Let's get this out of the way right off the bat: if you think Monta Ellis is a superstar quality player and you want him to be the star of your team, you are going to have a lousy team. Case and point: The 2009-10 Golden State Warriors.

This is a hard pill to swallow for Warriors fans. Monta seems like he should be an All-Star. He's scored a lot of points, he has played a ton of minutes. He is all the way back from his injury problems last year, and has regained the blazing quickness that he showed off his first 3 years in the league. Also, he has almost no support. Most nights he has to carry this offense by himself. Right?

I could write a thousand words about why Monta is not the star player your hoping for, but a picture is worth a thousand words

via hoopdata.com

This graph is pretty simple, it's plotting Monta's Offensive Rating (ORtg) on the Y-axis against the percentage of team possessions he is using (USG%) on the X-axis for each of his 5 seasons. What is shows us is that the more of the offensive load Monta has to carry, the less efficient he becomes. This is actually the norm for NBA players, and is pretty intuitive to understand. Economists would call this the law of diminishing returns. If you've read Dean Oliver's seminal Basketball On Paper, there is an entire chapter dedicated to this idea of "Skill Curves" (In the paperback version, it's chapter 19).

However, although this basically happens to everyone, what's important to realize about Monta is that his drop happens at a relatively low USG%.

One common claim is that the reason that Monta's efficiency has dropped off is because he is playing with such bad teammates. The same thing happened to Kobe with the post-Shaq, pre-Kobe Lakers. Right?

via hoopdata.com

Not quite.

Monta is actually a below average offensive player right now, despite his high counting stats, because his efficiency has dropped so far. The prescription for Warriors remains clear: Make Monta do less to help his team. He is taking too many shots, turning the ball over too often, playing too many minutes. He is a very good player, but he is not in the superstar stratosphere. He's not on the level of .

One of the main arguments against lessening Monta's usage is that the Warriors don't have anyone else who can score. This is a pretty argument.

The Warriors actually have several players who can find ways to score, whether it be Biedrins mopping up with high percentage shots, Morrow firing 3's at a high clip, or Maggette making a living at the free throw line. Spreading the ball around can only help the Warriors.

The point of all of this is not to say that Monta Ellis isn't any good. He is quite good. The point is that he should not be relied on like he's a superstar. He isn't, and if the Warriors continue to treat him that way, it would be borderline delusional. Here are the top 10 players in USG% and there ORating:

Probably all of these players are too far down their Skill Curve, and using too many possessions (except Lebron, but he barely qualifies as human). However, the league average ORtg is about 107, so all of these players are above average except Granger, who is pretty close to average, and Ellis, who is far below it.

So, let's all stop lying to ourselves. Monta is a very good player, but he's not a superstar in this league.

Note: The images above were done using Hoopdata.com's AWESOME NBA: Where Motion Happens tool. Read all about it and prepare to waste hours of your life here and here