With Or Without You: Kendrick Perkins
Posted by Neil Paine on June 17, 2010
At first glance, Kendrick Perkins' knee injury early in Game 6 may seem like only a minor setback for the Boston Celtics -- after all, Perkins has scored just 8.5 PPG in the Big Three Era, and while his rebounding and defense are solid, he's never finished in the top 10 in RPG or garnered All-Defense honors. However, despite his lack of box score presence, Perkins has been highly important to the Celtics' chances this season: when he plays less than 22 minutes in a game, Boston is 5-8, including Tuesday's ugly loss and a similar beatdown suffered vs. Orlando in the Conference Finals when Perkins was limited by foul trouble.
We can really illustrate Perkins' hidden importance by looking at the Plus/Minus numbers. When Perkins was on the court for Boston this season, the Celtics outscored their opponents by 7.2 points per 100 possessions; when he wasn't playing, that number was only +0.2, a difference of -7 pts/100 poss. The same story has been true since Perkins became a regular starter in 2006-07: putting together a rudimentary game-by-game "with or without you" regression (similar to what I did in this post) between team HCA-adjusted PPG differential and whether or not a player played 17 minutes in the game (the MPG that usually indicates a "contributor"), Perkins shows up as adding +2.18 PPG to Boston's differential when he's a contributor vs. when he plays limited minutes or less.
Meanwhile, Perkins' Game 7 replacements, Rasheed Wallace and Glen Davis, don't fare nearly as well by the WOWOY metrics. Despite Davis' heroics in Game 4, he and Wallace have been Boston's two worst players by net on/off rating during the playoffs. Wallace has been particularly toxic for the Celtics all season -- the team played 5 pts/100 poss. worse when he was on the floor, as evidenced by his recurring appearance in the Celts' worst lineup combinations. Davis & Wallace look better by the 4-year WOWOY regression (Davis is +1.36, Wallace is +0.47), but neither has the ability to positively impact the game the way Perkins does. Without his presence, and playing on the road (home teams win Game 7 80% of the time), the Celtics appear to be in dire straits tonight.
June 17th, 2010 at 11:28 am
It looks like we are royally boned tonight, but I'm feeling optimistic for some reason. Maybe it's the crack-cocaine talking, but I think Ray Allen is due for a comeback game, and this is what Doc has been saving KG and Pierce for all season.
June 17th, 2010 at 11:36 am
Interesting, Neil. Yet in the playoffs this year, Davis has an ORtg/DRtg of 110/101, and he has 0.148 win shares per 48 minutes, best on the Celtics. Wallace is at 105/100 and 0.120 WS/48 for the playoffs, while Perkins has 94/100, and Perkins's 0.076 WS/48 is worse than anyone on the team except Michael Finley and Shelden Williams. Why the disconnect between the Oliver stats and your plus-minus findings?
June 17th, 2010 at 12:13 pm
Yes Perkins has a Superior WS/48 in the regular season, but not the playoffs. Very interesting.
June 17th, 2010 at 12:21 pm
The Oliver stats do a really good job of capturing offensive ability (provided you look at %Poss in addition to ORtg), but defensively they're just a guess. My personal feeling is that because DRtg basically assumes the player is as good as the team's defense, it underrates Perkins' defensive contribution and vastly overrates Davis' (I don't think anything can get an accurate read on Wallace's D).
June 17th, 2010 at 1:20 pm
What I've heard that Perkins had some kind of injury he was playing with in the playoffs. His offensive skills are not at the same level in these playoffs.
His two year Adjusted +/- in the playoffs are also poor, that's why I don't really like to pay attention to that stat.
June 17th, 2010 at 1:26 pm
*What I've heard is that
Sorry.
Neil has mentioned the usage-efficiency discussion before but I always go back to Pau Gasol versus Kobe Bryant. Anyway is using a greater amount of possessions always the best thing for a team? Pau is much easier to fit onto a team and has a superior WS/48 rating. I know he is a capable defender. Assuming two players have a similar PER rating (and equal defensively), the one with lower usage rate generally has more Win Shares.
One last thing: http://basketballvalue.com/teamplayers.php?year=2010%20playoffs&mode=summary&sortnumber=94&sortorder=DESC&team=LAL
Wow that's crazy. Pau is in negative figures here? APM is so noisy.
June 17th, 2010 at 1:38 pm
... Why 22 minutes? Is there some significance to 22 minutes exactly? I could point out that number improves to 17-15 (53.13%) if you count games he played 24:30 or less, instead of 22. If Rasheed's back holds up, I would trust him to play good defence on Gasol and give them an option on offence.
June 17th, 2010 at 2:02 pm
Sheed can't stay out of foul trouble and hasn't been making shots. IF Boston can get GREAT teamwide effort, and the regular starters are all on, Sheed's (in)ability to contribute, probably is the deciding factor. Did I say I felt optimistic in my first comment?
June 17th, 2010 at 2:33 pm
Boston is going to have a hard time keeping Wallace out there for the same amount of time the Lakers will run with Gasol. That's what will hurt them some in my opinion. Pau will play 45 minutes easily tonight. Not to mention Bynum for a good 25 minute run. KG will play big minutes but I don't know if he can move over to center with Pau.
Having Perkins out give LA some size and lineup advantages they didn't have before. Boston will have to counter by going small.