Predicted points,
graded in public
| Club | Pos | ||||||
|---|---|---|---|---|---|---|---|
| De Cuyper | BHA | DEF | £4.6 | 8.1% | 3.9 | 17 | +13.1 |
| Hinshelwood! | BHA | MID | £6.0 | 2.4% | 3.0 | 16 | +13.0 |
| Mendy | HUL | DEF | £4.0 | 3.9% | 2.2 | 15 | +12.8 |
| Ajayi | HUL | DEF | £4.0 | 1.9% | 2.2 | 14 | +11.8 |
| M.Sangaré | BRE | MID | £5.6 | 8.6% | 1.8 | 14 | +12.2 |
| Palmer | CHE | MID | £9.5 | 16.0% | 3.9 | 13 | +9.1 |
| Kayode | BRE | DEF | £4.6 | 7.6% | 3.0 | 13 | +10.0 |
| Stach | LEE | MID | £6.0 | 2.4% | 2.7 | 13 | +10.3 |
| Gakpo | LIV | MID | £7.0 | 6.3% | 3.1 | 12 | +8.9 |
| White | ARS | DEF | £5.5 | 6.7% | 3.6 | 11 | +7.4 |
| João Pedro | CHE | FWD | £7.6 | 68.1% | 3.5 | 11 | +7.5 |
| Dewsbury-Hall | EVE | MID | £6.5 | 4.7% | 3.0 | 11 | +8.0 |
| Ødegaard | ARS | MID | £6.6 | 11.7% | 3.0 | 11 | +8.1 |
| Guéhi | MCI | DEF | £6.0 | 19.5% | 3.4 | 10 | +6.6 |
| Justin | LEE | DEF | £4.5 | 1.2% | 3.1 | 10 | +6.9 |
| Bijol | LEE | DEF | £5.0 | 0.9% | 2.8 | 10 | +7.2 |
| Tavernier | BOU | MID | £6.0 | 2.4% | 2.7 | 10 | +7.3 |
| Lewis-Potter | BRE | MID | £5.5 | 1.1% | 2.6 | 10 | +7.4 |
| Tzolakis | HUL | GKP | £4.5 | 2.9% | 2.2 | 10 | +7.8 |
| Saka | ARS | MID | £9.5 | 10.4% | 4.2 | 9 | +4.8 |
| Calafiori | ARS | DEF | £5.6 | 41.7% | 4.1 | 9 | +4.9 |
| Gvardiol | MCI | DEF | £5.5 | 13.7% | 3.5 | 9 | +5.5 |
| Ndiaye | EVE | MID | £6.0 | 16.1% | 3.0 | 9 | +6.0 |
| Janelt | BRE | MID | £5.0 | 0.5% | 2.5 | 9 | +6.5 |
| Trafford | LEE | GKP | £5.0 | 4.3% | 2.4 | 9 | +6.6 |
| Xhaka | SUN | MID | £5.5 | 4.7% | 2.3 | 9 | +6.7 |
| Elanga | NEW | MID | £6.0 | 2.1% | 2.2 | 9 | +6.8 |
| Emersonn | IPS | FWD | £5.5 | 1.8% | 1.5 | 9 | +7.5 |
| Cherki | MCI | MID | £7.5 | 10.6% | 4.0 | 8 | +4.0 |
| Barry | EVE | FWD | £5.5 | 3.3% | 2.9 | 8 | +5.1 |
| Szoboszlai | LIV | MID | £7.0 | 43.3% | 2.9 | 8 | +5.1 |
| Rogers | CHE | MID | £7.5 | 26.2% | 2.8 | 8 | +5.2 |
| Ajer | BRE | DEF | £4.5 | 4.1% | 2.8 | 8 | +5.2 |
| Armstrong | EVE | MID | £5.0 | 0.2% | 2.3 | 8 | +5.7 |
| Egan | HUL | DEF | £4.0 | 2.0% | 2.2 | 8 | +5.8 |
| Wieffer | BHA | DEF | £5.0 | 0.4% | 3.3 | 7 | +3.7 |
| Kelleher | BRE | GKP | £5.0 | 6.5% | 2.7 | 7 | +4.3 |
| Pickford | EVE | GKP | £5.5 | 8.8% | 2.7 | 7 | +4.3 |
| King | FUL | MID | £5.5 | 0.6% | 2.4 | 7 | +4.6 |
| Clarke | IPS | MID | £5.5 | 0.3% | 1.9 | 7 | +5.1 |
| Havertz | ARS | FWD | £7.5 | 7.5% | 3.6 | 6 | +2.4 |
| Mosquera | ARS | DEF | £5.5 | 5.8% | 3.2 | 6 | +2.8 |
| Tarkowski | EVE | DEF | £6.0 | 8.7% | 3.2 | 6 | +2.8 |
| Raya | ARS | GKP | £6.0 | 38.3% | 3.1 | 6 | +2.9 |
| Collins | BRE | DEF | £5.5 | 2.1% | 3.1 | 6 | +2.9 |
| Gomez | BHA | MID | £5.0 | 3.8% | 3.1 | 6 | +2.9 |
| Dunk | BHA | DEF | £4.5 | 1.2% | 3.0 | 6 | +3.0 |
| Boscagli | BHA | DEF | £4.5 | 0.3% | 2.9 | 6 | +3.1 |
| Lacroix | CHE | DEF | £6.0 | 9.5% | 2.9 | 6 | +3.1 |
| Bogle | LEE | DEF | £4.5 | 0.6% | 2.9 | 6 | +3.1 |
| Branthwaite | EVE | DEF | £5.5 | 1.8% | 2.8 | 6 | +3.2 |
| Mykolenko | EVE | DEF | £4.5 | 1.6% | 2.8 | 6 | +3.2 |
| Castagne | FUL | DEF | £4.5 | 0.4% | 2.7 | 6 | +3.3 |
| Rodon | LEE | DEF | £4.5 | 3.6% | 2.7 | 6 | +3.3 |
| Verbruggen | BHA | GKP | £4.5 | 21.5% | 2.6 | 6 | +3.4 |
| Jensen | BRE | MID | £5.5 | 0.2% | 2.6 | 6 | +3.4 |
| Vuskovic | BHA | DEF | £5.0 | 2.5% | 2.6 | 6 | +3.4 |
| Muharemović | LEE | DEF | £5.0 | 3.4% | 2.4 | 6 | +3.6 |
| Willock | NEW | MID | £5.0 | 0.2% | 2.2 | 6 | +3.8 |
| Coyle | HUL | DEF | £4.0 | 0.6% | 2.2 | 6 | +3.8 |
How these predicted points are produced
Onside projects expected points for every Fantasy Premier League player from the club's fixture difficulty, the player's expected goals and assists per 90, their share of available minutes, and the clean-sheet and bonus probabilities implied by the fixture. The projection for a gameweek is written once, before that gameweek's deadline, and committed to a public log — this page shows that frozen number, never a number recalculated after the result was known.
Accuracy is measured over players who reached 60 minutes. Unused substitutes are excluded on purpose: any model can look accurate by projecting zero for a bench that never plays. Mean error is the average gap in points between the projection and the official score, and it is shown next to FPL's own ep_next over the identical set of players, so the comparison is like for like.
Nothing here is curated. Every player we projected is in the table above, sortable, including the ones we called badly. The widest gap in GW1 was De Cuyper: projected 3.9, actually scored 17. A projection log you can only read forwards is the reason the hit rate above is worth anything.
— Onside (onsidearena.com), predictions graded in public.