Raya is the premium model leader at 12.82 xPts, Donnarumma follows at 12.15 and Leno is the best £4.5m option at 10.81. The default structure is one playing goalkeeper plus the cheapest credible backup; pay for rotation only when the second keeper's fixtures and starts justify the money and weekly decisions.
The price-band verdict
The model gap from £6.0m Raya to £4.5m Leno is 2.01 projected points across three matches. That is real but small enough that the £1.5m saving can win elsewhere. Donnarumma at £5.5m offers a City route at 12.15, while Lammens, Roefs, Kelleher and Sels fill the £5.0m middle.
A price band is a squad-construction tool, not a claim that every player at the same cost performs the same job. Start by defining the band and position filter, then compare xPts, xMins, start probability, ownership, role and fixture horizon. A chart sorted only by central projection can hide that the top name is a volatile substitute while the next name is a reliable starter.
The second test is replacement cost. Ask what happens if the pick loses his place or suffers an injury. A dense price point with several viable pivots is easier to manage than an isolated player who requires two transfers to replace. Money in the bank can protect that pathway, but unused money should have a documented purpose.
Reader-facing rankings should show the trade-off, not only the order. For each candidate retain official price, expected points, expected minutes and the factor most likely to invalidate the pick. A secure player can rank below an explosive one and still be the correct choice for a fragile squad. Conversely, a strong bench can make a calculated minutes risk worthwhile. This is why the article gives structures and review triggers alongside the chart.
Premium options
Raya combines 12.82 xPts with 87.3 expected minutes and an Arsenal defensive route. Donnarumma has 12.15 at the same xMins. Premium keepers reduce weekly rotation decisions and can capture clean sheets plus bonus, but they also consume money and a club slot that may be more valuable in defence or attack.
Premium spending buys some combination of reliable minutes, team strength and multiple routes to points. It should not be justified by price itself. Compare the premium with the best budget alternative plus the exact upgrade created elsewhere. If the combined alternative scores more over the planning horizon and preserves captaincy, the premium can be omitted.
Club limits create another opportunity cost. A premium defender or goalkeeper can block an attacker from the same team, while a cheap enabler from a strong club may consume a scarce slot. Rank routes within each club before comparing across price bands. This is especially important for Manchester City and Arsenal in the opening model.
Price-band analysis should be rerun after every official price, position or club change. Historical points can provide context but must not be substituted for the current role. When a player transfers, the previous projection becomes non-portable until team attack, defence, fixtures, set pieces and competition are recalculated. If that recalculation is unavailable, remove the player from the ordered shortlist and explain the evidence gap.
Opening goalkeeper projection
Three-Gameweek xPts with official prices.
Key takeaway: Three-Gameweek xPts with official prices.
View chart as a data table
| Item | GW1-3 xPts | Note |
|---|---|---|
| Raya | 12.82 pts | £6.0m |
| Donnarumma | 12.15 pts | £5.5m |
| Lammens | 10.83 pts | £5.0m |
| Leno | 10.81 pts | £4.5m |
| Roefs | 10.78 pts | £5.0m |
| Kelleher | 10.65 pts | £5.0m |
| Verbruggen | 10.39 pts | £4.5m |
Mid-priced options
Lammens leads £5.0m keepers at 10.83 xPts, narrowly ahead of Roefs and Kelleher. The choice depends on team defence, save volume and fixture path, not the first match alone. If a £5.0m keeper is paired with another starter, the structure quickly approaches premium spend without guaranteeing premium output.
Mid-priced players often provide the easiest transfer flexibility because adjacent options are plentiful. Their role is to deliver starter-level points without requiring captaincy. Prefer secure minutes and repeatable involvement over a pre-season scoreline. If two players project similarly, the one with clearer set pieces or a more stable starting position generally provides the better floor.
A mid-priced pick can still be structurally awkward when two teammates compete for the same FPL slot or when the next move requires selling a premium. Map at least one upward and downward pivot. The editor should show readers what new evidence would trigger each route instead of offering a static list.
