Notes · 1 October 2026
Five gameweeks in: who is running hot
The points table and the chances table do not agree. Groß has the most points in the game with seven goal involvements from 2.6 expected. Thiago has one from 3.3. Five gameweeks is not much, but it is enough to see who is living above their chances and who is living below them.
With no league football until 10 October, I wanted to use the break for something I have been meaning to do since GW3. Put every player's goals and assists next to what the chances say they should have had, and see where the gaps are.
The number I use is expected goal involvement, xG plus xA. It measures the quality of the chances a player took and the chances he created for others. It is not perfect, but over a few games it tells you more about what comes next than the goals do. Only players with at least 250 minutes are included.
Running hot
| Player | Owned | Pts | G+A | xGI | Gap | Points rank | Model rank |
|---|---|---|---|---|---|---|---|
| Groß (BHA, £5.8m) | 28.9% | 47 | 7 | 2.63 | +4.4 | 1 | 11 |
| João Pedro (CHE, £7.7m) | 65.1% | 33 | 6 | 2.36 | +3.6 | 12 | 36 |
| Semenyo (MCI, £8.4m) | 15.3% | 33 | 5 | 1.49 | +3.5 | 11 | 29 |
| Gvardiol (MCI, £5.7m) | 26.8% | 37 | 4 | 0.96 | +3.0 | 7 | 62 |
| Cherki (MCI, £7.8m) | 28.4% | 34 | 5 | 2.26 | +2.7 | 8 | 28 |
Groß is the obvious one. 47 points, 9.4 a game, top of the whole game, and the most bought player of the break. Three goals from 1.77 xG and four assists from 0.86 xA. He is a good player in a team that creates, and I would not be surprised if he keeps finishing a bit above his numbers. But not at this rate. The model has him at 4.68 for GW6, which is still 11th best of everyone, just about half his average so far.
Three of the five are City players. Semenyo has two goals from 0.33 xG. Gvardiol has four goal involvements from under one expected, and that is the biggest gap between points and projection in the list: 7th on points, 62nd in the model. A centre back does not keep scoring and assisting like that. The clean sheets and defensive points are real, the rest mostly is not.
João Pedro is worth a line too. Six involvements from 2.36. He is still in 65% of teams, and 387,000 people have already sold him this week. If they sold because of the flag, fair enough. If they sold because of the points, the points were flattering him anyway.
Running cold
| Player | Owned | Pts | G+A | xGI | Gap | Points rank | Model rank |
|---|---|---|---|---|---|---|---|
| Thiago (BRE, £7.8m) | 9.1% | 10 | 1 | 3.29 | -2.3 | 180 | 97 |
| Barry (EVE, £5.7m) | 8.3% | 20 | 2 | 3.65 | -1.7 | 75 | 24 |
| Le Fée (SUN, £5.7m) | 3.4% | 18 | 2 | 3.40 | -1.4 | 89 | 27 |
| Szoboszlai (LIV, £7.0m) | 32.8% | 20 | 1 | 2.24 | -1.2 | 77 | 41 |
| Saka (ARS, £9.5m) | 13.7% | 32 | 3 | 4.20 | -1.2 | 13 | 2 |
Thiago has had 3.13 xG and scored once. 10 points in five games for a £7.8m striker is awful, and nobody is going to buy him on that. The model still has him outside the top 90, so this is not me telling you he is a hidden gem. It is just that he has been far more unlucky than bad.
Barry and Le Fée are the ones I find interesting. Both £5.7m, both owned by almost nobody, both with at least 3.4 expected goal involvements and two actual. On points they are 75th and 89th. In the model they are 24th and 27th. If you are looking for a cheap player on a wildcard that the rest of your mini-league does not have, this is where I would start looking.
Saka is the quiet one. 32 points is fine, so nobody is calling him a flop, but he has been getting the chances of someone with a lot more. The model has him second over the next five, only behind Haaland.
Points tell you what happened. Chances tell you more about what happens next. After five games the two lists are further apart than most people think.
And Haaland
He has been the most captained player in every single gameweek so far. Five goals from 4.42 xG. 2, 13, 9, 9, 6. That is about as honest a return as you get. Nothing to regress, nothing he is owed. He is just very good, and the model still has him top for the next five.
How much of this the model believes
The projections lean on the chances, not the goals. With 450 minutes played, a player's actual goals and assists count for 17.5% of his attacking rate and the expected numbers count for the rest. That share grows with minutes and stops at 35% around ten full games, so a player who really is a great finisher gets credit for it over time. After that the whole thing is pulled toward an average for his position, harder when his output came in one or two big games. That part is in the one big score note.
So the model is making a bet here. It is betting that Groß and Gvardiol come down and that Barry and Le Fée come up. I think that is the right bet, but I do not know yet. Some players finish above their xG for years, and five games is a small sample for anyone. The scorecards will tell us whether it was right, and I will write it up either way.
What I would do with it
Do not sell a player only because he is on the hot list. Groß is still 11th in the model at £5.8m, which is great value. Just do not pay for him as if 9.4 a game is the new normal.
And do not buy a cold player only because he is due. Being due is not a thing. What matters is whether the chances keep coming, and for Barry, Le Fée and Saka they have. You can check any player's numbers on the players page, or lock them into the wildcard planner and see what it costs you.
Data from the official FPL API as of this morning, 1 October. Ranks are among every player in the game, by total points and by projected points over GW6 to GW10. João Pedro and Semenyo are both carrying a 75% flag, which pulls their projection down a bit for GW6. The last note has more on that.