FPL 2025 GW1 - Analysis of Prediction Websites

This year I decided to keep a track of the predictions made by popular FPL fansites to see which offering is the most accurate.

Here are the results from GW1

Average difference between real and expected points

So FF Scout was far away the closest on average, with FPL Form lagging some way behind. Although at least every one is within one point on average. FF Fix and FPL Team tended to overestimate points whilst everyone else underestimated points.

However this data is skewed by all the players who were predicted to play 0 minutes and did play 0 minutes. Lets look at the players who did play.

Average difference for players who played

Now FPL Form was far away the closest with FF Fix and FPL Team lacking behind. This suggests that FPL Form is poor at predicting who will play.

Average difference for players who scored more than 6 points (33 players)

Now the predictions seem alot more unreliable. The biggest problem is that the apps rarely predicted players scoring more than 6 points, so anyone who scored 12 would have a massive diff.

Accuracy of predicting haulers

Instead lets look at how accurate each source was at predicting that someone scored more than 6 points.

So FPL Form failed to predict any of top haulers, whilst the rest managed to predict 2 or 3 respectively.

Accuracy based on prediction

Most apps rarely predicted that anyone would score more than 6 points, lets look at what points most apps tended to predict and how accurate each one was.

The true test. What FPL Squad scored the most?

It’s hard to know exactly what is the most important factor when analysing these apps. Most people do not look at all 500 players in the game. But you can’t just look at the best 15 either because yo won’t be able to afford them. I would guess that most people would simply look at a sublist of the the players with the highest predictions and then choose their own team based on these.

So for the next assessment i’m going to take 4 GK’s, 10 DEF’s, 10 MID’s and 6 FWD’s with the highest predicted points from each app. This creates a 30 man squad from each app that most people who use the app might use to then build their actual team. I will then calculate their total actual points to see which app’s squad would have scored the most if you’d followed that app.

Great work from FF Scout! Far and away the best smashing the others out of the park, particularly FPL Form and FF Hub who are way behind with just 2/3 of FF Scout’s points tally.

FPL Review isn’t far behind whilst FF Fix and Draft Hound can consider themselves solidly mid table.

Best Budget 15

But one thing that’s really important in FPL is finding those budget players to enable us to buy the stars. So lets look at how well each app predicted the best 15 based whose price is 4.5 and below for defenders and goalkeepers, 6.5 and below for midfielders and forwards.

And FF Scout comes out on top again. Really impressive 60 points from all budget players.

2-4 and 5-7 are all the same, albeit in a different order, indicating there’s a real top and bottom half to this with FPL Form and FPL Hub really battling it out for the bottom spot as worst site predictor in GW 1.

The big difference here is that FPL Form is one of the only free predictors where as is one of the most expensive out of all the apps.

But it’s only one Gameweek and we know how much luck plays a part. I’ll be back next week to compare again.