FPLRogue Gameweek 1 squad draft: the first £100m model build

The squad structure

The optimiser spends exactly £100.0m and produces a 5-3-2 starting shape: Leno; Gabriel, O'Reilly, Calafiori, Dalot and Robinson; Bruno Fernandes, Cherki and Mbeumo; then Haaland with Kostoulas up front. The bench is Palmer, Reed, Destan and Yates. Bruno narrowly leads Haaland for the current Gameweek 1 captain projection, 7.96 to 7.91, so the armband is a provisional model output rather than a settled editorial call.

The important result is the budget shape. Haaland at £15.5m and Bruno at £12.0m can coexist with Gabriel at £8.0m because the draft accepts a cheap bench and one £5.5m starting forward. That concentration improves the projected first XI but lowers resilience. If Kostoulas or Calafiori lose their expected roles, the structure needs a deliberate downgrade elsewhere rather than a like-for-like replacement that may not exist.

The squad is legal under the official position, budget and three-per-club constraints, but legality is only the first gate. It spends every £0.1m, leaving no direct response to a late price correction or reclassification. Prices are locked before the first deadline, so that is not an immediate market problem; it is a flexibility problem. The second draft should test a version with £0.5m in the bank and measure exactly which projected points are surrendered for that optionality.

Goalkeepers and defence

Leno is the starting goalkeeper at £4.5m, with 10.81 projected points and 87.3 expected minutes per match across the opening three. Palmer is a £4.0m bench goalkeeper whose value is budgetary; this draft does not assume rotation. A safer pairing would spend an extra £0.5m on two credible starters, but that money must come from a named outfield slot.

The defence carries much of the draft's upside. Gabriel projects 17.99 points, O'Reilly 17.58, Calafiori 15.67, Dalot 13.45 and Robinson 12.98. The weakness is that projections are not equal to security: Gabriel and O'Reilly are above 84 expected minutes, Dalot 79.2, Robinson 68.8 and Calafiori 62.3. Starting all five maximises the model score, but the human review should determine whether one volatile defender becomes a safer £4.5m–£5.0m option.

A five-defender starting formation is also sensitive to late attacking-role evidence. If an inexpensive midfielder or forward gains a secure set-piece or starting role, the optimal formation can change without any defender becoming a bad pick. Compare a 4-4-2 and 3-4-3 version at the next run, and judge the fifth defender against the best extra attacker rather than against other defenders alone. Formation is an output of the squad, not a preference to protect.

• Start: Leno — £4.5m.
• Bench: Palmer — £4.0m.
• Defence: Gabriel £8.0m, O'Reilly £6.5m, Calafiori £5.5m, Dalot £5.0m, Robinson £4.5m.

Midfield and forwards

Bruno is the midfield anchor at £12.0m and leads the current GW1 captain projection. Mbeumo at £8.0m provides a second United route, while Cherki at £7.5m provides City attack without another premium price. The cheap midfield places are Reed and Yates at £4.5m. Reed has the stronger three-match projection and is first in the outfield bench order; Yates is purely a budget enabler in this version because the model gives him only 46.4 expected minutes per match.

Haaland is the starting forward and vice-captain, backed by 24.62 xPts over three fixtures. Kostoulas starts at £5.5m, while Destan at £4.5m is the second bench outfielder. Kostoulas projects 12.75 points but only 64 expected minutes, so this is the key compromise. A move to a more secure £6.0m–£8.0m forward would require money from Calafiori, O'Reilly, Mbeumo or Cherki; that alternative should be tested after the final friendly evidence.

The midfield is deliberately narrow because the optimiser values the fifth defender and second forward. That makes Bruno and Mbeumo especially important: an unexpected United role change affects two starting positions at once. The editorial comparison should include a diversified version with Gibbs-White or another secure-minute midfielder, even if its raw total is slightly lower. Correlated upside can be useful, but correlated selection risk should be visible rather than hidden inside the total projection.

Bench and flexibility

The bench order is Reed first, Destan second and Yates third, with Palmer the reserve goalkeeper. It costs the minimum practical amount and protects the premium core, but it is not robust. Reed and Destan have plausible appearance routes; Yates does not currently project enough minutes to be trusted for an automatic-substitution rescue. Managers who prefer resilience should turn £0.5m–£1.0m of defensive spend into a stronger fourth substitute.

Club slots also matter. The draft uses three City players—Haaland, O'Reilly and Cherki—and three United players—Bruno, Mbeumo and Dalot. Any late City or United breakout therefore requires a transfer out from the same club. The Arsenal allocation is Gabriel and Calafiori, leaving one slot open. That flexibility has real value while Saka's fitness and the forward hierarchy remain under review.

Bench quality should be evaluated by scenarios, not total projected points. Test what happens if one defender is unexpectedly benched, if a starting midfielder misses the opening fixture and if the cheap forward plays only a late cameo. Reed can cover one absence in principle, but two simultaneous doubts expose Yates or Destan. A modest downgrade from a premium defender may improve expected points in those failure states even if it reduces the base-case optimum.

Planned changes before deadline

Three checks can change the team. First, verify Calafiori and Cherki's expected starts; a reduction in either role should move the money to a secure replacement. Second, confirm whether Kostoulas has a genuine opening-day pathway or whether the draft needs a conventional second forward. Third, rerun captaincy after the final projection refresh, because Bruno and Haaland are currently separated by less than 0.05 expected points for Gameweek 1.

This version should not be copied blindly. It is a transparent baseline showing what the model buys when asked to maximise a weighted starting XI within the official rules. The editor should compare it with a safer-bench version and a one-premium version before publication. Prices are locked until the GW1 deadline, so evidence—not fear of a rise—should drive every final alteration.

The final sign-off should record four items: the fresh model timestamp, the total cost, the exact starting eleven and the bench order. Then verify every player's club against official FPL data, because a late transfer can breach the three-player limit or invalidate a fixture projection. If any of those checks fail, return the draft to Needs review rather than repairing the prose around an obsolete squad.

A second optimiser run should use the same constraints and a saved comparison receipt. Report which players changed, how much projected value moved, how the expected-minute floor improved and whether captaincy changed. Reconfirm two goalkeepers, five defenders, five midfielders and three forwards before sign-off, plus the three-player club limit, final total budget and remaining bank balance. That audit prevents a tiny projection update from being presented as a new strategy. The published team should be the version that balances current evidence and resilience, not simply the last combination returned before the deadline.