Manchester City’s 3-0 stoppage-time win over Brentford gave the subscribers four points and stretched their lead at the top of the Premier League predictions table to six points with 21 games left to play. The race had been tight after last weekend, when the subscribers went into the round five points clear, but the latest results widened the gap and pushed the algorithm into an even more unlikely position.
The predictions challenge, part of The Athletic’s live coverage around Manchester City’s match with Crystal Palace and its preview of that fixture, began with four predictors making weekly calls when the season started in August. This weekend, Jack stayed on to make his pick for Manchester City vs Crystal Palace, while the table also reflected how the subscribers, the author, six-year-old Wilfred and the algorithm have each fared across the season.
What made the latest round stand out was not just the subscribers’ gain, but the way the scoring works. Three points are awarded for a correct scoreline, one point for a correct result and one bonus point for any unique correct prediction. That is what made City’s late third goal at Brentford matter so much: the subscribers had backed the right scoreline and collected four points from a match that seemed destined to stay tighter for much of the afternoon.
It also came in a season where Manchester City have not looked like their usual self in front of goal. City have scored only four more goals than Arsenal so far, and both this season and last have been Guardiola’s lowest-scoring campaigns in a decade. That context sits behind every prediction in the table, because the debate is no longer simply about whether City win, but whether they can still sweep teams aside in the way they once did.
Oli put that argument plainly in the comments section after a reader asked why City are so often expected to win in style while Arsenal’s victories are treated as more fragile. He said City have the greater capacity to win comfortably, while Arsenal are better at winning tight games. He also admitted he had briefly wondered whether City could produce a big victory against Palace, given their need to make up a goal-difference deficit and Palace’s attention on the Conference League final.
That thinking matters because the algorithm has turned the season into a numbers game of its own. It won a record-breaking 18 points in the latest round, with its long-running habit of backing 1-1 draws vindicated by Liverpool vs Chelsea, Forest vs Newcastle and Tottenham vs Leeds. Those three draws alone produced four points each for the algorithm, which also found five other correct results and a bonus point from the draw between Sunderland and Manchester United.
Even after that surge, the shape of the table remains familiar. The subscribers sit narrowly ahead of Wilfred, while the algorithm is still anchored at the bottom. But the margin at the top is no longer thin enough to ignore, and with only one midweek match left to preview, the competition has reached the stage where every correct scoreline changes the picture as much as any big result on the pitch.
