Pick the tier you're competing with and the points gap you're working with. The page shows the best leverage picks, how risk scales with each extra swap, the per-GW captain landscape, and the full player pool.
For the selected gap. Stay level shows the template you can't miss; the bigger gaps rank players by the chance one extra differential alone closes the gap.
Hold a templated team and add bolder picks step-by-step. Each row stacks more differentials — expected points slip a little, volatility climbs, your catch-up probability climbs faster.
| Plan | Players (vs template) | Expected gain | Volatility added | Beat field by 10 | Beat field by 20 | Beat field by 30 |
|---|
Pick the players you'd own that the field mostly doesn't, and the popular players you'd skip. The readout shows how those picks shift your odds. Assumes the rest of your team mirrors the tier template.
Every player expected to feature. Right side = template (you're locked into them); left side = leverage (you take ground only if you own them). Up = high expected points, down = noise. Bubble size grows with ceiling potential.
Captaincy is a per-gameweek decision, so each gameweek is shown on its own. Field captain shares are the tier's predicted probability for that GW. The table ranks alternative captains by the chance of out-scoring the modal field captain by the selected gap — useful whether you're chasing rank or just deciding who to back this week.
| Player | xPts GW37 | Swing | Edge vs typical C | P(match) | P(+10) | P(+20) |
|---|
Every featured player, ranked by catch-up power for the selected gap and tier. Click headers to re-sort. Filter by position or search by name.
| # | Player | Pos | £ | 2GW xPts ↓ | Volatility | EO /GW | Pts vs field | P(close gap) |
|---|
Player expectations. 2GW xPts is Solio's projection for GW37 + GW38 points combined. σ is the standard deviation of that two-gameweek score, summed from per-fixture variance components — appearance (Bernoulli on 1–59 / 60+ minutes), goal/assist counts (Poisson), clean sheet (Bernoulli), saves (GKP only), cards, and the projected bonus-points variance term.
EO. Per-fixture effective ownership = lineup_share + captain_share + tc_share (so an all-tier-captained player approaches 2 per GW). The page displays "EO /GW" — the average of GW37 and GW38 EO. The math uses that per-GW number directly.
Relative variance. This page uses Solio's relative-variance decomposition: for every player×fixture,
μ_edge = (w − EO) · μ, var_edge = (w − EO)² · σ²,
where w is your slot weight in that fixture — 0 for not-owned, 1 for owned/lineup, 2 for captain. The solver expands the squared term to a MILP-friendly form: (lineup·(1−2·EO) + captain·(3−2·EO) + EO²) · σ², which checks out for all three cases. Total edge and variance are the sum across your picks; catch-up probability is 1 − Φ((G − μ_edge) / σ_edge).
Strategy combos. "1/2/3 leverage swaps" are picked greedily to maximise P(catch G) for your selected gap, drawn from the top differentials by edge mean.
Captain swing. The per-GW captain section compares a swap of captain Y → X within that gameweek alone. Pre-swap you own both at lineup, so only those two slot weights change: net delta is just X_X − X_Y (mean = μ_X − μ_Y, var = σ_X² + σ_Y²). The strategies-table "Captain swing" row optimises the two GWs independently — picking the best X for GW37 and the best X for GW38 separately, then combining the deltas (independent across GWs).
Caveats. Cross-pair correlation terms (same fixture, same team) are omitted — Solio's backend includes POSITION_CORRELATION_COEFFS for player pairs sharing a fixture, which we've left out for simplicity. The Normal approximation also smooths over FPL's long-tailed bonus point distribution. Treat the percentages as directional, not literal odds. Source data exported 2026-05-13 from Solio's optimisation backend.