FPL GW36: Which players are in form?

Today's topic

Just 3 gameweeks remain from the current FPL season, and the cards have been pretty much dealt.

Planning is super straightforward now, as 6 (mostly) prominent sides will double in GW37, and the vast majority of engaged managers will take advantage of this opportunity by playing their Bench Boost chip.

You're probably well-prepared for the season run-in, putting the last finishing touches on your team.

So, this time, I'll investigate which players are "in-form", meaning how well they are doing in some of the key metrics in the sample size of the last 15 games.

In other words, this calendar year.

Why not a shorter period of time? Well, if I did, let's say, 5-10 games, I would still need to adjust for the opponent's strength to make it unbiased.

This should be a sweet spot - the sample size is sufficient, yet it still accounts for more recent performances.

Getting into chances

Aka, the classic one - expected goals normalised per 90 minutes. Plus, non-penalty goals per 90 to showcase their short-term finishing ability when comparing these two:

As you can see, there's nothing wrong with Erling Haaland's (and Salah’s) movement in the box and ability to get into dangerous goal-scoring opportunities.

He doesn't convert his chances at his usual rate, but this slump won't last forever, and he will be indispensable in the culminating title race.

The funny thing is that players ranked 2nd and 3rd were victims of rotation lately.

However, Timo Werner is out for the rest of the season due to a hamstring injury, which opens the possibility of picking Brennan Johnson (£5.8m) as an enticing, cheap and possibly high-performing asset.

This table further underlines the importance of red-hot Alexander Isak and Phil Foden.

xG + final passes

Let's combine xG + xA/90 and non-penalty goals + assists/90.

For FPL purposes, I don't fancy merging these two metrics into one number without providing proper context, as the goals and assists are awarded differently (a 2-point difference in favour of goal for midfielders, for example).

However, it gives you a good idea of the final product. Remember that you must always emphasise xG (preferably its non-penalty variant).

And if you can't track the non-penalty version of xG, just be aware that the player is a designated penalty-taker.

Goal probability added

The GPA is a model that values each player's action (pass, run, dribble, etc.) based on its impact on the team's likelihood of scoring. The higher the index, the more dangerous areas a player moves the ball into.

How do we use this metric? If any of these players get injured, it will greatly impact the team's offensive abilities, which may help you make a tough decision (which defender to play, bench order, etc.).