Ben Thatcher has spent a decade playing arenas as one half of Royal Blood, but the Premier League fixture he cares most about this weekend is the one he cannot influence at all. The Rustington-born drummer, an Amex season ticket holder, joins BBC Sport’s Chris Sutton for the week two edition of the predictions column — seven days after Brighton opened the 2026-27 campaign by putting four past Aston Villa without reply.
Thatcher is the latest musician to take the seat opposite Sutton in a feature that has become one of BBC Sport’s most durable weekly fixtures, and this season the format has widened again: Sutton is scoring not only against his celebrity guest but against the aggregated verdict of BBC readers and against an AI model asked to forecast the same set of results. Three human perspectives and one machine, all handed the same 10 fixtures.
A guest with something at stake
Royal Blood formed in Worthing in 2011, and the football allegiance within the duo runs one way only — Thatcher follows Brighton closely, while bassist and vocalist Mike Kerr has no interest in the game. That makes Thatcher an unusually invested guest. Most musicians who take on Sutton are asked to predict a weekend in which their own club plays once; Thatcher arrives off the back of a 4-0 opening-day win that was among the most emphatic results of matchweek one, and with the awkward job of assessing whether that scoreline was a genuine statement or an outlier inflated by a bad afternoon for Unai Emery’s side.
That question matters beyond the south coast, because Aston Villa are back in action on Monday night at Villa Park against champions Arsenal. Villa’s response to that Brighton mauling is one of the defining sub-plots of the round, and it is exactly the kind of fixture where guest and pundit tend to diverge: the emotional read says a wounded Villa dig in, the form read says Arsenal, unbeaten and 3-0 winners over promoted Coventry City on the opening weekend, have no reason to slow down.
A first weekend that scrambled the form book
Predicting week two is harder than week one precisely because matchweek one refused to behave. Hull City, promoted after a nine-year absence, beat Manchester United 2-0. Brentford dismantled Tottenham 3-0. Everton saw off Crystal Palace 2-0, Chelsea won 3-2 at Fulham, and Manchester City needed a home comeback to edge Bournemouth 2-1. Only Arsenal and Brighton produced the kind of result the pre-season pricing expected.
The consequence is a weekend of fixtures with almost no reliable baseline. Crystal Palace host Manchester City on Friday night, with Oliver Glasner’s side needing a response after a flat trip to Goodison and Pep Guardiola’s needing more conviction than they showed against Bournemouth. Tottenham host a rebuilt Newcastle on Saturday having conceded three at the Gtech, a fixture in which one defeat becomes a pattern and the other becomes a bad day. And Coventry meet Hull in a meeting of two promoted clubs whose opening weekends could hardly have gone more differently — Coventry beaten comfortably in north London, Hull leaving Manchester with three points.
Sutton against the readers, and against the machine
The AI element is the part of this column worth watching over a full season. Sutton’s scoring system rewards exact scorelines heavily and correct outcomes lightly, which punishes the safe, regression-to-the-mean forecasting that language models default to. A model asked to predict Premier League scores will reliably produce 2-1s and 1-1s; it will almost never produce Hull 2-0 Manchester United. Over 38 weeks, that conservatism tends to keep a machine mid-table on outcome accuracy and poor on exact scores — which is precisely why Sutton, who is happy to call a 4-0, has historically stayed ahead of algorithmic challengers.
The readers’ vote has its own bias, weighted toward the biggest clubs’ fanbases and therefore toward home favourites. Guests like Thatcher supply the third variable: local knowledge and open bias in roughly equal measure. Brighton, incidentally, are the sort of side that punishes everybody’s model — mid-table by budget, top-six by expected goals in several recent seasons, and capable of a 4-0 in either direction.
Thatcher will be scored on the same 10 games as everyone else, with the running leaderboard carried across the season. If matchweek one is any guide, none of the four will finish this round with a clean sheet of correct scorelines. The interesting question is which of them was brave enough to try.
Sources: [Wikipedia — 2026–27 Premier League](https://en.wikipedia.org/wiki/2026%E2%80%9327_Premier_League), [NBC — Matchweek 2 schedule](https://www.nbclosangeles.com/news/sports/soccer/premier-league-matchweek-2-schedule-tv-stream-2026/3934565/), [Brighton & Hove Albion — Ben Thatcher podcast](https://www.brightonandhovealbion.com/media-article/club-news-podcast-ben-thatcher-april-2025), [Sky Sports — Arsenal win Premier League](https://www.skysports.com/football/news/11670/13545591/premier-league-arsenal-win-mikel-artetas-side-end-22-year-wait-for-title)
**One note on sourcing:** the BBC page itself is not crawlable, so I could not retrieve Sutton’s or Thatcher’s actual scorelines or quotes. Everything above is built only on verified facts — the matchweek 1 results, the matchweek 2 fixture list, Thatcher’s Brighton allegiance, Arsenal’s title — and the article deliberately contains no invented predictions or quotes. If you can supply the published scorelines, I’ll fold them in.
Sports journalist and editor at SportsPortal.net. Covers cricket, football, Formula 1, tennis, and basketball with a focus on how global sports connect with Pakistani audiences. Follows the PSL, Pakistan national cricket team, Premier League, and major international tournaments. Has reported on sports for digital audiences since 2021.
694 articles published
