How the numbers on KFT are meant to be used
What KFT is trying to do
Kiba’s FPL Treats is built to turn messy weekly FPL information into something easier to read. Expected minutes, player involvement, team context, fixture strength, clean sheet chances, and simulation-based ranges all feed into a view of what a player’s week could realistically look like. The point is not to claim certainty. The point is to make uncertainty more readable.
What sits underneath the outputs
The site leans on a mix of player-level and team-level inputs. At player level, the emphasis is on expected attacking involvement, role, and likely minutes. At team level, the emphasis is on projected attacking and defensive environment. Those pieces are then used to build views such as projected xGI, expected points, and point-range forecasts rather than just one blunt number with no context.
How to read each main tool
Use this to scan which players project well for underlying attacking involvement over the selected run. It is a fast way to see who is getting into the conversation before you even think about captaincy or transfers.
Use this to compare expected total value across the chosen span. It is better for ranking and shortlist building than for treating the top number as a promise.
Use this when you care about floor versus ceiling. It is the most useful view for risk appetite: safe picks, high-variance picks, and the kind of player who can swing a gameweek.
Use this to read runs, not one isolated fixture. The best use case is planning ahead rather than reacting only to the immediate gameweek.
How the work is meant to be used
If the work here helps you, use it. Build on it. Compare your own ideas against it. Use the numbers in discussion if you want. The only basic line is that you should not strip the work of its source and present it as if you made it yourself. If you are building on KFT or using the numbers publicly, say so.
Why the model keeps changing
FPL is not static. Roles change. Minutes expectations move. Injuries happen. Fixtures get shifted. Double gameweeks and blanks distort normal assumptions. Any serious model has to evolve with that reality. KFT is not trying to freeze football into a perfect system. It is trying to respond to the game as it actually changes.