The Receipts: Every Best XI Prediction We've Made This Season
Every FPL site claims their predictions work. Almost none of them show the full record measured against the only number that really matters — the gameweek average. Here's ours. Every Best XI we've published this 2025/26 season, every captain call, every player's return, and how the squad as a whole performed against the average FPL manager. No cherry-picking.
How we measure success
The benchmark we care about is the official FPL gameweek average — the mean total points across all FPL managers worldwide for that gameweek. If our Best XI is consistently above that line, our recommendations are helping managers finish above average.
Within that, we set ourselves a stretch target of 20% above the gameweek average. Hit that consistently and the squad isn't just above average — it's meaningfully better than the typical FPL manager every week, which is what builds rank.
The numbers below are the published FPL gameweek averages straight from the game's API. No filtering, no adjustments.
The Scorecard
Total Best XI Points
387
Total GW Average
301
+20% Target Total
361
Net vs Avg
+86
+14.3/GW
Beat GW Avg
5/6
1 matched
+20% Target Hit
2/6
stretch goal
Avg % Above
+36%
on the average
Across 6 logged gameweeks, the Best XI has averaged 65 points per week versus a GW average of 50 points. That's 36% above the average manager — comfortably past the 20% stretch target on a season-long basis, even though individual gameweeks vary.
The Standout Moments
🏆 Best week
GW34 — 72 vs avg 36
Exactly double the gameweek average. A 100% beat on the average manager, in a low-scoring week where most squads struggled. Bowen captain doubled, Gibbs-White hauled, Raya posted a goalkeeper 10.
⚠️ Flat week
GW35 — 50 vs avg 50
The one week the receipts go flat. Best XI matched the gameweek average exactly. A Bowen captain blank and a 1 from Rogers cost the squad its usual edge.
⭐ Best captain pick
B. Fernandes — GW31
13 actual doubled to 26 — the single biggest captaincy contribution of the season. In a week the average manager scored just 38, having a captain return 26 alone was decisive.
💎 Best differential
N. Williams (DEF) — GW32 captain
An attacking-defender captain call at heavy-differential ownership. Returned 11, doubled to 22 — the textbook differential play of the season, and the lift that kept the squad above the GW average.
🔥 Biggest non-captain haul
Gibbs-White — GW33 (20 pts)
Predicted 5.7, returned 20. The single biggest individual overperformance of the season — a goal, an assist, and bonus all in one gameweek. Carried the squad past the GW average single-handedly.
😬 Biggest individual blank
Thiaw (DEF) — GW31 (0 pts)
Predicted 5.2, returned a flat zero. A reminder that even strong defensive predictions can be undone by a single red card, an early concession, or a sub.
Gameweek by Gameweek
Every Best XI we've published, with the full eleven, predicted xP for each player, and the actual points they returned — measured against the official FPL gameweek average and our +20% stretch target.
Gameweek 31
✅ Top 20% target hit · +32 vs avgSnapshot taken March 2026 · team cost £68.4m
Best XI Total
70
incl. captain ×2
GW Average
38
all FPL managers
+20% Target
45.6
avg × 1.2
Captain
B. Fernandes
13 × 2 = 26 pts
| Pos | Player | Team | xP | Actual |
|---|---|---|---|---|
| GK | Darlow | LEE | 4.3 | 6 |
| DEF | Collins | BRE | 5.0 | 6 |
| DEF | Alderete | SUN | 4.9 | 1 |
| DEF | Rodon | LEE | 5.2 | 6 |
| DEF | Thiaw | NEW | 5.2 | 0 |
| MID | B. FernandesC | MUN | 7.7 | 13 (×2 = 26) |
| MID | Wilson | FUL | 5.8 | 9 |
| MID | Gibbs-White | NFO | 5.8 | 8 |
| MID | Iwobi | FUL | 5.7 | 1 |
| MID | Anderson | NFO | 5.6 | 5 |
| FWD | Bowen | WHU | 4.8 | 2 |
The headline week. Best XI returned 70 points against a gameweek average of just 38 — 84% above the average manager. Bruno Fernandes captain delivered 13 doubled to 26, and a bargain squad held up against a low-scoring round of fixtures.
