The budget
An NFL team runs roughly 60 to 65 plays in a game. Some are sacks and penalties that pay nobody, so what is left is on the order of 33 pass attempts and 26 carries — call it 59 chances for someone to touch the ball.
Those 59 get divided among one quarterback, one or two running backs, two or three receivers, and one tight end. Six or seven people. Nobody else is eligible for meaningful volume: the five offensive linemen never touch the ball, and the extra bodies rotating through are there to block.
These are league-average shapes, not measurements of a particular team — a fast pass-leaning offense and a slow run-leaning one are different animals. But the order of magnitude is durable, and it is the only arithmetic here that matters.
So there are two budgets, each summing to exactly 100%:
- Carry share — the ~26 handoffs, divided among the backs (and, on some teams, a running quarterback).
- Target share — the ~33 throws, divided among the receivers, the tight end, and whichever backs stay on the field on passing downs. A target is a throw in a player's direction, counted whether or not he catches it.
Neither budget can exceed 100%, and neither goes unspent. That is the entire concept. Everything below is a consequence.
The budget is set by the coach, not earned by the players
The part that resists intuition: the allocation is an assignment, not an achievement.
A coach decides who is on the field in which situation and which plays get called. A running back cannot decide to be handed the ball. A receiver cannot decide the quarterback throws to him. Talent determines who wins the job in August; after that the job determines the touches, and the touches determine the points. Playing well makes a player likelier to keep his allocation, but it does not conjure new plays.
This is why almost all useful in-season work is watching for role changes rather than for players getting better, and why a mediocre player in a large role beats a good player in a small one nearly every week.
Consequence one: adding a good player subtracts from his teammates
An offense that acquires an excellent receiver does not become an offense with more throws. It becomes an offense where the same ~33 throws are divided differently, and whatever share the new arrival takes comes out of the people who were already there.
So a strong supporting cast is not straightforwardly good for a player you own. "He plays on a great offense" cuts both ways: great offenses are usually great because they have several people worth throwing to, which is exactly when none of them commands a large share. The most valuable fantasy situation is often a merely decent offense with one obvious first option.
Consequence two: a breakout is usually someone else's collapse
When a previously irrelevant player starts producing, the honest question is not "what did he unlock?" but "whose share did he take?" Because the budget is conserved, the answer is almost always a specific teammate, who is now worth less than he was.
The clean version is injury: a role vacates and someone absorbs it, usually in a single lump rather than spread around — see The injury-inheritance asymmetry. The messy version is a coach quietly changing his mind, same arithmetic with no announcement. Either way, a breakout is a transfer, so a waiver-wire addition is also a sell signal somewhere, possibly on your own roster.
Consequence three: pace enlarges the whole budget
There is one way to get more touches that is not zero-sum: run more plays.
A team that plays fast, converts more often, and holds the ball longer might run 68 plays where another runs 58. That is ten extra chances across the same six or seven people, and it raises every share-holder at once — nobody has to lose for it to happen.
This matters far more in football than in basketball. Basketball possessions alternate: both teams get nearly the same number, so a fast team's extra possessions are matched by its opponent's. Football does not alternate that way — a team that sustains drives keeps the ball, and the possession it keeps is one its opponent never gets. Play volume is genuinely won, not shared. So "this offense runs a lot of plays" is a real and comparatively underpriced edge, and the first thing to look at in The 32 offenses.
Consequence four: the top of the budget is where the value is
Because the divisor is small, share differences are large in absolute terms. Going from a 15% target share to 25% is not a ten-point move on some scale — it is roughly three extra throws a game, which in full PPR is two extra catches, two guaranteed points, plus yards, plus a better chance at the six-point play. Concentration is what you are buying.
Should you model this yourself?
Modelling team budgets explicitly — projecting each offense's plays, splitting them into passes and runs, then allocating both budgets across every player on the depth chart — is a real technique, and it is how the better analysts find their edges. The payoff is specific: it surfaces mispricings, players whose draft cost does not match the share their situation implies. That is a genuine source of profit, not available any other way.
It is also not the right first move. Allocating two budgets across thirty-two teams means hand-judging on the order of a thousand-plus share rows, every one a guess about a coach's plan, and the errors compound rather than cancel. Doing it badly is worse than not doing it, because a bad model produces confident numbers.
So this build carries the concept without computing the rows. A first-time drafter should buy consensus — take the market's allocation, which already aggregates thousands of people doing this work, and spend the effort on structure, scarcity, and not making unforced errors. Model the budgets in a later season, when you have your own observations to check them against. The concept still earns its place, because it is what makes an ADP list legible: it tells you why the market has one player above another, even when you never compute a share yourself.
Open questions
- The 60–65 plays and the 33/26 split are durable league-wide shapes, not 2026 measurements; per-team pace and pass-lean belong in The 32 offenses.
- "Adding a good player reduces his teammates" holds for shares but is overstated for points, since a better offense scores more often and touchdowns are not conserved the way touches are.
- The claim that pace matters more in football than in basketball is argued structurally, not quantified; how much of a team's play-count edge persists year to year is uncertain.