Man, let me tell you why I even bothered diving into this FPL price change mess. It wasn’t about the thrill of predicting goals; it was pure, unadulterated frustration. Last season, Gameweek 15, I was sitting pretty. Had the obvious players, right? But I needed to dump a struggling midfielder and grab a rising star—I think it was Solanke back when he suddenly started hauling. I waited one damn day. One day! His price shot up 0.1, and that meant I had to take a minus four hit to restructure my bench just to afford him, or settle for a cheaper, less effective option.
I lost my mini-league match by three points that week. That’s when I snapped. I said, “Never again am I letting the market dictate my moves. I am going to figure out the stupid algorithm myself.” This wasn’t a complex, data-science project at first. It was a revenge mission fueled by rage and a lack of proper sleep.
Starting the Scrape: Where Do the Numbers Live?
My initial thought was simple: I needed the raw data. I knew the official FPL site and the popular tracking sites like FPLSTAT were always lagging, or only showing the cumulative Net Transfers (Nettos) once every few hours. That’s useless if you want to beat the morning price rise.
So, I grabbed Python—my go-to tool for brute force data collection. I focused entirely on the official API endpoints. These endpoints hold the real-time number of transfers in and transfers out for every single player. The key was not just getting the number, but getting it constantly.
I set up a crontab job that essentially ran my script every three hours, 24/7, starting immediately after the FPL deadline closed on Saturday morning. The script pinged the API, pulled the Nettos for the top 200 most-owned players, and logged the timestamp, the current price, and the Net Transfer count into a ridiculously long spreadsheet.
It was messy. The data was sometimes inconsistent, showing big jumps during the midnight hours, but I kept recording. My goal was just to map movement against the resulting price change. I needed to see exactly what count triggered that precious +0.1.
The Discovery: Defining the Thresholds
I spent maybe two full weeks just staring at this spreadsheet. I tagged every cell where a price change occurred. Then I backed up and looked at the net transfers for that player over the previous 24 hours and the previous 48 hours. This is where I started seeing patterns that contradicted what the online gurus were saying.
I realized that FPL uses different invisible thresholds depending heavily on the player’s starting price and their current ownership. It’s not a single magic number like “20,000 transfers in equals a rise.”
I created three simple buckets:
- Budget Players (Under 6.5m): These guys need the fewest transfers to rise. I found they often only needed around 12,000 to 15,000 net transfers in a quick 24-hour burst to trigger a rise. They move fast, especially if they are flagged as essential bench fodder.
- Mid-Priced Players (6.5m to 9.9m): This group required consistency. I tagged rises closer to 18,000 to 22,000 Nettos, but critically, they often needed to have a second good day of transfers lined up.
- Premium Players (10.0m+): These guys were stubborn. Salah, Haaland, etc. They needed massive movement—often 25,000 to 30,000 Nettos. If they dropped, they were also hard to bring back up.
I refined my script to not just log the total Nettos, but to automatically calculate the moving 24-hour total. This was the game changer. The official FPL price change usually happens either around 00:00 UK time or 02:00 UK time, sometimes later if things are volatile. If a player hit my calculated 24-hour threshold by 18:00 the day before, I knew the rise was almost guaranteed.
The Final System: From Data Junkie to Early Transfer Expert
My final setup is totally janky, but it gets the job done. I built a simple, private web dashboard using a basic framework—nothing fancy, just a table that displayed who was close to rising. Every three hours, the script runs, calculates the running 24-hour net transfers, and compares it against my manually determined thresholds for that price bracket.
If the player hits 85% of the threshold, they get flagged Yellow: Watch Closely. If they hit 100% or more, they are flagged Red: Rise Imminent. That red flag usually means I have maybe a 6 to 10-hour window before the official rise takes place.
I implemented a basic email notification—just a simple message sent directly to my phone when a player turns red. I don’t have to wait up until 2 AM anymore, but if I wake up at 6 AM and see those red flags, I immediately pulled the trigger on the transfer, knowing I had beaten the rush.
Was all this necessary just to save 0.1m? Maybe not, but it solved my problem of losing team value to market timing. The biggest lesson I learned is that predicting price changes isn’t about stats; it’s about predicting mass human movement and panic buying. And once you reduce human panic to a threshold number, it becomes surprisingly predictable. I’m constantly adjusting the percentage points slightly because FPL definitely tweaks their formula year-to-year, but the core mechanic remains solid. I’ve saved several million in cumulative team value just by getting ahead of those damn price rises, and that feeling of winning the market game is better than any goal scored.