Gameweek 32
Beat GW avg by 9 (20%)Snapshot taken March 2026 · team cost £69m
Best XI Total
55
incl. captain ×2
GW Average
46
all FPL managers
+20% Target
55.2
avg × 1.2
Captain
N. Williams
11 × 2 = 22 pts
| Pos | Player | Team | xP | Actual |
|---|---|---|---|---|
| GK | Pickford | EVE | 4.9 | 0 |
| DEF | N. WilliamsC | NFO | 7.8 | 11 (×2 = 22) |
| DEF | Milenković | NFO | 6.2 | 2 |
| DEF | O'Brien | NFO | 6.1 | 4 |
| DEF | Van Hecke | BHA | 6.1 | 5 |
| DEF | Kayode | BHA | 5.8 | 4 |
| MID | B. Fernandes | MUN | 7.3 | 4 |
| MID | Rice | ARS | 5.6 | 4 |
| MID | Szoboszlai | LIV | 5.2 | 5 |
| MID | Gordon | NEW | 5.2 | 3 |
| FWD | João Pedro | CHE | 5.3 | 2 |
Beat the GW average by 9, but came up 0.2 points short of the +20% stretch target. The N. Williams captaincy call landed (11 doubled to 22 — the textbook differential pick of the season). The rest of the XI was quieter — a Pickford zero hurt — but the squad still finished comfortably above the average manager.
Gameweek 33
Beat GW avg by 4 (6%)Snapshot taken April 2026 · team cost £68.4m
Best XI Total
70
incl. captain ×2
GW Average
66
all FPL managers
+20% Target
79.2
avg × 1.2
Captain
N. Williams
2 × 2 = 4 pts
| Pos | Player | Team | xP | Actual |
|---|---|---|---|---|
| GK | Donnarumma | MCI | 4.0 | 8 |
| DEF | N. WilliamsC | NFO | 6.4 | 2 (×2 = 4) |
| DEF | Virgil | LIV | 5.1 | 8 |
| DEF | Senesi | BOU | 5.0 | 6 |
| DEF | Guéhi | CRY | 4.9 | 8 |
| DEF | Mitchell | CRY | 4.9 | 7 |
| MID | Gibbs-White | NFO | 5.7 | 20 |
| MID | Anderson | NFO | 5.2 | 2 |
| MID | Semenyo | BOU | 4.6 | 5 |
| FWD | Bowen | WHU | 4.9 | 2 |
| FWD | João Pedro | CHE | 4.6 | 0 |
A high-scoring gameweek across the whole game — average was 66 points. Best XI returned 70, so we beat the average but didn't reach the stretch target. Gibbs-White's 20-point haul kept us above the line even with a blank from the N. Williams captaincy call.
Gameweek 34
✅ Top 20% target hit · +36 vs avgSnapshot taken April 2026 · team cost £74.7m
Best XI Total
72
incl. captain ×2
GW Average
36
all FPL managers
+20% Target
43.2
avg × 1.2
Captain
Bowen
10 × 2 = 20 pts
| Pos | Player | Team | xP | Actual |
|---|---|---|---|---|
| GK | Raya | ARS | 4.5 | 10 |
| DEF | Virgil | LIV | 5.6 | 4 |
| DEF | N. Williams | NFO | 5.3 | 5 |
| DEF | Tarkowski | EVE | 5.3 | 3 |
| DEF | Milenković | NFO | 5.0 | 8 |
| MID | André | WOL | 5.0 | 1 |
| MID | Rogers | AVL | 4.8 | 2 |
| MID | B. Fernandes | MUN | 4.8 | 5 |
| MID | Gibbs-White | NFO | 4.7 | 13 |
| FWD | BowenC | WHU | 6.0 | 10 (×2 = 20) |
| FWD | Gyökeres | ARS | 4.7 | 1 |
The best week of the season. Best XI returned 72 points against a gameweek average of just 36 — exactly double. Bowen captain delivered (10 doubled to 20), Gibbs-White hauled again, and Raya posted a goalkeeper 10 in a low-scoring round where most managers struggled to break 40.
Gameweek 35
Matched the GW avgSnapshot taken April 2026 · team cost £76.2m
Best XI Total
50
incl. captain ×2
GW Average
50
all FPL managers
+20% Target
60.0
avg × 1.2
Captain
Bowen
2 × 2 = 4 pts
| Pos | Player | Team | xP | Actual |
|---|---|---|---|---|
| GK | Verbruggen | BHA | 4.4 | 2 |
| DEF | Senesi | BOU | 5.1 | 14 |
| DEF | Struijk | LEE | 5.1 | 1 |
| DEF | Guéhi | CRY | 5.1 | 4 |
| DEF | Bogle | SHU | 5.0 | 5 |
| DEF | O'Reilly | MCI | 5.0 | 1 |
| MID | Rogers | AVL | 5.4 | 1 |
| MID | B. Fernandes | MUN | 5.1 | 5 |
| FWD | BowenC | WHU | 6.0 | 2 (×2 = 4) |
| FWD | Haaland | MCI | 5.6 | 7 |
| FWD | Thiago | BRE | 5.0 | 6 |
The one week the receipts go flat. Best XI matched the gameweek average exactly — neither above nor below. A Bowen captain blank and a 1 from Rogers cost us the upside, even though Senesi's 14 was the standout individual return.
