2026/27 ·
Ownership: FPL API
Ownership vs Projection Matrix
Every player plotted by how many managers own them against how the model rates them
over the next five gameweeks. Ownership is cut into template, popular and under-owned,
which splits the chart into six cells: players you need to own, players who could buy
you a lead, and players the field owns that the model rates poorly.
Scatter plot of FPL ownership percentage against projected points per game over the next five gameweeks
Table view: every plotted player
Why two ownership lines. Traps live in the band where a player is
popular but far from universal, while “essential” only applies at the very top
of ownership. A single cut-off misses one or the other, so ownership is split twice:
template, popular, and under-owned.
The cut-offs are fixed thresholds. The two ownership lines default
to 50% (template) and 30% (popular) and stay put as you filter, so a player sits in
the same cell whether you are looking at the whole field or just forwards. The
projection line starts near the top sixth of the field. All three are adjustable
with the sliders.
Who is on the chart. Players the model expects to start less than
60% of the time are left off, so the plot is not crowded with names carrying no
minutes. Everything the model gives a real chance of starting is here.
Reading the x-axis. Ownership is on a square-root scale because FPL
ownership is heavily skewed (the median player here is under 1% owned), and a linear
axis would crush every differential into the left margin.
One model. Figures come from