Gameweek 8: do not chase the 5-0

The GW8 deadline is Friday 23 October at 18:30 UK time, earlier than usual because Ipswich play Forest at 20:00 tonight. If you normally do your transfers on Saturday morning, this week you can't.

One thing first, after Tuesday's GW7 review. The model has been too generous at the top of its list for three weeks running: the top twenty came in 1.45 points a player under projection last week. So read the numbers below as an order, not a promise. A 4.9 is better than a 4.4. It is probably not going to score 4.9.

The captain: Haaland, and it isn't close

Haaland is 0.92 clear, almost exactly the same gap as last week, when he scored twice and got 13. Villa away on Saturday lunchtime is not an easy game. It is still the best option by a distance, and he is in 77% of teams. I would not overthink it.

If you don't own him, the next three are close enough that it doesn't matter much. I would pick João Pedro or Mbeumo over a defender, because a captain needs a high ceiling, and Thomas's projection is mostly defensive contribution points, which max out at 2.

Brighton: the fixture changed

Groß had 241,000 transfers in this week and De Cuyper 218,000. Both are among the five most bought players in the game after the 5-0 against Palace. The model has Groß 29th for GW8 at 4.20 and De Cuyper 96th at 3.60. Brighton go to Anfield.

This isn't a new view. The midweek piece said GW6 and GW7 were the two good Brighton fixtures and that the problem comes after. GW6 and GW7 went well: De Cuyper got 15 and then 6, and Vuskovic got 11. Now the run turns. Over the next five, Vuskovic is still joint second in the whole game on 23.5, so Brighton are not a sell. But if you are buying one of them today for this week, you are buying last week's score.

I could be wrong. Last time I backed a cheap defender against a City clean sheet I lost. But at least this time the money is going against the model, not with it.

The cheap defender, again

Thomas is second for the whole week at £4.0m, at home to Fulham. He got 4 in GW6 and 3 in GW7. Last week he made 10 defensive actions at Spurs, which earned 2 points even though Coventry conceded three. That is the floor the model is betting on.

I have been wrong about cheap defenders before, so I will keep it simple: if you need a £4.0m defender to start, start him. I wouldn't bench a City or Arsenal defender for him.

Chelsea, after a week of 2s

João Pedro, Rogers and Palmer all got 2 last week, at Everton. The model still has all three in the top eight at home to Spurs, and Neto, who scored both goals, is 18th. Rogers (216,000 in) and Palmer (183,000 in) are also among the most bought. That is the rare week where the crowd and the model agree.

Bruno, for the third week

I said twice that I didn't see why people were selling Bruno. He got 1 point both times. This week he is the third most sold player (318,000 out), and the model has him 28th at 4.21 at home to Bournemouth. Mbeumo, in the same game, is 5th. I'm not going to argue for Bruno a third time. If you are moving him, Mbeumo is the obvious replacement, and almost £4m cheaper.

The flags

Semenyo and Szoboszlai are the two most sold players of the week, and with no return date for either, that is fair enough. Semenyo played 13 minutes last week before the knee went on the list.

Saka is the hard one. At 50% he is 154th for the week. If he plays, Everton at home is a good fixture. With the deadline on Friday evening you will not know before you have to decide. If you have a playable bench, keeping him is fine. If not, this is the week to move him.

How this gets scored

The scorecard locks the model's top twenty before the deadline, as it does every week, and I will score it after the Monday games, Brighton included. If De Cuyper gets another 15 at Anfield, it goes in the review.

If you want to see how this looks for your own team, put your ID into My Team. Good luck this week.