Tactical Tankfoolery -- When Perverse Incentives Aren't
Posted on Fri 21 March 2014 in Uncategorized by Alex Dewey
A few months ago, Zach Lowe unveiled Mike Zarren's "Wheel" proposal as an alternative to the current draft lottery. In essence, every team's draft order would be set in stone for the next 30 years. This would utterly eliminate any incentive to lose and therefore eliminate any semblance of tanking. Now, there are some problems with this, and with the less-stony updated proposal as Lowe explained it today. But first, let's talk about incentives.
Why incentives? Because something totally mystified Aaron and I when the Wheel first came out. The absolute biggest critique of the Wheel in comments and on Twitter seemed to center on, say, a generational star like Dante Exum choosing his team by waiting for the right draft class. "Oh, I can get to the Lakers if I wait a few years! I'll stay on the amateur circuit until then."
The reason this mystified us is that that scenario_ freaking crazy_.
For several reasons, actually. Unless a prospective rook is genuinely on the fence about something that affects their lifetime earning potential (an upcoming CBA change, whether or not to go to college another year, or if someone has some hidden medical condition where he can only play in Los Angeles), no rational player is going to forgo a huge part of his earning career and his development curve (or risk injury!) to maybe get to a better, more-marketed team. The way the CBA is structured right now is highly unfavorable to productive players on the rookie scale (both in terms of getting market wages and in having free movement between teams), and rational players want to get out of that stage of their careers as soon as they can. What's more, the realities of age mean that the rookie scale is least problematic when a player is 18-20 when they're starting out. Players in that 18-20 age group are far less physically developed and have more uncertainty in their true NBA level; hence, getting less money for their services makes a whole lot of sense for everyone involved. Plus, these players hit free agency just when they're hitting their early physical primes. It works out really well.
On the other hand, if a player waits until he's 22 (no matter how you feel about college ball), then that player is stifling his development as an NBA player and hurting his lifetime earning potential a whole lot, typically. Oh, sure, that player might end up playing for his preferred team, but he's drastically increased the odds that he won't be in the league at all 4 years later years_._ I could elaborate on all of these things, but my point is: I just don't see the Dante-Exum-staying scenario as being a real problem with the Wheel at all, and it totally mystified us that this was even an issue. There would just be too many incentives going the other way for the Exum Apocalypse Scenario to be even 1% of the discussion.
And yet, it struck me: The reason that we're focusing on this ridiculous hypothetical is that the anti-tanking discussion is seemingly built entirely on these worst-case perverse incentives. Present-day tanking itself being just one example. Sure, Lowe may have given us a very nuanced look at the tanking problem, but in the end, he's viewing the problem almost entirely through these absolute worst-case scenarios. Check out how Lowe goes after a couple of alternative scenarios:
• People around the league like the idea of returning to the unweighted lottery, where every lottery team has the same chance of nabbing the no. 1 pick. Go that route, and I’m tanking the hell out of the no. 8 seed and into the lottery every time — and I might even tank my way through Bill Simmons’s Entertaining As Hell Tournament, if that’s what it takes.
• Thinkers have also kicked around ideas that would make getting into the playoffs a more desirable outcome on its own. One idea would be to place 22 teams into the lottery, excluding only the top four seeds in each conference, and to guarantee some juicy picks — perhaps two picks in the 5-10 range — would go to playoff teams. But that would introduce a tank race into the no. 5 spot, and hold the potential for sending multiple impact rookies to teams that are already strong.
Lowe is presenting his critiques almost entirely in terms of the perverse incentives of a) the current system, b) the simpler alternatives, and c) the Wheel. Now grant(land)ed, he's doing an exceptional job of presenting all of these critiques (and the unintended consequences of several systems). But, for his immense holistic understanding of the league and its people, I'd argue that Lowe (and other anti-tanking folks) has fallen into a Venus Fly Trap of reasoning here in fixating entirely on these perverse incentives over everything else.
• • •
The language of economics is really attractive for constructing narratives and for teasing out the good and bad incentives in every system. If you can find the right dataset and the right interpretation, it's powerful as an aid to understanding. But, it's also prone to giving misleading impressions if a small group of incentives (in this case, the draft lottery system) are falsely assumed to be the whole of an agent's decision-making processes.
There's a crucial and subtle and persistent error in reasoning behind most of these anti-tanking articles. Essentially, it's a mass conflation of "having this one perverse incentive" with "being consistently incentivized to the point of the perverse course of action". Why is this an error? Here are dozens of different incentives for an NBA franchise at any given time. "Having an incentive" just means you're pulled in one direction by one hand, while dozens of hands pull you in their own directions. Clearly, the tanking incentive can turn into action, even accounting for organic rebuilding. But that's just one of many incentives driving teams' actions. What about just plain wanting to develop and find a young, non-contending core (hey, Utah)? What about showing that their franchise is respectable, and not just Tank City (*cough* Philly)? What about desperately wanting a 8th-seeded playoff run to show fans more than distant hope (...Sorry, Cavs.)? Those are all real incentives that have manifested in real outcomes over the last few years, but because most of them aren't anything to get outraged about (or, often, even to notice), we don't even bother with them.
