FPL GW1 Tips: Three Names, Then a 1.05 Cliff

Squads lock at 18:30 tonight. The Algorithm's Gameweek 1 board has an unusual shape: three players sit clear of the field, and then the numbers fall off a cliff. That shape decides how much a late transfer is actually worth, so it is where these FPL GW1 tips start.

Every projection below is The Algorithm's expected points for Gameweek 1, as of 21 August 2026. You can check all of them yourself on the free FPL predicted points table.

Who should I captain in GW1?

Haaland (£15.5m) projects 6.58. Bruno Fernandes (£12.0m) projects 6.22. The gap is 0.36.

Two days ago that gap was 0.13, and we called it a coin toss in print. It is no longer a coin toss. It has also widened toward the player 69.4 per cent of managers already own, which makes the safe pick and the model's pick the same pick this week.

Fernandes is the better contrarian play if you want one, and Man Utd at Hull is the round's second-best clean sheet number at 47.3 per cent with the third-highest attacking number at 2.24, so the fixture backs him. But you are chasing 0.36 points of projection to differentiate against half the game.

Verdict: captain Haaland, vice Fernandes. If you own only one of them, the other is not worth a transfer tonight.

The tier break nobody is pricing

Gabriel (£8.0m) is third on the board at 5.77. Fourth is Mbeumo at 4.72.

That drop of 1.05 is larger than the entire gap between fourth and twentieth, which is 0.65. In other words, once you are past the top three, forty-odd players are separated by rounding errors.

This is the practical consequence, and it is the most useful thing on this page: a one-week upgrade inside that flat middle band is worth well under a point, and a hit costs four. Own the top of the board, then stop tinkering.

Which fixtures should I load into?

Arsenal at home to Coventry is the best fixture in the round at both ends: 2.67 expected goals for and a 60.6 per cent chance of a clean sheet, both the highest of any side this week. Coventry, facing them, sit at 6.9 per cent.

Man Utd away at Hull is second on clean sheets at 47.3 per cent. Man City at home to Bournemouth carry 2.37 expected goals against a Bournemouth side at 9.3 per cent to keep it out.

The sides to avoid backing this week are the ones facing those three. Coventry (0.50 expected goals), Hull (0.75) and Bournemouth (1.08) hold three of the four weakest attacking numbers in the round.

Liverpool are the interesting middle case. They project 1.99 expected goals away at Newcastle, fourth-best in the round, but only a 25.6 per cent clean sheet. Buy their attackers, not their defence.

Three transfer calls

Gabriel (£8.0m), Arsenal Fixtures: Coventry (H)

The highest-projected defender in the game this week by a distance, and third overall across all positions. He scored 209 points in 2,750 minutes last season, the sort of return that comes from a centre-back who attacks set pieces rather than one who waits for clean sheets. Arsenal's 60.6 per cent is the only clean sheet number in the round better than a coin toss. At 29.7 per cent ownership he is not a differential, he is simply correct.

Semenyo (£8.5m), Man City Fixtures: Bournemouth (H)

Fifth on the board at 4.68. Last season he returned 202 points in 3,200 minutes on 14.21 xGI, and he now does it in a side projecting 2.37 goals against the division's third-weakest defensive number this week. Rotation is the honest risk in a City attack, but the fixture is close to the best available. Owned by 26.4 per cent.

Calvert-Lewin (£6.0m), Leeds: the fade Fixtures: Forest (A)

This is where The Algorithm disagrees with the crowd. Almost a third of the game, 30.1 per cent, is starting him. Leeds away at Forest is one of the five weakest attacking numbers in the round at 1.21 expected goals, and Forest are 29.9 per cent to keep a clean sheet. He projects 3.89, which is 28th this week.

He scored 142 points in 2,721 minutes last season, so this is not a verdict on the player. It is a verdict on the week. Bench him if you have a credible alternative, but do not take a hit to remove him, because the tier break above says that trade cannot pay.

One more for the anxious: João Pedro (£7.5m) is the second-most-owned player in the game at 63.2 per cent and projects 4.18, thirteenth this week. That is fine. You are not behind for owning him, and he is not your captain.

The differential: Saka at 9.2 per cent

Saka (£9.5m) is 9.2 per cent owned and plays in the best attacking fixture of the round. Mikel Arteta confirmed on Thursday that he and Declan Rice are available, saying they are "ready to be involved".

The Algorithm projects him at 4.10, eighteenth this week. If he starts, it has him at 5.31. The difference between those two numbers is a 22.7 per cent chance he does not start at all, the highest minutes risk of anyone in this week's top forty.

That is the entire case, stated honestly. "Ready to be involved" after a summer break is not the same sentence as "plays 90 minutes", and our number says so. If you want a low-owned route into the round's best fixture and you can live with a one-in-four chance of a blank, he is the pick. If you need certainty tonight, Gabriel is the safer way into the same match.

Availability: what actually matters tonight

Saliba and Timber are both out for Arsenal, with Arteta saying Timber is "not training with the group yet".

Arsenal's Bruno Guimarães (£7.0m, 6.6 per cent owned) is a genuine doubt with a thigh problem and is officially flagged at a 75 per cent chance of playing. Note the name carefully: this is not Bruno Fernandes, who carries no injury note and remains our second-ranked captaincy option.

Elsewhere: Fofana misses Chelsea's first two gameweeks through suspension. Leeds are without Gudmundsson and Gnonto. Forest are missing Yates and Savona. Brighton remain without Ferguson, Mitoma and Tzimas. At Man Utd, Sesko is back in training and Mount is described by Michael Carrick as "touch and go".

FAQ

Who is the best captain for FPL GW1? Haaland, on 6.58 projected points against Bruno Fernandes on 6.22. The 0.36 gap is small but it is real, and it points at the 69.4 per cent owned option.

Should I take a hit before the GW1 deadline? Almost certainly not. Across this week's forty highest-projected players the board runs from 6.58 down to 3.61, a total spread of 2.97 points. A hit costs four. A single-week upgrade mathematically cannot pay for itself unless you are replacing a player who will not feature at all.

Who is a good GW1 differential? Saka at 9.2 per cent ownership, in the fixture with the round's highest attacking number. The trade-off is a 22.7 per cent chance of no minutes, so treat it as a punt with a ceiling rather than a safe pick.

Want a second opinion on the whole squad before the deadline? Score it free, no signup, with Rate My Team, or read how The Algorithm builds a gameweek. If you are still weighing chips, our triple captain breakdown and vice captain guide both apply from tonight.