Using underlying data from this and previous seasons, we’ve tasked our advanced AI model with creating Premier League predictions with correct score picks on every match of the 2026/27 campaign.
Our AI predictor combines data from matches in this season, previous seasons, bookmakers and more to determine the most likely correct score outcome from the game.
As the tournament progresses, the model will focus more heavily on the games that have already taken place within the tournament and it will look beyond results to input shot, possession and general match momentum data to see where value may lie.
Crystal Palace vs Manchester City
AI Prediction: 1-3
The Eagles opened with a 2-0 defeat at Everton and the numbers say they deserved better: 1.60 xG, four clear-cut chances created, all four spurned. That profile has been Palace’s problem for a while — they scored just 41 league goals last season, fewer than any other side that stayed up. City won 2-1 against Bournemouth from 1.97 xG and five big chances, wasting four of them, and were still comfortable. The recent history here is brutal, too. The Citizens have won the last three meetings 11-2 on aggregate, including a 3-0 stroll at Selhurst last December. Palace should get on the board this time. Stopping City is the harder job.
Liverpool vs Nottingham Forest
AI Prediction: 3-1
Liverpool were arguably the best team in the division on opening weekend and have a single point to show for it. The Reds mustered 3.07 xG from 27 attempts at St James’ Park, hit the target seven times and drew 2-2, with the Newcastle goalkeeper making five saves. Volume like that usually pays eventually. Forest offer a stark contrast: beaten 1-0 at home by Leeds from 0.75 xG and 12 shots, only two of them on target. The caveat is a real one, mind — the Tricky Trees won 3-0 at Anfield last November. But that was one afternoon, and Forest have yet to score this season. Liverpool by two.
Coventry City vs Hull City
AI Prediction: 0-0
A first Premier League meeting between two of the promoted sides, and recent history points firmly at a stalemate. Their last three encounters all finished level: 0-0 at Hull last April, 0-0 at Coventry the previous August, and 1-1 before that. Two goals in three games. Opening weekend did nothing to change the picture. Coventry were picked off 3-0 at the Emirates, managing four shots and 0.30 xG — the lowest attacking output of anyone in the round. Hull saw off Manchester United 2-0 but were outshot 21-8 and out-created 1.79 to 1.27 on xG, with their keeper making five saves. Two cagey sides still finding their feet. Take the draw, and take the unders with it.
- Enhanced Offers
- Exceptional Mobile App
- Uniquely Designed Sportsbook
AD 18+ New UK customers only, Place a £10 fixed odds bet at minimum odds of 1/2 and get 3 x £10 free fixed odds bets, 3 x £5 Total Goals football spread bets and 3 x £5 Winning Favourites spread bets on consecutive days and a £1 racing Race Index spread bet. The promotion will run from 11th March 2025 until discontinued by Spreadex Ltd. using its reasonable discretion, (“Promotion Period”). Offer can only be used once per customer. Full T&Cs apply.
Bournemouth vs Everton
AI Prediction: 1-1
The Cherries finished sixth last season on the back of 18 draws — no side in the division shared the points more often — and that habit feels relevant here. Bournemouth were beaten 2-1 at the Etihad but offered next to nothing going forward, registering five shots and 0.54 xG. Everton arrive with a 2-0 win over Palace banked, though the Toffees rode their luck: 1.06 xG created, 1.60 faced, four saves from their goalkeeper. Last season’s meetings split neatly, each side winning away from home by a single goal. Two well-organised teams, neither carrying much attacking momentum into the weekend. A share of the spoils looks the sensible read.
Tottenham Hotspur vs Newcastle United
AI Prediction: 1-2
Spurs were dismantled at Brentford and the underlying numbers were uglier than the 3-0 scoreline. They were outshot 26-9, faced 3.81 xG and produced 0.62 of their own. That followed a 17th-place finish last season, when only the three relegated clubs conceded more than their 57. Newcastle drew 2-2 with Liverpool in a game they were fortunate to take anything from, riding 3.07 xG against. But the Magpies have Spurs’ number — a 2-1 win in north London last February and a 2-2 draw at St James’ Park in December, four points from six. Both defences look vulnerable, so goals should come. Newcastle’s greater conviction in front of goal ought to settle it.