If you have some time to kill, listen to some conservative political commentary on poverty some day, say, on talk radio. There are some attractive arguments that they make, after all, and they're at least worth hearing. They usually go something like this: Welfare encourages poverty and unemployment. If you know you'll be taken care of, then you have less incentive to take care of things yourself. And isn't that right? It's good reasoning, it's well-considered thought, and the incentives driving the discussion are pretty much accurate. Economically, everything that they're saying is totally valid, if not exactly sound. And you know what? You can't just wave it away with 'compassion'. There are fundamental economic truths underlying our society, and, according to those economic truths, welfare does encourage poverty in the ways the commentators explain.
Now, a lot of people buy into these arguments, and they're not wholly without merit. The only problem is that - in their treatment of a complex system - these commentators have fixated wholly on the one positive thing on earth (namely, welfare, low-income housing, and food assistance) that's harder to get if you're rich. Never mind the increased incidence of nearly every malady possible for the poor over the rich, the worse access to health care, education, good infrastructure, and financial literacy. Never mind that being poor means you're always one false or unlucky step away from total financial ruin. Never mind that even for the most motivated poor people, these same commentators have already written them off as lazy.
Much of the poverty discussion in this country, unfortunately (and cynically), is essentially to fixate on the one or two mitigating factors like welfare checks, and to wholly ignore the huge structural disadvantages of, say, living in a poor neighborhood without effective transportation. And ignoring the unseen (in this case, the non-financial, mostly) factors is not actually economics - it's intellectual (and possibly moral) negligence wearing the mask of economics.
And that's roughly how I feel when I listen to the anti-tanking discussions, even in their most nuanced form. As much as any red-blooded sports fan, I dislike seeing a slate of NBA games poisoned by three teams in three games that are tanking. It doesn't quite ruin a diehard's night, but it's rough. I hate tanking and everything that goes with it. I hate teams when they are entirely cynical, and tanking is probably basketball at its most cynical. If we can ameliorate tanking itself, or at least flex out the worst offenders off our national television slate, that would be to everyone's benefit.
But when you're having this discussion, you have to be aggressively honest: if you fixate only on the incentive gap in the lottery between the very worst and the 8th-worst team in basketball, and you don't take into account the massive revenue-building incentives to win those extra games on a nightly basis? If you fixate only on the 18-win team with a potential of adding an All-Star instead of the 38-win team that already has one? Then you're doing the Dante Exum thing -- you're the guy who's saying "Well in this exact circumstance, Exum could stay another year and go to the Lakers instead of the Bobcats. That would be awful; this Wheel thing is such a problem."
No. That minor Exum hypothetical (even if he is that good) isn't nearly enough for the league to completely restructure its incentives around for the other 450 players. Who cares - in the grand scheme of things, mind you - if one player harms his own brand and potential while the league around him does what it does well and advances its personnel, its training methods, and its assets without him? Tanking is a complex issue that can ruin a lot of potentially good basketball. But I've not heard one anti-tanking system that really, honestly engages the economics of the situation - despite that I've heard now hundreds of anti-tanking systems that scold teams or diss their incentives. And, until I do hear that sound economic analysis, I'm not sure that I've heard anything but well-intentioned, half-baked noise. And just as we wouldn't want to restructure our system to avoid the Dante Exum Apocalypse, we probably shouldn't do much more than tweak if our biggest problem is that Thursday is a little more depressing for diehards and professionals.
If that all sounds like an unfair statement, then consider that no one has particularly effectively argued whether or not tanking even works (and if it doesn't, well, it's hardly a problem of incentives at all!). And no one is really controlling for intent vs. results. See, for example, the overperforming tank-intended teams like Phoenix and the underperforming, playoff-intended teams like Cleveland and New York. Phoenix was a cynical, evil team until they overperformed and, hey, look, Hornacek is doing awesome! Look the other way! If they didn't have a bunch of uncertainty coins flip their way (i.e. Jeff Hornacek's head coaching ability), they'd be doing things just like Philly and we'd be just as mad. Howard Beck's great article totally deconstructs the general anti-tanking sentiment by focusing on the particular: what's actually going on with the bad teams.
So here's my beef, in a nutshell: If you're going to critique an economic phenomenon, you need at least a solid, cohesive theory of what's actually going on with these struggling teams; the costs, the benefits, and the alternatives. We don't really distinguish who's trying to tank -- just who ends up tanking. We don't distinguish whether tanking is a good thing for the team engaging in it -- just that it ends up happening. We don't really know the basic facts of the situation -- we've just established (quite strongly) that it's incredibly unfortunate when tanking ever happens. In quite a few years, we've seen a lot of anti-tanking proposals (most without mention of potential downsides), but only a few clearly-framed problems, even fewer nuanced explanations of tanking, and, as far as I can tell, nothing that would really sway me in the slightest if I were the commissioner of an 11-digit organization. And, if you're making a serious solution, then that has to be the standard.
Continue reading