The planning horizon should match the intended transfer cycle. Opening squads can look across six fixtures while the embedded model chart covers three; state that difference and avoid presenting the shorter ranking as a season forecast. A player with a difficult first match may still be the better hold, and a one-week bargain may create an immediate transfer booking. Future moves have value because free transfers are scarce.
Budget options
Leno leads the £4.5m bracket at 10.81 xPts, with Verbruggen at 10.39, Perri at 9.11 and Petrovic at 8.89. Palmer and Dovin are £4.0m keepers with current full-match projections, but promoted-team and role assumptions need final verification. A non-playing £4.0m backup is acceptable only when the starter is secure and the saved £0.5m is useful. Note: Perri and Dovin could be replaced by new signings.
Budget value is created by credible minutes and the upgrade funded elsewhere. A minimum-price player with no route to the pitch is not an enabler when he is first off the bench. Separate starters, rotation options and pure placeholders. Expected minutes is the fastest first filter, but current injury, transfer and friendly evidence can override a stale model input.
Ownership can reveal where a failed assumption will be widely shared, but it is not proof of quality. A highly owned cheap player may protect rank while creating the same weakness in many squads. A low-owned option may be a genuine discovery or simply less secure. Keep the decision grounded in role, minutes and price pathways.
Bench order is part of the price decision. The first substitute needs the strongest combination of appearance probability and points when called; the cheapest player is not automatically last. Avoid pairing a volatile starter with a non-playing substitute in the same position. One unexpected absence should still leave a legal formation, and two absences should not force an early hit unless the structure consciously accepted that risk.
Best squad structures
The simplest route is one secure £4.5m starter plus a minimal backup. A premium plus minimal backup buys defence and fewer decisions. A rotating £4.5m pair can improve fixtures but costs £0.5m and creates bench-point risk. Compare the modelled rotation gain over six weeks with the best outfield use of that money before committing.
A complete structure must satisfy the £100.0m budget, position quotas, three-per-club rule, legal starting formation and four-player bench. Test it with one unexpected absence and one planned transfer. If a single benching breaks the formation or forces a hit, the savings have been overused. A modestly lower projected starter can be worth selecting when he protects several future decisions.
The final review should refresh prices, positions, team assignments and availability from official FPL, then rerun the model under the current fixtures. Confirm that chart values, prose and social summaries use the same snapshot. Name the premium or flexibility bought by each enabler; if the answer is unclear, select the stronger player rather than the cheaper one.
For the final editor pass, rebuild the example allocation from zero rather than trusting arithmetic carried through revisions. Check all fifteen positions, total cost, club counts, starting formation, captain, vice-captain and bench. Then compare the final names with the chart snapshot and current official data. This mechanical check catches duplicate club slots, stale prices and recommendations whose written rationale no longer matches the selected structure.
Finally, separate editorial confidence from model rank. High confidence requires current official team and availability data, a credible role, a legal example structure and a chart generated from the same snapshot as the copy. Medium confidence is appropriate when minutes or tactical usage still needs final verification. Low confidence means the name belongs on a watchlist rather than in the ordered recommendations. Recording that state helps later automation refresh only the claims whose evidence has actually changed, while leaving stable rules and structural guidance untouched. Keep an explicit review timestamp beside that confidence so readers and editors can tell whether a late team-news change has already been incorporated. If confidence differs by player, show it on the individual recommendation rather than applying one blanket label to the entire band.
How this article was checked
Goalkeepers are compared on official price, frozen three-Gameweek xPts and expected minutes. The recommendation includes outfield opportunity cost and does not assume rotation always adds value.
- FPL Rogue data and models — Frozen three-Gameweek xPts, xMins, start probability, price and ownership.
- FPL Rogue data and models — Accepted engine, recipe and methodology status.
- Fantasy Premier League — Official player, team, price, position, status and GW1 deadline data.
- Fantasy Premier League — Official 2026/27 fixture list.





