FPL 26/27: AI vs Humans — Building the Perfect Squad
Published on HivePostify by @talesfrmthecrypt · Mon Aug 10 2026
If you're anything like me, you've probably spent far too much time tinkering with your FPL squad already. And with the GW1 deadline of Friday 21 August fast approaching, the big question remains - what does the perfect FPL squad actually look like?
This season, rather than relying solely on gut instinct, I've decided to pit FPL managers against the machines.
On one side, we've got the template squad — the players that millions of managers are currently backing. On the other, we've got AI and a whole load of data, crunching the numbers to identify the players it thinks will deliver the most points.
I'll also throw a few other approaches into the mix, including the best points per million picks (a favourite from previous years) and a squad made up of players who didn't feature in the Premier League last season.
So, who should you put your faith in to deliver that all important initial squad — the FPL community, the algorithms, or ......me?
Let's find out...
The Template Team
Here then are the most popular picks (at time of writing) in FPL 26/27. Obviously this is not a viable squad both in terms of budget (£12mil over) and the fact that it contains 4 Man City players.
In goal, FPL managers are once again favouring Arsenal's David Raya despite being the most expensive keeper in the game. He certainly offers you a strong chance of picking up clean sheet bonuses behind that Arsenal defense but would the money be better spent elsewhere?
As back-up goalie, players have so far gone for Dubravka who was a popular pick last season too. Of course, last year he was Burnley's number 1 while this season he's likely to be warming the bench at Spurs.
If you have David Raya in your side then you probably don't care who is on the bench because he will start every game. However, there is the strong possibility that Dubravka's price will fall if and when managers realise he's not getting game time this season so personally I'd be inclined to look elsewhere for a back-up keeper.
In defence, Gabriel is £1.5mil more expensive than any other defender but that hasn't stopped him appearing in a quarter of teams.
Only just behind him is Luke Shaw who is £4.5mil and enjoys a nice run of fixtures in the early part of the season. What is somewhat surprising in regards Shaw as well as O'Rilley and Porro is that managers are still favouring full-backs who might give you an attacking return over CBs who will almost certainly provide def con bonuses.
Just as was the case last year, I think defence is going to be absolutely key to getting a high finish in 26/27. You should certainly be investing in that area of the pitch but I'm not sure I fancy some of the picks that the FPL community is making to date.
Fernandes and Semenyo were key performers last year and managers are banking on them to perform again. Meanwhile, Szoboszlai, who had a similarly good 25/26 seems to be the man that almost everyone wants in their squad due to his £7mil price tag.
The other 2 picks are very curious ones. Declan Rice was not fit during England's World Cup campaign and I'd be surprised if he plays quite as many minutes in the first part of the season. The signing of Bruno G from Newcastle will help to facilitate that.
Morgan Rogers to Chelsea was a bit of a shock move. We wait to see exactly how Alonso will use him in a side with a good number of attacking midfielders already in it. Personally, I don't think he's worth the risk at this stage.
Up front, almost 3 quarters of FPL managers have Haaland despite his £15.5mil price tag. Are you one of those brave managers not to put him in?
He's joined in the majority of sides by Joao Pedro who is the 2nd most popular pick across all players to date. The Brazilian is very likely to be Chelsea's primary striker this season, has started the pre-season well and has a bit of a point to prove after missing out on the World Cup squad.
He's also increased his goal and assist return season on season since coming to the Premier League. If he manages to maintain that trend then £7.5 mil will be a bargain.
It's very close between Calvert-Lewin and Brobby for the 3rd striker spot but the Leeds man is just shaving it at the moment. He's looked sharp in pre-season so far and Leeds do have a decent start to the campaign.
Fitness is always the key for him. If he's up to speed and starting week in week out then he's a top pick. If not, well I'd avoid him in favour of several other £6mil players who could do the job.
The AI Picks
I've gone for 2 teams that AI favours in terms of predicted points over the first 3 game weeks.
The first team below is the 15 players that it believes will score the most points during those opening few rounds but of course this isn't a viable team.
It really loves Manchester (The Andy Burnham effect apparently extends to AI) with 4 United and 4 City players.
