A Good Week From Your Receiver Is 10× a Bad One. That’s Where the Season Is Decided.
A fantasy matchup is not won by whoever averages the most points. It is won where the two rosters differ — and they differ where the scoring swings. A slot that reliably delivers the same number every week cannot win or lose you anything; it cancels out. The whole game lives in the volatile positions, and most managers spend their energy everywhere else.
The question: which player on your fantasy roster actually decides whether you win?
The answer: not your best one. A matchup is decided where the two rosters differ, and they differ where the scoring swings. Your quarterback averages seventeen points and both of you will get roughly seventeen.
Why that is surprising. We draft and fret over the biggest scorers. A number that never moves cannot separate two teams, however large it is.
Here is a claim that sounds wrong and is not. The highest-scoring player on your roster is often the least important one you own.
A number that never moves can’t predict anything
Notice that this is a general truth about prediction, not a fantasy quirk. It holds anywhere you compare two things. A number can only tell two outcomes apart to the extent that it moves.
The floor and the coin flip
Read the chart and the strategy reads itself. A starting quarterback averages around seventeen points and almost never strays far; we can pencil him in for both sides.
Where the touchdowns hide
Do the arithmetic on a typical starter. Touchdowns are only about a third of his scoring, and they carry almost all of the swing.
A number that never moves can’t tell you who’s going to win. Variance isn’t noise to be smoothed away — it’s where the outcome lives.
The ProfessorWhat to take home: spend your attention where the swings are
Your decisions — a waiver claim, a start-or-sit, a trade — can only change the outcome where the outcome is genuinely in doubt.
Whether we chase that swing or run from it depends on something the average hides: whether we walked in ahead or behind.
Notes & sources
Figures are 2024 regular-season weekly fantasy points (standard scoring) for genuine starters — players with at least twelve games — from nflverse weekly data. “A good/bad week” is the mean plus or minus a typical player’s week-to-week swing (one standard deviation): a quarterback runs about 17 points with a swing near 7; a running back about 7 with a swing near 5; receivers and tight ends carry the largest swing relative to their scoring. The touchdown decomposition uses the same starters, splitting each back’s points into yardage and touchdowns.
The general principle — that a predictor only carries information to the extent it varies, and that in an additive score the variance of the total is the sum of the parts’ variances — has its own primer (Variance). A companion caution applies in real prediction: variance is necessary for a predictor, but not sufficient — a variable also has to actually relate to the outcome. Fantasy is the clean case because the categories are the outcome by construction. See the archive for the companion pieces on communicating uncertainty and rates of change.
