HartfordWhalers wrote:2011 Dallas SRS for entire season (8th)
Amount of games Dirk missed 9.
Therefore we should take Dallas's SRS for just those 73 games Dirk played in and compare it with the other 29 teams in the league ignoring that those other teams had injures also.
Which "other teams" are you talking about? And yes, the Mavericks with Nowitzki had a higher SRS than 25 other teams in the league with their respective best player playing as well. The same goes for the Heat with Wade and O'Neal playing. And you better not ask for the Pistons with Rasheed Wallace. ;)
HartfordWhalers wrote:That doesn't really seem like an unbiased comparison to me.
What bias is included? Did the Mavericks in 2011 play the playoffs with Nowitzki or without him?
HartfordWhalers wrote:The argument is not just do you need good players, you need one of the best players.
No, good players is enough here, because, as I pointed out (and you seem to agree with) the MVP voting is not a sufficient method to identify the best players.
HartfordWhalers wrote:So, pick your metric.
I actually use a merged metric of SPM and RAPM, where I use a regression analysis on the team's game-by-game performance in order to calculate the coefficients needed. The used RAPM is 1yr no-prior informed, the SPM is my own boxscore-based metric (quite similar to what is still listed on the blog).
HartfordWhalers wrote:APM of the single best player on a team actually does better than SRS of a whole team (unless maybe you do some weird manipulations) at pinpointing each season's winner -- 10 of the last 20 champs have had the single highest APM in the league. The only teams without a top 5 APM for instance where Jordan's Bulls (when Pippen had a higher APM and was still at 6 for one year) and the Lakers (when Odom was 6th and one year when Kobe was highest at 18). {To me those examples show some of the limitations of APM but that is a really separate discussion; the more you look at different metrics the more it seems like it should be obvious that you need not just your 'good' player but a true 'top' player.
I don't think that you actually know what APM even means ... but it is at least a start. :)
And adjusting for the fact that a team plays worse while the best player/s sit/s out due to injuries is hardly a "weird manipulation". ;)