I followed NBA free agency this summer with great interest. Beyond the Sixer moves, what most intrigued me was the significant number of players who signed for far more or less than seemed justified to me by their play. When I mentioned this to friends, the common response was that I was just a nerdy blogger eating Cheetos in my Mom’s basement, and the folks doing the signing were big-time professional basketball experts. Well, they didn’t say it quite like that; my friends are nice! Still, they questioned whether my faith in on-off stats, which after all can be easily influenced by luck, might be misplaced. Perhaps I was overrating high-efficiency, low-volume defensive players and underrating bucket-getters, etc. It’s a fair question.
So, let’s run an experiment! It’ll just be one season, and one group of guys. But, still, it’s better than nothing. And, after all, the teams put their money where their mouth is, why shouldn’t I?
So, here’s my comparison list. The list contains a pair of centers, 5 pairs of wings, and a pair of point guards. I made the list two weeks into free agency and only made adjustments to deal with surprises like the Bjelica thing. Each list contains only players signed in free agency this offseason, and doesn’t contain max players or players who were perceived as signing for a big discount to stay where they were or to chase a ring (e.g. I left off Boogie Cousins). The two groups are similar in age and experience; I made sure not to pull some trick like comparing pro-signed players being paid for their long-run potential to old one-shot mercenaries, or anything like that. Let me add that the Professional picks are not a bunch of guys signed by awful franchises like the Kings. They are from San Antonio (perhaps the best franchise of all), the Clippers (overseen now by Jerry West, among the best executives ever), Indiana (very highly regarded), Chicago (mediocre but not terrible), Dallas (very good franchise), the Knicks (bad), and Utah (excellent). Overall, an above-average group of franchises. And yet I claim they made a set of obviously-erroneous decisions.
In each case the Professional pick makes more money, in some cases vastly more money. In the aggregate the Professional picks make almost twice as much. Will they be twice as good?
Professional Picks vs. Nerd Picks
- C DeAndre Jordan $24M vs. Kyle O’Quinn $4.5M
- Wing Zach LaVine $19.5M vs. Tyreke Evans $12M
- Wing Avery Bradley $10M vs. Kyle Anderson $9M
- Wing Doug McDermott $7M vs. Nemanja Bjelica $6.5M
- Wing Marco Belinelli $6M vs. Luc Richard Mbah a Moute $4.3M
- Wing Mario Hezonja $6.5M vs. Yogi Ferrell $3M
- PG Dante Exum $11M vs. Fred VanVleet $9M
If Jamal Crawford signs somewhere as expected, we can add another pair; I’m quite sure I’ll be able to find someone who signed for as little as Jamal will and who will play better than he does!
How will we judge the experiment? My own preference would be to use RAPM, the adjusted plus-minus statistic. As mentioned above, there’s luck in RAPM, but, first, across so many players luck will probably wash out, and, second, for an experiment like this luck is a lot more tolerable than bias. I can’t see any way the on-off numbers could be meaningfully biased toward some players and against others. Almost all these guys are joining new teams, each of whom expect them to fit well and fill a hole; there shouldn’t be anyone who plays great but hurts their team because they’re stuck in an untenable situation.
Now, if the professionals got it right, the best predictor of performance here should be, more or less, the salaries players signed for. I say “more or less” because of subtleties like the fact that not all the contracts are one-year deals. But basically, guys should get paid for helping their team win. Alternatively, it might be the case that the truth is halfway between the nerd view and the pro view; i.e. the two groups might play about equally well. That, of course, would be intensely embarrassing to the pros; after all, they paid their group far, far more, so if they don’t get any additional production out of them, that’s — what is it the kids say? — oh, yeah, an “epic fail.” And of course if the Nerd All-Stars actually play better than the guys that got all the money, that’ll be even more, um, epic of an, um, fail.
My view is that the truth is much closer to the nerd opinion than the professional take. So, let’s find out. At the end of the season we can see how these players did and see if the Pros or the Nerds are closer. As I say, I’ll tend to focus on adjusted plus-minus as a measure of performance, but if folks think that adjusted +/- is somehow biased toward guys like Tyreke and against guys like DeAndre, please go ahead and suggest other measures of performance we can use for the final evaluation, and also please try to explain what the bias could possibly be!