Here are some numbers to ponder:
- This season (as of Wednesday night) Monta Ellis averages 21.9 points per game, a mark that ranks 9th in the NBA.
- Carmelo Anthony (again, as of Wednesday night) averages 21.3 points per game, a mark that ranks 12th in the NBA (of all players who played at least 30 games so far this season).
- Last season Ellis ranked 8th in the NBA with a per-game average of 24.1; while Melo averaged 25.6 points per game (a mark that ranked 3rd).
What do all those numbers tell us? If you follow the conventional basketball wisdom, these numbers scream: "Ellis and Melo are great!!" And when a team acquires great players, the team must get better. In other words, the Milwaukee Bucks -- who just acquired Ellis -- must now be better. And we all know that the Knicks have been amazing since Melo came to town.
Okay, unfortunately -- and contrary to this conventional wisdom -- if we delve just a tiny bit (and I mean tiny bit) deeper, we see that there is a reason to think maybe Ellis and Melo ain't so great.Let's start with something obvious. How many points a player scores is determined by the following two factors (and I really hope this is obvious):
- how many shots a player takes
- how efficiently the player shoots (or how often those shots actually go in the basket)
When we look at Monta Ellis, we see that this season he has taken about seven more field goals per 48 minutes than an average shooting guard. So Ellis is "good" at "taking" shots (and "good" and "taking" are in quotes for a reason).
But again, scoring isn't just about shot attempts. Shooting efficiency also matters. And when we consider true shooting percentage (a measure of efficiency that takes into account two-point, three-point, and free throw shots), we see that Ellis in 2011-12 has a mark of 51.6% while an average shooting guard has a mark of 54.3%. So when we look at efficiency, Ellis isn't "good".
Okay, so what matters? Should we consider Ellis "good" because he "takes" shots? Or is it a problem that these shots -- relative to an average player at his position -- are often missing the mark?
Well, let's think about some additional numbers. On Tuesday night -- without Ellis (or Stephen Curry) in the line-up - the Warriors hoisted 91 field goal attempts against the Sacramento Kings. About a month earlier -- with Ellis and Curry in the line-up -- the Warriors attempted 90 field goal attempts against the same Kings. In other words, with and without Ellis (and Curry) the Warriors took about the same number of shots.
And this is not a story confined to just two games. Consider the case of Carmelo Anthony. Melo is another player who tends to be above average with respect to taking shots but is not always above average in terms of shooting efficiency. And just like we saw with Ellis, it doesn't seem like his teams need help "taking" shots.
Melo began the 2010-11 season with the Nuggets. But after 57 games, the Nuggets sent him to the Knicks. With Anthony, Denver averaged 80 field goal attempts per game. Without him Denver averaged 82.2 field goal attempts per game. A similar pattern has been seen this year with the Knicks. With Anthony in the line-up the Knicks have averaged 81.2 field goal attempts per game. Without him, the Knicks are averaging 81.6 field goal attempts per game.
One can see essentially the same story when we look at LeBron James and Allen Iverson (two more players who have historically been "good" at "taking" shots). When LeBron left the Cavaliers the team's shots from the field actually increased (77.9 field goal attempts with LeBron in 2009-10, 81.1 without him in 2010-11). And when Allen Iverson left the Sixers -- the first time - the team's shots from the field stayed at about 78 per game.
If such anecdotes are unconvincing, in Stumbling on Wins we reported a study of how a player's shot attempts related to the shot attempts of their teammates. What we found was that a player's field goal attempts were almost entirely "taken" from his teammates. In other words -- as the above anecdotes indicate -- players do not "create" shots. Shots are merely "taken"; and by "taken" we mean "taken" from the player's teammates.
Given all this, should we be impressed by scoring totals? In the case of LeBron, one should be impressed. Like Ellis, Anthony, and Iverson, LeBron is above average with respect to taking shots. But James is also well above average with respect to getting his shots to go in the basket. And getting those shots to go in the basket is what leads to wins (along with getting rebounds, accumulating assists, etc...). So far this season, LeBron's numbers have been worth more than twelve wins to the Miami Heat (for those who are interested, here is how the number of wins LeBron produces is calculated). No other player in the NBA has produced more this season.
In contrast, Ellis -- despite all that scoring -- only produced about 1.1 wins for the Warriors this season. And as of Wednesday night, Melo has also only produced about one win for the Knicks this season.
Again, wins in the NBA are not about taking shots. Win are about getting shots to go in the basket. And because scoring totals are determined by how many shots a player takes from his teammates, when people focus on scoring totals they can easily get fooled.
This isn't just true for the casual fans. Academic studies for decades have noted how scoring totals dominate player evaluation in the NBA (with respect to free agent salaries, awards, etc...). Despite the problems Anthony and Ellis have with getting the ball to go in the basket, people in the NBA still think these players are immensely valuable.
But when we keep things simple, we see that both players are not really "stars". For teams to win they need to get the ball to go in the basket (and really, shouldn't this be obvious?). This was illustrated by the Knicks on Wednesday night. After a long losing streak, the Knicks blew the Blazers off the floor. How did this happen? One big key was the ability of the Knicks to get the ball to actually go through the net. Melo hit 50% of his shots from the field (including an amazing 50% from beyond the arc). Amare Stoudemire was even better, hitting 80% of his shots from the field; while Steve Novak and J.R. Smith hit 13 of 24 of their three-point shots. If the Knicks could get their shots to go in the basket like this every night, then just about anyone could probably coach this team to a championship.
Unfortunately for the Knicks, this one game was not consistent with the historical performance of these players. In other words, when we look at what both Melo and Stoudemire have done this season, we don't see these players hitting shots at this rate. So although both players have scored at an above average rate this season, neither has accumulated these point totals because they have generally been "good" at getting shots to go in the basket. And that means that neither player has actually been "good" this season.
So next time you hear an "expert" tell you how good a player is because he averages so many points per game, ask yourself: Is this because that player just "takes" shots from his teammates (like Ellis, Anthony, and Iverson)? Or is this because that player gets his shots to go in the basket (like LeBron)? And more importantly, why isn't the "expert" making this distinction?