What's interesting is that only 5 of the players currently in the most popular picks by humans feature in the side that AI thinks would be the best 15 players you could pick from for the opening 3 rounds.
The likes of Shaw, O'Riley and Guehi are ranked 25th, 26th and 29th respectively amongst defender picks with uncertainty over the availability of the 2 City players who went the distance with England in the World Cup.
Porro meanwhile ranks 68th with similar concerns over his immediate availability while AI remains suspicious on picking Spurs defenders after 2 woeful Premier League seasons.
Szoboszlai just misses making the cut by 0.1 of a point over 3 gameweeks. Rogers is ranked 21st but behind Cole Palmer who is expected to deliver a point a game more than his new teammate. Declan Rice is way down in 50th for midfielders, again with AI concerned about his game time early in the season.
Up front, Calvert-Lewin is ranked 4th and Joao Pedro 6th behind Erling Haaland who is the number 1 ranked forward and player overall with a predicted 22 points over the first 3 GWs.
The 2nd AI Team is the viable 15 that the algorithm suggests you go for.
As I suggested above, defence is likely to be an area that managers will invest in again this season and AI agrees going for the expensive Gabriel alongside VVD + a £6.0mil Lacroix and £5mil Thiaw. These are all players who are likely to put up good def con numbers as opposed to full-backs who probably won't!
I agree with the tactic but maybe not the personnel. Gabriel at £8mil is fantastically expensive while I think there might be better options out there than Chelsea's new signing Lacroix and Newcastle's Thiaw.
By going big at the back, the AI hasn't left budget for Fernandes despite suggesting he will be the highest point scoring midfielder. Instead it's gone for United teammate Mbeumo who benefits from his side's decent fixtures at the start of the season and everyone's favourite midfielder (human and AI), Szoboszlai.
At £6.5mil it likes Palace's Sarr who certainly had a very good 25/26 campaign but like so many players this season, will find himself in a new system under a new coach.
Elsewhere it's 2 budget Hull picks in the midfield which again, means the formation it favors is 4-3-3. In 24/25 and earlier seasons, almost no one would pick a back 4 but it's now becoming the norm to invest in that area of the pitch over midfield and attack.
Upfront it's gone fairly predictable with its picks although I'm not as convinced that Thiago will be good value at £8mil. While refs always seem to give Brentford twice as many penalties as any other side, can you really bank on him propping his numbers up on that front again?
Best points per million squad
Finally a team you could actually select (if you wanted to) with a cool £9.5mil budget to spare!
The equation for this one is simple divide the player's total points from last season by their cost this season to get a points per million valuation.
The idea is that it's a good way to see where the value buys are at the start of this year. The limitation is clearly that it's all based on historical data which is no guarantee of future success (although you could probably say the same of any squad in this blog).
Still, I think it throws up a couple of contenders for spots that might be getting overlooked thus far.
Tyrick Mitchell is the best value defender using this method at just £4.5mil and that to me seems a fair judgement. He's been an almost ever present for Palace the last 3 seasons and is remarkably consistent.
If you're looking for that '4th' defender then you could do a lot worse. Palace play Man City 3rd game of the season but otherwise they don't have any really tough looking matches until November!
In terms of the more premium end of thing, Senesi and Guehi will both set you back £6mil. While Guehi could prove a solid investment as one of the first names on the team sheet at City, it's unlikley Senesi will be an automatic pick for Spurs.
You could instead go for Senesi's new Spurs teammate, Van Hecke who seems far more likley to get starts for his old boss De Zerbi. Alternatively you've got Senesi's old teammate Truffet who despite rising to £5.5mil is still a better value pick than a whole host of bigger name players at that price point.
Into midfield and the best value pick is Elliot Anderson. He was an absolute must have last season where he scored a 180 points from a starting price of just £5.5mil. It remains to be seen if he can have quite the FPL impact now that he's moved to City.
James Garner is another one who flew under the radar for much of last season. He has had minor groin surgery over the summer and so is set to miss the early rounds of the 26/27 season but might be one to keep an eye on as the season progresses.
Harry Wilson was absolutely brilliant for Fulham last year. Can he replicate that form for new club Leeds?
Tags: #football#soccer#premierleague#epl#hivefpl#fantasy#sportsblock#footballexchange#ocd#sports