The Fantasy Football Encyclopedia
Encyclopedia · Concepts · Opportunity share

Concepts

Opportunity share

Snaps gate opportunity but are not opportunity — a player has to be on the field to be eligible for the ball, and then the ball still has to come to him.

as of 2026-08-02 · status seed · format espn-ppr1.0-10tm-std

Up: Fantasy Football · Related: The conserved team touch budget · Reading an NFL depth chart · Player archetypes · Sample size in a 17-game season

Fantasy points come from touching the ball, and nothing else on a play pays anything. A receiver who runs a perfect route the quarterback never looks at scores zero, and so does a back standing on the sideline while his team scores.

So "how much will this player score?" decomposes into three links, and the links are not interchangeable:

  1. Is he on the field? Every play begins with a snap — the ball being put into play. His snap share is the fraction of his team's offensive plays he is out there for.
  2. Is he eligible? Being on the field makes him one of five or six people who could be handed the ball or thrown to. Nothing more.
  3. Does the ball actually come to him? A carry is a handoff. A target is a pass thrown in his direction. Together these are his touches, and touches are what the scoring rules pay for.

Only the third link produces points; the first two are preconditions. That is the whole thesis: snap share is a ceiling, not a quantity. It tells you the largest opportunity a player could have had, not the opportunity he got.

The case that breaks the naive model

Take two running backs on the same team, same game. One plays 60% of the snaps and finishes with 8 touches. The other plays 40% and finishes with 18. Round illustrative numbers, not a projection — but the shape is ordinary, and any model that reads snaps as opportunity says it cannot happen.

It happens because which snaps you take is a different fact from how many. The 40%-snap back is on the field in run-heavy situations — first down, short yardage, a lead late — where a large fraction of his snaps convert straight into carries. The 60%-snap back is out there for passing downs too, and on a passing down there is one ball and it usually goes to a receiver. He accumulates snaps that were never going to become touches.

Same field, same game, and the more-used player by minutes is the less-used player by opportunity. This is why the twelve Player archetypes describe situations — early-down committee back, passing-down back, goal-line back — rather than sorting backs by how much they play.

Two currencies, and they are separate

There are two ways to get the ball, so there are two shares worth tracking. Carry share is a running back's percentage of his team's handoffs. Target share is a receiver's or tight end's percentage of his team's pass attempts thrown his way. Each sums to 100% across the roster — see The conserved team touch budget.

Most players draw on one: a move TE or an alpha X lives entirely on target share, an early-down back entirely on carry share. The rare and expensive thing is the three-down bell cow, who collects both — two income streams on one roster spot.

A target counts even on an incompletion

The unit of receiving opportunity is the throw in your direction, not the catch. If the ball is badly thrown, or dropped, or knocked away, the target still happened and still counted. This has no basketball analogue — nobody records passes intended for you that never arrived.

It matters because the target is the part the offense chose and the catch is the part that happened to work out. Catch rate bounces around with accuracy, coverage and luck; target volume reflects a standing decision about who the offense wants to throw to, and standing decisions persist. In full PPR, where every reception pays a full point before a single yard, a player drawing ten targets a week has a floor a four-target player cannot reach no matter how well he plays.

Target share is the most stable number in fantasy football

Of everything you can put in a spreadsheet before the season, target share forecasts best. It is set by role and alignment, and roles change slowly.

Efficiency does the opposite. Yards per carry swings year to year enough to be nearly useless as a forecast, because it depends on the five linemen in front of him, the defences he drew, and two or three long runs that may not repeat. Receiving efficiency is stickier, but not much. Volume forecasts; efficiency mostly explains what already happened. So when a player's cost is justified by how good he was, be suspicious; when it is justified by how much of his offense he gets, pay it.

The basketball model, and exactly where it breaks

Two instincts transfer badly here.

Minutes are a near-direct proxy for opportunity in basketball. Snaps are not. A basketball player on the floor is continuously eligible — possessions flow past him at a steady rate and he participates in most of them. A football player is eligible for one ball per play, and whether it can come to him is decided by the play call before the snap. Minutes measure exposure to a stream; snaps measure attendance at a lottery whose tickets someone else hands out.

Usage rate is partly self-driven in basketball. It is not here. A hot scorer takes more shots; the game reallocates toward whoever is producing. Football skill players cannot do this. A back cannot decide to be handed the ball, and a receiver cannot decide the quarterback throws to him. Usage is an assignment, not an achievement — stickier than a basketball usage rate, and far less responsive to merit. See Reading an NFL depth chart for where the assignment is visible.

What to do with this

Buy share, not talent. Ask what fraction of the carries or the throws a player is paid to receive, and treat everything else as commentary. When something changes in-season the snap share moves first, which makes it a genuine early warning — but confirm it in touches, because a rising snap share with flat touches is a player being asked to block more. And be slow either way: fourteen weeks cannot distinguish a role change from three odd afternoons. See Sample size in a 17-game season.

Open questions