Captain

Who takes the armband

Every figure here is already doubled. Ranking by expected points alone is the answer every free tool gives, and it discards what actually decides the week: two players on the same projection are not the same bet if one has a ceiling of 24 and the other of 14. The armband doubles the variance too.

Highest expectedHaaland13.8 ptsBest average outcome. The default answer.

Least likely to blankGabriel23% blank riskSteadiest route to points. Protecting a rank.

Best haul chanceHaaland19% chance of 20+Biggest upside. Chasing a rank.

The shortlist, with its spread

Bars run from the 10th to the 90th percentile of 6,000 simulated gameweeks; the notch is the median. A long bar is a swing pick, a short one a floor pick.

1HaalandMCI · £15.5m · 74.6% ownedFDR 3Template13.8 EV
Floor 4Median 12Ceiling 20Haul 20+ 19%Blank ≤4 41%
2B.FernandesMUN · £12.0m · 48.7% ownedFDR 2Template13.4 EV
Floor 4Median 10Ceiling 20Haul 20+ 13%Blank ≤4 41%
3GabrielARS · £8.0m · 26.3% ownedFDR 2Popular12.4 EV
Floor 4Median 12Ceiling 18Haul 20+ 9%Blank ≤4 23%
4SakaARS · £9.5m · 10.7% ownedFDR 2Popular11.8 EV
Floor 4Median 8Ceiling 20Haul 20+ 11%Blank ≤4 47%
5CherkiMCI · £7.5m · 9.3% ownedFDR 3Differential10.8 EV
Floor 4Median 8Ceiling 18Haul 20+ 9%Blank ≤4 46%
6Gibbs-WhiteNFO · £8.0m · 11.6% ownedFDR 2Popular10.4 EV
Floor 4Median 4Ceiling 14Haul 20+ 7%Blank ≤4 62%
7RiceARS · £7.5m · 22.4% ownedFDR 2Popular10.0 EV
Floor 4Median 8Ceiling 14Haul 20+ 3%Blank ≤4 44%
8SemenyoMCI · £8.5m · 22.9% ownedFDR 3Popular10.0 EV
Floor 4Median 4Ceiling 14Haul 20+ 6%Blank ≤4 61%

What this does not know

Not true effective ownership. Real EO needs captaincy share, and FPL publishes ownership but not captaincy anywhere in its data. Template / popular / differential is an ownership proxy, labelled as one rather than dressed up as EO.
No rotation certainty. The simulation assumes a fit player starts. Rotation is the silent killer and our start estimate is a model, not a team sheet.
Components only. Bonus points, cards and saves are not simulated, so ceilings are conservative rather than optimistic.

Deterministic: the same data always produces the same shortlist, which is what lets these picks be scored later on the report card. Not a guarantee.