Our Accuracy Record Every call we have made, scored on FPL's own results
How good are our calls? Here is every one, hits and misses. The FFIQ Team's captain matched or beat the most-captained player in 10 of 10 gameweeks (beat him in 2, matched him in 8, behind in 0).
One squad, scored on FPL's own results
One squad of fifteen, filed before each deadline and scored on FPL's own results. It has no FPL entry, so its place in the table is always "would rank": the rank those points would hold among 11,284,859 managers.
| Week | Points | Average manager | Overall |
|---|---|---|---|
| GW1 | 866 bench points | 50 | would rank 31,193rd |
| GW2 | 146 | 81 | would rank 25th |
| GW3 | 554 bench points | 51 | would rank 50th |
| GW4 | 8915 bench points | 69 | would rank 66th |
| GW5 | 602 bench points | 48 | would rank 34th |
| GW6 | 9216 bench points | 58 | would rank 5th |
| GW7 | 84 | 59 | would rank 2nd |
| GW8 | 512 bench points | 45 | would rank 2nd |
| GW9 | 838 bench points | 47 | would rank 1st |
| GW10 | 6712 bench points | 55 | would rank 2nd |
10 of 10 gameweeks above the average manager. A week still being played joins the list when FPL has finished checking it.
Every gameweek, with the fifteen and what each player scored →
Our captain, every gameweek
The benchmark is the crowd: the player most FPL managers made captain that week. Our captain beats it when he scores more points than that player, is level when the scores match, and is behind when he scores fewer. Points are before the armband doubles them.
| Week | Our captain | Most captained | Result |
|---|---|---|---|
| GW1 | João Pedro11 points | Haaland2 points | Beat |
| Top scorer that week: De Cuyper, 17 | |||
| GW2 | B.Fernandes23 points | Haaland13 points | Beat |
| Top scorer that week: B.Fernandes, 23 | |||
| GW3 | Haaland9 points | Haaland9 points | Level |
| Top scorer that week: Mitchell, 15 | |||
| GW4 | Haaland9 points | Haaland9 points | Level |
| Top scorer that week: Groß, 17 | |||
| GW5 | Haaland6 points | Haaland6 points | Level |
| Top scorer that week: Brobbey, 17 | |||
| GW6 | Groß6 points | Haaland6 points | Level |
| Top scorer that week: Henderson, 18 | |||
| GW7 | Haaland13 points | Haaland13 points | Level |
| Top scorer that week: Neto, 15 | |||
| GW8 | Haaland8 points | Haaland8 points | Level |
| Top scorer that week: Gabriel, 17 | |||
| GW9 | Haaland2 points | Haaland2 points | Level |
| Top scorer that week: Van Hecke, 15 | |||
| GW10 | Haaland2 points | Haaland2 points | Level |
| Top scorer that week: Barry, 16 | |||
FPL does not publish a most-captained player for Gameweek 1, so the comparison starts at Gameweek 2. The honest limit of that benchmark: it only asks whether we beat the crowd, not whether we found the week's best captain. Our captain was the week's top scorer, or level with the top scorer, in 1 of 10 gameweeks.
Our forecasts, scored after the deadline
Before each deadline we freeze our forecast for every player, so it cannot be changed after the results are in. Once the gameweek is over we take the 10 players we ranked highest and average what they actually scored. The comparison is the 10 players you would pick by ranking everyone on points so far this season. Same players to choose from, same gameweek.
| Week | Our top 10 | Points so far top 10 | Result |
|---|---|---|---|
| GW2 | 8.6average points | 3.6average points | Ahead |
| GW3 | 3.4average points | 2.4average points | Ahead |
| GW4 | 7.2average points | 2.6average points | Ahead |
| GW5 | 5.4average points | 4.9average points | Ahead |
| GW6 | 5.0average points | 5.2average points | Behind |
| GW7 | 3.5average points | 3.9average points | Behind |
| GW8 | 3.8average points | 2.5average points | Ahead |
| GW9 | 4.3average points | 4.2average points | Ahead |
| GW10 | 4.8average points | 6.1average points | Behind |
Across 9 gameweeks: our top 10 averaged 5.1 points a player against 3.9 for the points-so-far list. The forecasts are not printed here. Only gameweeks FPL has finished checking are scored.
What was measured, over what
Every result on this page is a gameweek FPL has finished checking, scored on FPL's own points. Nothing is graded while a gameweek is still being played, and a row is never removed after the fact. Every call combines 34 calculations with an algorithm that keeps updating as the scenario changes.
The numbers are small because the season is young. A handful of gameweeks cannot prove anything on its own, so each section says how many it rests on and the lists grow every week.
Where this stops
Price Watch went live on 26 August 2026. It is not a crystal ball and we do not call it one. The nightly record of hits, misses and false alarms is printed on the price page itself.
Backtests are not guarantees. Football has variance; a projection is a probability statement, not a promise. What we promise is honest measurement: the same test, misses included, every time.