2025/26 is a reconstructive replay (cutoff = first kickoff − 90 minutes, not a live two-hour freeze). 2026/27 pages are an illustrative sample of the live format. What that means

FPL Replay

FPL Gameweek 2 (2026/27): Team, Captain and Transfers

For FPL 2026/27 Gameweek 2 (deadline 24 Aug 2026, 10:00 UTC), the FPL Replay optimiser picked Haaland as captain and Palmer as vice, a 3-5-2, and one free transfer (Eze out, Jota in), projecting 75.9 points. The plan is an illustrative sample frozen at 24 Aug 2026, 08:00 UTC in the live-product format. It scored 98 net points after hits.

Cutoff 24 Aug 2026, 08:00 UTCIllustrative samplesha256:d41d8cd9…bc8fDownload snapshotSettledIllustrative sample

Key facts

Captain
Haaland (10.4 xP)
Vice-captain
Palmer
Formation
3-5-2
Transfers
Eze → Jota
Bank
£0.0m
Hits
None
Chip
None
Projected
75.9 xP
Points
98
Status
Settled

Squad

Starting XI and bench with projected and realised points
PosPlayerClubFixtureFDR£mxPPtsRole
GKPRayaARSAVL (A)35.54.18
DEFSalibaARSAVL (A)36.04.06
DEFGvardiolMCIIPS (H)16.05.86
DEFMuñozCRYWHU (H)25.04.32
MIDSakaARSAVL (A)310.06.26
MIDPalmerCHEWOL (A)210.57.417Vice-captain
MIDSonTOTEVE (H)210.07.216
MIDJotaLIVBRE (H)27.56.45
MIDRogersAVLARS (H)45.03.82
FWDHaalandMCIIPS (H)115.110.434Captain
FWDIsakNEWBOU (A)28.55.92
Bench
GKPValdimarssonBRELIV (A)54.00.00Sub #1
FWDJoao PedroBHAMUN (H)35.54.57Sub #2
DEFFaesLEIFUL (A)34.03.18Sub #3
DEFBarcoBHAMUN (H)34.00.00Sub #4

Why this plan

GW2 demonstrated the payoff of keeping £0.5m in the bank from GW1. While the baseline heuristic rolled the transfer, the MILP solver spent the FT on Jota (H vs BRE), generating an immediate +4.1 xP model delta. Realised score hit 98 points as Haaland delivered 17 points (34 with armband), Palmer exploded for 17, and Son registered 16.

Transfers

Eze out (£7.0m) → Jota in (£7.5m). Direct route into Liverpool's Brentford home fixture. Jota out-expected Eze by +1.56 xP in single-week horizon and +3.2 xP over the 3-GW rolling window.

  • Roll transfer vs Eze to Jota: The solver calculated an EV threshold of +1.4 xP needed to justify spending a transfer rather than rolling. Jota's projection cleared this hurdle at +1.56 xP immediate advantage.
  • Captaincy consensus: Haaland vs Ipswich (H) yielded an overwhelming 10.45 xP projection with a 58% clean sheet concession probability for the opposition, rendering alternative captain arms mathematically irrational.
  • Bench sacrifice: Faes's 8-point haul on the bench was unavoidable under rational pre-match models (£4.0m Leicester defender vs Fulham away had lower EV than Arsenal and City defensive starters).

What extra resources were worth

An extra £1.0m was worth +0.51 projected points.

An extra free transfer was worth +2.15 projected points.

Expected points from auto-subs: 1.48.

Agent arm attempted an aggressive -4 hit to bring in Minteh alongside Jota. The -4 hit was not recovered (+3 pts gained minus 4 hit = net -1 pt), reinforcing the MILP arm's conservative hit policy.

Three approaches compared

Template (hold transfers), optimiser (selected plan), and high-ceiling (more risk).

ApproachFormationCaptainTransfersProjectedvs TemplatePointsNoteDetail
Template
Rolling 1 Free Transfer into GW3 to accumulate 2 FTs for the international break buffer.
3-5-2Haaland (Son)071.889Top 12% percentileDetail
Optimiser (selected)
Single FT deployment: Eze to Jota using £0.5m liquidity; locks Haaland captaincy vs Ipswich (H).
3-5-2Haaland (Palmer)175.9+4.198Top 1.8% percentileDetail
High-ceiling
Double transfer with -4 hit: Eze + Rogers to Jota + Minteh, hunting high delta against the template.
3-5-2Haaland (Palmer)2 (−4 hit)74.2+2.491Top 8% percentileDetail

Chip check

No chip is in the selected plan this week. Triple Captain: Hold. Best projected window: GW34 and GW37. Bench Boost: Do not play this week. Best projected window: GW34 and GW37. Free Hit: Do not play this week. Best projected window: GW29 and GW34. Wildcard: Optional this week. Best projected window: GW4 and GW6.

  • Triple Captain: Hold. Best projected window: GW34 and GW37.
  • Bench Boost: Do not play this week. Best projected window: GW34 and GW37.
  • Free Hit: Do not play this week. Best projected window: GW29 and GW34.
  • Wildcard: Optional this week. Best projected window: GW4 and GW6.

What-ifs for this gameweek

  • Triple Captain Early Burn vs Baseline Preservation

    Deploying Triple Captain on Haaland against newly promoted Ipswich at the Etihad (projected 10.45 xP) produces higher certainty than waiting for spring Double Gameweeks where rotation and fixture congestion degrade per-minute output.

    Control 75.9 xP · Alternative 86.3 xP · Delta +10.4 xP

How to verify

This Gameweek 2 plan was written to a JSON snapshot at 24 Aug 2026, 08:00 UTC and hashed with SHA-256 (sha256:d41d8cd98f00b204e9800998ecf8427e324a18b76e1a90c0ef49a9987812bc8f). Download the snapshot to inspect the frozen inputs. The freeze rule is described in Methods.

Download the JSON snapshot · Methods · Freeze and hash

Who should I captain in Gameweek 2?

The selected plan captains Haaland (10.4 projected points) against IPS (H). Vice-captain is Palmer.

Should I use a chip in Gameweek 2?

No chip is in the selected plan this week. Triple Captain: Hold. Best projected window: GW34 and GW37. Bench Boost: Do not play this week. Best projected window: GW34 and GW37. Free Hit: Do not play this week. Best projected window: GW29 and GW34. Wildcard: Optional this week. Best projected window: GW4 and GW6.

Is Gameweek 2 a live freeze or a replay?

An illustrative sample of the live page format, not a live two-hour freeze.