Are All Minutes Created Equal?

Kevin Love and the Paul Millsap Doctrine

When I was in the 8th grade, it was a requirement for every kid to complete 'the gym class mile', which meant running around our track four times. Some kids were worried about being able to finish the mile. Not me. I was already the fastest sprinter in our class and I figured my speed would be just as valuable on a long run. The teacher yelled, "Go!" and I was off, running as fast as I could and easily going past the rest of my classmates.

It was about third-quarters around the first lap that it hit me. My legs didn't want to go anymore. Some of my classmates - the same ones who seemed to be going so slowly only a couple of minutes ago - were passing me by. Although I was one of the fastest 8th graders in the school in short bursts, I couldn't keep that speed up over a longer period of time.

The Paul Millsap Doctrine - a term coined by Tom Ziller - essentially attempts to show that in basketball, among players who are in far better shape than the average 8th grader, production levels tend to improve when given more regular minutes. Millsap, one of  the standout performers of Ziller's study, was able to translate good performance in limited minutes into even better production as a starter for the Utah Jazz.

Last season on the Timberwolves, much to the bewilderment of fans and analysts alike, Kevin Love started only 22 games and averaged only 28.6 minutes per game. This season, Love has started every game and is playing an average of 36 minutes. His numbers have obviously increased as well and the increased playing time has certainly not harmed his ability to score or rebound the basketball.

So, does this prove the accuracy of the Paul Millsap Doctrine? Love and Millsap certainly have had their production levels improve with more regular minutes, but are these two examples of the rule or exceptions to the rule?

On BasketballProspectus.com, Kevin Pelton gives a brief recap of the theory and takes a deeper look at Paul Millsap's improvement. Pelton argues that, "there's a perfectly logical explanation for why [improved performance with more minutes] would be the case. Given more playing time, players have a chance to get warm and play through their mistakes instead of constantly watching the bench and waiting for their coach to pull them from the game." 

Fair enough. But what about players who aren't in great basketball shape? What about guys who foul so often that they couldn't play extended minutes, since they would inevitably foul out? What about the fact that minutes against Kobe Bryant are much harder than minutes against Shannon Brown, for example? What about guys who simply get lucky or are on hot streaks during limited minutes, but don't get the opportunities or don't capitalize on those opportunities when they get more minutes?

There are certainly examples of those, too. Trevor Ariza, for example, received almost 12 more minutes per game after going from the Los Angeles Lakers to the Houston Rockets, but only improved his scoring by a little over a point per game; saw drops in rebounding, steals and shooting percentage; and forced more turnovers.

I wanted to know more information. After tracking down the original article, I was able to dive a little deeper into the Paul Millsap Doctrine and look at Ziller's study. Ziller, with the help of Pelton, compiled a list of players in 2006-2007 "who got stretches of starter-level playing time due to injuries to starting teammates". There were 17 players who fit that criteria. Both of the analysts made sure to note that the increased minutes for these players came from necessity - where starters had been injured or suspended - and not cases where players were given more minutes precisely because they were playing better.

The results:

Shooting percentage: On average, effective field goal percentage and True Shooting percentage both increased by .003. Eight players saw their shooting percentages improve with the extra minutes; nine saw declines.

Points per 40 minutes: +1.82 on average; 15 improved, 2 declined.

Rebounds per 40 minutes: -0.01 on average; 11 improved, 6 declined.

Assists per 40 minutes: +0.48 on average; 10 improved, 7 declined.

Steals + Blocks per 40 minutes: -0.11 on average; 4 improved, 13 declined.

Turnovers per 40 minutes: -0.10 on average; 9 improved, 8 declined. (Of course, improved means lower turnovers here.)

Fouls per 40 minutes: -0.92 on average; 13 improved, 4 declined.

And finally... PER: +2.38 on average, 15 improved, 2 declined.

