Signal vs Noise

The Career Year Is a Loan, Not a Raise

Of the 112 skill players since 1999 who jumped 40% or more in yards per game, fewer than one in four held that level the following season.

Every August, some share of a fantasy draft board is made up of players being priced off the best season of their lives. This is what happens to them.

The median career year is a jump from about 50 yards a game to about 86, which is enormous, the kind of leap that wins leagues and gets a player drafted three rounds earlier the next summer. The median follow-up is 76. Not a collapse. But a giveback of roughly a quarter of the gain, and well short of what the price implies.

The rate is the harsher number. Twenty-three percent held or improved. Better than three in four went backwards.

Fewer than one in four career years survive contact with the following September.

Why the giveback is partial rather than total

This is the part that gets missed by people who overcorrect into pure regression pessimism. The year-after median of 76 is a long way above the year-before median of 50. Most breakouts are a mix of two things: a genuine change in role or ability, which persists, and a run of good fortune, which does not.

The trap is paying for all of the breakout, not believing in it. The market tends to price the career year as the player's new baseline, when the data says the new baseline sits somewhere between the old level and the peak, closer to the peak than to the floor.

How to tell the durable ones from the rest

The 26 who held share a pattern, and it is the same pattern our other work keeps landing on: their jump was built on opportunity rather than on efficiency. A back who went from twelve touches a game to twenty because the starter left has a new job, and jobs persist. A receiver whose yards per catch spiked while his target share stayed flat had a lucky year, and luck does not persist.

So the diagnostic question for any breakout on your board is simply: what changed? If the answer is a depth chart, a scheme fit, or a quarterback upgrade that is still in place, the gain has a reason to stay. If the answer is that everything went right at once, or worse, if the answer is a touchdown rate that doubled, you are buying the peak of a distribution.

And be careful about the version of this that shows up in auction rooms and first rounds, where a career year plus a new contract plus offseason hype compound into a price nobody would pay if the player had simply been good for three straight years instead. Consistency is cheaper than a breakout and it projects better. That is close to a free edge, and it is available in every draft.