awooga83 as Quincy stated City were 40% to win the title with arsenal 30% so the model still had city as the more likely team before these results.
You've only gotta scroll back a couple of dozen posts in this very thread to find me pointing out that the Opta model for outright winner is going to swing like a weathervane in response to every setback in a close title race.
The fact it's doing that week by week, match by match is a testament to the inutility of this kind of model under short range conditions and in response to unpredictable, and even non-analytic conditions and circumstances, and little else.
Leaning on the Opta percentages at this point means using a season-long analytic model, skewed by design to the club that's won five of the past six titles, to rule on questions such as whether Trossard buries that presentable chance.
It's an irony that the less salience such a model has, the more interest there is in its prognostications. Might as well ask Paul the octopus—RIP.
