FPL 2025/2026 Season Review: A Thread
Final OR: 309 ๐
Total points: 2469
Out of 3 seasons playing FPL, this was clearly my best ever. The headline is obviously the final rank, but the journey was probably the most satisfying part: from around 5.8m OR after GW1, into the top 1m by GW7, breaking into the top 5k by GW13, and then staying there all the way until GW38.
Reflecting back, I believe the final rank was not built from one miraculous GW, but a season-long accumulation of edges: strong squad structure, good xMins and rMins, no hits, and optimal transfer decisions landing at exactly the right time.
#FPL #FPLCommunity #FPLIndo #FantasyPremierLeague
@Mike_Metz66 He didn't say he use solvers to make 90% of his decisions, though? He said 90% of the projection data he uses comes from FPL Review. You are just twisting words at this point.
๐ ๐ฌ ๐๐ฃ๐ฃ๐ฅ๐ข๐๐๐ ๐ง๐ข ๐๐ฃ๐: ๐๐๐๐๐ฆ๐๐ข๐ก-๐ ๐๐๐๐ก๐
2025/26 was my first full FPL season. After sitting at 300k in GW10, I discovered the analytical side of FPL on X and got hooked.
The analytical community helped me finish 13k, and this season Iโm currently 1795 overall (GW10). Hereโs what Iโve learned and how I approach decisions. Still learning, of course!
๐จ๐ฆ๐๐ก๐ ๐ ๐ข๐๐๐๐ฆ
Around 90% of the projections I use come from FPL Review. I occasionally check other models for close calls. (FFM, Solio, TFT).
I trust the model rather than trying to outsmart it. Where I do make adjustments is xMins, particularly around injuries, rotation and players returning to the XI.
Itโs not about predicting every minute perfectly. Itโs about understanding uncertainty around playing time and making better-informed decisions.
๐๐ข๐ก๐-๐ง๐๐ฅ๐ ๐ฃ๐๐๐ก๐ก๐๐ก๐
Last season Iโd spend ages comparing players with almost identical EV. Now I care more about how each route affects my team over the next 6โ8 GWs.
Does a transfer restrict future moves? Will it need reversing? Am I buying for the long term or just chasing fixtures?
A good example from last season was selling Thiago (ยฃ6.8m, on penalties) for Ekitike (ยฃ9m, no penalties), mainly because Isak got injured and Ekitike had just scored twice and the next 3 fixtures looked good.
I reacted too quickly without thinking enough about the long term. I eventually had to take a hit, while Thiago went on to comfortably outscore Ekitike.
Getting the big decisions right matters more than squeezing out every marginal EV gain.
๐๐๐๐ซ๐๐๐๐๐๐ง๐ฌ & ๐๐ก๐๐ข๐ฅ๐ ๐๐ง๐๐ข๐ก
Sometimes the best transfer is no transfer.
Rolling gives me another GW of information on minutes, roles, fitness and team strength. I aim to keep at least 2 FTs available at all time (inspired by @FPL_Brandon ).
Itโs a guideline, not a rule Iโd follow blindly. A good move can be worth making now, but keeping transfers in hand gives me options when new information arrives.
Also, information over team value, always.
๐ฅ๐ข๐๐จ๐ฆ๐ง๐ก๐๐ฆ๐ฆ
When EV is close, I look at minutes security, injury risk and long-term role.
I sold Calafiori on WC3. His GW8 benching, while Gabriel scored 17 points for me, was a reminder of the value of reliable minutes (not proof the decision was right).
Iโve also held De Cuyper longer than expected because new information improved my view of his minutes.
Robustness shouldnโt always beat EV, but uncertainty matters.
๐ฃ๐ฅ๐ข๐๐๐ฆ๐ฆ ๐ข๐ฉ๐๐ฅ ๐ข๐จ๐ง๐๐ข๐ ๐๐ฆ
FPL has huge variance. Good decisions can fail and bad ones can pay off.
I try to judge decisions by what I knew at the time, not the points that followed.
My relative variance is at the 98th percentile, so I know Iโve been very lucky. A 1.8k rank doesnโt automatically make me better than someone ranked 1m.
๐ฆ๐ฃ๐๐ก๐ ๐ง๐๐ ๐ ๐ช๐๐๐ฅ๐ ๐๐ง ๐ ๐๐ง๐ง๐๐ฅ๐ฆ
I still overthink close calls, but Iโm trying to spend less time on differences of under 1 EV and more on team structure, transfers, captaincy and chips.
Some calls are simply too close to separate. Make a decision and move on.
๐๐๐ก๐๐ ๐ง๐๐ข๐จ๐๐๐ง๐ฆ
My aim is to keep improving my process, not just my rank. Iโve learned loads from top managers like @elevenify , @FPL_Bronx and @FPL_Brandon and many more!
FPL is incredibly complex. Finding ways to simplify decisions has helped me enjoy it more.
I donโt have it all figured out, and Iโm always keen to learn how others approach the game.
