Gameweek eight Premier League predictions vs ChatGPT

Last time out, ChatGPT correctly predicted seven results, but didn’t get any of the scorelines. Meanwhile, I only got four results right - and nearly got the Brighton v Tottenham scoreline correct, just the wrong way around.

Thankfully, since that was only the first week we’ve done this, we don’t have a big enough sample size to claim that data-driven AI is categorically better than the simple eye test (which this reporter prefers much more than the woke nonsense of expected goal involvment and other mundane stats).

The gloves are well and truly off this week; here is what ChatGPT has predicted for this weekend’s Premier League games, and what myself and the AI think the scorelines will be.

Tottenham v West Ham

ChatGPT: 3-1 | David George: 2-0

Tottenham have been scoring goals for fun recently, and will be keen to put the misery of bottling the win against Brighton behind them. Because of that - and West Ham’s poor form - I think the Hammers will be entirely shut out at the Tottenham Hotspur Stadium.

Meanwhile, ChatGPT expects them to score, saying that “West Ham, although competitive, might struggle away from home against a well-organised Spurs side.”

Fulham v Aston Villa

ChatGPT: 1-2 | David George: 0-1

Aston Villa has statistically been one of the best defences in the league this season - although sometimes the eye test doesn’t support that. Ezri Konsa’s injury is cause for concern, but Unai Emery’s sides have always been sound along the back line.

At the same time, Fulham have been a surprise package over the first seven games, and after Aston Villa have a relatively easy run of fixtures. ChatGPT claimed that “Villa's attack should give them the edge”.

Ipswich v Everton

ChatGPT: 2-1 | David George: 2-0

Liam Delap is in the form of his life. The AI says that “Ipswich's high-tempo football at home could be too much for Everton,” and I could not agree more. But because of this, I’m perplexed as to why ChatGPT thinks Everton will score.

Defensively Everton are a decent team, but they really do lack offensively, and I think they will really struggle at Portman Road.

Manchester United v Brentford

ChatGPT: 2-1 | David George: 1-1

Some might call it an anti-Man United agenda, but once again I just don’t see them getting the result ChatGPT expects them to. Last week it forecast a draw, but a 2-2 rather than the 0-0 we ended up with.

Personally, I think Brentford’s attack, coupled with United’s defensive woes, will lead to a 1-1 draw this weekend. ChatGPT said: “Man Utd's inconsistency has been an issue, but they generally perform better at Old Trafford. Brentford, known for their disciplined play, could give them a tough challenge, though United might edge it with their individual quality.”

Newcastle v Brighton

ChatGPT: 2-2 | David George: 1-2

Newcastle are seriously struggling without Alexander Isak upfront, and I think Kieran Trippier’s injury will expose them on the wings too. Meanwhile Brighton are in freescoring form at the moment - expect Mitoma to pop up with at least one goal contribution, while Welbeck’s bizarre resurgence continues in force.

Writing about this matchup, ChatGPT said: “This could be a high-intensity game between two attack-minded teams. A close and exciting contest seems likely.”

Southampton v Leicester

ChatGPT: 1-2 | David George: 2-3

A relegation six-pointer is always a great watch. Both teams have been woeful defensively this season, but Leicester have far more attacking threat; I think this one goes the distance, with a late winner handing the victor the spoils.

ChatGPT claimed that “Leicester's efficiency might get them the win,” and I’m inclined to agree.

Bournemouth v Arsenal

ChatGPT: 0-3 | David George: 0-4

Bournemouth’s attacking threat has been their biggest asset so far this season, but their defensive record has been middling against worse opposition. Arsenal meanwhile are rock solid at the back, and even without Bukayo Saka will probably run riot at Dean Court.

According to ChatGPT “Arsenal's firepower and control in midfield should see them win comfortably,” which, even as a Bournemouth fan, sounds about right to me.

Wolves v Manchester City

ChatGPT: 0-3 | David George: 1-4

Without Rodri, it’s difficult to see Manchester City keeping a clean sheet, especially away from home. Even Fulham put two goals past them last time out, and I think Wolves has the strength in attack to get a goal too.

But upfront, Man City are firing on all cylinders - and the best way to turn attention from Pep Guardiola’s future is to absolutely hammer the team in front of you. The AI wrote that “Even without Kevin De Bruyne, Man City have shown their depth and should dominate.”

Liverpool v Chelsea

ChatGPT: 2-1 | David George: 2-1

A bizarre moment where ChatGPT and I are in total agreement. The AI suggested that “both sides have been inconsistent, but Liverpool, especially at Anfield, tends to have the edge” - and while I don’t think the sentiment is true (Liverpool are top of the league for a reason) I do think that is the outcome we will see here.

Cole Palmer will inevitably contribute for the Blues, but Liverpool certainly have the edge here.

Nottingham Forest v Crystal Palace

ChatGPT: 1-1 | David George: 2-1

The AI claimed that “Palace's counter-attacking ability could make this a tight match, but Forest might just nick it,” - and then forecast a draw. Forest are in electric form and in the top half of the Premier League table, while Crystal Palace might end up in a relegation scrap.

If I were a betting man, I’d put money on Forest going 2-0 up, with Palace getting a consolation goal late into the match.