Gameweek 36
Beat GW avg by 5 (8%)Snapshot taken May 2026 · team cost £80.3m
Best XI Total
70
incl. captain ×2
GW Average
65
all FPL managers
+20% Target
78.0
avg × 1.2
Captain
Haaland
11 × 2 = 22 pts
| Pos | Player | Team | xP | Actual |
|---|---|---|---|---|
| GK | Verbruggen | BHA | 4.9 | 6 |
| DEF | Van Hecke | BHA | 5.2 | 6 |
| DEF | F. Kadıoğlu | BHA | 5.0 | 6 |
| DEF | Pedro Porro | TOT | 4.8 | 1 |
| DEF | Gabriel | ARS | 4.7 | 11 |
| MID | Rogers | AVL | 5.2 | 2 |
| MID | B. Fernandes | MUN | 5.0 | 3 |
| MID | Rice | ARS | 4.7 | 3 |
| FWD | HaalandC | MCI | 6.8 | 11 (×2 = 22) |
| FWD | Igor Jesus | NFO | 5.4 | 1 |
| FWD | Watkins | AVL | 5.1 | 9 |
A solid recovery week. Best XI returned 70 against a high-scoring gameweek average of 65 — a 5-point beat. Haaland captain delivered (11 doubled to 22), Gabriel hauled 11 from a 4.7 xP prediction, and Watkins added 9 up top. The +20% stretch target was a tall order in such a high-scoring landscape, but the squad still finished above the average for the fifth time in six gameweeks.
What the Data Tells Us
6 gameweeks in, the pattern is clear.
The Best XI consistently beats the average manager. In 5 out of 6 gameweeks the squad finished above the official GW average — and the one week it didn't, it tied. There's no gameweek so far where the recommendations would have left a manager below the average.
The season-long edge is 36% above average. Averaging across all 6 logged gameweeks, the Best XI returns are 36% above the average FPL manager. That's comfortably past the 20% stretch target on the season level, even though individual weeks vary. That level of outperformance, sustained, is what climbs rank.
Big-edge weeks come from finding signal in chaos. The two best weeks — GW31 and GW34 — were both low-scoring gameweeks where the average manager struggled (38 and 36 respectively). The Best XI's ability to find captain returns and bargain bench depth was what created the gap. In those weeks, the squad was 84% and 100% above the average.
High-scoring weeks compress the edge. GW33 and GW36 were both high-scoring rounds (averages of 66 and 65). The Best XI beat the average in both — but the margin shrank to single digits. When everyone hauls, the ceiling for beating the average is naturally lower. The edge tends to widen when the field gets it wrong, not when the field gets it right.
The differentials really worked. N. Williams as captain in GW32 (22 doubled) and the broader squad construction in GW31 and GW34 — these are the same kinds of picks described in our differentials guide. Not theoretical — these are the receipts.
The Honest Reflection
6 gameweeks is still a small sample, and we're not pretending otherwise. We'll keep building this gameweek by gameweek as more data comes in. The shape we'd expect over a longer horizon: more gameweeks at the +20% target, occasional weeks at the line, very few below it. The shape so far is consistent with that.
There are also weaknesses the data exposes. GW32's Pickford zero was a template goalkeeper pick that didn't pay out. GW33's N. Williams captaincy call blanked even though the underlying profile was strong. Rogers has been selected three gameweeks running (GW34-36) and returned 2, 1, and 2 — a persistent miss the model needs to recalibrate around. These are calls we'll keep working on.
The bigger picture, though, is that the squad has done what it's designed to do — finish above the average manager every week, and well above the average on the season — exactly the kind of pattern that climbs rank over a full FPL year.
📌 This post will update
The 2025/26 season has gameweeks remaining. As each gameweek's actual points and official GW average are confirmed, we'll add the receipt here too — beats and flat weeks included.
Get the same Best XI for your squad
The same model that produced these receipts will produce a personalised Best XI for your team every gameweek. Free to start, no credit card required.