What is xPts, and how does FPL Intel forecast it?

A forecast, not a promise

xPts is our estimate of the points a player could return in an upcoming gameweek. It is not their FPL form, their official points total, or a guarantee about the next match.

The estimate is useful when two players have similar recent returns but very different minutes, availability or opponents. It gives you a common starting point for a decision, not an instruction to make a transfer.

What the model looks at

The current forecast starts with recent form and season points, then considers expected goals and assists, likely playing time, clean-sheet chances, saves, bonus and defensive-contribution history. A player flagged as doubtful is adjusted for their reported chance of playing the next round.

For a double gameweek, the model adds a projection for each fixture. For a blank gameweek, it projects zero. Future gameweeks use their own fixtures rather than repeating the next-round estimate.

Why opponents matter

Official fixture difficulty is only one input. The forecast also looks at recent opponent results, with early-season results pulled towards the league average so a short streak does not dominate. For attackers, how often the opponent concedes is particularly relevant; for defenders and goalkeepers, how often it scores matters too.

That is why a green fixture can still produce a modest xPts value against a side that has allowed few goals recently.

A worked comparison: two forwards

Imagine you are choosing between Erling Haaland and Alexander Isak for one free transfer. These names make the example concrete; every number in this paragraph is illustrative, not a current player statistic or live recommendation. Suppose Haaland has a recent form figure of 7.0 and season points per game of 6.0. The starting blend is 60% of 7.0 plus 40% of 6.0, or 6.6. Suppose Isak has form of 5.0 and points per game of 6.0. His starting blend is 5.4. That gap does not automatically become the gap between their final xPts values.

Now imagine Haaland is expected to play around 65 minutes against an opponent that has conceded little, while Isak is expected to play 85 minutes against a side conceding more often. The minutes and opponent adjustments can narrow, erase or reverse the initial difference. A lower starting blend is therefore compatible with a higher one-week forecast. The real player pages show the current inputs and fixture-by-fixture estimates; use those rather than these made-up figures when making a decision.

Minutes change more than appearance points

A player who usually completes a match has more time to score or assist and a better chance of reaching the minutes threshold for a clean sheet. The model blends recent playing time with season playing time, then uses the result in its appearance, attacking and defensive estimates. That is why a regular starter can forecast better than a substitute with a striking points-per-minute record.

When a player has only a few appearances, the sample can be misleading. A substitute may have scored during a short cameo; simply multiplying that return to 90 minutes would overstate the likely result. Check starts and minutes alongside xPts, especially after a transfer, injury return or managerial change. The number is a model estimate of opportunity, not a promise about the team sheet.

Expected goals and assists add another view

Goals and assists are valuable outcomes, but they arrive unevenly. Expected goals and expected assists describe the chances a player has been involved in, whether or not those chances became returns. The forecast blends season rates with recent rates and applies the opponent adjustment to attacking potential. This helps a player with repeated good chances remain visible after a quiet scoring run.

The reverse matters too: a goal from one low-probability shot does not mean the same scoring rate will continue. Expected involvement is still imperfect. It cannot know the next tactical assignment, penalty order or a late substitution. Read it together with minutes, fixtures and actual role instead of treating one statistic as a verdict.

Blank and double gameweeks

The forecast is calculated fixture by fixture. A blank gives a player zero projected points for that gameweek because there is no scheduled league match. A double adds two separate match estimates, each with its own opponent and home or away context. A double can lift the total even when neither individual fixture looks especially attractive.

That does not mean a player will play 90 minutes twice. Rotation risk, recovery time and competition for places can make a double less certain than the sum looks. Use the five-week view to see whether the advantage lasts beyond the immediate round. If a transfer costs points, compare that cost with the forecast gain over your actual planning horizon.

Availability is a probability, not a headline

An available player is forecast at full availability. A doubtful player is scaled using the reported chance of playing the next round; unavailable players receive no projected return. A 75% chance is not a statement that the player will score 75% of their usual points if selected. It reflects the possibility of getting no match at all.

News can change after a forecast is produced. Check the timestamp on the data and the latest reliable team news before the deadline. An uncertain starter may still be reasonable if your bench is strong, while the same risk can be painful when your bench is empty. The forecast provides a consistent number; your squad context decides how much uncertainty you can accept.

What an accuracy result can and cannot prove

A single gameweek is too noisy to judge a forecast. A penalty, an early red card or a deflected goal can swing one player far from expectation. FPL Intel saves predictions before deadlines and compares them with actual points after completed rounds. Mean absolute error shows the average size of the misses, while the number of comparisons shows how much evidence sits behind it.

Even a well-calibrated forecast will miss individual outcomes. Use the accuracy page to understand its track record as results accumulate. For a transfer decision, compare candidates, inspect the reasons for the difference and allow for uncertainty. An estimate of 5.2 rather than 4.9 points is a small edge, not a meaningful guarantee.

How to use xPts

Compare players in the same position over the number of gameweeks you actually plan to hold them. Check the underlying minutes and availability, then consider your budget, free transfers and the strength of your bench. The published forecast is saved before a deadline so its later accuracy can be measured without rewriting history.