@JJWiafe I mean Harris wasn’t a particularly strong candidate but her high level platform was reasonably left for this country. Turnout was fine in urban areas.
This can be boiled down to: cost of living, misogyny, and many voters vote on vibes rather than any understanding of policy
Democrats losing Hispanic voters as bad as they have the last 10 years should be fucking studied by future historians. Hopefully Hari Seldon will see to that.
@GraceOnFootball Xhaka isn’t as good but a similar profile. Think guys like Kroos are so unique. There was Xabi too but a good bit before. Think guys with this skillset will find a way. A one player ball progression machine will always have value.
@Jon_Mackenzie@NathanAClark I don’t know what to say other than I don’t think any statistician would approve of just removing parts of the sample for no good reason. Are you removing the best match of Pool, City, Chelsea, etc in these adjusted calculations you’re doing?
@Jon_Mackenzie@NathanAClark Fair enough. I’m perhaps incorrectly interpreting your post but I took your post to mean that romp should be discarded or regressed in some way, which I disagree with. United aren’t good but beating them by 4-5 xG actually tells us quite a bit
@Jon_Mackenzie@NathanAClark Pretty much all studies on xG predictiveness I’ve seen show “game state” is incredibly overrated and trying to account for it can decrease predictability because you lose a lot of your sample. Also, good teams blow teams out, bad teams get blown out
@NathanAClark xGD is definitely better than xPoints in terms of predictiveness, game state mostly doesn’t matter. Also worth noting imo it stabilizes quickly. Get an r2 of >0.4 at like 5 matches. About .5 at 10 and maxes out below .6. Anyways, we’re good. Some receipts: https://t.co/jdOba5Lhk7