I haven't read this entire thread, and I'm just assuming K19's game-by-game breakdown is correct, but there is one serious flaw in the statistic that has been cited (wins against teams that finished 10-5-1 or better).
The problem is that the other team's final record includes the game Bledsoe played in. If he's playing a team that's 10-5 going into the final season and wins, that team ends up 10-6 and isn't included in the stat. If he loses, that team ends up 11-5 and IS included in the final stat. Basically, the stat counts ONLY the games Bledsoe lost against teams that were otherwise 10-5 -- thereby skewing that stat against him. The proper way to look at such a stat would be to remove the games Bledsoe played from the other team's final record, then look at how Bledsoe did against those teams.
How much would the stat change if done properly? If we take K19's list and change the opponents' records to take out the result of each game against Bledsoe, we see that he played 42 games against teams that were otherwise 10-5 or better. (To be more accurate, you'd have to account for the teams Bledsoe played twice in a season, but I'm not doing that here.) He went 13-29 in those games against teams that were otherwise 10-5 or better, for a winning percentage of 31 percent. In their other games, those 42 opponents had a winning percentage of 74 percent (466-164) -- so Bledsoe's teams actually did 5 percent better than the average team did against those opponents.
|
06-14-2005
Linear Mode
