Every cell in the grid is one team's next fixture, coloured 1–5 — 1 is the easiest fixture and 5 the hardest. Nothing here is hand-set: the colour is computed from how strong the two teams are, so it updates as the underlying ratings do.

Where the difficulty comes from

We compare the two sides' attacking and defensive strength, add a bump for playing at home, and turn the result into a win probability for the team whose row you're looking at. That probability is then banded into the 1–5 colours. Hovering a cell shows the underlying probability, which is more precise than the band.

The two ratings we use

FPL strength is the rating the official Fantasy Premier League game assigns each club. It's a reasonable starting point, but it has a real weakness: in practice it is set at the start of a season and then doesn't move. A team can go on a long run of bad form and its fixtures will still look exactly as hard — or as easy — as they did in August.

Elo (from the public ClubElo project) fixes that. It's the same idea used to rate chess players: every club has a number, and after each match the winner takes points from the loser — more points for beating a stronger side. Because it updates after every single game, it reflects current form rather than a preseason guess.

Blend combines the two. We tested all three against a full season of real results, and the blend predicted actual outcomes — goals, clean sheets and the points players went on to score — better than either rating did alone. If the Elo data is ever unavailable, the grid quietly falls back to FPL strength rather than showing you nothing.

Overall, Attack and Defence

The same fixture can be easy in one sense and hard in another — a leaky but dangerous opponent is a good match for your forwards and a bad one for your defenders. The lens switch lets you ask the question you actually care about:

• Overall — balances both sides of the fixture. The default.
• Attack — only your attack against their defence. Use it when picking forwards and attacking midfielders.
• Defence — only your defence against their attack. Use it when picking defenders and goalkeepers, where clean sheets drive the points.

One honest limitation: Attack and Defence always use FPL team strength, even when you have Elo or Blend selected. Elo gives each club a single overall rating with no attack/defence split, so it simply cannot answer those two questions — and we would rather tell you that than show you an FPL number with an Elo label on it. It matters because FPL's strength ratings are set before the season starts and, in practice, are not updated during it. So treat those two lenses as a preseason view of a team's shape, and use Overall when you want the up-to-date picture.

Adjusting it yourself

Click a team, or its pencil icon, to change the strength values we feed into the calculation. This is useful when you know something the ratings don't — a key injury, a new manager. Your changes are saved in your browser only, re-colour the grid immediately, and never affect anyone else's view. "Reset to default" restores the real rating.

What this is not

This is a model, not a prediction. It knows about team strength, home advantage and recent results — it does not know about injuries, suspensions, rotation, weather, or a manager resting players before a cup tie. A fixture coloured 1 is one where the odds are in your favour, not one you're guaranteed to profit from. Treat it as one input among several.