Comments on: Estimating team chemistry http://arbitrarian.wordpress.com/2008/07/14/estimating-team-chemistry/ Anything but arbitrary. Tue, 23 Jun 2009 01:22:44 +0000 http://wordpress.com/ hourly 1 By: d sparks http://arbitrarian.wordpress.com/2008/07/14/estimating-team-chemistry/#comment-393 d sparks Tue, 15 Jul 2008 13:21:31 +0000 http://arbitrarian.wordpress.com/?p=208#comment-393 Q: Certainly it is useful to distinguish between scorer types, and adding in three point attempts is something I have considered doing. The problem is that it shifts almost all guard-types waayyy over to the Pure Perimeter position--very few players are left as Scoring types at all. The addition of three-pointers only seems to reinforce differences between "bigs" and "guards" rather than distinguishing amongst scorers very well. Also, FWIW, on the SPI graphic, which has a lot more nuance than these seven arbitrary categories, you can see a small difference between for example, Rice (more on the perimeter side) and Wilkins (more on the interior side), which perhaps attests to their primary function as scorers, and secondary nature as a particular type of scorer. As for the three player sets, I did run that regression, and I might post it, but I didn't find the results all that interesting, to tell you the truth... Thanks for your comments, per normal. Q: Certainly it is useful to distinguish between scorer types, and adding in three point attempts is something I have considered doing. The problem is that it shifts almost all guard-types waayyy over to the Pure Perimeter position–very few players are left as Scoring types at all. The addition of three-pointers only seems to reinforce differences between “bigs” and “guards” rather than distinguishing amongst scorers very well.

Also, FWIW, on the SPI graphic, which has a lot more nuance than these seven arbitrary categories, you can see a small difference between for example, Rice (more on the perimeter side) and Wilkins (more on the interior side), which perhaps attests to their primary function as scorers, and secondary nature as a particular type of scorer.

As for the three player sets, I did run that regression, and I might post it, but I didn’t find the results all that interesting, to tell you the truth… Thanks for your comments, per normal.

]]> By: Q http://arbitrarian.wordpress.com/2008/07/14/estimating-team-chemistry/#comment-392 Q Mon, 14 Jul 2008 23:53:43 +0000 http://arbitrarian.wordpress.com/?p=208#comment-392 Great work as usual. Two questions arise for me: 1) In looking at this, I also wonder if it would be useful to distinguish between types of pure scorers. For example, Glen Rice and Dominique Wilkins were very different types of pure scorer (3 pt threat vs. slasher). Do you think there would be much difference in value between the two? Could they be distinguished by adding 3pa to the equations? 2) Since conventional wisdom holds that you need three impact players to win, I also wonder about the effects of adding a third player to the combinations (aside from adding more work for you). ;) For example, scorer's opposite (Detlef Shrempf/Kareem) and (Shawn Kemp/AC Green) might not be very good pairs, but as a trio with a pure perimeter player (Gary Payton, Magic Johnson) they can be extremely successful. I suppose in each case there was a strong pair anyway, pp/ii or pp/pi though.... A lot to chew on here... Great work as usual.

Two questions arise for me:

1) In looking at this, I also wonder if it would be useful to distinguish between types of pure scorers. For example, Glen Rice and Dominique Wilkins were very different types of pure scorer (3 pt threat vs. slasher). Do you think there would be much difference in value between the two? Could they be distinguished by adding 3pa to the equations?

2) Since conventional wisdom holds that you need three impact players to win, I also wonder about the effects of adding a third player to the combinations (aside from adding more work for you). ;) For example, scorer’s opposite (Detlef Shrempf/Kareem) and (Shawn Kemp/AC Green) might not be very good pairs, but as a trio with a pure perimeter player (Gary Payton, Magic Johnson) they can be extremely successful.

I suppose in each case there was a strong pair anyway, pp/ii or pp/pi though….

A lot to chew on here…

]]>