I Let My FPL AI Pick a Free Hit. It Dropped Haaland.

I built an AI assistant for FPL and let it pick my Free Hit team. And it dropped Haaland.

My first thought was: well, that can't be right.

But then I looked at who it picked instead… and now I'm not so sure.

What I asked it to do

The brief was simple: a hundred-million-pound budget, a full legal squad, and this gameweek only. That's the Free Hit approach: your usual team comes back afterwards. Here it is in my app, FPL Cat. It spent ninety-nine point nine million and picked a four-five-one. So the money is concentrated in defence and midfield, with one budget striker starting.

My real fantasy team

And for context, here's my team loaded in FPL Cat, alongside the assistant. The app starts from my last-deadline lineup. And here's how I've currently set it up on the official Fantasy site — Haaland still has the armband. The hundred-million-pound Free Hit squad is a separate experiment. Let's go through the planner's picks.

Why no Haaland?

Haaland costs fifteen point six million, and City go to Liverpool. He's still expected to score six point one points. The planner hasn't decided he's a bad pick.

In yesterday's comparison, I also ran a version that had to include him. That team came out at sixty-one point five expected points. Without him, sixty-two point six. Just one point one between them. That's a tiny edge, not proof Haaland will blank.

This is a bet on the whole squad, and with almost seventy-four percent ownership, it could be an uncomfortable watch. But let's roll with the result.

Pickford in goal

In goal: Jordan Pickford, five point five million, away at Hull. Three clean sheets and fifteen saves from five games give him more than one route to points. He's on three point seven for this week. That's a sensible starting keeper, although we're doubling up on Everton's defence, so one Hull goal can hurt twice.

The Arsenal double-up

The defence starts with Gabriel and Calafiori, both at home to Leeds. Gabriel is expensive at eight million, but comes out at five point nine expected points. He's played every minute so far, and offers clean-sheet points, defensive contributions and a goal threat. Calafiori costs five point nine, with four point six expected points. Two assists already, and five starts from five. This is a deliberate Arsenal double-up, with attacking potential on top of the clean-sheet bet.

Tarkowski and Hall

James Tarkowski completes the Everton double-up. He's six point two million and expected to score five point two points at Hull. He's already on forty-three points this season, with seven bonus points. The prediction includes roughly one point one from defensive contributions alone, so his case goes beyond a clean sheet.

Then Lewis Hall, five point three million, away at Coventry. One goal, one assist and six bonus points so far. He's on four expected points: a mix of attacking involvement, defensive work and clean-sheet potential.

Bruno gets the armband

Here's the captain: Bruno Fernandes. Eleven point nine million, Spurs at home, and six point five expected points before we double the captain's score. He's played all four hundred and fifty minutes, scored three goals, supplied an assist, and is United's first-choice penalty taker.

Alongside him is Bryan Mbeumo at seven point nine million. Two goals and three point zero seven expected goals already. He's on five expected points. We're backing two routes into United's attack.

Saka and Gibbs-White

Bukayo Saka is vice-captain. Nine point six million, Leeds at home, three goals in the first five games, and Arsenal's first-choice penalties. He's on six expected points.

Morgan Gibbs-White costs eight million and goes to Palace. A less inviting fixture, but he has one goal, two assists and six bonus points already. He's also on first-choice penalties. Four point seven expected points keep him in this midfield.

The cheaper midfielder

The fifth midfielder is Enzo Le Fée, five point seven million, at home to Brighton. He's started every game and has two point zero six expected goals plus one point three four expected assists. Only one actual goal and one assist so far, so the appeal is partly in the underlying chances. Four point two expected points make him the budget option that helps this midfield fit.

Gonzalo leads the line

Up front, it's Gonzalo García for Fulham, six million, away at Ipswich. Just one goal so far, but two point three three expected goals and five starts from five. He's on three point five expected points. This is the compromise: a cheaper striker lets us afford that midfield. We're backing the chances to turn into returns, rather than buying a forward who's already hauled.

All four bench players

First substitute: Thierno Barry, five point seven million. Two goals, three point five expected goals, and three point three expected points at Hull.

Second: Leif Davis, four million. Two goals and two assists already; three point one expected points against Fulham.

Third: Emersonn, five point five million, also against Fulham. Two goals and an assist, but he often comes off early. Three point one expected points, with minutes the concern.

Backup keeper: Tzolakis, four point seven million. Three clean sheets and seventeen saves already. He covers Pickford; his points don't simply get added to our score.

The final team and the question

So that's the team: four-five-one, Bruno captain, Saka vice, and no Haaland. The live planner now shows sixty-two point seven expected points, including captaincy and estimated bench cover. It's a prediction, not a promised score. And that tiny edge in yesterday's Haaland comparison is exactly why this is an interesting experiment. I'm rolling with the planner's pick for this video. Would you trust this team, or would you find a way to bring Haaland back?