Gameweek 5 Roundup

Ten games, one genuine form line (Manchester City), and almost everything else went against what Predicts10 players actually called. Seven of the ten results went against the crowd's favourite, which makes this the second-toughest week of the season so far to call: only 35% of predictions got the right result, and just 9% nailed the exact score.

Brighton 3-0 Arsenal was the shock of the week. Nobody backed it - not one player among the 12 who predicted the fixture gave Brighton anything, three-quarters had Arsenal winning and the rest backed a draw. Brighton won by three.

Leeds 0-0 Crystal Palace did something similar in miniature. Three in four players had Leeds beating Crystal Palace at home. Instead, nothing happened at all - no goals, no result, and 75% of the gameweek's predictions were wrong from the first whistle.

Brentford 3-0 Chelsea, on the opening night, set the tone early. More than half the pool actually fancied Chelsea to win that one. Brentford didn't just win, they won comfortably.

Not everything went against the crowd. Manchester City's 5-3 win over Sunderland was the week's one point of agreement - 92% of players called it correctly, and City delivered. Everton's 1-0 win over Ipswich was similarly straightforward, called by three in four players.

At the top of the table, sjd took the gameweek with 29 points and the £25 prize, with heatherlauren second on 26 points for £10.

Last Man Standing had already gone quiet - the pool ended in a total wipeout the week before, so there was no round running alongside this one. Both games are back on 6 October, for that weekend's fixtures.

A week like this is a reminder of why the site keeps a tiebreak chain nine steps deep. When only about a third of the pool gets the headline result right, closeness across all ten games ends up mattering more than any single correct call.