Thursday, October 23, 2008

3pt Rating Nonsense, Part I

I'm collecting 3pt data because I'm wicked cool, not because I have a player edit addiction. Since 2k is not rating based solely on percentage, we need to figure out their new standards/criteria in order to rate players accurately now.

3PM36 = Threes made per 36 mins, as per basketball-reference.com
3PA36 = Threes attempted per 36 as you probably guessed
3PT% = If you don't know, you're probably lost.
Rate = 2k9 default rating (NOT MY RATING)
JmpAst = % of 2pt + 3pt jumpers that were assisted, as per 82games.com.
3ptAst% = % of 3pt jumpers that were assisted. Not available for all players (hence the need for JmpAst)

Blogger.com by the way is the joy of my life and does not allow for the data to be presented in any other way than mush. Persevere.

EXAMPLE #1
Player 3PM36 3PA36 3PT% Rate JmpAst% 3ptAst%
R. Allen 2.5 6.2 .398 94 70% 80%
S. Stoudamire 2.6 7.6 .341 91 73% 82%
M. Bibby 2.3 6.1 .369 89 65% 92%
E. House 2.8 7.2 .393 87 76% 90%

Ray Allen of course, has a high rating. Why not, hey, the guy is a good shooter and 2k fans cried en masse when Ray didn't have a 90 3pt rating last year. Few people picked the Celtics online last year, meanwhile when Ray had 90+ 3pt rating the Sonics were one of the most popular online picks even though their front court consisted of Nick Collison, Danny Fortson, and your pick of garbage centers.

But should he really that high? 5 pts from the max is a lot for a guy who only hit 40% of his threes. That's not such a huge accomplishment anymore. And he did this with KG and Pierce drawing quite a bit of attention, one would have hoped that no longer being the #1 guy would have raised his percentages substantially... they did not.

Eddie House put up a similar 3pt%, and hit slightly more threes per minute. That he was assisted 10% more on his threes, does that really warrant a 6 point difference given their comparative similarities?

Mike Bibby does not compare favorably to Eddie House, as he shot a lower percentage, made fewer threes, and was even assisted on more threes. Yet he has a higher 3pt rating than House.

Salim Stoudamire, I don't think we need to talk about really. His percentage compared to House's speaks for itself. He is only a career .366 3pt shooter and his career high is .380 so there is obviously no way he should have a higher 3pt rating than House. If anything Salim is still trying to prove he can cut it in the league as a House type of player. And it should be fairly clear Salim does not deserve to be within 3 pts of Allen.

So the question becomes: Which of these players is overrated? Or is House underrated? To answer, we need more data. Tomorrow I'll post a followup with some more data, I just want to collect a wee bit more for discussion purposes.

Thoughts?

5 comments:

Anonymous said...

good job once again, rashidi.

of course it should be allen>house>bibby>stoudamire

only god knows what 2K was thinking...

Mario said...

From personal experience, just a theory :
When you play the games the height-factor comes into play as well, a 6´6 SG with the same ratings as a 6´1 guy will hit more 3s since he´s less bothered by defenders running at him or contesting his shot.

Just a theory, but imo it is definitely a factor in played games, no testing done about simmed games though.

The type of release also makes a huge difference for both your players and CPU players.

Thanks for your good work, greets from germany.

Anonymous said...

now that the season is almost here will you be going through and getting the 12 mans perfect? When do you expect to have the all the teams done because I want to start my chise and your rosters are seriously the best ever.

Anonymous said...

I think that kobe has a lot of problems with regards to tendencies. I think he drives the lane too frequently and takes too many threes. BTW pretty much all the other star plyers are fine in this respect.

pragpro said...

Ratings inflation is always an issue, at the same time the capacity of the player has to be respected. The problem arises even though they haven't proved it in NBA regular season settings, their numbers could be retarded due to their system, playing time, garbage time plays go up on garbage time players. Depending on the situation and who's guarding you will color your 3's %. Only players that avg. 30 mins. plus see the full gamut of situations, IMO.

So I don't think it's so bad to give higher numbers for players who haven't had the chance to play, to me capacity has to be put parallel to history.