Using data, odds and statistical methods to analyse FPL.
Norway
Joined July 2019
- Tweets135
- Following243
- Followers728
- Likes243
FPL Research retweeted
A few smaller updates added over the past week or so
- Page load should be much snappier (especially on mobile)
- WC/FH diversity added as a solver setting
- Arrows keys can control both GW (←→) and draft (↑↓) selection in planner
- Solver gain/loss in evaluation score now displayed
- Significant backend upgrade
FPL Research retweeted
Excuse my language but this is the best f*cking fpl news ever 🥝
A significant optimisation upgrade has made it's way into the fplreview solver!
The fuller, non-linear model of the FPL problem has been sufficiently linearised, effectively merging the old Transfer Solver and Linear Optimiser
This has enabled faster solves, richer evaluation and all new & existing features/improvements will be shared by both solve modes
FPL Research retweeted
A significant optimisation upgrade has made it's way into the fplreview solver!
The fuller, non-linear model of the FPL problem has been sufficiently linearised, effectively merging the old Transfer Solver and Linear Optimiser
This has enabled faster solves, richer evaluation and all new & existing features/improvements will be shared by both solve modes
FPL Research retweeted
Forgot to mention on Thursday that you can now click on the price progress sparklines for an enlarged image
A few examples for players with notable market movement from todays matches
FPL Research retweeted
Elite 1000 GW5 Summary
- 9.3% used WC, 4.5% went FH
- 79% captained Haaland (99.9% of those who own him)
- 13% captained Bruno and 6% went with MGW
- Barry and MGW the main transfers in
- Joao Pedro, O'Reilly, Sangare, Szobo and Wirtz on the way out
- ~50% of non FH/WC managers rolled
- 1 of the 1000 has rolled their way up to 5FTs already
FPL Research retweeted
Price Bonus is now live on FPL Review as a solver setting!
This acts as an incentive to buy or sell players based on forecasted progress towards an SV change
Getting these market effects integrated into the solver is a significant step, something many of us had been estimating manually on top of solver outputs for years
This is currently a beta feature and is defaulted off, as it's quite a big change for many. You can default it on from the settings sidebar.
FPL Research retweeted
Said this before, but FPL creators should be transparent about the tools they actually use, especially when promoting alternatives.
If you regularly use tools like Review or LiveFPL, I think it’s only right to acknowledge and credit them.
Despite never relying on predictive models myself, I have a lot of respect for @fplreview. He built and maintained a very strong product year after year, and that, I know, is very tough.
Really wished people who use it and rely on it would give proper credit. Same with LiveFPL.
FPL Research retweeted
Despite never relying on predictive models myself, I have a lot of respect for @fplreview. He built and maintained a very strong product year after year, and that, I know, is very tough.
Really wished people who use it and rely on it would give proper credit. Same with LiveFPL.
FPL Research retweeted
Some interesting takeaways on WC4, viewing the fplreview.com/elite - Elite 1000 sample.
• 187/1000 people wildcarded this week (compared to 427/1000 in WC3).
• 63% with Haaland, 36% with Bruno, 1% with both.
• 3 players are 100% owned: Konsa, Palmer, Joao Pedro.
• Rogers and Szoboszlai are sitting at 99% and 90%.
• 100% captained Palmer.
• Quite a spread among Arsenal players apart from Konsa: Raya, Calafiori, Gabriel, Saka, Odegaard all popular.
• Slater is the most popular 4.5M mid at 24%, with Olusesi at 2%.
• Dubravka at 40%, by far the most owned 4.0M GK.
• O’Shea is the favored 4M def, with Egan, Thomas, and Mendy all equal behind him.
My team:
• The starting XI covers 65% of max Elite ownership (711/1100%), but only 30% of General ownership (329/1100%).
• Full squad Elite coverage: The 15-man squad holds 806% out of 1500% max (~54% of total theoretical Elite ownership).
Conclusion:
• I’m surprised to see this many Haaland teams honestly. I get that it’s a matter of risk preference and most people didn’t want to take the risk, but it was still the suboptimal path for most inputs and xMins views.
• Many more “grassy” picks appear too: Calafiori, Odegaard, De Cuyper…
FPL Research retweeted
Elite 1000 GW4 Summary
- 43% played TC, 19% WC and 3% FH
- 95% went with Palmer as captain (99.6% of those that own him)
- Rogers, Palmer and Konsa were the most common transfers in
- Cherki, Bruno, Ndiaye, Mbeumo and Maguire to most common to make way
- 2 of the 1000 will have 4FTs available for the upcoming GW
FPL Research retweeted
For anyone who didn't immediately realise this was a joke, here's my annual reminder that @fplreview is basically just one person (and he is constantly killing it 💪🫡🥝)
FPL Research retweeted
The Elite 1000 data page is back!
- 42.7% played WC this GW, 26.7% played FH and 29.7% played TC. 5.8% did not use a chip
- 99.9% of managers who owned Haaland, set him as captain
- Ndiaye, Mbeumo, Tzolis and DCL most common sales for non WC/FH
- Cherki, Wissa, De Cuyper and Groß the most common transfers in for non WC/FH
fplreview.com/elite
FPL Research retweeted
FPL Elite 1000 GW2 Update: The Bruno Effect! (and BB)
We’ve just witnessed the best start ever for the Elite 1000, hitting an incredible median rank of ~138k following GW2. 📈
The main catalyst? Bruno Fernandes: hat-trick + assist against Ipswich. But the connection between the Elite 1000 and Bruno goes much deeper.
As you can see in chart, the Elite group's fortunes seem directly tied to the Manchester United talisman:
✅ Current Best Start (26-27): Bruno hat-trick vs IPS
✅ Previous Best Start (21-22): Bruno hat-trick vs LEE (GW1).
❌ Worst Start Ever (25-26): Bruno misses a penalty vs FUL (GW2).
FPL Research retweeted
Price changes are now integrated into the fplreview planner toolset!
- Visualise price change progress over the GW with sparklines
- Ability to toggle between price changes alongside transfer paths on main planning page
- Progress % is settable as a viewable slot in pitch view
- Can also be used for the helper column in the player projections table
This is the initial implementation and there will be elements added/refined over time, but hopefully a useful start!