FPL AI Transfers: Recommended Transfers for Your Squad
What separates a useful AI transfer suggestion from a confident guess, how The Gaffer's Pulse builds its recommendations, and where you still need to make the call yourself.
Ask a general chatbot "who should I transfer in?" and you'll get a fluent answer about players it remembers. It won't know that you have £0.4m in the bank, that your midfielder only sells for £7.8m, that you already own three Arsenal players, or that the striker it's recommending was flagged on Friday. A transfer suggestion is only worth acting on if it's built from your squad, your money and this week's data.
What an AI Transfer Suggestion Has to Get Right
- Your actual squad, including the three-per-club limit and who's injured or suspended.
- Your real budget. That's your bank plus the selling price of the player going out, not his current market price.
- Your free transfers, because the answer changes completely when the second move costs four points.
- A forward-looking ranking. Points expected over the next few gameweeks, not points already scored.
- The whole player pool, not just the dozen names everyone talks about.
How The Gaffer's Pulse Suggests Transfers
The Gaffer's Pulse is the AI assistant inside FPL Pulse. When you ask it for transfers, this is what happens behind the answer:
1. It reads your team, bank and free transfers
Your squad, bank and squad value come straight from your FPL team ID. Your free-transfer count for the coming gameweek is calculated from your transfer history and handed to the AI as a fixed fact, including a warning when you're at the maximum and about to lose one. If the count can't be worked out, the AI is told to ask you rather than guess.
2. It ranks candidates by predicted points
The main signal is our predicted points model, summed over the next three gameweeks for transfers (single-gameweek xP is used for captaincy). Fixture attack and defence ratings explain why a player projects well, and form is only a tie-breaker. It never overrides a clearly higher xP. We grade the model against FPL's own expected points every gameweek, in public.
3. It checks the money with a calculator, not a guess
Language models are bad at adding up prices, and FPL makes it harder: you sell a player for his purchase price plus half of any rise, rounded down. Buy at £7.5m, watch him climb to £8.1m, and he sells for £7.8m, not £8.1m. Before the Gaffer says a move is affordable, it runs the swap through a calculator that uses your real bank and selling prices. It may only call a transfer affordable, or quote what's left in the bank, if that check says so.
4. It can reach every player, not just a shortlist
The AI starts with a shortlist of the top projected players at each position, but it can search the full pool by position, club and price (for example "a cheap Everton defender who actually starts") and look up any player you name. It isn't allowed to dismiss a player you ask about until it has looked him up, and when it does say no, it has to give the reason with real numbers: injured, low xP, poor fixtures or unaffordable.
Hits: When Is a -4 Worth It?
The question AI suggestions are most often wrong about is not who but whether. A rule of thumb that holds up:
| xP gain over the horizon | Verdict |
|---|---|
| Under 4 points | Roll the transfer. The hit costs more than it earns. |
| Around 4 points | Break-even. Only worth it if it also fixes a long-term problem. |
| Well over 4 points | Take it, especially to replace an injured or non-starting player. |
The biggest gains usually come from replacing a player who won't play. Zero expected points is easy to beat, which is why hits for injured players pay off far more often than hits for "upgrades".
Where AI Transfer Suggestions Stop
- Late team news. The AI sees FPL's injury and availability flags, not a press conference from an hour ago. Check the lineup news before you confirm.
- Your long-term plan. It optimises the next few gameweeks. If you're building towards a Wildcard or the mid-season chip reset, tell it, because that changes which moves are worth making now.
- Mini-league context. Protecting a lead calls for safer picks than chasing one. The mini-league simulator tells you which situation you're in.
From Suggestion to Decision
Treat the AI's answer as a shortlist, then test it. The Transfer Planner shows injury flags, start %, xP and points next to every player, lets you map moves up to five gameweeks ahead, and puts the two players side by side before you confirm a swap. A suggestion that still looks right there is one worth making.
Free and Pro
The "What transfers should I consider making?" quick prompt is free, along with the captain, bench and team-analysis prompts, up to three a day. Pulse Pro (£2.99/month or £24.99/year, 7-day free trial) lets you ask your own questions, such as a specific swap, a two-gameweek plan or a hit versus a roll, and unlocks the five-gameweek predicted points horizon.
FAQs
- Is there a free AI that suggests FPL transfers?
- Yes. In FPL Pulse, the "What transfers should I consider making?" quick prompt in The Gaffer's Pulse is free, up to three quick-prompt questions a day. It reads your actual squad, bank and free transfers. Asking your own follow-up questions, such as a specific swap or a two-gameweek plan, needs Pulse Pro (£2.99/month, 7-day free trial).
- Can ChatGPT suggest FPL transfers?
- It can talk through transfer ideas, but on its own it doesn't know your squad, your bank, the price you'd get for each player, or this week's injury flags, and it tends to recommend players from memory who may have moved clubs or lost their place. An FPL-specific assistant connected to your team ID works from live data instead.
- How does the AI know if I can afford a transfer?
- It doesn't add up prices itself, because language models are unreliable at arithmetic. Before it says a move fits, it calls a calculator that uses your real bank and the selling price of the player going out (purchase price plus half of any rise, rounded down), and the AI may only say a move is affordable if that check passes.
- Should I take a -4 hit for an AI-suggested transfer?
- Only if the gain clearly beats four points over the gameweeks you'll own the player. A 1.5-point xP upgrade for one week does not cover a hit; a 4-point gain over the next three gameweeks roughly breaks even; anything less than that usually means rolling the transfer is the better play. The Gaffer knows your free-transfer count and that each extra transfer costs 4 points, so ask it to compare the hit against rolling.
- How accurate are AI transfer suggestions?
- They are only as good as the projections underneath. The Gaffer ranks candidates by our predicted points model, which we grade publicly against FPL's own expected points every gameweek. It still can't see late team news, so always check the lineup before the deadline.