On Saturday night, Paul Millsap did not record a double-double, and its absence was the story. The Utah Jazz forward had a relatively short 22-minute night in his team's easy 17-point win over the Detroit Pistons. The lack of run snapped a 19-game double-double streak, the longest in the NBA since 2006. In 21 starts in place of All-Star power forward Carlos Boozer--who will miss at least another month after undergoing arthroscopic knee surgery last week--Millsap has averaged 18.0 points and 11.5 rebounds, earning some All-Star talk of his own.
Millsap's success as a starter is an interesting case study in how players respond to different roles. Over the last two seasons, Millsap's per-minute numbers as a reserve behind Boozer were excellent. However, critics of statistical analysis are quick to point out the ways in which per-minute stats can be misleading in the cases of reserves like Millsap.
A little over a year ago, in the midst of one such debate, I helped Tom Ziller compile numbers for BallHype.com looking at the performance of reserves who, like Millsap, suddenly saw their playing time increase dramatically because of the loss of the player ahead of them on the depth chart. Conceptually, this study (based on a similar one done by John Hollinger in the original Pro Basketball Prospectus book) avoided the usual pitfalls of assessing performance based on playing time, namely that players naturally tend to play more when they play well.
Lo and behold, the data showed that these players tended to improve their performance when given more regular minutes, contradicting the critics' naysaying. There's a perfectly logical explanation for why this 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.
To name the theory derived from the study, Ziller used one of its standout performers, a player who translated good performance in limited minutes into even better production as a starter. Fittingly enough, that player was Paul Millsap. Thus was born the Paul Millsap Doctrine.
The funny thing about the issue of per-minute stats is that we've largely advanced beyond them. I'll use per-minute stats here and there to make a point, as with per-game stats, but for the most part I lean on per-possession numbers to add another layer of contextualization. For whatever reason, per-possession stats don't seem quite as offensive to the old school as per-minute numbers--maybe because there isn't the same kind of implication that a deep reserve could play 48 minutes (or 40 or 36 or whatever other number you prefer).
Anyways, let's take a look at how Millsap's key metrics compare this season as a reserve and as a starter to his performance during his first two NBA seasons.
Role ORtg DRtg Win% WARP TS% Usg Reb% Pass
0607 105.8 102.3 .615 6.1 .571 .179 17.4 .06
0708 105.9 104.0 .563 5.2 .548 .182 16.5 .11
Reserve 104.9 104.5 .513 0.7 .576 .202 14.4 .08
Starter 109.9 102.8 .716 4.6 .605 .215 19.1 .30
Role 2P% FT% OR% DR% Ast% Stl% Blk% PF% 2A% FTA% TO%
0607 .527 .673 15.1 19.8 1.9 2.1 2.3 7.5 .68 .16 .16
0708 .507 .677 12.2 20.9 2.1 1.9 1.9 7.2 .72 .15 .13
Reserve .543 .757 11.3 17.5 2.3 0.9 4.0 6.6 .70 .11 .17
Starter .570 .694 15.5 22.6 3.1 2.0 2.6 5.1 .73 .15 .12
Quite simply, Millsap's performance as a starter has been incredible. Saying he's played at an All-Star level over the last month and a half would be something of an understatement. If Millsap maintained this level of play for an entire season, we'd have to start talking about him as one of the best big men in the game. Seriously.
Millsap's defense and rebounding made him a very valuable reserve during his first two seasons, but the numbers show how he has taken a giant step forward on offense as a starter. Not only is Millsap playing a larger role in a Utah attack that has needed players to score, with injuries not only removing Boozer from the lineup but forcing Mehmet Okur and Deron Williams to play at less than full strength, he's doing so while improving his efficiency. Millsap is even starting to become a threat as a passer, handing out nine assists over the last two games and seven in an early-December win at Toronto.
It's hard to find much evidence that playing a larger role has hampered Millsap. He has been rebounding better than ever as a starter, while his blocks and steals are similar to where they were his first two seasons (I'm willing to write off the extra blocks and lost steals early in the season to small sample size). In fact, the only sign Millsap has altered his game to accommodate additional minutes is in terms of personal fouls. One argument against per-minute stats is that bench players often have high foul rates and would, in theory, foul out well before they could rack up the projected per-40 minute numbers. However, in practice most players tend to foul much less often when playing more minutes, responding to the fact that foul trouble is now a real issue for them. Millsap is no exception in this regard.
Boozer and Millsap are both scheduled to become free agents at season's end (Boozer has indicated he will opt out of the last year of his contract, while Millsap will be restricted), and with the Jazz already in danger of entering luxury-tax territory, the most likely scenario is Utah will have to choose between the two power forwards. Even before the season, there was a certain school of thought the Jazz would be better off saving money by letting Boozer walk and replacing him in the lineup with a combination of Millsap and more minutes up front for Andrei Kirilenko. Now that theory is gaining traction.
It's unfair to compare Utah's performance with the two different power forwards on the court this season because Boozer was injured shortly after Williams returned to the lineup. Last year, the Jazz played nearly as well with Millsap at power forward as with Boozer on the floor, the latter holding an advantage of about two points per 100 possessions. Utah scored better with Boozer, but defended better with Millsap.
The Jazz's differential this season with Millsap in the lineup--+6.2 points per 100 possessions--is very impressive considering the other injuries that have plagued the team. Based on that, it's hard to argue Utah has suffered any drop-off by moving Millsap into Boozer's spot. The bigger issue for the upcoming offseason would be finding another post player to supplement Okur, Kirilenko and whichever power forward is re-signed; opponents have played even with the Jazz when neither Boozer nor Millsap has been on the floor. Center Kosta Koufos, who has held his own in spot minutes during his rookie season, may be able to step in to that role.
Whether in Utah or elsewhere, Millsap will surely get a chance to continue his strong run as a starter next season. People around the league are beginning to take notice of what Millsap is doing, and Jerry Sloan even threw out the famous name of Wally Pipp as a comparison to Boozer being replaced by Millsap. Just don't call it a total surprise, because Millsap's production in a smaller role portended this all along.
Kevin Pelton is an author of Basketball Prospectus.
You can contact Kevin by clicking here or click here to see Kevin's other articles.