What is bust probability?

Bust probability is the chance the model gives that a prospect's career ends up in the bottom tier of outcomes for his position. Not that he fails, not that he is bad — that he lands in the worst quarter of the players who play his job.

Position is the whole game here. Tiers are cut by percentile of career value within position, so a tackle is judged against tackles. This matters more than it sounds: the base rate of busting is genuinely different at running back than at guard, and a single league-wide threshold would tell you mostly which position a player plays.

Read it as a frequency, not a prophecy. A 30% bust probability means roughly three in ten players who looked like this one, at this position, at this point on the board, went on to bust. It is not a claim about this specific player's character or work ethic, neither of which the model can see.

We publish the number's own limits alongside it. The model's skill at separating busts from non-busts varies a lot by position — strong in the trenches, close to a coin flip at some skill positions — and where it is weak we say so rather than printing a confident-looking decimal. For quarterbacks we do not publish a bust probability at all, because the sample of first-round quarterbacks is too small to support one.

Very low numbers get compressed rather than shown raw. A model that says a prospect has a 0.4% chance of busting is claiming a certainty no public-information model has earned; the floor keeps the tail honest without flattening the players who genuinely are safe bets.