There are always changes going into any Premier League season. But on the eve of the 2026-27 campaign, there is more uncertainty than ever. 

The first big change is that Arsenal are defending champions for the first time since 2004, with Mikel Arteta hoping to oversee the Gunners' first successful defence of the top-flight crown since 1934-35.

Manchester City, Liverpool and Chelsea are all under new management, with Enzo Maresca, Andoni Iraola and Xabi Alonso taking those respective jobs.

And the remaining members of the 'Big Six' – Manchester United and Tottenham – are entering their first full seasons under Michael Carrick and Roberto De Zerbi. 

So, might stability be the key for Arsenal as they attempt to hold onto the trophy?

Overall, nine of the Premier League's 20 clubs have new head coaches, including Bournemouth – who moved quickly for Marco Rose after Iraola's move to Anfield – and Newcastle United, who will hope Matthias Jaissle can build on Eddie Howe's fine work.

Nine is also the number of representatives the Premier League will have in Europe in 2026-27, with Bournemouth, Crystal Palace, Sunderland and Brighton all having continental football to deal with this term. Will the additional demands prove too great for any of those sides?

There is also a novel feeling surrounding two of the promoted clubs, with Coventry City back at the top table for the first time since 2000-01 and Hull City returning following a nine-year exile. Ipswich Town bounced back at the first attempt, but even they have a new man in the dugout after Gary O'Neil replaced Kieran McKenna. 

With so much still up in the air, we have turned to the Opta supercomputer to predict the 2026-27 season. 

And after 10,000 data-led simulations of the campaign, here are the predictive model's results.

Gunners tipped to achieve rare title feat

The supercomputer did not predict Arsenal's title win in 2025-26, though it was not too far off, with the Gunners listed as second favourites. Their 24.3% probability was only marginally below Liverpool's 28.5%, but the Reds underwhelmed following a £450m transfer outlay. 

This time around, the Gunners are seen as the team to beat, with 37.9% of our simulations resulting in Arteta's team retaining the trophy.

Defending the Premier League title is notoriously tough. 

Only three clubs – under three managers – have ever accomplished that feat. They are Manchester United under Alex Ferguson (four such streaks, including two instances of three in a row), Chelsea under José Mourinho, and Manchester City under Pep Guardiola (two streaks, including four in a row from 2020-21 to 2023-24).

Arsenal's rock-solid defensive play and reliance on set-piece goals meant they may not have been the most popular champions.

But this is a points game, not a popularity contest, and their 2025-26 defence was one of the best the Premier League has ever seen, judging by the underlying numbers. 

Their 27 goals conceded were comfortably more than Chelsea's frankly ridiculous single-season record of 15 from 2004-05, but Arsenal's average of 0.74 expected goals against (xGA) per game was the fourth-lowest figure on record (since 2012-13) after Man City's title-winning teams from 2017-18, 2021-22 and 2018-19.

The back injury sustained by William Saliba at the World Cup will be a concern, though, as Arsenal's Premier League win percentage drops from 68.7% to 47.6% when the Frenchman is absent, since his 2022 debut. Arsenal also failed with an audacious transfer swoop for Real Madrid's Vinicius Junior recently.

But bringing in Bruno Guimaraes to complement Declan Rice and Martin Zubimendi suggests Arteta's team will be combative again – the former Newcastle United midfielder led all players in the Premier League for duels won (6.1) and tackles won (1.6) per 90 last term.

Guardiola may be gone, but the supercomputer considers City to be Arsenal's nearest rivals, as they are assigned a 20.5% chance of winning the title.

It was difficult to know what to make of Maresca's time in charge of Chelsea. In just 18 months, he led the Blues to Champions League qualification and lifted the Club World Cup and Conference League trophies, all while working with a bloated squad.

But his average of 1.74 points per game was the worst record by anyone to lead Chelsea for at least 30 Premier League games this century, and tensions with the club's hierarchy led to his exit.

However, the Italian inherits a squad that is superbly equipped for his possession-oriented style, spearheaded by three-time Golden Boot winner Erling Haaland, who could now match the record of four scoring crowns (held jointly by Thierry Henry and Mohamed Salah). 

City's £116m swoop for Elliot Anderson should give them a key midfield cog for years to come, though Rodri continues to be linked with Barcelona and Bernardo Silva has already left for Real Madrid.

No other team is assigned more than a one-in-10 chance of winning the title, with Liverpool's 9.2% probability putting them third. The Reds, of course, finished as champions the last time they went into a season under new management, when Arne Slot took charge ahead of 2024-25.

They need Alexander Isak and Florian Wirtz to take big second-season jumps. Injuries and poor form restricted the former Newcastle striker to four goals overall last term, and while Wirtz showed flashes of his quality in 2025-26, more is expected of the £116m creator.

The arrivals of Jeremy Jacquet and Ronald Araujo should give Iraola two viable options to partner Virgil van Dijk in defence, but Liverpool are yet to replace Salah, their all-time leading Premier League scorer (191) and assister (93).

Man Utd are the supercomputer's fourth favourites for the title, at 6.2%. The Red Devils have now gone 13 years without a Premier League title and have only been runners-up twice in that time, but their form in the second half of last season was tremendous.

Following Carrick's return to the Old Trafford dugout on January 12, they won 39 points from their final 17 games (W12 D3 L2), the most of any team in the division, three more than Arsenal.

