GW9 forecast report
The latest saved GW9 checkpoint from before the deadline, scored against official FPL points. The forecast baseline stays fixed when the model changes later.
The final deadline capture was missed, so this report uses the latest saved GW9 checkpoint from before the deadline. The forecast fingerprint is 3c5429b87f4aae9cdb012ba6d2942ca1841b8fe1db9963b1a0c2b030c38cab3a. FPL has marked every fixture finished.
Headline figures cover only players who played, since forecasting zero for a benched player is easy and would flatter the numbers. Across all 669 players the average miss was 0.97 pts.
Pre-kickoff comparison is unavailable: 7 of 10 fixture forecasts were saved. We only publish that comparison when every fixture has a verified pre-kickoff capture.
Against a flat 2-point guess
Both rows use the same 310 players who recorded minutes. A hit-rate can hide how far misses are, so compare average miss and RMSE too.
| Forecast | Within 2 | Avg miss | RMSE | Bias |
|---|---|---|---|---|
| FPL Prism | 73.2% | 1.76 | 2.75 | +1.00 |
| Always forecast 2 | 78.7% | 1.80 | 2.79 | +0.83 |
How the xPts rankings performed
Top picks are ranked by forecast xPts. Actual averages include zero-minute picks; the field average uses the same 669 forecasted players. This is selection signal from the frozen forecast pool, not a guaranteed squad result.
| Highest xPts | Total actual | Avg per pick | Field avg / player | Uplift / pick | Played |
|---|---|---|---|---|---|
| Top 10 | 47 | 4.70 | 1.31 | +3.39 | 10/10 |
| Top 20 | 88 | 4.40 | 1.31 | +3.09 | 20/20 |
| Top 50 | 221 | 4.42 | 1.31 | +3.11 | 49/50 |
By position
| Position | Avg miss | Bias | Within 2 |
|---|---|---|---|
| Goalkeepers | 2.20 | +1.37 | 65.0% |
| Defenders | 2.17 | +1.45 | 63.4% |
| Midfielders | 1.40 | +0.67 | 80.4% |
| Forwards | 1.91 | +0.93 | 75.0% |
Best calls
High-output players scoring at least 4 points where the pre-deadline forecast was within 1.5 points of the official result.
| Player | Pos | Forecast | Actual | Gap |
|---|---|---|---|---|
| Dewsbury-Hall | MID | 3.3 | 4 | +0.7 |
| Xhaka | MID | 3.1 | 4 | +0.9 |
| Anderson | MID | 3.1 | 4 | +0.9 |
| Wharton | MID | 2.7 | 4 | +1.3 |
| Anthony | MID | 2.5 | 4 | +1.5 |
Largest misses
| Player | Pos | Forecast | Actual | Miss |
|---|---|---|---|---|
| Van Hecke | DEF | 3.1 | 15 | 11.9 |
| Fofana | DEF | 1.5 | 11 | 9.5 |
| McBurnie | FWD | 2.8 | 12 | 9.2 |
| Verbruggen | GKP | 2.1 | 11 | 8.9 |
| Van den Berg | DEF | 1.2 | 10 | 8.8 |
| Konsa | DEF | 2.1 | 10 | 7.9 |
| Costinha | DEF | 0.4 | 8 | 7.6 |
| Ayari | MID | 2.5 | 10 | 7.5 |
| Awoniyi | FWD | 1.5 | 9 | 7.5 |
| Acheampong | DEF | 1.5 | 9 | 7.5 |
| Branthwaite | DEF | 1.5 | 9 | 7.5 |
| Adli | MID | 0.7 | 8 | 7.3 |
What these measures mean
- Average miss (MAE): the typical gap between forecast and actual points, ignoring direction.
- RMSE: like the average miss, but big misses count for more.
- Bias: actual minus forecast on average. Positive means players outscored their forecasts.
- Within 2: share of players whose actual score was within 2 points of the forecast.
Snapshot and data source
- Model version
- 1.4.5
- Forecast source
- Latest saved pre-deadline checkpoint
- Training window
- Results through GW8; GW4–GW5 walk-forward validation.
- Checkpoint captured
- 31 Oct 2026, 08:50 UTC, before the deadline
- Report generated
- 3 Nov 2026, 09:44 UTC
- Points source
- https://fantasy.premierleague.com/api/event/9/live/
- Result status
- 10 of 10 fixtures provisionally finished when scored
- Snapshot fingerprint
- 3c5429b87f4aae9cdb012ba6d2942ca1841b8fe1db9963b1a0c2b030c38cab3a
This report is historical evidence for GW9, not a test of the current model. See the live GW10 forecasts.