions.

David Hess United States

1/10/2011 10:37:47 PM #

Nathan Walker

I plotted turnover rate (aka turnover percentage,etc) versus average tempo, and found roughly zero correlation.

Here's the graph (all NBA teams from 04-05 through 09-10):

Nathan Walker United States

Comments are closed
Archive | FeedSubscribe | Log in