What I Got Wrong About Tanking

For a long stretch I held what I considered the sophisticated position on tanking. Deliberately losing to improve draft position is unpleasant to watch but structurally rational, I argued, and the people objecting were confusing sport with morality. The incentives were built by the league. Teams responding to them were behaving sensibly. Blame the rules, not the players of the game.

I no longer believe this, and the reasons I abandoned it are not the ones its critics usually offer.

The Argument I Used to Make

The logic went like this. In a sport where a single elite player transforms a franchise, and where those players enter almost exclusively through the draft, a team without one has essentially two options: spend years as a mediocre side that never picks high enough to acquire one, or accept a short period of deliberate awfulness in exchange for the picks that provide a genuine chance.

Framed that way, the second option is obviously correct. The alternative is a permanent middle, which is the worst place a franchise can be: too good to draft well, not good enough to compete. Every argument against tanking seemed to me to be an argument for permanent mediocrity dressed up as integrity.

I still think that reasoning is internally sound. What I got wrong was everything around it.

The First Mistake: It Does Not Work Very Often

The empirical case for tanking is much weaker than its advocates, myself included, assumed. The strategy requires several independent things to break correctly: the lottery has to cooperate, the available player has to actually be a franchise talent, the developmental environment has to be good enough to realise that talent, and the front office has to build competently around them before the rookie contract expires.

Each of those is uncertain and they compound. A team that executes a deliberate rebuild perfectly still faces a fairly modest chance of emerging with a contender, and the failure mode is severe: years of losing, a damaged organisational culture, and nothing at the end of it.

I was treating tanking as a plan when it is closer to a lottery ticket bought with several years of institutional health.

I mistook a bet with poor odds for a strategy with a mechanism.

The Second Mistake: The Culture Cost Is Real

I used to dismiss talk of culture as the sort of thing people say when they have run out of arguments. I was wrong about this in a fairly specific way.

An organisation that spends three years optimising for losing develops habits. Young players learn the game in an environment where the outcome does not matter, where losing carries no consequence, and where the standards that separate professionals from talented amateurs are never enforced because enforcing them would be counterproductive.

Then the rebuild ends and those same players are asked to compete. The evidence suggests that transition is far harder than the model assumes. Development is not a process that happens to a player independent of context, and a context built for losing produces players who have learned things that then need unlearning.

The Third Mistake: I Ignored the Audience

This is the one I am least comfortable about. My position treated supporters as a rounding error, an inconvenience that would be compensated later by success.

But a season is a year of someone's life. People organise their weeks around these teams, take their children, spend money they could have spent elsewhere. Telling them that three of those years are a deliberate write-off in exchange for a probabilistic shot at contention is a substantial thing to do, and I waved it away because it did not fit in the model.

The dismissal was intellectually convenient rather than correct. A sports franchise is not only an asset being optimised. It is an institution with obligations to the people who sustain it, and a strategy that treats those obligations as costless is not sophisticated. It is just incomplete.

What I Think Now

I still believe the incentives are badly designed, and I still think blaming individual front offices for responding to them is misplaced. Lottery reform, pick protections and other structural fixes are the right target, and leagues that have adjusted these mechanisms have generally seen the behaviour reduce.

But I no longer think the rational-actor defence is sufficient. Something can be a response to bad incentives and still be a bad decision, badly executed, at a cost you underestimated. That is roughly where I have landed, and it took me an embarrassingly long time to get here.