This is neat, even for non-fans.

[theweaselking]

A detailed analysis of shot locations for two NBA teams, broken down by player, displayed graphically.

Seriously, check it out. (Requires that you enable Javascript)

Comments

[rdmasters.livejournal.com]

Maybe, but looking at the crunched numbers reveals no statistically significant difference between them at the standard 5% confidence level, and barely significant differences at the 1% level. Indeed, the point-per-shot difference is just 2% - an amount that most statisticians would regard as noise, or barely above it.

So why are these 'Sports Analysts' demonstrating such a complete inability to actually understand statistics?

(Gets grumpy about bad statistics almost as often as he does about bad surveys...)

[theweaselking.livejournal.com]

But the images are *pretty*.

[spartonian.livejournal.com]

I doubt any serious statisticians care much for shot distribution in an NBA game anyway...

This is useful to scouts and enjoyable for basketball enthusiasts. It gives good information as to where the players in question shoot consistently.

[miz-geek.livejournal.com]

This isn't a sample of data - it's the actual population - so statistical significance isn't relevant. It's not like the article is using the numbers to claim that the Thunder are going to win the series by 2%.

And the fact that the two best teams in the league seem to be evenly matched overall (that there is very little practical difference between them as a whole) doesn't seem surprising - I don't think that's really the point of the article. The point of the article seems to be to explain how the parts that make up that whole are different within and between the two teams.