LST's playoff projection model
Posted: Tue Mar 04, 2014 5:48 pm
I expect that Tim will be updating his KOM thread now that playoffs seem to be a possibility, but since he and I use different models, I thought I would start my own thread. Interestingly enough, we both have the Wolves and Suns tied for the 8th spot (I also have Memphis tied), but he has them at 48 wins and I have them at 46.
My model treats Dallas, Golden State, Memphis, Phoenix and the Wolves equally in predicting wins and losses. Without getting into nauseating detail, it predicts that all five will win home and away against the league's many doormats, lose home and away to 4 elite teams (Miami, Indy, OkC, and San Anton), and win at home and lose on the road against everyone else. It does not factor in any impact of BTBs.
I thought about including Portland, Houston and especially the Clippers as elite teams, but decided not to. Thus the five teams I am tracking are projected to beat these near-elite teams at home. This assumption comes into play tonight, as the Suns are hosting the Clippers and projected by the model to win, even though Vegas has them as a 4 point underdog (perhaps overvaluing the impact of the Clippers recent roster additions). If the Suns lose, they will fall into 10th place in my model with 45 wins.
Here is the latest projection:
6.Golden State 51 wins
7. Dallas 50 wins
8.Memphis 46 wins
9.Suns 46 wins
9.Wolves 46 wins
The Wolves have a remaining home game with the Suns, and if they win, will hold the tiebreaker against Phoenix. Minnesota has two games left with Memphis, and if they split them, the season series will be 2-2 and the next tiebreaker is record within the conference, where unfortunately Memphis currently hold a 2 1/2 game advantage over us. As I understand it, Memphis also holds the 3-way tiebreaker because of their superior head to head record among all three teams, which is why I have them in the 8th spot above.
I think this model is very realistic because the five contending teams are all fairly equal to me. The fact that the Wolves make up 5 games over the rest of the season reflects how much easier their schedule is compared to the other four teams.
Go Clippers, well, at least tonight!
My model treats Dallas, Golden State, Memphis, Phoenix and the Wolves equally in predicting wins and losses. Without getting into nauseating detail, it predicts that all five will win home and away against the league's many doormats, lose home and away to 4 elite teams (Miami, Indy, OkC, and San Anton), and win at home and lose on the road against everyone else. It does not factor in any impact of BTBs.
I thought about including Portland, Houston and especially the Clippers as elite teams, but decided not to. Thus the five teams I am tracking are projected to beat these near-elite teams at home. This assumption comes into play tonight, as the Suns are hosting the Clippers and projected by the model to win, even though Vegas has them as a 4 point underdog (perhaps overvaluing the impact of the Clippers recent roster additions). If the Suns lose, they will fall into 10th place in my model with 45 wins.
Here is the latest projection:
6.Golden State 51 wins
7. Dallas 50 wins
8.Memphis 46 wins
9.Suns 46 wins
9.Wolves 46 wins
The Wolves have a remaining home game with the Suns, and if they win, will hold the tiebreaker against Phoenix. Minnesota has two games left with Memphis, and if they split them, the season series will be 2-2 and the next tiebreaker is record within the conference, where unfortunately Memphis currently hold a 2 1/2 game advantage over us. As I understand it, Memphis also holds the 3-way tiebreaker because of their superior head to head record among all three teams, which is why I have them in the 8th spot above.
I think this model is very realistic because the five contending teams are all fairly equal to me. The fact that the Wolves make up 5 games over the rest of the season reflects how much easier their schedule is compared to the other four teams.
Go Clippers, well, at least tonight!