Are the most expensive players in the game worth it?

Every season opens with the same argument: someone costs £15.5m, the rest of the squad has to live on what's left, and you need to decide whether that price is worth paying. We've tried to answer it properly, using model expected points from the 10 August snapshot (10,000 simulations, odds-backed team rates, gameweeks 1–6), the same MILP that runs the site's optimiser, and seven seasons of finished results from the archive.

Only four players in the game cost £9.5m or more, and third place is a tie: Erling Haaland at £15.5m, Bruno Fernandes at £12.0m, then Bukayo Saka and Cole Palmer level on £9.5m. Strictly speaking that's three price tags shared between four names, and we'll cover all four.

What “worth it” means here

Scoring a lot of points doesn't make a premium worth it on its own. He earns his price only if he outscores the alternative use of the same money — and in FPL the alternative is never a single player, it's a chain of small upgrades spread across the fourteen other slots. With a £100.0m budget that's the only definition that holds up, so it's the one we test against.

Start with the crudest version: expected points per £m over the opening six gameweeks. Among the 111 players the model expects to average 60+ minutes, the median is 3.46 xP per £m.

On its own that proves nothing, because points per £m is the wrong objective: you have fifteen slots rather than infinite ones, and four of them form a bench you mostly don't score from. If money were the only constraint, the answer would be a squad full of £4.5m defenders. The real question is whether the slot constraint bites hard enough to justify concentrating funds in one player — so the next step is to look at what the money buys elsewhere.

The cheaper players who out-project them

Now for the awkward comparison: six players in the pool project higher than Haaland over the same six gameweeks, and five of them cost £8.0m or less — roughly half his price. The sixth is Bruno Fernandes.

Saka faces the same problem inside his own dressing room: he projects 21.8 xP, while Declan Rice projects 22.8 from the identical fixture list at £7.5m and Gabriel 29.2 at £8.0m. Talent has little to do with the gap — it comes down to minutes. The model gives Saka a 0.68 chance of starting and 65 expected minutes per game against Rice's 0.85 and 78, a discount earned by starting 25 of 38 league games last season for 2,218 minutes.

The squad test: what owning them costs

Points per £m ignores the slot constraint, so we ran a test that respects it: the site's own MILP over the six-gameweek horizon on a £100.0m budget, 2/5/5/3, maximum three per club, best legal XI plus captain. Once with a free hand, then once more for each premium forced into the squad. The difference between builds is the true cost of ownership, in points, with every knock-on effect priced in.

Given a free hand, the optimiser spends £99.5m and scores 327.0 over six gameweeks. Bruno Fernandes makes the squad; Haaland, Saka and Palmer don't.

It's worth noticing what the Haaland row doesn't say. Forcing him in doesn't hollow out the XI — spend on the starting eleven is £82.5m either way. What changes is that Bruno drops out and the goalkeeper slot slides from Raya to Leno, so the money comes off the edges of the squad rather than its middle. The total damage runs to two points over six gameweeks, or about a third of a point a week.

Break-even: how wrong does the model have to be?

Two points is well inside the model's own error bars, so the honest way to close the argument is to ask how much each player would have to beat his projection by before the optimiser buys him voluntarily. We scaled each premium's expected points up until the constrained squad matched the free one.

Those three numbers carry most of the argument. Haaland needs the model to be 7% light on him, which is entirely plausible — if you believe the simulation is underrating Manchester City's route to goal, buy him with a clear conscience. Saka needs 22%, a big ask when the shortfall comes from minutes rather than quality. Palmer needs 39%, and nothing in the data gets close to supporting that.

The captaincy defence, and why it fails

The strongest argument for a £15.5m striker is the armband. Captaincy doubles a player's points without doubling his price, so someone who wears it most weeks is worth more than his raw expected points suggest. The logic is perfectly sound; it runs into trouble the moment you check the specifics.

Haaland isn't the model's top captain pick in any of the first six gameweeks. Worse, when we took the squad that was forced to buy him and asked who it would captain each week, it never once chose him.

So buying the most expensive player in the game improves your captaincy over the opening six weeks by precisely nothing. Even at home to Coventry — the fixture that's supposed to make his whole case — his own teammate O'Reilly out-projects him.

What seven seasons of history says

The model only covers six gameweeks, so it helps to check a much longer record. The archive is blunt on the subject: since 2019-20, the game's most expensive player at the opening deadline has finished as the season's top scorer exactly once.

Once in seven seasons, and the misses weren't close: in 2023-24 the priciest asset in the game returned 217 while a £5.0m Cole Palmer returned 244, and last season Salah cost £14.5m and delivered 123. Season after season, the best value has come from the bracket just below the top one — the place where price hasn't yet caught up with performance.

The case for each of them

So, are they worth it?

• Bruno Fernandes: yes. He's the only one the optimiser buys unprompted, and he's the highest projection of the four.
• Haaland: not on the numbers, but the gap is two points over six gameweeks and 74% ownership. If you want to own him, own him — just don't expect the armband, because the model has someone better every single week.
• Saka at £9.5m: no. It's the 25 starts from 38 that sink him, not the talent.
• Palmer at £9.5m: no, and by the widest margin of the four.
• The real finding is structural: the money is in the £5.0–8.0m band this year. Gabriel, O'Reilly, Szoboszlai, Watkins and Thiaw all project within 2 points of Haaland, and Mbeumo beats him outright by 2.9. None of the six costs more than £8.0m.