Fixture difficulty is the first thing most managers look at and the last thing most managers get right.
The mistake isn't looking at fixtures. It's treating a difficulty rating as a verdict on a match, when what you actually need is a forecast for a *player* — and those are not the same question.
What a difficulty rating is really answering
When you look at a fixture and think "that's a good one", you're compressing several different questions into one colour:
- How many goals is this team likely to score?
- How likely are they to keep a clean sheet?
- Is this a game where the manager rotates?
A single 1-to-5 rating can't separate those, and they often point in opposite directions. A team playing a leaky side that also scores freely has an excellent fixture for its attackers and a poor one for its defenders. Reduce that to one number and you have thrown away the only part that mattered.
That's why I don't rate matches. I project the two things that actually pay you: goals scored and clean sheet probability, for each team, in each fixture. A defender and a forward from the same club get different answers for the same match, because they should.
Why the official ratings mislead
The difficulty ratings built into the game are broadly reasonable and stubbornly slow. They lean on where teams finished last season, and they don't move much once the season is under way — so they're least accurate in exactly the window where your decisions matter most: a promoted side that turns out to be solid, a big club whose defence has fallen apart, a team that just changed manager.
They're also symmetrical, which reality is not. The same fixture is rated the same for both clubs' attackers and defenders, when the whole point is that it usually isn't.
I tested my own fixture model against the official difficulty ratings across three full seasons, out of sample — fitted on one period, scored on another it had never seen. It came out ahead on both clean-sheet accuracy and goals accuracy in every one of the three seasons. Not by a landslide. But consistently, and in the direction that changes transfers.
Where my numbers come from
The starting point is the betting odds, because they already carry a great deal of information: team strength, injuries, home advantage, form, and real money standing behind them. I don't copy them across — I translate them into fantasy terms. How many goals is this team likely to score? What's the chance it keeps a clean sheet? Odds price the match; I need the match priced *per team, per outcome you get points for*.
For fixtures nobody has priced yet — which is most of the ones you're planning around — the projections come from a team-strength model fitted on historical results. That's why the numbers for the coming gameweek are always firmer than the numbers six weeks out, and why I'd treat a six-week-out projection as a direction rather than a decision.
How to actually read a run
A few things I'd hold onto:
- Plan around runs, not single gameweeks. One fixture is noise. Four in a row is a signal, and it's where transfers actually pay off — you're buying several weeks of a player, not one.
- Read the run for the position you're buying. The same run can be green for a team's attackers and amber for its defenders. If you take one number from a fixture table, you've merged two answers that disagree.
- Beware the run that's already priced in. When a good run is obvious, ownership climbs before it starts, and by the time the fixtures arrive you're paying the premium without the differential. The value is in the run the table hasn't got to yet.
- A hard fixture is not a reason to sell a great player. Elite attackers score in hard fixtures. Fixture swings matter far more for the players whose ceiling depends on their opponent — cheap defenders, rotation-risk midfielders — than for the ones you own because they're simply the best.
- Watch the schedule, not just the opponents. Blanks and doubles reshape a run more than any single opponent does, and they're visible weeks in advance if you're looking.
The honest limits
Projections are probabilities, not promises. A red card, an injury, a surprise line-up or one deflected shot can make a "good fixture" worthless, and no model gets around that. Over a season the numbers pull in your favour; over one gameweek they frequently don't, and being right about the process is not the same as being right about the week.
The other limit is minutes. The best fixture in the league is worth nothing to a player who starts on the bench — which is why I'd always check who's likely to play before I let a fixture run talk me into a transfer.
You can see the live version of all of this on the Fixture Difficulty and goal projections board — every team's projected goals and clean sheet chance for the coming gameweeks, and the fixture run behind them. If you want the full picture of how the projections are built and tested, that's in how my expected points model works.