Notes · 30 September 2026
Three weeks off: what actually changes
Nothing has happened on the pitch since GW5, so the projections have barely moved. Prices have moved by a tenth at most. What has changed is a pile of 75% injury flags, and they are hiding a lot more than they look like. That includes in my own model.
This is the first of the new longer breaks. The international window runs from 24 September to 6 October, a lot of countries play four games instead of two, and the Premier League does not start again until Saturday 10 October. The GW6 deadline is 11:00 UK time that morning.
Three weeks feels like it should change a lot. I went through the data this morning to see what actually has.
The projections have not moved
On Monday I wrote about the GW6 to GW10 fixture run. Haaland was projected 5.47 for GW6. Stach was 4.86. Today they are 5.47 and 4.86. Not roughly the same. Identical to two decimals.
That is how it should be. The model is built on minutes, expected goals and assists, defensive actions and fixtures, and none of those have changed since 21 September, because nobody has played a league game. A break does not make a player better or worse. It only changes what we know about whether he will play.
So when someone says the break has changed their view on a player, it is worth asking what actually happened. Usually the answer is news, a price, or three weeks of staring at the same team.
Prices: a tenth here and there
Since the GW5 deadline, 13 players have gone up and 88 have gone down. None of them by more than £0.1m. Around 4.6 million transfers have already been made for GW6, with the deadline still ten days away.
The most bought player is Groß, 779,000 transfers in and now £5.8m. The most sold is João Pedro, 387,000 out, down to £7.7m. He is still in 65% of teams.
A tenth is not nothing. It can be the difference between two players on a wildcard. But chasing it now means making a transfer with ten days of news still to come, and for a lot of the players people are moving, that news matters far more than £0.1m.
The flags are where it gets interesting
29 players are flagged as doubtful right now. 26 of them are at exactly 75%, and 15 of the 29 flags were added after GW5 finished. Isak left the Sweden camp with a thigh problem. Gakpo pulled out of the Netherlands squad with his ankle. Havertz went back to Arsenal with a hamstring after Germany's draw with the Netherlands. Semenyo was withdrawn from the Ghana squad with an ankle injury.
75% is FPL's standard flag for a knock. It is a placeholder, not a medical opinion, and it will almost certainly change before 10 October in one direction or the other.
Here is the part I need to be honest about. Until today my model took that 75% and applied it to every one of the next five gameweeks, not only GW6. A player with a small knock was marked down for GW10 as well. That was too harsh, and it made several of the most owned players in the game look a lot worse than they are.
That is now fixed. A flag counts in full for the next gameweek, halfway the week after, and not at all from there. A player ruled out with no return date still projects zero for the whole run, because guessing a return is the more expensive mistake. Here is what the change did:
| Player | Owned | GW6 | GW6 if fit | 5 GW before | 5 GW now | Rank before | Rank now |
|---|---|---|---|---|---|---|---|
| João Pedro (CHE, £7.7m) | 65.2% | 3.59 | 4.55 | 17.7 | 21.0 | 97 | 36 |
| Palmer (CHE, £9.7m) | 25.7% | 3.99 | 4.60 | 19.7 | 21.8 | 59 | 26 |
| Isak (LIV, £9.1m) | 22.4% | 3.97 | 4.84 | 19.8 | 23.0 | 55 | 16 |
| Semenyo (MCI, £8.4m) | 15.3% | 3.64 | 4.13 | 19.6 | 21.5 | 60 | 29 |
| Gakpo (LIV, £7.2m) | 12.4% | 3.48 | 4.35 | 17.5 | 20.5 | 105 | 44 |
| Rice (ARS, £7.4m) | 11.6% | 3.09 | 3.71 | 15.4 | 17.7 | 160 | 101 |
| Havertz (ARS, £7.6m) | 8.3% | 3.27 | 3.81 | 16.1 | 18.0 | 142 | 93 |
| Brobbey (SUN, £5.7m) | 7.0% | 3.68 | 4.51 | 17.8 | 20.6 | 94 | 42 |
| White (ARS, £5.5m) | 4.9% | 3.47 | 4.32 | 17.0 | 19.8 | 122 | 62 |
Rank is among every player in the game by projected points over GW6 to GW10. "GW6 if fit" is the same model with the flag removed and nothing else changed. Ownership and prices as of this morning.
Isak is the clearest case. Before the fix he was 55th, behind a lot of players nobody would pick over him. Now he is 16th, and if the flag comes off completely he is 8th. João Pedro goes from 97th to 36th, and he is the player 387,000 people have already sold.
GW6 itself has not changed, and it should not have. For next week the 75% is still the best information there is. I am not saying all of these players will be fit. Some of them will not be. The point is that GW6 is a blend of "fit" and "not fit" that will not be true on the day. Within a week each of these players becomes one or the other, and the gap between those two is often bigger than anything the fixtures do.
The break changes almost nothing on the pitch. It changes what we know about who will be on it, and most of that arrives in the last three days.
One game to watch
Liverpool against City on Sunday 11 October has Isak and Gakpo on one side and Semenyo on the other, all three flagged. Chelsea, with Palmer, João Pedro and James all at 75%, play Bournemouth on the Saturday. Between those two games you have most of the widely owned doubts in the game.
So what would I actually do
Wait. The international games finish on 6 October, most players are back at their clubs by midweek, and the press conferences are usually on the Thursday and Friday before the deadline. That is when the flags turn into real information.
If you want to move now to catch a price rise, move a player who is not in doubt for one who is not in doubt. Do not sell a flagged player just because he is flagged. A sale today costs you a tenth at most. Selling a player who turns out fit costs you the player.
If you are planning a wildcard, the wildcard planner lets you lock players in or leave them out, and shows what each choice costs in projected points. It is a quick way to see what your team looks like with and without Isak before the news settles it.
And if you looked at the players page earlier this week and wrote off Isak or João Pedro on the five-week number, look again. That number was wrong, it was my mistake, and I would rather tell you here than have you find it.