Throughout the summer, I've been firmly planted on the negative side of the trade Iguodala for Melo debate. My logic has always been when you take the complete game into account, Melo is probably a slightly better player, but the difference between the two on the floor isn't worth the difference in salaries. I'd like to test the theory a bit, today.
Here's the theory I'm operating on: Iguodala is a better defender and a better playmaker than Melo. Much better. Melo is a better scorer than Iguodala. Efficiency-wise, the gap between the two isn't that great at all, but when you take into account that Anthony has proven he's more than capable of keeping his efficiency at decent levels while accounting for a huge percentage of his team's offense, the gap grows quite wide. The question is, how wide is that gap, and would it make up for Iguodala's distinct advantages in other areas of the game.
I thought of one way we could try to quantify the scoring disparity between the two. What if we could take Iguodala away from the Sixers and sub in Anthony, using last year's statistics?
We're going to have to make some crazy assumptions to get some workable numbers:
- Melo would use the same percentage of possessions in the Sixers offense.
- The Sixers would play at the same pace with Melo instead of Iguodala.
- Melo would play the same number of possessions as Iguodala played last season.
Basically, what we're trying to do here is normalize the numbers so we can wind up comparing like items. In reality, Iguodala played over 500 more minutes than Anthony, and nearly 1,000 more possessions. (Durability is another great advantage for AI9 over CA, but for the purposes of this exercise, we'll ignore it).
Here's the exercise, we take the Sixers team stats, remove Iguodala from the equation, figure out team per-possession averages. Then we extrapolate Melo's production to the number of possessions Iguodala played for the Sixers, then we figure out how many possession would be left over for the rest of the Sixers roster, and we come up with team stats to compare with the actual stats from 2009-2010.
As a team in '09-10, the Sixers had 7,560 possessions and scored 8,014 points. That's your baseline. The number of possessions will not change, but the points scored will.
Here are the raw numbers:
- Possessions played: 6,114
- Field Goals (m/a): 496/1120
- 3PT (m/a): 94/330
- FT (m/a): 315/430
- Turnovers: 218
- Points: 1,401
- Possessions played: 6,114
- Field Goals (m/a): 811/1,771
- 3PT (m/a): 70/221
- FT (m/a): 599/722
- Turnovers: 246
- Points: 2,291
I'm going to jump right to the results here, because they're significant and I don't want to bore everyone with the math behind them (if you care about the math, you can check out the explanation at the bottom
Keep in mind, we're assuming Melo plays at the same level, with less talent around him. We're also assuming his supporting cast plays at the same level, which I consider a stretch when you factor in Iguodala's playmaking abilities. That being said, these numbers are telling us adding Melo to last year's team would've moved them from the 20th most-efficient offensive team to 10th. Over 3 points-per-100-possessions is extremely significant.
Now, would I expect this kind of jump if the Sixers somehow made the move? I don't know. Collins certainly has a history with high-usage wings (Jordan and Grant Hill). In some ways, the Sixers might even be a better environment for Melo's scoring skills. Jrue and Turner are both willing passers, something Melo hasn't really had in Denver since Andre Miller left (Chauncey has always been too much shoot-first for me). Plus, if Melo is taking shots away from the right people (Hawes, Brand, maybe Thad), his positive effect would be even greater.
The one question I simply can't answer is how much this trade off would cost on the defensive end. If you're gaining 3.4 points in OFR and losing 2.4 on the defensive end, it's probably not a deal that's worth making. I'm also not sure how you put a dollar value on this type of improvement, especially when you're talking about a young team coming off a 27-win season.
If I had my druthers, I'm still in the stick with Iguodala camp, but I can see some logic behind trading for Melo. I've been swayed a bit when you look at the overall effect he could have on the entire team's offensive efficiency. I shudder to think of the results an exercise like this would yield if I plugged a true superstar into the equation, like Kevin Durant. Thoughts in the comments, as usual (the math is sort of explained below, but it's late so if you care and have any questions, let me know).
The key to the math behind this post was using plays rather than possessions. A play being a field goal attempt, a turnover, or a trip to the foul line (this is because offensive rebounds throw off the possession figures), here's the formula I used for plays: FGA + (FTA * 0.44) + TOV = plays. Melo had 2,334 plays for 2,291 points in 6,114 possessions played (after I extrapolated his stats to equal Iguodala's minutes/possessions/plays). Iguodala had 1,527 plays for 1,401 points in 6,114 possessions played. The entire Sixers team had 8,693 plays for 8,014 points in 7,560 possessions played.
As I said above, the difference between plays and possessions is offensive rebounds (and the math involved in calculating free throws), so adding Melo in for Iguodala would also add to the number of plays, because he was a much better rebounder on the offensive glass.
To make a long story short, what we're looking for is a total for the number of points scored on 8,810 plays (the estimated number of plays the team would have in 7,560 possessions. I used a ratio of OREB * 1.20 to keep the ratio in line with '09 results, and added in Melo's 97 extra OREBs over Iguodala's total.). We already know Carmelo is responsible for 2,334 plays and 2,291 points, leaving 6,476 plays for the rest of the roster.
From this point, I used ratios to get the numbers for the entire team. I left the offensive rebounds static, and used the ratios for FGA, FTA and TOV. Then I added Melo's numbers to the numbers for the rest of the team to come up with the chart above.