FPL GW2 Preview: The Armband Flipped Overnight

Two days ago The Algorithm had Bruno Fernandes ahead of Erling Haaland for the armband. As of this morning, 28 August 2026, it does not. That is the honest starting point for this FPL GW2 preview: the numbers moved, the order reversed, and the gap is still small enough that either answer is defensible.

Deadline is 18:30 tonight. Here is what changed, and what it should change in your team.

Who to captain in FPL GW2

Haaland (£15.5m) projects 6.63 this week. Fernandes (£12.0m) projects 6.52. That is a gap of 0.11, which is a coin toss dressed up as a decision, and anyone telling you otherwise is selling certainty they do not have.

Wednesday's captain piece had Fernandes narrowly in front on the numbers as they read then. They have since moved. We would rather say that out loud than quietly publish a different answer.

The tiebreak is not this week. It is next week. In GW3 Haaland reads 7.02 at home to Coventry, while Fernandes drops to 5.13 at Everton. Over the two rounds together it is 13.65 to 11.65, a full two points, and the captaincy of GW3 is not close at all.

Verdict: Haaland. Not because this week separates them, but because captaining Fernandes now means handing the armband back in seven days anyway. Haaland is 68.0 per cent owned, so this is the safe answer as well as the marginally better one.

The GW2 board is a Man Utd board

Six of the 25 highest projected players in the round are Man Utd, at home to Ipswich. Fernandes at 6.52 and Bryan Mbeumo (£8.0m) at 5.40 are the obvious two. The interesting part is the other four: Matheus Cunha (£8.0m) at 4.35, Senne Lammens (£5.0m) at 3.97, Harry Maguire (£5.0m) at 3.96 and Luke Shaw (£4.5m) at 3.91.

That is a whole defensive unit priced at £5.0m or less appearing inside the top 25 of a 616-player game. It is the cheapest route into the best-projected fixture on the card.

One caution before you buy the cheapest name in that list. Cunha carries a 14.9 per cent chance of no minutes this week against 7.5 per cent for Fernandes, so the same club and the same fixture do not mean the same start.

Three transfer calls

Luke Shaw (£4.5m), Man Utd Fixtures: Ipswich (H), Everton (A)

Shaw projects 3.91 this week. Riccardo Calafiori (£5.6m) projects 3.85. Calafiori is owned by 41.5 per cent of the game after nine points in the opener and Shaw by 19.8 per cent, so the crowd is paying £1.1m more for a number that is fractionally lower.

Shaw also did real attacking work in GW1: 0.26 xA and an ICT of 9.0, the highest of any Man Utd defender. Arsenal go to Villa this week, Man Utd host a promoted side, and the money you free up plays somewhere else.

This is the call where we disagree with the template. Fewer than one in five own Shaw.

Cole Palmer (£9.5m), Chelsea Fixtures: Brighton (H), Arsenal (A), Hull (H)

Palmer was the best attacking midfielder in the opening round on the eye and on the numbers: 13 points, 0.47 xGI and an ICT of 16.1, the highest in the game. He is owned by 14.9 per cent.

The catch is the shape of his next three. He reads 4.58 this week, 3.07 at Arsenal in GW3, then 5.45 at home to Hull in GW4. Buy him, by all means, but do not buy him as a captain option, because the middle week of that run is the worst of his next four.

Alexander Isak (£9.0m), Liverpool Fixtures: Forest (H), Ipswich (A)

Isak posted 1.09 xG in GW1, the highest of any forward, and did not score. He took two points and a lot of abuse. He is 16.5 per cent owned and projects 4.71 this week, second among all forwards behind Haaland.

Selling a striker who created the most and converted none of it is how managers spend transfers to buy last week. Hold. If you have not got him, this is the cheapest he will look for a while.

The differential: Cody Gakpo

Cody Gakpo (£7.0m), Liverpool Fixtures: Forest (H), Ipswich (A)

Gakpo took 12 points in the opener off 0.45 xGI and an ICT of 11.5, and is still owned by only 5.8 per cent. He projects 4.28 this week and 23.7 across the next six, which puts him ahead of several £8.0m-plus midfielders on our board.

Liverpool are at home to Forest and then travel to a promoted side, which is the softest attacking pair Gakpo gets in the opening two months. Diogo Szoboszlai at the same £7.0m price is 43.0 per cent owned and projects 5.22, so the premium version of this pick is not the secret. The point is that you can own the fixture at 5.8 per cent instead.

Verdict: the best sub-10 per cent pick on the board this week.

The minutes risk that actually matters

Ignore the doubt flags for a moment and check something simpler in your own app: whether your player has played at all.

18.7 per cent of managers own Dubravka, who has 0 minutes this season. Antonin Kinsky, same club, same position, is 22.9 per cent owned and played all 90 in GW1. Pedro Porro at 13.7 per cent and Ezri Konsa at 11.1 per cent are also on zero minutes.

None of that is an injury story we can confirm for you. It is a starting-eleven story, and it is checkable in thirty seconds before 18:30.

FAQ

Who should I captain in FPL GW2? Haaland. He edges this week 6.63 to 6.52, which is inside the noise, but he wins GW3 by 7.02 to 5.13, so he wins the pair.

Is it worth taking a hit in GW2? Almost never for a one-week gain. The full spread between a median starter and the best player on the board is smaller than the four points a hit costs, which we worked through in our piece on free transfers and when a hit pays.

Who is the best GW2 differential? Gakpo at 5.8 per cent owned, with 12 points already banked and the friendlier of Liverpool's two fixtures first.

All projections are The Algorithm's own, as of 28 August 2026, and you can check every one of them on our free predicted points table. If you want a second opinion on your whole squad before the deadline, rate my team scores it free, no signup.