There is still work to be done after the signings of midfield duo Youri Tielemans and Andrey Santos, but United's supporters have far more reason to be excited in 2026-27 than they did in 2025-26 – when the supercomputer assigned them a 0.6% title chance.

Things start to get interesting from here on, as 13 teams are assigned at least a 1% chance of winning the title, showing how difficult it is to predict the 2026-27 table solely on data. Villa are fifth favourites at 4.8%, followed closely by Chelsea at 4.1%, while Tottenham's chances are rated at 2.93%. 

Villa tasted silverware in Europe last term, but with key players departing in the transfer window, perhaps Chelsea and Spurs are the most interesting outsiders – both have highly-rated coaches at the helm and no European football to contend with.

Chelsea and Spurs the ones to watch in European race?

In 2025-26, for the second season in a row, the performances of English clubs in Europe earned the Premier League a fifth Champions League qualification spot.

Even if we assume a top-five finish will be enough again, only Arsenal (83%) and City (69.4%) are assigned a greater than 50% chance of making the cut, showing just how competitive the European race should be.

Liverpool (48.6%), Man Utd (40%), Villa (33.8%), Chelsea (30.3%), Newcastle (26.2%), Spurs (24.1%) and Brighton (21.5%) are the other teams with at least a 20% top-five probability. 

Villa have finished fourth in two of the last three seasons and, even after the sales of Tielemans and Morgan Rogers, they will have similar ambitions again after completing exciting deals for Johan Manzambi and Joao Gomes.

Newcastle's place in the supercomputer's predictions may surprise some, given they have lost the most successful head coach in their recent history, as well as three key players (Anthony Gordon, Sandro Tonali and Guimaraes).

Jaissle comes in with a big reputation, though, and in the last four seasons, Newcastle have made the top five on the two occasions when they did not have European football, missing out twice when they had to juggle those demands.

Like the Magpies, Chelsea and Spurs will be free to plan week-to-week without worrying about continental commitments. 

Since the start of 2022-23, new Blues boss Alonso has averaged 2.2 points per game in the Bundesliga and LaLiga, while in charge of Bayer Leverkusen and Madrid. 

The Spaniard has been heavily backed to the tune of almost £350m already this summer, and he could field Rogers, now the most expensive British player ever, alongside Cole Palmer and Joao Pedro at the tip of his 3-4-3 system.

Spurs have also gone all out in the transfer market, with £185m spent on a new midfield pairing of Tonali and Mateus Fernandes.  

Free from the relegation fears that plagued them in 2025-26, Spurs should be a fun watch. During De Zerbi's time in charge of Brighton, the Seagulls scored 116 goals and conceded 110 in 70 Premier League matches, at an average of 3.23 per game. 

Brighton, Bournemouth (18.9%) and Brentford (14.6%) all overachieved last year and will have European aspirations again in 2026-27, while Nottingham Forest (13.7%) could be outsiders under former Crystal Palace boss Oliver Glasner – his old Eagles team have a 12% top-five likelihood under Pierre Sage.

Promoted trio may all struggle

Last season bucked the recent trend of promoted teams struggling, with both Sunderland and Leeds United extending their top-flight stays into a second campaign. 

In both 2023-24 and 2024-25, all three promoted clubs went straight back down, and if the supercomputer's predictions are accurate, that could be the case again in 2026-27.

Hull City's promotion came as an enormous surprise, even more so if you glance at their underlying numbers – their 1.7 expected goals against (xGA) figure last season was the second-worst in the Championship, only better than bottom club Sheffield Wednesday. 

Sergej Jakirovic's Tigers are the supercomputer's favourites for the drop at 38.7%. That is not an enormous figure by any means, but it is a touch higher than fellow promoted sides Ipswich (31.2%) and Coventry (27.9%). 

A fast start will be vital for the new boys. In 34 Premier League seasons, none of the 16 promoted teams who won at least eight points from their opening five games have gone straight back down.

But, according to the Opta Power Rankings, Coventry have the second-toughest start in the division across the opening five matchdays, while Hull's schedule is the sixth-toughest. Ipswich are in the middle of that pack, and their opening game against Sunderland could be a good test of their credentials. 

Despite finishing seventh last term, the Black Cats are candidates to suffer a case of second-season syndrome in 2026-27, with the supercomputer assigning them a 25.9% probability of the drop. 

Sunderland's transfer activity has been limited after an exceptionally busy summer ahead of 2025-26, but the supercomputer's pessimistic forecast is most likely related to their poor underlying numbers last term. Only Burnley (30.94) and Wolves (31.58) managed a lower open-play expected goals (xG) tally than Sunderland (34.95) last season.

Fulham may also be looking over their shoulders, with a 23.4% chance of relegation. Under Marco Silva, the Cottagers never finished lower than 13th in four seasons following their return to the top flight in 2021-22, but replacing the now-Benfica boss with Alvaro Arbeloa is a gamble.

Arbeloa has drawn on his Real Madrid connections to swing deals for Cesar Palacios and Gonzalo Garcia, with the latter a particularly exciting acquisition. However, with just nine LaLiga starts under his belt, Garcia is raw and may need to build some early confidence if he is to adequately replace Raul Jimenez.

Leeds (23.4%), Palace (20%), Forest (18.1%), Brentford (16.3%), Everton (13.6%) and Bournemouth (13.1%) are also deemed at risk of getting dragged into a relegation fight. 

After West Ham dropped to the second tier last season, having scrapped with Tottenham until the final day, nobody in the bottom half can rest easily.