Once again, I'm using the python script I wrote to parse the play-by-play data from our games to help us analyze our team better. LVBruin was the first to come up with the idea, and I expect he'll provide a more detailed set of data when he has time. In the meanwhile, you can check out the individual +/- for each lineup after the jump. Does anyone know why there are so many errors in the substitution data though? Sometimes the play-by-play will say a player has been subbed out, while no corresponding sub in has been made. I can't get detailed play-by-plays from anywhere else except our athletics website, and if it has errors, it's very problematic for me to go correct the data.

Below are the +/- for each possible rotation that Howland played against Cal and Stanford. Only rotations that played at least 2 minutes together are shown.

vs Cal:
Honeycutt, Jones, Lee, Nelson, Stover: 1.16, 10, 8.6 minutes
Honeycutt, Jones, Lane, Lee, Nelson: -0.13, -1, 7.68 minutes
Anderson, Honeycutt, Jones, Lane, Nelson: -2.71, -8, 2.95 minutes
Jones, Lamb, Lee, Nelson, Stover: -1.83, -4, 2.18 minutes
Anderson, Honeycutt, Jones, Lee, Nelson: 2.35, 5, 2.13 minutes

vs Stanford:
Honeycutt, Jones, Lee, Nelson, Stover: 0.27, 3, 11.32 minutes
Anderson, Honeycutt, Jones, Lee, Nelson: 0.73, 6, 8.23 minutes
Anderson, Honeycutt, Lane, Lee, Stover: 0.76, 4, 5.27 minutes
Honeycutt, Jones, Lane, Lee, Nelson: -0.98, -4, 4.1 minutes
Anderson, Honeycutt, Lamb, Lee, Stover: 1.63, 4, 2.45 minutes

As we have been observing lately, the 3-guard lineup of Anderson, Jones, and Lee tends to be effective, especially with Honeycutt at the 4. Would anyone be interested in seeing aggregate +/- data for pac-10 conference games this season?