FPL captain picks — Gameweek 8
The highest projected scorers in the game this gameweek, whoever owns them. Captaincy is the one decision where the question really is just "who scores most this week", so this is ranked on the model's projection alone — not on form, not filtered by ownership.
The shortlist
Ranked by projected points for Gameweek 8. Where two players project the same, the more widely owned one is placed first — they are the same bet on points and a different bet on rank, and the more-owned one is the one that does not cost you places when it goes wrong.
Haaland
6.1 projected
Haaland is the model's highest projection this gameweek at 6.1 points against AVL (A), owned by 76.8%.
- Price
- £15.7m
- Owned
- 76.8%
AVL (A)BHA (H)NFO (A)
B.Fernandes
5.5 projected
B.Fernandes projects 5.5 points against BOU (H), 0.6 behind the leader, owned by 29.2%.
- Price
- £11.8m
- Owned
- 29.2%
- Behind
- 0.6
BOU (H)CHE (A)AVL (H)
Rice
4.7 projected
Rice projects 4.7 points against EVE (H), 1.4 behind the leader, owned by 11.2%.
- Price
- £7.3m
- Owned
- 11.2%
- Behind
- 1.4
EVE (H)LIV (A)HUL (H)
The rest of the shortlist
The next 7 names, in the same order. The behind column is the gap to the top projection: a tenth of a point is the model saying "either", and it is worth reading before treating this as a ranking rather than a group.
| # | Player | Club | Pos | Projected | Behind | Price | Owned | Next |
|---|---|---|---|---|---|---|---|---|
| 4 | Anthony | Brentford | MID | 4.6 | 1.5 | £5.9m | 0.3% | HUL (A)NFO (H)BHA (A) |
| 5 | Rogers | Chelsea | MID | 4.5 | 1.6 | £8.0m | 48.0% | TOT (H)MUN (H)SUN (A) |
| 6 | Lacroix | Chelsea | DEF | 4.5 | 1.6 | £5.9m | 6.3% | TOT (H)MUN (H)SUN (A) |
| 7 | Murillo | Nott'm Forest | DEF | 4.5 | 1.6 | £5.5m | 2.1% | IPS (A)BRE (A)MCI (H) |
| 8 | Saka | Arsenal | MID | 4.4 | 1.7 | £9.6m | 16.7% | EVE (H)LIV (A)HUL (H) |
| 9 | Iwobi | Fulham | MID | 4.3 | 1.8 | £5.4m | 1.1% | COV (A)AVL (A)NEW (H) |
| 10 | N.Williams | Nott'm Forest | DEF | 4.2 | 1.9 | £5.0m | 6.3% | IPS (A)BRE (A)MCI (H) |
How to use this
The gap matters more than the order. Two players projected within a couple of tenths of each other are, as far as this model is concerned, the same pick. The behind figure is there so that reads as "either of these" rather than as first and second place.
Ownership is a separate question. Captaining the most-owned player is the low-variance choice: when he blanks, so does everyone else. Captaining a differential is how you gain rank and how you lose it. This page ranks on points only, so if rank is what you care about, read the ownership column as the other half of the decision.
The projections are wrong about individuals most weeks. They are useful for comparing players, poor as a forecast of any single score — and how wrong they have actually been is published on the accuracy page rather than described.
Rebuilt nightly from the current squad prices, form and team news, last updated 2026-10-21. Unlike the dated gameweek briefing, which is frozen at its deadline, this page follows the news: if the top pick is ruled out on a Friday, the shortlist changes.
How the model's captain pick has actually fared. Each week its top pick is set beside the player who really scored most — one gameweek at a time, newest first. Every one of these is read from a frozen edition, so the projection shown is the number the model published before the deadline, not one worked out with hindsight.
Loading…