Good decisions donโt always lead to good outcomes. That doesnโt make them bad decisions. #FPL
4/4 โ GW4: WC Initial Draft
๐ด Raya + Gabriel + Konsa: defensive CoVar from the best defence in PL with secure xMins. A no brainer pick.
๐ต Palmer (C) + Joรฃo Pedro + Rogers: HUL(H) + strong medium-term run + the deadly trio of CHE attack. Another no brainer pick on WC4, coming from the 3 players with the highest EV projection on GW4.
๐ฃ Bruno stays despite MCI: WC is built for GW4+, not one fixture. A self-explanatory pick for analytics manager.
๐ด Szobo + Groร: xMins + set-piece routes at sensible prices.
๐ก๏ธ Vuskovic + Van Hecke keep the defence cheap enough to spread value across the XV. Also really good xMins from both of them with decent EV projection.
Bench: Wissa / Tarkowski / Gonzalo. Still playable and provides good rotation in the future GWs.
#FPL 1/4 โ GW3 Review and GW4 Current Draft | FH
55 pts | OR 20,641 ๐ข
Very belated GW review since I was busy with real life work. Apologies!
Small green from 22.8k โ 20.6k. Definitely not the explosive FH score I was hoping for, especially after going quite aggressive with CoVar.
But 55 pts in isolation doesn't tell me much about whether the FH was good or bad.
#FantasyPremierLeague #FPLIndo #AnalyticsFC #AnalyticsFPL
3/4 โ Was FH3 actually worth it?
If I had simply rolled my GW2 team into GW3 with the same XI + Bruno (C):
No FH: 30 pts
FH3: 55 pts
โ +25 pts realised FH gain
And even with perfect hindsight: best possible XI from my original 15 + Mbeumo (C), that team would've scored 44.
So even against an unrealistic hindsight benchmark, FH3 still gains +11.
The comparison is fair since if I decided not to FH3 (and not to WC3), I'll certainly roll my FT.
And on that basis, FH3 did its job.
#FPL โ 1/2 FH3 Draft ๐ | Getting closer
Market data is lining up quite nicely with the structure:
๐ต MCI โ 2.97 xGoals | 49% CS
โ Haaland (C) + Foden locked. Best attacking AND defensive projection of GW3.
๐ด LIV โ 2.35 xGoals
โ Wirtz + Szobo + Isak. Clear #2 attack in the market, so happy embracing the offensive CoVar vs IPS(A).
๐ BRE โ 39% CS
โ Kelleher + Collins + Ajer. Second-highest CS probability in my data, so again I'm happy concentrating the outcome rather than diversifying.
Saka stays since his EV projection is really good despite pretty bad fixtures. Make sense since based on the past GWs, CHE defense are so leaky.
Structure close. One decision still bothering me though.
#FantasyPremierLeague #AnalyticsFPL #FPLIndo #AnalyticsFC
1/3 โ GW2 Review | 116 pts | OR 23k ๐ข
Was busy in real life, so late posting for my GW2 review. Sorry about that!
โญ Bruno (C) 46
โญ Tarkowski 12
โญ Mbeumo 11
โญ Joรฃo Pedro 9
โญ Gabriel + DCL 8
Obviously a huge outcome, with Bruno captaincy doing most of the damage. Insane generational points!
Neco's 6 pts on the bench hurts visually, but I'm not going to hindsight a decision to bench him at Anfield for MCB at Coventry.
Onto the numbers ๐
3/3 โ What did GW2 actually change? ๐๐
โฌ๏ธ MUN โ Bruno + Mbeumo
Bruno's 23-pointer came from a hat-trick + assist, with 8 shots and 3 chances created. Mbeumo hauled but also missed a 0.92 xG chance. He's encouraging beyond the 11 pts.
โฌ๏ธ CHE โ Palmer + Joรฃo Pedro
Still very happy here. JP could've scored much more than his 9 and Palmer returned again. The original Chelsea thesis looks stronger after two GWs.
โ ๏ธ Neco
The 6-pointer on my bench isn't the important information. Muรฑoz joining Forest is. That introduces genuine atk share drop uncertainty for Neco going forward.
๐ฏ Ndiaye
This was probably the most questioned decision of my GW.
My data had MCB only +0.08 EV ahead of Ndiaye.
So yes, the decision was basically a coin flip, but if my process says MCB by 0.08, I'm happy to follow it.
Ndiaye scored 4, MCB 1. That doesn't suddenly make the decision wrong.
If MCB had hauled, everyone will praise my benching decision. That's exactly why I don't want to evaluate picks with hindsight.
The outcome was bad. The decision and the process was still fine. Trust the data and the process, always ๐ง
2/3 โ Expected vs Realised ๐
My xP: 60.91 โ 116
Avg xP: 63.43 โ 111.23
Pre-GW expected edge: -2.52
Realised edge: +4.77
I was expecting a red arrow in GW2 since there are a lot of players going BB. Thanks to Bruno the variance monster, it carries my posvar to green arrow. I gladly take it though!
Unlike GW1, where I basically landed on expectation, GW2 was massively above EV for both me and the field.
The chart below, taken from @FPLOptimized shows how crazy the variance was in GW2!