Reading the price change page

FPL never publishes its price-change thresholds. Everything on /prices comes from lib/predict/price.ts, which estimates them.

Rises and falls work differently

This is the part most trackers get wrong, and the first version of this one got wrong too.

• A rise fires on a roughly fixed number of net transfers in, about 216,000 in the gameweek, almost regardless of how many people own the player.
• A fall scales with owners, at around a quarter of them leaving since the player's last price change, and triggers on far smaller absolute numbers. Injured, suspended and doubtful players don't follow it at all.

Treating both as a share of owners makes a heavily-owned player shedding 66,000 net look "stable" while an obscure player with a few thousand transfers looks about to rise. Both readings are wrong.

Counted over the gameweek

FPL moves a player at most once a gameweek, and its count starts again at each deadline. Of 473 changes we recorded from August to October 2026, 471 were the player's only one between two deadlines, and nobody fell twice in one.

So rise progress counts the gameweek's net transfers, and anyone who has already moved this gameweek is locked until the next deadline. Counting rises from the last change instead carried deadline-day rushes forward that FPL never acted on: one player sat 418,000 transfers up for three weeks without a rise, called likely every night.

Falls are the other way round. The gameweek's count told falls apart barely better than a coin toss; the count since the player's last change did it well. So fall progress keeps counting across deadlines until the price moves.

The progress number

Progress is a percentage of the estimated threshold: . It is comparable between players in a way a raw transfer count is not.