Grantland's Kirk Goldsberry has given us a chance to stop overreacting to a bad preseason performance by coming up with a new metric to help evaluate scorers. It's called Shotscore, and it's essentially the number of points a player scores based on the shots they take compared to the number of points an average NBA player would take on those shots.
Curry is a better perimeter shooter than both James and Durant, but he can't match their abilities near the basket, where he's actually a below-average NBA scorer. Still, thanks to his insane jump shot, Curry accumulates points at unusually effective rates.
Read the whole article here.
The nice thing about this statistic is that it combines usage and efficiency in a non-trivial way. It's a counting shot in that it rewards high usage, so long as that usage comes with high efficiency. If you get your extra points by being hyper-efficiecnt, or by taking more shots with slightly-above-average efficiency, this stat still rewards you.
That being said, I'm not sure that it's the be-all, end-all scoring stat, or even better than Wages of Wins' "Net points" evaluation. The problem is that, as I understand it, Shotscore rewards a player who's a slightly above average midrange shooter for that shot, while punishing a player who's a slightly below average at-the-rim shooter for his layup attempts - but the latter player might still be more efficient, on average, and you'd probably rather have the slightly-below-average layup attempt than the slightly-above-average midrange jumper.
In other words, this stat basically normalizes to, "Given your shot distribution, how much do you help your team?" when the simple truth is that many players can help their team more by changing their shot distribution. (eg, Lee is better when he's not taking too many long jumpers, even if he's well-above-average at making them). Eliminating bad shots is one of the best things a player can do to help their team win, but this stat takes shot selection as a given.
On the other hand, this shot says Curry is #3, after Durant and Lebron, so, uh, BEST STAT EVER! SHOTSCORE FOR LIFE!
That being said, Goldsberry's shot chart for Curry makes clear how much room for improvement there is in Curry's game. He's a below-average rim scorer, and if the rumors that he's been working on that part of his game are true, we could see some real improvement from him this season, which would be scary for Warrior opponents. (KD, Curry's coming for you!)