DeAndre Jordan
Ersan Ilyasova
Matt Barnes
Jason Kidd
Brandan Wright
It is worth noting that wins produced is a metric of efficiency more than anything else. All of these players are very, very efficient because they have high shooting percentages and low turnover rates. Furthermore, there is a major conceptual shift between "normal" thinking and the way they calculate player efficiency: most casual fans don't penalize a player for missing shots so long as they score points. However, the wins produced model actively does this: it essentially counts missed shots as turnovers (you can argue about this if you like, but there is a certain logic to it: after all, missing a shot and not getting an offensive rebound often start fast breaks and have the same end effect on a possession as a turnover).
Thus, due to the parameters of the model, volume shooters get penalized whereas players that play completely within their means and don't take bad shots/make bad passes get higher scores. The metric doesn't speak to the "skill" of a player, per se, but rather it grades the degree to which what that player does on the court contributes to winning basketball games. Like it or not, a player like Deandre Jordan DOES contribute to winning basketball games because he only takes high percentage shots and doesn't turn the ball over, and those make up for his poor free throw shooting. (In fact, statistically, as long as you shoot over around 53-55% from the free throw line you are positively contributing).
This model isn't saying that Matt Barnes is a better basketball player than, say, Carmelo Anthony - rather, it is saying that Matt Barnes is a more efficient player, which contributes to his team winning games. That is a subtle difference, but an important one.
I know they also gave the Warriors an "F" for drafting Harrison Barnes, and the Pistons a D+ for Andre Drummond.
And they've obviously admitted they were wrong on those counts. There is no model of player performance that won't be wrong sometimes. It isn't possible.
Those models sure do like shooting guards, huh?
Actually, the Wins Produced model generally finds shooting guard to be the least important/productive position on the floor. It tends to like point guards and power forwards.