FPLai Methodology
This is the proof surface behind FPLai’s recommendations. It explains the data inputs, the decision frame, and the limits of the output so managers can use an AI assistant as a second opinion rather than treat it as a guarantee.
Editorial record
Author: FPLai Editorial · Reviewer: FPLai Product & Data · Updated:
Change log: Clarified the input boundary, removed unsupported numeric weighting claims, and documented model limits and source links.
Inputs: Official FPL public data fields, squad context used by the analyzer, product behaviour, and the linked official API documentation.
Limits: The page describes the decision framework, not a published accuracy study. Exact data freshness, feature availability, and recommendations can change with the live game state; no outcome is guaranteed.
- Official Fantasy Premier League — Public game context and team data entry point.
- FPL bootstrap endpoint — Public source for current player, price, availability, and points fields.
- FPLai Team Analyzer — Public product workflow where squad-specific analysis is applied.
- FPLai Team Rating — Public product workflow for the squad-rating decision.
The decision this model is built to support
FPLai is designed for a bounded question: given a manager’s current squad and the current FPL state, which decision deserves attention before the next deadline? It is not a universal player-ranking table and it is not an autopilot. The output is intended to help rank transfer, captaincy, and structural questions that a manager still owns.
Inputs used in the analysis
| Input family | Examples | Decision use |
|---|---|---|
| Live FPL state | Prices, points, fixtures, availability flags, and ownership fields. | Keep the recommendation tied to the current game state. |
| Player availability | Minutes and status signals available in the source data. | Surface rotation, injury, and playing-time risk for review. |
| Fixture context | Upcoming opponents and home/away fixture difficulty. | Compare the near-term opportunity and downside of a move. |
| Squad context | Current 15-player squad, budget shape, positions, and transfer constraints. | Reject attractive moves that do not solve the actual team problem. |
| Manager objective | Rank protection, upside, chip timing, or a draft/planning question. | Keep the final choice with the manager rather than hide the trade-off. |
The public bootstrap endpoint is a useful way to inspect the raw feed fields. Derived scores and recommendations are an interpretation layer over those inputs, not a claim that the source publishes FPLai’s internal calculations.
How the decision logic is applied
- Define the squad problem. Identify the weakest position, captaincy concern, fixture turn, or structural constraint.
- Filter for feasible actions. Keep budget, team slots, transfer availability, and chip context in view.
- Compare expected value and risk. Balance fixture opportunity, recent output, playing-time uncertainty, ownership pressure, and exit flexibility.
- Return a ranked next step. Present the clearest action and the reasons a manager may want to disagree.
FPLai does not publish one fixed numeric weight for every signal on this page because the live decision context changes. The useful invariant is the frame: live inputs, feasible squad context, explicit trade-offs, and a human final call.
What the model can and cannot tell you
- Prioritising a squad-specific transfer question.
- Comparing captaincy options with fixture and risk context.
- Spotting structural problems worth checking before deadline.
- A particular player’s points or minutes.
- Perfect late team-news coverage.
- A better rank, mini-league result, or season outcome.
Use the proof with the product
Start with the FPL Team Rating for a compact squad verdict, open the FPL Team Analyzer for prioritised actions, and use the Captain Matrix or Fixture Swing planner when you need supporting context. If a recommendation conflicts with your objective or late team news, investigate the disagreement before you act.
Current Form Leaders — GW5
Updated for the 2026/27 season. Data refreshed each gameweek.
Groß
BHA
Schade
BRE
Tarkowski
EVE
Kostoulas
Bogle
LEEFrequently Asked Questions
What data does FPLai use?
Does FPLai publish the exact model weights?
Does the methodology guarantee the right transfer?
Where can I see the methodology applied?
Apply It To My Team
Signed-in managers go straight into analysis. New users can create an account and continue into the same flow.
Apply It To My Team