Page 1 of 1
Looking for Material on:
Posted: Wed Feb 08, 2012 5:56 pm
by beardsnotbombs
Anyone familiar with material on whether/how much a player changing teams affects his statistical outputs? Specifically, discussion on the, "Player A and player B produce similarly, Player A's team is better than Player B's team, so if Player A was put on Player B's, he'd produce more" conversation that seems to pop up too often. I'd be obliged if you could point me in the right direction.
Re: Looking for Material on:
Posted: Wed Feb 08, 2012 6:54 pm
by xkonk
This might fall under the umbrella of diminishing returns if you want to search for it. The only specific numbers I've seen for an overall measure are in Stumbling on Wins, and you can see a chart version here:
http://arturogalletti.wordpress.com/201 ... ing-teams/ .
Re: Looking for Material on:
Posted: Wed Feb 08, 2012 9:52 pm
by Mike G
Every year that I've done comprehensive lists of
most improved players, the bottom of the list -- most dis-improved -- is disproportionately populated by players who have changed teams.
I count 98 players with new teams this year. Here are their unweighted average per36 rates, after adjustment to team rates.
Code: Select all
year mpg Eff% Sco Reb Ast PF Stl TO Blk e484
2011 19.6 .509 11.95 5.97 2.62 3.15 .97 1.74 .59 .59
2012 17.9 .492 11.62 5.68 2.40 3.30 .97 1.88 .58 .50
ratio .91 .97 .97 .95 .91 1.05 1.00 1.09 .99 .84
These are mostly below-average players: avg eW/484 = 1.00
The net eWins for team-changers is -4.5 (relative to 2011), and for those who have stayed put, it's +14.4
It's even worse for above-avg players. Here's the same breakdown for those .90 (last year) or above:
Code: Select all
year mpg Eff% Sco Reb Ast PF Stl TO Blk e484
2011 28.6 .546 16.90 7.07 3.84 3.01 1.07 2.24 .62 1.24
2012 25.7 .500 14.51 6.41 3.25 3.02 1.02 2.12 .64 .90
ratio .90 .91 .86 .91 .85 1.01 .95 .95 1.03 .73
Someone might remind me to do this again at the end of the year. Curious to see whether players tend to adjust eventually.