YouTuber finds out if AI can win the Premier League on Football Manager

Can AI really outsmart one of the most demanding football simulations on the planet? YouTuber Zealand set out to answer that question by handing over the tactical reins to ChatGPT in Football Manager 2024, with one clear goal: win the Premier League with Manchester United in a single season. No mods, no cheats, no shortcuts. Just raw GPT advice.

This would be impossible in real life, because the Manchester United squad is diabolically bad. But in Football Manager, it’s a bit easier. So, can AI do it? Or will it take a few more years until it’s caught up with us.

Step 1: The AI Assessment

ChatGPT’s plan starts sensibly: assess the squad. The bot quickly flags Bruno Fernandes, Rashford, Casemiro, and Varane as key assets, while pointing out clear squad weaknesses at right back, in midfield depth, and backup striker options.

Already, it’s making more sense than Rio Ferdinand.

Step 2: Financial Meltdown

ChatGPT instructs Zealand to sell Donny van de Beek (already on loan), Martial, Maguire, McTominay, and Brandon Williams to raise funds. And it’s got big plans for that cash.

Recommended signings include:

• Jean-Clair Todibo or Giorgio Scalvini (CB)
• Manu Koné, João Neves or Romeo Lavia (CM)
• Jeremie Frimpong or Tino Livramento (RB)
• Mohammed Kudus or Jesper Lindstrøm (versatile attackers)

But there’s just one problem: ChatGPT didn’t check the bank balance. The ambitious wishlist forced Zealand into creative accounting, racking up over £100 million in future debt just to satisfy the bot’s shopping spree. Well, you can’t say that ChatGPT isn’t realistically emulating Manchester United in real life.

Step 3: Tactics by AI

Next, the AI delivered a full tactical setup:

• Formation: 4-2-3-1 Gegenpress (because, of course)
• Mentality: Positive
• Team Instructions: High tempo, short passing, counter-press, high defensive line

It even suggested training regimes, friendlies against top-tier teams like PSG and Napoli, and the importance of praising players after good training performances.

While basic, it wasn’t wrong. It just wasn’t… imaginative.

Mid-Season Check-in: The Fall to 10th

By January, the cracks were showing. Despite Champions League progression and a deep Carabao Cup run, United sat 10th in the league.

Cue the mid-season GPT pep talk:

• Use the Data Hub
• Switch to a more cautious 4-3-3
• Bench underperformers
• Focus on clean sheets
• Short-term tricks like team meetings and rest days

But with morale low, injuries stacking, and finances crippled, Manchester United fans are getting a serious case of deja vu.

The Final Push: Did It Work?

With some tactical tweaks and a crucial January loan for Daichi Kamada, things did improve… slightly.

• Final League Finish: 8th
• Champions League: Quarter-final exit vs. Barcelona (respectable I guess)
• FA Cup: Semi-final loss to… Bolton Wanderers
• Carabao Cup: Final defeat to Man City

A valiant effort. A few moments of brilliance. But not the success United fans (or Zealand) were promised. Although, granted, it is still a lot better than Manchester United this season under Ruben Amorim (providing they don’t beat Tottenham on Wednesday)

Verdict: D+ for the AI Gaffer

Despite a promising start and the occasional tactical gem, ChatGPT ultimately fell short. While it gave solid foundational advice and showed flashes of understanding, the lack of nuance, financial realism, and adaptive in-game management proved costly.

“As developed as it might be,” Zealand concludes, “ChatGPT still can’t handle the heat of Football Manager.”

For now, Football Manager remains a human-dominated domain. But give it a few versions—and a few financial fair play modules—and who knows? You can watch the video and full experiment here: Can Chat GPT Win The Premier League?