The match has not been played. This is not a tip: the model turns the chances both sides create and concede into probabilities, and it is wrong regularly.
A slight edge to Arsenal: 44% against 32%. Attack against defence: level, 1.57 and 1.72 expected goals a match. Read the table with care: Manchester City have taken 6.5 points more than their play deserved, so their position flatters them.
| Club | Place | P | Points | xPTS | Δ | xG/match | xGA/match |
|---|---|---|---|---|---|---|---|
| Arsenal | 2 | 7 | 18 | 13.5 | +4.5 | 1.81 | 1.06 |
| Manchester City | 1 | 7 | 21 | 14.5 | +6.5 | 2.38 | 1.33 |
Probabilities come from a Poisson model over both sides' xG with a home advantage applied. We know nothing about injuries, suspensions or rotation, and we do not give betting tips.