How FPL Apex works
Where the numbers come from, and how to read each part of the solver. About five minutes to read.
What Apex does
FPL Apex is a free planner for Fantasy Premier League. It projects how many points every player should score over the next ten gameweeks, then works out the transfers, captain and chips that should score your team the most.
It does not pick for you. It shows the best options, how far apart they are and how sure it is, so the final call stays yours.
Where the numbers come from
Every projection starts with the betting market. We take match odds from a bookmaker known for accurate prices, remove its profit margin, and read off how many goals each team should score and concede. Where the market has not priced a match yet, team ratings built from past market prices fill the gap.
Each team's goals are shared among its players by expected minutes, each player's record of chances, and the betting prices on who will score where those exist. Everything is then scored with FPL's own rules, bonus included.
Getting started
- On the home page, enter your FPL Team ID. It is the number after /entry/ in the address of your Points page on the FPL site.
- Press Solve my team. Your squad, bank and free transfers load from FPL. No sign up.
- On the Squad page, choose a strategy and press Solve.
Reading the planner
The pitch
Your starting eleven and bench for the week, with each player's expected points on his shirt. The arrows above the pitch move between weeks. The tiles show squad value, money in the bank and free transfers.
Strategy: Most points, Safe, Bold
Most points goes for the highest total. Safe moves popular players up and counts points hits more heavily. Bold moves players few managers own up. The strategies change the order of the options, never the numbers.
The cards
After Solve, the planner shows a fan of cards for this week: your best move and the alternatives.
- The move: the player sold is struck through, the player bought is in bold.
- The big number: a score for the whole plan that starts with this move, hits taken off and later moves included. Use it to compare cards.
- The smaller total: that plan's expected points across the weeks shown.
- Top in 41%: how sure the planner is. See the note on uncertainty.
Cards within half a point of each other are marked level: a coin flip. A fourth card appears when the next transfer is within one point of the first.
Tap a card to take that move, and the planner moves on to the next week. Press Undo, or tap a decided week, to go back. Apply plan to squad puts the plan on your pitch.
The options list, saved plans and prices
Show every option as a list opens every move the planner checked, each with a Take button. Under Safe or Bold a smaller number can sit above a bigger one, and the list says why.
Save this plan keeps a plan. Compare saved plans scores your saved plans side by side, week by week, hits taken off, and marks the best.
This week uses FPL's real prices. Later weeks use forecast prices, from how many managers are buying and selling each player, and FPL's rule that you keep only half of any rise.
Chips
Wildcard and Free Hit give unlimited free transfers for a week; after a Free Hit your old squad returns. Bench Boost scores your bench too. Triple Captain triples your captain's points.
Each chip box gives a verdict. Best GW14 means that week clearly stands out. Best GW9, GW14 close means play it in the coming week, with a later week close behind.
No clear week means no week beats the others by more than good models usually disagree. A chip can also appear as a card in the planner. Apex never plays a chip for you.
Squad Brief
Your team against everyone else's: how many managers own your captain, how many of your starters are in the most owned eleven, your differentials (starters owned by under 15%), and a warning when three or more starters come from one club.
The top 1000 managers at the GW deadline shows what the best ranked managers did at the last deadline: most owned players, captains, transfers, chips and hits, and how much of your team they share.
Mini League
Apex plays out the rest of the season thousands of times for every manager in your league, chips included.
- Chance to win: how often each manager finishes top at GW38, plus the chance of being top after a nearer gameweek.
- Your league's template: what your rivals own and captain, your differentials against them, and the players you lack that would gain them points.
- Chip watch: rivals holding a chip with a standout week soon. A warning, not a prediction.
- Beat a rival this week: you against one rival, and the players that decide it.
Other pages
Fixtures: each team's expected goals for and against, and its clean sheet chance, match by match, plus team strength ratings.
Decision report: good decisions and good luck, kept apart. Your picks are judged on what they were projected to score; luck is what came above or below that.
Model scorecard: every week the projections are saved before the deadline, then checked against real points. See the Model scorecard.
Insights: the season's leaders in expected goals, expected assists, defensive contributions and goalkeeping, on the Insights page.
Making it yours
Edit minutes. Know something the model does not? In the Player Pool, tap a player's expected minutes, set your own (0 for out, 90 for a full game) and press Save & recompute.
Every projection, card and plan updates, and your edit follows your Team ID between devices.
Lock and exclude. Each Player Pool row has two buttons. Lock means never suggest selling him; Exclude means never suggest buying him. They stay until you clear them.
A note on uncertainty
A projection is an average, not a promise. One goal or one red card can swing a week, so even the best plan will lose to a worse one some weeks.
That is why each card shows top in a percentage. We shake every projection by about as much as good models usually disagree, run the plan again 300 times, and count how often it still comes out best.
A high share means the plan wins under most versions of the numbers. Similar shares mean the choice is close and your own judgement counts.
Judge Apex over many weeks, not one. The Model scorecard shows how close the projections have been, week by week.