FPL Predicted Points 2026/27: The Best Value in the Game

The 2026/27 projections went live this week, which means that for the first time this pre-season there is a number to argue with instead of a vibe. All figures below are FPL predicted points from The Algorithm, as of 6 August 2026, covering Gameweeks 1 to 6.

The short version: the best players in the game and the best value in the game are almost entirely different lists. One of them is the one everybody is building.

What predicted points actually means

Predicted points is one number per player per gameweek: what The Algorithm expects him to return, already discounted for the chance he does not play at all. That last part matters more than people think, and we will come back to it when we get to Cole Palmer.

You do not need to take any of this on trust. Every figure in this post is on our free FPL predicted points table, all 568 players, eight gameweeks ahead, updated daily, no signup. Check the numbers against the argument.

Who has the most predicted points in FPL 2026/27?

Ranked on raw projected points across the opening six gameweeks:

Nothing shocking at the top. Haaland first, Bruno Fernandes second, and the crowd has both: 74.8 per cent and 48.7 per cent owned.

The number worth staring at is third. A centre-back, projected above Saka, above Isak, above every midfielder in the game bar one. Gabriel is 26.2 per cent owned, which means nearly three in four managers are leaving the third-best projected asset in the game out of their squad.

Points per million turns the list upside down

Divide those projections by price and the order collapses.

Haaland finishes last. Not last in his price bracket, last of the entire top 21 by projected points, on 2.39 per million against Raya's 3.87.

That is not an argument against owning him, and I want to be careful here. Points per million is the wrong objective if you take it literally, because you cannot field fifteen goalkeepers and a squad of 3.8s would have no ceiling at all. Haaland is still the single biggest points source in the game and there is a real case for paying for that, which I went through in full in our look at whether Haaland is worth £15.5m.

What the per-million column tells you is where the rest of your money should go once you have made the premium decision. And on that measure the crowd is badly positioned.

The four best-value picks on the projections

Raya (£6.0m) — Arsenal

Fixtures: Coventry (H), Aston Villa (A), Chelsea (H), Sunderland (A), Brighton (A), Leeds (H).

The dull pick is the right pick. Arsenal have the best defensive opening six in the league on our own fixture ratings, and it is not close: 5.12 expected goals conceded across the window against 6.42 for the next best, with an average clean-sheet probability of 43 per cent where nobody else clears 36.

Raya projects 4.20 in GW1 and never drops below 3.59 in the window. His chance of missing any given gameweek is 7.5 per cent, the joint-lowest in the top 20, which is what you are actually buying at goalkeeper. He returned 162 points in 3,330 minutes last season.

At 30.9 per cent owned he is not a differential, he is a consensus pick that happens to also be correct. Those exist.

Tarkowski (£6.0m) — Everton

Fixtures: Crystal Palace (H), Bournemouth (A), Manchester United (H), Tottenham (A), Ipswich (H), Hull (A).

This is the one. Tarkowski is 9.9 per cent owned and rates second in the entire game on points per million, at 3.83.

Everton's window reads seventh-best in the league defensively on our ratings, 7.70 expected goals conceded with a 29 per cent average clean-sheet probability, and the shape of it is friendly: the two fattest clean-sheet spots are GW5 at home to Ipswich (41 per cent) and GW6 away at Hull (33 per cent), which is exactly where his projection peaks at 4.33 and 4.10. He opens at 4.57 in GW1 against Palace.

Minutes are the reason to trust it rather than hope. He played 3,330 minutes last season for 170 points, and The Algorithm puts his chance of missing GW1 at 8.0 per cent, lower than any outfield player in the top 20.

The verdict: at £6.0m he does the job a £6.5m or £7.0m defender is supposed to do, and fewer than one in ten managers own him.

Gabriel (£8.0m) — Arsenal

Fixtures: Coventry (H), Aston Villa (A), Chelsea (H), Sunderland (A), Brighton (A), Leeds (H).

Expensive for a defender and still fourth in the game on value, which tells you how far ahead of the position he is projected. 30.1 points across the window, 5.79 in GW1, and a 7.5 per cent chance of missing it. He scored 209 points at 6.5 per game last season.

He is also the defender in the GW1 Pack, where The Algorithm put £8.0m into one back-line slot. That is a real commitment from a squad build that only has £100m to spend, and it is the clearest signal in the reveal.

At 26.2 per cent owned, owning him is still a mild differential. Barely.

Bruno G. (£7.0m) — Newcastle

Fixtures: Liverpool (H), Tottenham (A), Bournemouth (H), Leeds (A), Hull (H), Coventry (A).

The forgotten one. 9.2 per cent owned, 25.8 projected points, level with Isak on the raw list at £2.0m less.

His shape is the opposite of the template's. He opens at 3.90 against Liverpool, the hardest fixture on his card, then climbs: 4.50 in GW3, 4.85 in GW5, and his three best gameweeks all land after the opener. Newcastle rate sixth in the league both ways on our ratings, 9.95 expected goals scored and 7.53 conceded. He managed 154 points at 5.3 per game in 2,456 minutes last season.

Rotation risk is genuinely low for a midfielder at 9.1 per cent in GW1. The verdict: if you are squeezing a premium forward in, this is the £7.0m midfielder that lets you do it, and nine in ten managers have not noticed.

The rotation flag the raw list hides

Palmer sits eighth on projected points, which makes him look like a straightforward £9.5m buy. The per-gameweek numbers say otherwise.

The Algorithm puts a 26.4 per cent chance on him playing zero minutes in GW1, roughly double every other name in the top 20. By GW2 that falls to 14.0 per cent and stays there. His GW1 projection is 3.97, or 5.40 if he starts, and the gap between those two numbers is the entire story.

That was priced before the news arrived. He sat out Thursday's friendly against Juventus with a knock, which Chelsea described as discomfort rather than an injury and explicitly not the groin problem that disrupted his last season.

So: not a fade. A GW2 buy at 12.1 per cent owned, and if he starts on the opening weekend you will have lost nothing by waiting a week to find out.

What would change my mind

Three things, honestly.

Saliba is reported as Arsenal's main defensive absence for GW1, and both the Raya and Gabriel cases lean on an Arsenal clean sheet. That is the largest live question mark in this post.

The fixture ratings also carry a promoted-side assumption rather than a measurement, which flatters anyone playing Coventry, Hull or Ipswich early. And one real gameweek of football will move these projections more than the whole of pre-season did.

If you want to see what The Algorithm does with all of this rather than read my read of it, the demo is a real plan with no signup. If you are still deciding when to tear the whole thing up, our wildcard timing piece covers the GW4 turn.

FAQ

Where can I see FPL predicted points for free? On our predicted points table. Every player in the game, eight gameweeks ahead, refreshed daily, no account needed.

Who has the most FPL predicted points for 2026/27? Haaland, on 37.1 across GW1-6 as of 6 August, ahead of B.Fernandes on 33.7 and Gabriel on 30.1. On points per million the order reverses and Haaland finishes last of that top 21.

Which cheap FPL defender has the best predicted points? Tarkowski at £6.0m, on 23.0 across the window for 3.83 points per million, the best of any defender in the game. He is 9.9 per cent owned.

Should I captain Haaland in GW1? He is the highest projection in the game for GW1 at 6.77, narrowly ahead of B.Fernandes on 6.57, with the lower chance of missing the game of the two. Captaincy is the one place where paying for the ceiling is unambiguously right.

Numbers move daily. The table is free, so check them before the 21 August deadline rather than trusting a post you read a fortnight earlier.