No man is an island…with the possible exception of Monta Ellis.
We can say that because we have the data to prove it. Synergy Sports has been tracking NBA offenses since 2004. According to their database, NBA teams have utilized isolation sets on 12 percent of their offensive plays during that time period.
Keep that 12 percent figure in mind as you digest the following information: Ellis attacked on offense from isolation sets on a whopping 22.6 percent of his possessions as a Warrior last year—nearly twice the league average.
There are a couple of obvious caveats that must be mentioned with that information. First, Ellis only played 37 games for Golden State last season, so the sample is small. But, we know from looking at past seasons that Ellis has always been a very heavy isolation attacker. In 2009-10, his offense came from isolation plays nearly 28 percent of the time. In 2010-2011, his rate was 24 percent. In every recent season, Ellis was used with great frequency in isolation sets.
And sure, those isolation plays are more likely to yield a highlight than most. Ellis was known for the spectacular play as a Warrior, and most of his SportsCenter moments came from one-on-one situations. Isolation plays made Ellis a fan favorite in the Bay Area.
Here’s the thing, though: Isolation plays are the worst way a team can use an offensive possession.
A team running an isolation play can expect, on average, .78 points per possession. No other possession type yields a lower result.
So last year, the Warriors used Ellis almost a quarter of the time in those situations. That’s an illogical strategy to begin with. But what’s even worse is the fact that Ellis was actually less efficient than the league average in isolation situations.
Remember, the NBA as a whole averages .78 points per possession in isolation sets. Ellis averaged .73 points per possession on 34 percent shooting in isolations last year. That's not good.
For his career, he's been slightly better than he was last year as an isolation player—near the league average. But it's very interesting to note that after he joined the Bucks last season, he was even worse than he was as a Warrior. He averaged just .70 points per isolation play for Milwaukee.
Here, it might be fair to argue that Ellis was forced to finish broken possessions and take tough shots when the shot clock was winding down—those factors definitely contribute to inefficiency. But it's worth noting that as an overall offensive player, factoring in every type of offensive situation, Ellis ranked in the NBA's bottom half as a Warrior last year with a points per play average of .89.
Ellis was much more efficient as a spot-up shooter and, surprisingly, in the post. Of course, those types of possessions have always yielded higher-than-average points per play figures for all players—not just Ellis.
Expanding the analysis to cover the team as a whole, the Warriors were very near the league average in isolation sets last year—both in terms of usage frequency and effectiveness. That means that despite Ellis’ unique inefficiency, the rest of the Warriors were far enough above average in isolation situations that the team was halfway decent in isolation plays overall.
The final upshot: Monta Ellis was used twice as often in isolation plays as the rest of the league and he was a below-average player in those situations.
Where’s the sense in that?
Fortunately, the Warriors came to their senses and moved Ellis last year. The move looked even smarter as time passed because, as I mentioned earlier, Ellis played even worse for Milwaukee. By simply subtracting his inefficient offensive game from the team, the Warriors improved their overall offensive efficiency.
When you also consider that Ellis was a lazy defensive player who rarely showed any commitment or effort, the Warriors’ decision to jettison him to Milwaukee only gets better.
And before Ellis apologists call me unfair for killing Ellis' defense, let's take a look at the facts. Last year, Ellis was in the bottom quarter of all defensive players in total efficiency, according to Synergy. And ESPN's John Hollinger has repeatedly mentioned that Ellis is a very poor defender. Hollinger noted:
The bigger issue right now is that he just doesn't try very hard—as vulnerable as he was on the post, he also was routinely beaten off the dribble by slower players and often didn't get in anything even resembling a defensive stance.
So, statistically, Ellis was a poor defender. He also failed the eye test from an effort perspective. But this is an article about Ellis and isolation offense. So, back to the point...
Overall, there's just no getting around the fact that Ellis (especially when cast as the primary scoring option) was an inefficient offensive player for the Warriors. When his offensive efficiency rates are measured against the Warriors' team numbers, it's clear that he actually dragged down the team's offensive effectiveness.
The evidence is overwhelming.
Part of the blame belongs to the coaching staff for using Ellis inefficiently, but just as much rests on Ellis himself for stubbornly continuing to attack when better options were available. After all, in the years since Baron Davis left, Ellis has played the same inefficient way for three different coaching staffs.
With a largely revamped roster, the Warriors’ offense should look vastly different going forward. Andrew Bogut and David Lee are excellent passers out of the post, which should yield an even larger number of open spot-up jumpers from the Warriors’ stable of young gunners. That will be an undeniably positive change because the Warriors were the league’s most efficient spot-up shooting offense last year.
Monta Ellis may be an island unto himself, but the numbers clearly show that the Warriors were wise to vote him off of theirs.