In "Every Play Counts," Kevin Pelton focuses on one player, team or matchup in a single game, looking to explain how and why they succeed or fail. Naturally, one game isn't everything, but the results can be illuminating.
There are plenty of NBA players who have had better seasons than Monta Ellis. However, the Golden State Warriors' guard may be having as interesting a campaign as anyone in the league. Ellis started training camp by questioning his ability to play alongside rookie Stephen Curry, and was supposedly on the trading block. But when Golden State dealt Stephen Jackson and was hit with an unprecedented number of long-term injuries, Ellis suddenly become the NBA's marathon man. He is averaging 41.8 minutes per game, second in the league, and has played a full 48 minutes (53, in one case) a remarkable 10 times.
The heavy workload, combined with the Warriors' fast pace, has produced some odd statistics. Ellis ranks sixth in the league at 26.5 points per game, third in steals (2.2) and 15th in assists (5.4). Those numbers made Ellis a candidate for the All-Star Game, but others are more dubious. Ellis averages a league-leading 4.1 turnovers and rates by WARP as a below-average player on a per-minute basis, though the value he provides by being on the floor means he has been worth nearly three wins over a replacement-level player. The most bizarre stat of all is this: The Warriors have been an incredible 14.0 points worse per 100 possessions with Ellis on the floor, per BasketballValue.com.
The combination means Ellis is worthy of more observation, so I decided to make him the subject of my first Every Play Counts since the early days of the season. This plan was actually hatched several weeks ago, so the game I broke down was one I've already written about--when the Warriors lost to the Milwaukee Bucks at home on Jan. 15 and finished the game with Curry staying on the floor with six fouls because no other players were available. Both guards played all 48 minutes, making this a good case study for Ellis' heavy usage this season.
First, I wanted to track Ellis' involvement in the Golden State offense. On the season, he's averaging 29.2 possessions used per game, which is a close second in the league to Denver's Carmelo Anthony (29.3). His 30.0 percent usage rate is high, but again the per-game statistics are inflated by pace and minutes. Using Bob Chaikin's formula for touches estimates 61.4 touches per game for Ellis, 11th in the league and third among non-point guards behind LeBron James.
For this game, I tracked Ellis' touches--counting only meaningful opportunities to score in the half-court offense. I came up with 59, pretty similar to the calculation and more than one every other possession. Beyond that, Ellis brought the ball up 28 times--a little more than one every four possessions as he shares the role of point guard to some extent with Curry. (One thing I failed to note before specifically watching--Corey Maggette brings the ball up quite a bit for the Warriors, nearly as much as Ellis.)
No matter how it's measured, suffice it to say that Ellis is expending a lot of energy at the offensive end. His level of effort is not commensurate on defense. I tracked a defensive box score for Ellis, but this is one of those situations where it's not telling. He allowed four made field goals and forced 4.5 misses while sending an opponent to the line once, but on multiple other occasions Ellis' lax defense allowed his man (usually Charlie Bell) to get into the paint and draw a foul from another Warrior.
Don Nelson lightened Ellis' load by putting him on Bell or the weakest perimeter scorer on the floor for the Bucks. For some players, like Kobe Bryant or LeBron James, that would have been a license to roam and offer double-team help or play the passing lanes. Ellis is very good at generating steals overall, but I didn't see much of that from him in this game. He mostly floated defensively, allowing Bell to beat him backdoor on multiple occasions. When he did go to help, he tended to overcommit and give up open looks on the perimeter.
The thing is, I actually think Ellis can be a pretty good defender (and this is the key here) when motivated. He actually first caught my eye as a rookie by the way he played Ray Allen defensively. Ellis uses his strength well against bigger players, which is why I'm a believer that he can be part of a successful backcourt with the smaller Curry (who got to spend his 48-minute night chasing around a fresher Brandon Jennings). We've seen that at times in 2009-10, especially early in the season. That skill was not on display in this game, not even in the proverbial "spurts."
That all said, the plus-minus numbers tell an unexpected story about Ellis' impact above and beyond the fact that it is negative. Golden State has been only slightly worse defensively with Ellis on the floor, but dramatically less successful at the offensive end. Saying this doesn't square with Ellis' reputation would be an understatement, and even his less impressive stats hardly suggest that he is such a drain on the Warriors' offense.
In watching him play, I saw no particular evidence of problems Ellis might be creating on offense. Two things stood out. First, I was pleasantly surprised by his ability to distribute the ball in the paint. He had eight assists in the game, but the quality was more impressive than the quantity, as Ellis drew the defense and hit teammates at the rim in position for easy buckets. According to 82games.com, 44.5 percent of Ellis' assist this season have resulted in scores in the paint, which is slightly above average.
Less impressive was Ellis' shot selection on the perimeter. He launched several contested jumpshots, most of them long twos. While the degree of difficulty might make those shots look good when they go in, they are the basketball equivalent of hitting on 17. Hoopdata.com shows Ellis attempting 7.1 shots per game from 16-23 feet, the league's fourth highest average, and making them at a below-average 38.0 percent clip. These long twos help offset the fact that Ellis is also fourth in the league in total attempts at the rim and do as much to explain his below-average 52.7 percent True Shooting Percentage as his role in the Golden State offense, as I see it.
As for the plus-minus, I'm inclined to write it off as a fluke. The flip side of Ellis playing so many minutes is he simply hasn't been off the floor that much--393 minutes all season, and those haven't necessarily randomly distributed. Funny things can happy in such a small period of time, like the Warriors shooting the ball extremely well. Their differential with Ellis on the floor (-6.7 points per 100 possessions) seems much more representative of their true ability than what they've done with him on the bench (+7.3 points per 100 possessions).
Even in the midst of a season where circumstances and Nelson's whimsy have forced him to take on a relatively unique role in the NBA, I think we can cut through the context and fairly assess his value. Ellis has not played as well as he did during his last healthy season, 2007-08, and may be somewhat better suited for a secondary role in the offense than the go-to status he currently holds. Playing so many minutes seems to be responsible for his drop off in rebounding percentage as well as some of his half-hearted defense. Ellis isn't an All-Star because he's scoring 26 points per game, but nor is he the detriment his plus-minus and poor efficiency would seem to indicate. Look for him to return to the middle ground as an above-average starter whenever normalcy is restored in the Bay Area.
Follow Kevin on Twitter at @kpelton.
Kevin Pelton is an author of Basketball Prospectus.
You can contact Kevin by clicking here or click here to see Kevin's other articles.