A deep FPL model
with nothing hidden
Sideline Quant is a Fantasy Premier League planner built on a deep expected points model with all its inner workings accessible. Use it to build your team, forecast transfers and win your mini-leagues.
No account needed. Public FPL data only - I never ask for your login.
Chelsea 1.9 goals expected · 29% chance of clean sheet
8% chance this player scores in the match
It rates the haulers above the rest, seven times in ten
Hauls decide gameweeks and mini-leagues. I’m proud of what the model forecasts, so much so I publish all its results openly. Nothing here is edited after the fact, and the weeks it got wrong are published with the rest.
- 65%
- Spotting a 6+ haul
- 66%
- Spotting a 10+ haul
how often it rated a player who went on to score 6+ above one who didn't across 1,057 player-weeks and 254 hauls.
the same read on the returns that decide a captaincy, across 65 of them.
Upcoming gameweek
Max xP is the best fifteen the model can find for the next gameweek, three per club as usual. The Differential XV goes after the same points with players hardly anyone owns, and Elite Ownership is ranked on overall ownership until a top-10k snapshot lands.
These are projections with no budget: the best fifteen by expected points, three per club, ignoring the £100.0m limit. They’re a different thing from the legal squad the model picks inside the budget before each deadline, which then gets marked on what actually happened. That one’s on each settled week’s own page.
Where the numbers come from
I turn each team's expected goals, each player's share of them and his minutes risk into expected points for every player and gameweek, and re-run it all before each deadline.
- 01Read this week's board
Expected points, minutes risk and haul chances for the next few gameweeks.
Browse predictions → - 02See the fixture map
Every match is priced in expected goals, so you can see which runs are actually soft.
Browse fixtures → - 03Open any number
Click any xP and it breaks down into the four things it's built from: team goals, the player's share, minutes and the sims.
How it works →
What makes the numbers different
- I model minutes as four outcomes - plays 60+, subbed off before 60, a cameo or not playing - rather than one average, because the second appearance point and clean sheets only come at 60 minutes.
- Fixture difficulty comes from de-margined bookmaker odds while the market is posted, then my own team ratings - both finer than FPL's 1-5 FDR.
- Each player also gets a chance of a 6+ and a 10+ haul from the same simulations, as well as the average.
- I've put last season's backtest up in full, every gameweek, transfer and chip.
What you can do with them
The optimiser plans your transfers off the board, and Rivals shows what that means for your mini-league.
The optimiser runs a mixed-integer solve over the whole player pool, up to 6 gameweeks at a time. It picks transfers, captain, bench order and chips, and works out whether a -4 is worth taking. Each plan comes with its chance of beating just holding.
Open the optimiser →Import your mini-league and the simulations play out the run-in: where you'll probably finish, your chance of winning it, who's realistically catchable and which differentials swing most against the managers above you. Scoring your own league is for supporters, and there's a full worked example anyone can look at.
Score your rivals →The board, the optimiser and Rivals read off one set of simulations, so they shouldn't ever tell you different things. If a number surprises you, open it up and see where it came from.
See how it fits together →The predictions are free, and so is the optimiser
You don’t need an account to see any of the predictions. A free account adds the optimiser, 6 gameweeks ahead. Supporters keep the site running and get Rivals scoring, a weekly email with extra research, new posts early and a vote on what I look into next.
The predictions are free without one. An account adds the optimiser.
- The optimiser, planning 6 gameweeks ahead
- All the settings: locks, bans, preferences, hit cost, minutes
- The chip planner
- 5 optimiser runs a day · 4 synced drafts
Helps pay for the servers and buys me time to build and research.
- Rivals, which scores your mini-league week by week
- 20 optimiser runs a day
- A weekly email from me summarising what the model is saying for that GW, player and captain recommends and match previews
- New research posts early
- A vote on what I research next
- Everything a free account gets
How billing works
- Cancel on your account page
It's one button, with no form to fill in and nobody to ring. You keep supporter access until the end of the period you've paid for.
- Refunds if you forget to cancel
Email support@sidelinequant.com within 7 days of a Supporter renewal and I'll refund it in full, no questions asked.
- I never ask for your FPL password
All I use is your public FPL Team ID. I don't need your Premier League login for anything.
They're written into the Terms as well.
Last season’s performance
This is a new model I built for 2026-27. To test it I ran it back through 2025-26 one GW at a time, only letting it see what was known before each deadline when it picked the team and made transfers, and this is what it would have scored. The chip weeks are the exception - they were chosen with hindsight, so treat them as a best case. Bear in mind it's a simulated replay, the live record is the real test, and past performance does not guarantee future performance. It also used last season's bonus point rules, and this season's changes are only in the live model.
+991 vs the best set-and-forget baseline, over 38 gameweeks with hits netted off.
Every backtested gameweek, transfer and chip →- This model242763.9/GW
- Set-and-forget template143637.8/GW
- Set-and-forget by price108428.5/GW
The baselines are set-and-forget teams that never make a transfer, so this isn’t a comparison with a typical manager.
What I’ve been looking into
I matched every European tie English clubs have played since 2022/23 to the league game straight after it. The knockouts cost starts but hardly any points, and the real cost is spread across the whole season.
I tested 20 ways of picking a captain over 339 gameweeks. The one that won is pretty basic, and my own model didn't beat it.
I assumed the answer was 'as long as you can afford to wait'. It isn't, or at least not in ten seasons of data.
Transfers, captain and chips planned around the squad you've already got.
Any xP opens up into the team goals, minutes and sims behind it.
Good weeks and bad ones both go up on the track record.