Leeds sign £15m Fulham star and wonderkid full-back in £125m monster transfer window prediction

With the summer window open once more and plenty of business already getting done, one wouldn’t blame any Leeds United fans if they were starting to develop an itchy transfer finger ahead of their return to the Premier League.

Daniel Farke’s newly-promoted side have already confirmed one impending signing, with VfL Wolfsburg forward Lukas Nmecha set to join on a free transfer at the beginning of July, but so far they haven’t spent a penny of their supposed £100m-plus transfer budget.

We’re expecting the club to complete the purchase of Slovenia international defender Jaka Bijol from Udinese soon, so reinforcements are on the way, but keeping track of everything Leeds are planning has been made all but impossible thanks to deluge of transfer rumours that have been falling off the rumour mill incessantly over recent weeks.

Given that actually working out which transfer stories are true and which are simply not going to happen is nearly impossible for any human being, we thought we’d see if a computer could do any better – so we asked a well-known AI learning model what it predicted Leeds’ transfer window would look like come the start of September.

Is AI able to cut through the noise and offer some insight into what might actually happen? Or is it, as we strongly suspect, going to prove to be no more able to get inside Farke’s mind than we are? We’ll know for sure in a couple of months’ time, but for now, here’s how the machines see the summer window panning out…

Spurs, Brighton & Fulham stars on Leeds’ predicted shopping list

The good news right off the top is that the algorithm is confident that Leeds will indeed secure the signing of Bijol pretty soon, apparently for a fee of £14m – but that isn’t the only defensive addition predicted.

Apparently, Leeds will sign not one new left-back – a necessity given that Junior Firpo is leaving on a free transfer – but two. Gabriel Gudmundsson, the Lille and Sweden left-back who has been linked with a move to Elland Road by a few media outlets, is one, with a £10m price tag on his head, but the other purchase would be a little more audacious.

The AI reckons that Leeds will sign Brighton full-back Valentín Barco on a permanent deal after a frustrating first season in English football for the Argentine. After failing to make an immediate impact after a £7.5m move last summer, he spent most of the season on loan at Sevilla and Strasbourg, and now he could apparently be sold altogether. Leeds would pay £12m for the privilege, so Brighton turn a quick profit, but given that he was touted as one of the most promising left-backs in the world at Boca Juniors, we wonder if the Seagulls will really give up on him so easily.

Another player whose hasn’t really been linked to Leeds much in the gossip columns is Fulham midfielder Sander Berge, signed for £15m apparently. Given that he played 31 matches for Marco Silva’s side last season, we’re not sure that one’s all that realistic, but adding a new midfielder seems sensible enough – the AI also thinks they’ll get Strasbourg wonderkid Habib Diarra, a target for several sides, for around £20m.

Rounding the shopping list off are two players that very much have been linked with a move – striker Nikola Krstović of Lecce and Spurs winger Manor Solomon, who was a hit on loan with Leeds last season, signed for a combined £30m.

Illan Meslier among those leaving according to the algorithm

On the sales front, the AI reckons that seven players will be leaving Elland Road this summer, including Firpo. A few names won’t come as a surprise and don’t require any great predictive powers on the computer’s part: Joe Gelhardt, Darko Gyabi and Sam Greenwood are all sold in our simulation for a combined fee of a little over £10m.

Max Wöber gets sold off to for £6m – the AI doesn’t comment on who he’ll be sold to – while Jack Harrison is also off for £10m. Few alarms or surprises there, but at least Leeds don’t seem likely to lose any key players. The final player on his way is controversial goalkeeper Illan Meslier, apparently leaving for Strasbourg for £15m after making one too many mistakes for Farke’s comfort. Once touted as being among Europe’s most promising shot-stoppers, it seems his time with the club is set for an ignominious end.

Now, one problem that may be immediately observed is that the AI seems to have Meslier leaving without a replacement. As it happens, it did suggest that Leeds would buy a new goalkeeper – but in the process, helpfully demonstrated that you really shouldn’t use AI to do your homework.

That’s because it has Leeds dropping £25m on Caoimhín Kelleher, which would be a fine idea if he hadn’t already signed for Brentford. Why an apparently powerful and advanced learning algorithm can’t pick up on something like that, we don’t know, but it rather undermines the limited faith we had in its ability to determine the shape of the summer transfer window.

So in the end, Leeds spend just over £125m, buy a player who’s only just signed somewhere else, and get a shade of £40m back to ensure that they don’t have to fuss about PSR problems. A realistic assessment of the transfer window? Probably not. But we’ll find out just how good these machines really are in a short while…