Tuesday, October 28, 2008

A Bold Prediction

I'm going to go out on a limb and say that any model that suggests Derrick Rose will not be as productive an NBA player as Mario Chalmers has some faulty assumptions and probably needs to be chucked and rebuilt from scratch.

Time will tell, I guess, but come on. Statistical models are supposed to deepen (and, yes, even complicate) the more conventional knowledge that we receive from watching games, but not completely contradict what we see.

3 Comments:

Anonymous Anonymous said...

Actually, statistics aren't "supposed" to do anything. If you think a model is flawed, test its ability to predict. Don't just complain because it doesn't satisfy your confirmation bias.

1:42 PM  
Blogger BenGo07 said...

Okay, fair enough. I guess we'll see whether this model's ability to predict confirms my bias.

2:02 PM  
Anonymous bullshooter said...

The problem with those numbers is that Rose's college sample size is much smaller than Chalmer's and therefore the numbers based on that have less power. Rose didn't really come on until the second half of the season last year, so his numbers aren't very representative. Plus he played next to 4 different guys in the preseason. Berri even noted that Oden's numbers were skewed by an injury in his one year. All I'd take out of this is that Chalmers projects to be a good player, not that Rose doesn't.

I wouldn't be surprised if it took Rose a while to get adjusted like it did for CP and DW, but he looked good tonight and I am confident like you are that injury and attitude are the only things that could stop him from being an all-star.

10:45 PM  

Post a Comment

<< Home