I’m obsessed with the fact that line-ups featuring three or four bench players and Ray Allen have drastically out-performed line-ups featuring three or four back-ups and Paul Pierce since Allen arrived in Boston. The trend is holding true again this season; the line-up of Ray-Sheed-House-Daniels-Williams (the second most-often used line-up of the season for Boston) is scoring 127 points per 100 possessions and allowing just 93 in 72 minutes of play, according to 82games.com.
The same four back-ups with Pierce (the 4th most-often used line-up) has scored just 101 points per 100 possessions and allowed 116 in 57 minutes together.
To put those numbers in context: The C’s average about 110 points per 100 possessions and allow about 100; the Ray line-up is performing well above average on both sides of the ball.
The trend was similar (though less pronounced) last season.
So does Ray Allen possess that most unique basketball trait—the ability to make others around him better? If he does, shouldn’t that have a huge bearing on the decision (which isn’t being discussed in reality but is in blog comments) of whether to dangle Ray’s expiring deal at the trade deadline?
It’s far from a final conclusion, but there are strong indications that Ray indeed possesses this magical skill. I went back into Ray’s three final seasons with Seattle to see if the same trend was borne out, and though the data is shaky, it definitely leans toward confirming that Ray has some ability to carry line-ups populated with back-ups.
The five players who started the most games for the ’07 Sonics were Allen, Rashard Lewis, Chris Wilcox, Luke Ridnour and Nick Collison. I wanted to look at line-ups composed of four back-ups and Ray, but Bob Hill (Seattle’s coach that season) did not favor this substitution pattern; instead, he mixed starters and back-ups a bit more liberally than Doc Rivers has.
So I looked at line-ups in which Ray played with two or fewer starters. Ray played at least 48 minutes with five such line-ups, according to 82games.com. (They are line-ups #3, 5, 6, 8 and 10 on that page).
Two of those line-ups had positive plus/minus numbers for the season (+22 and +6), two had close to neutral negative plus/minus numbers (-2 and -3) and one bombed (-18).
This is probably not enough to make any definitive conclusions, though it’s worth noting that the two line-ups that got by far the most playing time among these five both outperformed the starting line-up offensively in terms of points per minute and put up the positive plus/minus stats.
The 2005-06 Sonics gave at least 20 starts to eight different players, so it’s difficult to separate the back-ups from the starters.
The 2004-05 Sonics are a cleaner case, since five players (Allen, Ridnour, Lewis, Reggie Evans and Jerome James) started at least 71 games.
Lucky for us, Allen played heavy minutes with two line-ups featuring just one other starter and a single line-up featuring Ray and four back-ups.
And guess what? All three of these line-ups significantly outperformed the Sonics starters and blew opponents away offensively, according to 82games. (The line-ups in question are numbers 3, 4 and 7).
Specifically, the starting line-up for Seattle that season was +28 in 789 minutes and scored 1.9 points per minute.
And the three Ray/back-up units in question did this:
The line-up of Allen-Lewis-Vlad Radmanovic-Danny Fortson-Antonio Daniels was +46 in 160 minutesand scored 2.35 points per minute.
The line-up of Allen-Daniels-Vlad Rad-Fortson-Collison (Ray plus four subs) was +56 in 147 minutes andscored 2.2 points per minute.
The line-up of Allen-Daniels-Lewis-Vlad Rad-Collins was +46 in 86 minutes and scored 2.9 points per minute.
So there’s no question that in this season—plus the last two with Boston—something good happened when Ray played with the back-ups.
Is this a definitive conclusion? No. More research is still necessary, though the stats clearly lean one way.
But I wanted to do one more thing. I saw this post on The New York Times’ Off the Dribble blog by Jon Nichols (of basketball-statistics.com and occasionally Orlando Pinstriped Post) showing how Kobe Bryant and LeBron James impact the shooting percentages of their teammates. Nichols wanted to know this: Does being on the court with Kobe or LeBron get teammates better looks, which result in better shooting percentages? (Short answer: Yes).
I asked Nichols if he could pull out similar data for Ray Allen and Paul Pierce, and he was kind enough to do so. The results are below (The positive and negative percentages represent the difference between the players’ collective overall field-goal percentages and their FG % while playing with Ray or Pierce).
First, the last three seasons of data for Ray Allen:
Only one significant negative number on the entire chart (mid-range shots this season), with the rest positive—and especially huge numbers for close shots the last two seasons. In an email, Nichols tells me it’s “no small feat” for any player to have positive numbers across the board in any season. (Kobe and LeBron each did it once in the three seasons Nichols analyzed).
Now the numbers for Pierce:
The numbers aren’t nearly as good for the captain. I’d caution against reading too much into these numbers, since they swing pretty wildly from year to year. But the pattern is there, and it probably goes a long way toward explaining why Allen/back-up units have consistently produced better numbers than Pierce/back-up units.
Of course, this doesn’t explain why teammates shoot better with Ray Allen on the court. Does his outside shooting draw more attention from the defense than Pierce’s herky-jerky isolation game? Does Ray create inside shots for the C’s back-up bigs because opposing big men have to jump out on Ray’s classic baseline-to-foul line curl cuts?
Is Ray just a more natural passer?
I have no clue. This is mystical basketball stuff—the ability to make others better.
And it’s definitely something to consider if you are one of the proponents of trading Ray Allen.
(Update: And Jeff Clark at CelticsBlog says now is not the right time to deal Ray. Agreed).