> guide · updated Jul 2026

FPL captaincy explained: rules, strategy and how a model picks a captain

In FPL your captain scores double points every gameweek: whatever the player you name returns, it counts twice. It is the single biggest swing on your teamsheet — one call usually worth more than a transfer. If your captain does not play, the vice-captain's score is doubled instead.

How captaincy works

Before each gameweek deadline you assign one player the captain's armband and one the vice-captain. The captain's gameweek score is multiplied by two; every other starter scores as normal. The rules that govern it are stable season to season:

  • Double, not bonus. Doubling applies to the player's entire gameweek total — goals, assists, clean-sheet points, bonus, appearance points, the lot.
  • One armband, set before the deadline. You can change your captain any number of times up to the gameweek deadline; after it locks, the pick is fixed.
  • The vice-captain is insurance. If your captain plays 0 minutes, the double transfers to your vice-captain automatically. If the captain plays even one minute, the vice never triggers — the mechanics and edge cases are covered in the vice-captain guide.
  • Both must start. The armband only pays out on players in your starting XI; a captain on your bench earns you nothing extra.

Deadline timing (typically ~90 minutes before the first kickoff of the gameweek) and any chip interactions are the standard 2025/26 rules — exact 2026/27 deadlines and chip details await confirmation at game launch.

The captaincy maths: the biggest swing on your teamsheet

A transfer changes one squad slot at the cost of points hits and price risk. The armband is free and moves far more: it doubles the output of your best-projected player, so its value is roughly a second copy of that player's gameweek. That is why the captain choice, over a season, decides more rank than any single transfer.

The 2025/26 numbers make the scale concrete. Across the completed season the highest-scoring player in an average gameweek returned 17.2 points34.4 doubled — against an average manager score of 49.9. In other words, a single perfectly-timed armband was worth around 69% of a whole team's typical weekly haul. The best single captain week of 2025/26 was J.Timber's 24-point gameweek 2 for Arsenal — 48 to anyone who captained him.

2025/26's highest-scoring captain assets (final season totals, Official FPL API)
PlayerTeam · posPointsCaptained (×2)Selected
HaalandMCI · FWD23947862.5%
B.FernandesMUN · MID23547048.0%
GabrielARS · DEF20941845.4%
SemenyoMCI · MID20240446.2%
Gibbs-WhiteNFO · MID1883769.2%
BowenWHU · FWD18737417.5%

Template vs differential captaincy

The right captain is not just the highest projected score — it is the highest projected score relative to your rank goal. Two frames:

  • Template captaincy — captain the most-owned premium (the crowd pick). If they haul, you keep pace with everyone above you; if they blank, so does the field, so you lose no rank. It is the low-variance choice and the correct default when you are protecting a lead.
  • Differential captaincy — captain a lower-owned player. When they haul you gain rank on the whole field; when they blank you drop. It is the tool for chasing rank, not protecting it, and the maths only favours it when your projected edge clears the ownership risk.

The concentration in 2025/26 shows how heavily the crowd leans template: across all 38 gameweeks only two players ever led the armband — Haaland (the most-captained pick in 34 of 38gameweeks) and Salah (the other 4). The honest twist is in the next section: leaning on a name is not the same as catching the week's ceiling.

How a model turns probabilities into a captain ranking

A captain pick is an expected-points question, and a calibrated model answers it the same way for every fixture. Bawler's approach, disclosed as a projection rather than a promise:

  • Per-fixture probabilities. The model prices each match into win / draw / clean-sheet and goal-expectation terms from rolling expected goals — see the full methodology.
  • Player expected points (xPts). Those probabilities map onto each player's likely FPL returns — goal, assist, clean-sheet and appearance points — weighted by minutes and role.
  • Captain expected return. Because the armband doubles the score, a player's captain expected return is simply 2 × their projected xPts. Rank every candidate on that and you have a captain shortlist, not a hunch.
  • Ownership overlay. Pair the expected-points ranking with ownership to separate the template-safe call from the differential swing, so the pick fits your rank goal.

This is the exact engine behind Bawler's weekly captain shortlist, which publishes on /fpl/captain-picks from Gameweek 1 of 2026/27 — a ranked table of candidates by model captain expected return, with the vice-captain call alongside. Every projection is settled against real results on the public track record, including the misses.

2025/26 receipts: the crowd's captain vs the week's ceiling

No captain method hits every week, and honesty means showing that. The headline receipt from 2025/26 is blunt: the most-captained player was the week's single highest scorer in 0 of 38 gameweeks. The crowd pick is a floor, not a ceiling — reliable, rarely optimal. A sample of the opening weeks (all figures doubled, i.e. as captained):

Crowd (most-captained) return vs the gameweek's top haul, captained points, 2025/26
GWCrowd captainReturnedWeek's top haulGap
1Salah16Ballard 3418
2Salah10J.Timber 4838
3Haaland18Guéhi 3012
4Salah18Zubimendi 3214
5Salah10Stach 3020

The lesson is not "chase the ceiling" — hindsight is free and nobody catches the top haul reliably. It is that the safe template pick and the optimal pick diverge constantly, and the value is in a repeatable, calibrated read of expected return rather than following the herd or your gut.

Triple captain, vice-captain and where to go next

The Triple Captain chip takes the same logic further — it triples your captain's score for one gameweek instead of doubling it, which is why it is best saved for a fixture with the highest projected captain return (often a double gameweek). Chip availability and counts are the standard rules and await confirmation at the 2026/27 launch.

From here: read the vice-captain guide for the insurance mechanics, check the model's fixture difficulty to see which captains have the softest run, and start at the FPL hub. Bawler's ranked captain shortlist lands on /fpl/captain-picks from Gameweek 1 of 2026/27.

All figures are final 2025/26 numbers computed from the Official FPL API. Model outputs are projections, not guarantees. FPL is a free game — this is pure fantasy-points analysis for entertainment.