fpl model · best team · 2026/27 launch preview

Best FPL team: the model-built XI, with projected points

The best FPL team is not the eleven biggest names — it is the eleven players a model projects to score the most points, inside the rules. Below is exactly that: Bawler's model XI with per-player projected points — the full team, free, no paywall — followed by the top projected-value picks and, further down, the best points-per-£m XI of the completed 2025/26 season as the receipts. The full per-player table and the published backtest behind these numbers live on predicted points.

The model XI: most projected points per gameweek

Pos	Player	Club	Price	xPts	Owned
GKP	Raya	ARS	£6.0m	3.80	31.0%
DEF	Gabriel	ARS	£8.0m	5.06	26.9%
DEF	Virgil	LIV	£6.5m	3.82	16.8%
DEF	Thiaw	NEW	£5.0m	3.62	2.0%
MID	B.Fernandes	MUN	£12.0m	5.04	48.1%
MID	Palmer	CHE	£9.5m	4.60	11.2%
MID	Mbeumo	MUN	£8.0m	4.21	25.3%
MID	Wirtz	LIV	£7.5m	3.92	14.4%
FWD	Haaland	MCI	£15.5m	5.26	73.5%
FWD	Isak	LIV	£9.0m	4.05	13.5%
FWD	Thiago	BRE	£8.0m	3.90	16.1%

How it's picked: purely on projected points, in a legal formation with no more than three players per club — no budget cap at all. Projections are model estimates, not guarantees; the method and the full walk-forward backtest are published on predicted points.

Notice what an unlimited budget buys: an XI costing only £95.0m, with a back line of cheap defenders (£19.5m for the whole defence). That is not a glitch — it is the model reading the new defensive-contribution rule, which pays defenders 2 points for racking up tackles, blocks and interceptions. Cheap, busy defenders now project almost as many points as premium ones, so even with money as no object there is nothing at the back worth paying up for. The premiums that do make the XI are there because their attacking projection genuinely clears the field.

The value XI angle: most projected points per £m

The XI above ignores price; a real squad can't. Divide each player's projection by his price and these are the ten best returns in the game right now — the enablers that fund the premiums you'll actually captain. Almost all of them are defenders, for the same defensive-contribution reason.

#	Player	Club	Pos	Price	Owned	xPts	xPts / £m
1	Thiawnew mgr	NEW	DEF	£5.0m	2.0%	3.62	0.72
2	Richardsnew mgr	CRY	DEF	£5.0m	0.8%	3.56	0.71
3	Verbruggen	BHA	GKP	£4.5m	18.1%	3.09	0.69
4	Vuskovicnew club	BHA	DEF	£5.0m	2.7%	3.41	0.68
5	Canvotnew mgr	CRY	DEF	£5.0m	0.6%	3.41	0.68
6	Maatsen	AVL	DEF	£4.5m	0.7%	3.03	0.67
7	Lenonew mgr	FUL	GKP	£4.5m	3.3%	3.00	0.67
8	Collins	BRE	DEF	£5.5m	2.1%	3.58	0.65
9	Kelleher	BRE	GKP	£5.0m	5.7%	3.26	0.65
10	Rodon	LEE	DEF	£4.5m	2.0%	2.93	0.65

xPts / £m = projected points per gameweek divided by price. The full projected-points table for every player is on predicted points.

What a “model-built XI” actually means

Most “best FPL team” pages give you either an opinion or a description of what everyone already owns (an effective-ownership table). A model-built XI is different: it scores every player's expected return, then solves for the highest-scoring legal team — the same optimisation a serious manager does by hand, run over every player in the game at once. It is prescriptive, not descriptive. Bawler discloses these as projections, not guarantees, and publishes how the model actually performed on the predicted-points backtest and methodology page — we never invent an accuracy figure.

One promise sits underneath all of it: the whole answer is free. The best-known predicted-XI tools reveal a handful of rows and paywall the rest. Every table on this page is complete and rendered straight into the page — nothing hidden behind a sign-up.

Last season's receipts: the 2025/26 best-value XI (actuals, not projections)

Everything from here down is what actually happened in 2025/26 — settled points, final prices, no modelling. It is the backward-looking proof of the same idea the model XI runs forward. This is the highest points-per-£m player at each position that still forms a legal XI — one goalkeeper, a valid outfield shape, and no more than three players from any single club, with a 1,500-minute floor so the list is regular starters, not one-week flukes. The optimiser landed on a 5-4-1: eleven players for just £57.9m, returning 1,796 Fantasy points across the season.

Source: Official FPL API, 2025/26 season, all 38 gameweeks settled. Value = total points ÷ final price. Legal XI: 1 GK, valid outfield shape, max three per club, ≥1,500 minutes. Ownership shown for context — note how thinly some of these were owned (E. Le Fée just 1.6%), which is exactly where a value model finds an edge.

2025/26 value leaders by position (every list, complete)

The XI above is one legal shape; the real toolkit is the full value board. These are the top six points-per-£m returns at each position over 2025/26 (actuals, 1,500-minute floor). Defenders dominate FPL value — the best of them, Guéhi (Man City), returned 35.1 points per £m, more than double what any premium forward managed.

Source: Official FPL API, 2025/26 final. Read alongside the bonus-point magnets study — several of these value defenders were also quiet bonus earners, which compounds the return.

Why the best team isn't just the biggest names

Erling Haaland was the top raw scorer of 2025/26 with 239 points — but at £14.7m he converted only 16.3 points per £m, less than half of Guéhi. Mohamed Salah, at £14.0m, returned just 8.8 — a premium price for a mid-table value season. That is the whole reason a value-first model exists: the same £14m spent on two or three efficient starters buys more total points than one superstar.

The honest counterpoint — and the reason the value XI is a lesson, not a literal team-sheet — is that a real season needs a captain, and captaincy multiplies. A £14.7m striker who scores triple on his big weeks can be worth his price through the armband alone, which raw points-per-£m does not capture. The best actual squad usually pairs a couple of premiums you'll captain with a spine of these value picks. The model's job is to find that balance; see how it ranks the armband on the captaincy guide.

Squad-building principles the model follows

• Spend for points-per-£m, not for names. Every non-captain slot is a value decision. The value board above is where the search starts.
• Fund one or two premiums you'll captain. Concentrate your budget where the ×2 (or ×3) multiplier lands, then find efficiency everywhere else.
• Respect the constraints. £100.0m, 15 players (2 GK, 5 DEF, 5 MID, 3 FWD), max three per club — a legal squad is the first filter, not an afterthought.
• Build the bench to actually play. Cheap, nailed-on starters (the £4.5–5.0m defenders above) beat expensive non-players — they matter most on a Bench Boost.
• Weight the picks by fixtures. Value is a season average; the next six gameweeks are what you buy. The model tilts toward good runs using the fixture-difficulty model.
• Plan the rebuild. A squad is never finished — the Wildcard and Free Hit exist to reset it around fixture swings and blank weeks.

All figures on this page are computed from the Official FPL API — pure fantasy-points analytics, no betting angle, and projections are model estimates, not guarantees. Looking for the season's top raw scorers instead? The FPL hub ranks the highest points-scorers by position. From here, plan the rest of the squad: predicted points (the full table and backtest), fixture difficulty, chips, Wildcard and captaincy. Then give it a team name worth reading on the mini-league table.