I'll be honest, some of  the numbers went against my original instincts. For one, foul rate not only went down, it went down quite substantially. Our own Nikola Pekovic is a shining example of this. Although Pekovic could hardly stay on the court during his first few games due to foul trouble, when Darko was hurt and Pekovic received more minutes as a result, his foul rate did not increase. Darko left the game against the New York on 12/6 with an injury and did not play in the next game against Oklahoma City. Pekovic played 25 minutes against the Thunder, and only had 4 fouls, despite racking up 2 fouls in 9 minutes in the previous game against the Knicks.

Scoring, assists and turnovers all improved in the study. Let's look at our own case of Sebastian Telfair to see if it held true for the Timberwolves. Luke Ridnour was injured against the Houston Rockets and Sebastian Telfair had to start the next five games. Telfair went from averaging 20.9 minutes per game as a backup to 37.2 minutes as a starter. His scoring went down from .432 points per minute to .311 points per minute. His assists, however, went up from .151 per minute to .172 per minutes.

The most surprising part of the study's results, however, involved rebounding rate. Rebounding actually went down ever so slightly on average and 6 of the 17 players saw their rebounding rates decline. Given that much (of course, not all) of Kevin Love's improvement - and Paul Millsap's improvement - has to do with rebounding numbers, this seemed to go against what the study originally attempted to prove: that production increases with more playing time.

Kevin Love's minutes have seen a 26% increase from last season. His rebounds per 36-minutes, however, have only increased by 11% from last season and his total rebound percentage has only increased by 9%. In other words, his increase in minutes has not had a corresponding increase in rebounding production. How about scoring? With that 26% increase in minutes, Love's scoring has increased by 17.5%; certainly more than his rebounding, but again, that increase does not correspond with his increase in minutes.

However, the study, and the Paul Millsap Doctrine, only states that production will improve. It does not state by how much. So far, from looking at the study's own data and even some of our own examples of the Wolves, the Doctrine holds up.

Of course, there are still questions to be asked and a couple of larger problems to address from this study. For one is the fact that teams are not only limited to one option to replace an injured or under-performing player.

Using our example of Telfair from above, if Telfair's production started dropping as a replacement starter for Ridnour, Kurt Rambis had the option of using Mo Ager/Wayne Ellington/Sundiata Gaines as a replacement. Coaches tend to let players play who are playing well - or are playing well enough - so its hard to tell if productivity increases with the numbers of minutes played, or if it works the other way around.

Silverbird at the site FreeDarko.com had similar concerns and tried testing for the relationship between Teammate and Opponent-PER with the number of minutes that a person plays. They concluded that:

"We find that yes, the longer a player stays on the floor, the higher the quality of both teammates and defenders. Thus, given the positive effect of minutes-played on match-up quality, and the negative effect of match-up quality on individual production, it seems plausible that – all things really being equal – an increase in minutes will lead to (slightly) decreased productivity, on average. And that this is especially true for above-average bench-players who get a large bump in mpg."

This, of course, leads to the counter that perhaps a player who is playing well with increased minutes will face better defense since the opposing coach will strategize to stop him.

If anything, it seems that players who get increased minutes, in general, will not see their production levels decline substantially as a result. Kevin Love is in shape enough to play 36 minutes a game. Nikola Pekovic has not fouled out of every game where he's received more than 20 minutes. And Sebastian Telfair improved his assists, while seeing a slight dip in scoring. 

What I think the Paul Millsap Doctrine, and the discussion of its validity, really shows us is the complexity of the game of basketball and how hard it is to truly find any definitive statements. Unlike baseball, where stat analysis and sabermetrics are less messy due to the limited number of interactions, basketball is a free flowing game that is next-to-impossible to divide into the sum of its parts. I'll admit that stats can be valuable, but like this analysis shows, only when you can add context to them by looking at the individual scenarios.

Are all minutes created equal? Certainly not. The Paul Millsap Doctrine is helpful to show, however, that when bench players are performing well, there may be cause to want their minutes to increase, since its unlikely that their production will dip substantially. I'm glad I could look at the study and the numbers in depth to get a better understanding of this theory.