Sunderland vs Fulham
AI Prediction: 2-1
This is the contrarian call on the card, because the head-to-head is emphatically Fulham’s. The Cottagers completed the double last season, winning 3-1 on Wearside in February and 1-0 at Craven Cottage in November, 4-1 on aggregate. Sunderland, though, finished seventh — two points and four places above Sunday’s visitors — and their opening-day 2-1 defeat at Ipswich came with two clear-cut chances created and both missed. Conversion like that tends to correct itself. Fulham were beaten 3-2 at home by Chelsea and shipped 51 league goals last term. The Black Cats were below their best at Portman Road but had enough openings to take something from it. At home, they carry the greater threat.
Leeds United vs Brentford
AI Prediction: 1-2
Brentford produced the performance of the opening weekend and it barely registered outside west London. The Bees put 26 shots on Tottenham, eight on target, and generated 3.81 xG — comfortably the highest figure of the round — to win 3-0. Leeds got the job done at the City Ground, but 1-0 from 0.59 xG and 11 attempts was a grind rather than a statement. Both meetings last season finished level, 0-0 at Elland Road and 1-1 in west London, so there is a case for another tight afternoon. The gap in current form is too wide to ignore, though. Elland Road will make it awkward and Leeds should score. Brentford’s numbers point to them scoring more.
Chelsea vs Brighton & Hove Albion
AI Prediction: 2-2
The most watchable fixture of the weekend on paper, and both sides opened up. Chelsea won 3-2 at Craven Cottage from 1.87 xG and four clear-cut chances, while Brighton battered Aston Villa 4-0, racking up 21 shots and 2.82 xG. The history is all Seagulls, mind. Brighton did the double last season, winning 3-1 at the Bridge in September and 3-0 on the south coast in April — 6-1 across the two games, with Chelsea managing a single goal. That said, the hosts look sharper than the side that finished tenth, and home advantage counts for something. Both defences leaked freely in matchweek one and neither side looks minded to sit in. A high-scoring draw is the call.
Manchester United vs Ipswich Town
AI Prediction: 3-1
United’s 2-0 defeat at Hull was a jolt, but it was not the performance the scoreline suggests. They had 21 shots to Hull’s eight, posted 1.79 xG against 1.27, and created two clear-cut chances without taking either. Third place and 69 goals last season was no fluke. Ipswich made a decent fist of their return, edging Sunderland 2-1 at Portman Road from 1.20 xG, and the Tractor Boys will not be overawed — they drew 1-1 with United on their own patch in 2024 and lost only 3-2 at Old Trafford that season. Even so, a wounded United at home to a newly promoted side has a familiar look. Expect a response, and expect it to be emphatic.
Aston Villa vs Arsenal
AI Prediction: 0-2
Villa’s opening day was as bad as it gets. Beaten 4-0 at Brighton, they had six shots and failed to register a single one on target, finishing with 0.31 xG. For a side that finished fourth on 65 points, that is a worrying afternoon’s work. The champions were rather more convincing, seeing off Coventry 3-0 with 20 shots while restricting the visitors to 0.30 xG — the same defensive control that brought them the title with just 27 conceded. The one flag is that the Villains beat Arsenal 2-1 here last December, so Villa Park is no gimme. But the Gunners won the return 4-1 and look in far better shape. A clean sheet for the visitors is the likeliest route.
The stats so far
Of course, we wouldn’t keep you in the dark in terms of how our AI model is doing with it’s weekly predictions. Here’s the latest:
| Matches Predicted: | 10 |
| Correct Scores: | 7 |
| Correct Outcomes: | 1 |
| Correct Score hit-rate: | 10% |
| Correct Outcome hit-rate | 70% |
