There's a big problem with that logic, too. For examle, let's create a hypothetical 4 team league...two teams are good (Teams A & B) and two stink (C & D). If a player on Team A goes nuts in games against C&D and plays OK against B...he has pretty decent stats. This is one reason I shy away from many stats (like +/-) wherein the whole spectum is considered. What's important IMHO is not the games against ALL teams...but the games against playoff caliber teams.
I can dig that. I think I posted those numbers somewhere recently (or at least I was looking at it) -- record against playoff-caliber teams. The drawback is you reduce the sample of games you look at. Another thing I think is worth looking at is +/- in more critical stretches of the game. For example, the Blazers all season have been very good in 4th quarters, and the Rockets much less so. This was the case 2 seasons ago, to perhaps a lesser degree, when we matched up against the Jazz. Most analysts that picked the Rockets did so based on our supposed edge in experience and better record. But the Jazz were a better executing team in fourth quarters for the most part, and in the playoffs being able to out-execute the other team in crunch time becomes much more important. <hr> I calculated some splits for the Rockets and Blazers -- looking at how they performed against playoff teams, top 16 teams based on b-r's simple ranking system (SRS), top 16 teams in offensive efficiency, and top 16 teams in defensive efficiency. A lot of overlap between playoff teams and top SRS teams. Rockets win loss off def net vs playoffs 23 18 107.6 106.3 +1.3 vs good teams 23 19 107.3 106.0 +1.2 vs off 26 17 109.3 105.9 +3.4 vs def 23 17 107.7 106.2 +1.5 Blazers win loss off def net vs playoffs 22 19 112.7 111.0 +1.7 vs good teams 26 17 113.6 112.2 +1.4 vs off 26 18 114.1 112.6 +1.5 vs def 22 18 112.0 110.1 +1.9 One thing going for the Rockets is they did pretty well against above average offensive teams (26-17 win-loss record, +3.4 net efficiency). In a match of really offense (Blazers) against really good defense (Rockets), maybe we have an edge.
I like to look at Sagarin's ratings vs. top 10 and top 16 teams as a measure of how competitive a team is against quality opponents. You'll note the rockets are one of the best of the non-favorites, and better than portland: http://www.usatoday.com/sports/sagarin/nba0809.htm
I'm surprised there's such a difference between W-L versus top 16 according to his rankings, and looking at it based on SRS. Blazers were 26-17 against top 16 SRS teams. Actually, when I look at his rankings, his top 16 teams are identical to the top 16 by SRS. I doubled checked, and the numbers I posted are correct -- Portland was 26-16 against those teams. I don't know why he's reporting they were 21-23.
According to computer simulations, we had a better chance to be the 4th seat than the 5th seat. We can throw that right out of the window now. Somebody should do a stat analysis of the various prediction models
I think both Morey and Hollinger were saying that 5th seed was the most likely outcome before our last game.
Before the last game that is. It wasn't hard to predict - we were on the road, both Spurs and Blazers were at home. But if you go back a few games, I believe the prediction by computer simulation was Rox at 4th seed.
simulations dont take into account that the nuggets could afford to blow their last game, and portland couldnt.
