Disclaimer: If you're not statistically minded (or I would argue open-minded), feel free to ignore my shenanigans.

Last night I made some modifications to my PBP code so that I could look at stats with or without any single player. I'll definitely be posting much more of this type of analysis on my blog, but I figured I would give a taste of it here - starting, of course, with Monta Ellis. The advantage of using play-by-play data is that the "with" or "without" is actually determined at the lineup level. So, even though Ellis only missed a couple of games, I have data for every possession during the season where he was on the bench, too. So, from that perspective, these are pretty unique data you are about to see.  :D 

With no further adieu, here are the ezPM100 ratings for each Warrior when Ellis was "in" or "out" (along with number of possessions):

In/Out NAME POSS EZPM100

in Vladimir Radmanovic 1606 -3.62

out Vladimir Radmanovic 628 3.24

in Stephen Curry 4183 -0.50

out Stephen Curry 684 1.86

in Rodney Carney 451 -6.79

out Rodney Carney 201 -2.28

in Reggie Williams 1975 -0.80

out Reggie Williams 1147 1.80

in Monta Ellis 6343 1.62

in Louis Amundson 933 -2.45

out Louis Amundson 387 -0.03

in Jeremy Lin 188 -3.37

out Jeremy Lin 340 3.49

in Jeff Adrien 226 -0.84

out Jeff Adrien 146 -8.09

in Ekpe Udoh 1432 -5.32

out Ekpe Udoh 544 -1.43

in Dorell Wright 5409 -0.10

out Dorell Wright 785 2.90

in David Lee 4576 -0.68

out David Lee 567 -0.40

in Dan Gadzuric 431 -4.11

out Dan Gadzuric 138 -5.49

in Charlie Bell 142 8.15

out Charlie Bell 160 -3.65

in Brandan Wright 222 -2.11

out Brandan Wright 158 2.99

in Andris Biedrins 2575 -1.62

out Andris Biedrins 177 -4.36

in Al Thornton 319 -4.00

out Al Thornton 252 5.69

in Acie Law 704 -2.46

out Acie Law 481 -2.82

Like I said, I'll be posting more analysis of these data on my blog, but the main thing you can see here is that almost every player on the roster has a better rating when Ellis is OUT. In fact, in many cases, the differences are quite dramatic. It is important to take into account the sample size for each case, and it will also surely be argued that Ellis is often on the bench in garbage time. Even with those caveats, it's hard to imagine such a universal trend not having some meaning. In time, as I look at other players around the league, I will hopefully be able to put this in better perspective.