blaaklint

interestingly, their xga is worse than leeds - but burnley have an excellent keeper (and have gotten quite lucky), whereas leeds have meslier (who is decidedly meh)

A lot of those players won't have made enough appearances - I assume the stat requires a certain percentage of minutes played.

Ruud van Nistelrooy

The finishing is just natural variation, not something that's really controllable. There are very few players in the world who consistently score more than their xG.

The pass to Vardy deflected off a Brighton player I think.

Fotmob have us with an xG of 1.67. As far as I can tell, a big chance is > 0.1xG, so really not very big.

I was amazed Dalot didn't get a yellow card, brought Abdul down four or five times after he'd gone past.

Yeah I was convinced play was being called back to give them a penalty - no idea how it wasn't given.

Vestergaard was our most succesful player last season (with other 180 mins) for aerial wins: 72.3% vs Coady's 66.7% and Faes 55.7%. Very small sample numbers but this year Okoli has only won 37.5%. I think Vestergaard just looks worse when he gets beaten because he's so tall.

Nearly £13k per minute

I think the comment about Mads is a concern for me. It's clear he loves being involved in possession - if he doesn't get to do so, that might be an incentive for him to seriously consider offers.

For challenges against players dribbling, he's at 62.8% success (45th in the league). Decent, but behind Doyle (71.1%), Faes (69.0%) and JJ (64.0%). Sidenote: didn't expect Doyle's success rate to be so high, and thought JJ would be higher since both Enzo and Rodgers thought he was the best 1-on-1 defender at the club.

You're right, I put this the wrong way around. In situations like this, they actually calculate the probability that you didn't score. In the example I put, that would be (1 - 0.8) * (1 - 0.5) = 0.1 probability of not scoring. xG would then be given as 0.9 Your three tap-ins example would be 1 - 0.6^3 = 0.78 xG

No, most xG models take this into account. If you have a 0.8 chance then a 0.5 rebound, the total xG given (at least by Opta) would be 0.8 * 0 5 = 0.4

It's not far off par, in the Championship about 40-45% of big chances are scored on average. So maybe an assist or two more would be fairer, but it's within the right ballpark.