RealGM

http://82games.com/0708/0708WASP.HTM

Very nice data set for those who like to play with the numbers.

As always, be careful in evaluating this numbers. Data isn't an easy thing to draw finite conclusions. For example, data can be skewed by limited mins but usually you can compare relative data with in a group.

Example. The GA and Mason numbers fly off the charts but its only based on
18 minutes so while the numbers could hold up with more minutes that would be a very speculative prediction based on the sample size. But it would be worth giving them a few more minutes together to see what happens. Maybe they are a good match.

Also the data is backward looking and only shows to player combinations.

Haywood/Blatche at +19 practically jumps off the page, Hands.

Just looking at that data,

Stevenson/Mason/Jamison/Blatche/Haywood

could be a dominating lineup.

Chocolate City Jordanaire wrote:Haywood/Blatche at +19 practically jumps off the page, Hands.

45 minutes practically jumps off the page. I think the player with the best +- for the first 45 minutes of the season last year was Richie Frahm.

Does anyone know which teams leads the league in bench scoring or where I can find that statistics at?

Interesting to examine, but considering we're a .500 team who scores a lot, it's not surprising the best combo is our current starting lineup.............

Maybe this helps pair guys off the bench, or exclude garbage players completely but I don't see that set of stats doing much more than exposing the obvious.

The stats can be very useful when massaged the right way. Last year, I'd use the player pairs data to eliminate the "Arenas Effect".

Arenas had such a dramatic effect on the team's performance, it was unfair to compare a guy like Deshawn Stevenson, who always played alongside Arenas, to a guy like Daniels, who often played without Arenas.

The fair way to compare the players was to compare Stevenson without Arenas versus Daniels without Arenas, and to compare Stevenson with Arenas verus Danies with Arenas.

floppymoose wrote:-= original quote snipped =-



45 minutes practically jumps off the page. I think the player with the best +- for the first 45 minutes of the season last year was Richie Frahm.

floppymoose, the Warriors looked tremendous beating the Spurs last night, BTW.

As for Frahm, stat enthusiast that I am, the guy sure puts up good numbers for somebody that can't stay in the league. IIRC, he had 30 something in a game for Seattle but barely played again after that.

As for my comment about Blatche/Haywood, I noticed it because Eddie Jordan has been critical of the pairing and says he tries to avoid it. (Yet he hasn't avoided Mason/Young--which he said he would and the player pair data strongly suggests he should avoid). Not that I would ever criticize Eddie Jordan ...

floppy, most fans on this board would like to see more of Blatche with Haywood, but last game it was tried they had a bad minute or two against the Pistons.

nate33 wrote:The stats can be very useful when massaged the right way. Last year, I'd use the player pairs data to eliminate the "Arenas Effect".

Arenas had such a dramatic effect on the team's performance, it was unfair to compare a guy like Deshawn Stevenson, who always played alongside Arenas, to a guy like Daniels, who often played without Arenas.

The fair way to compare the players was to compare Stevenson without Arenas versus Daniels without Arenas, and to compare Stevenson with Arenas verus Danies with Arenas.

nate, do you think those early stats indicate Jamison plays well with everyone on the team?

Chocolate City Jordanaire wrote:-= original quote snipped =-



nate, do you think those early stats indicate Jamison plays well with everyone on the team?

Yes.