The Emirates test
Arsenal sit second on 18 points after seven league games; Everton are eighth on 10. That eight-point gap frames the 24 October visit to the Emirates in round eight, but the sharper tension is stylistic: Arsenal have won all three home league matches, while Everton have drawn all three away. One record has to bend.
Arsenal's current table line is 6-0-1, 11 goals for and five against; at home it is 3-0-0 with seven scored and two conceded. Everton are 2-4-1 overall, eight for and seven against, and 0-3-0 away with three scored and three conceded. The away column is the counterweight: Everton have not lost on the road, but they have not won there either.
What the stored run says
The stored recent runs add texture. Arsenal's newest five are a 2-2 draw at Bayern München on 21 October, a 1-0 win at Nottingham Forest, a 3-0 home win over Lille, a 2-1 home win over Leeds and a 0-3 defeat at Brighton on 19 September. That is three wins, a draw and a loss: strong, not spotless.
Everton's newest five are a 0-2 home loss to Chelsea on 17 October, a 2-2 draw at Hull City, a 1-0 home win over Ipswich, a 0-0 draw at Tottenham and a 2-2 home draw with Manchester United on 6 September. Three draws in five, and the Chelsea result is the outlier loss. The pattern is competitive, low-margin football rather than a collapse.
Across the stored eight-match samples, Arsenal's points-per-game is 2.38 and Everton's is 1.25. Those samples can include older seasons, so they are not a clean current-season form table; they are a broader efficiency check. The current table still points the same way: Arsenal have 18 points from seven, Everton 10.
The model's read
Nutmegly's AI read leans home. The model gives Arsenal a 65.8% win probability, the draw 19.6% and Everton 14.6%; it also puts over 2.5 goals at 58.2%. Those are estimates, not guarantees, and the read was generated on 24 September, before Arsenal's 21 October draw with Bayern and Everton's 17 October loss to Chelsea.
The case for the home lean is straightforward. Arsenal's home table line is perfect, their overall points total is eight clear of Everton's, and the stored eight-match points-per-game gap is wide. Everton's away record is resilient but draw-bound: three away games, three draws, three goals scored and three conceded. A draw is a real outcome, but the model makes it a 19.6% outcome, not the central one.
The counterevidence is equally clear. Everton are unbeaten away in the current table, and Arsenal's stored run includes a 0-3 defeat at Brighton and a 2-2 draw with Bayern. The model's date matters: it has not seen Arsenal's latest match or Everton's Chelsea defeat. No confirmed lineups or injury reports are in the supplied evidence, and there is no odds summary or external forecast to calibrate the number. The model's own counterpoint is that the starting XI is not fully settled, so in-game selections could change the result picture.
The watchpoint
That leaves the match as a test of two records rather than a coronation. If Arsenal's home scoring carries, the 65.8% home lean has support. If Everton keep the game level and their away draw habit holds, the 19.6% draw becomes live. The over 2.5 number is only 58.2%, so the model is not calling a shootout; the recent scorelines are mixed enough to leave the goal line uncertain.
The watchpoint is specific: can Arsenal score before Everton's away resilience turns the game into another draw? If the game is level deep, Everton's three away draws are the evidence to trust; if Arsenal's home output arrives, the home lean is justified. The question to watch: can Everton's away draw habit survive Arsenal's home scoring, or does the home lean finally break it?