Thank you for the added level of nuance to aid my understanding of xG.
No worries, like I say it's not a perfect system but it does give some indication but there will always be some outliers. Last season with all the data they had a top 7 of Ipswich, Sheffield Wednesday, Peterborough, Derby, Bolton, Plymouth and Barnsley who made that up, although not in that order. They had Ipswich as the best team which certainly backs up what most thought. At the other end they would have had Accy 18th, MK 21st, FGR 23rd and Morecambe 24th so not a million miles off there either.
Just so we're clear:
Overperforming vs xG, just a bit = good.
Overperforming vs xG a little bit more = not so good (unless you're Barnsley or Bolton, then it's quite good again)?
Obviously any overperformance is
good, we'd all be happy if we scored every shot we had most people have said it's about being sustainable. It's a bit disingenuous to not see the difference between some of those numbers and what we were doing early in the season. I'm not sure why it would be any different for Bolton or Barnsley and I never suggested it was, although it's pretty unarguable that they have better forwards than us so you'd probably expect them to overperform more than us. For reference this season the biggest overperformers in each league are:
Arsenal 26.5%
Norwich 26%
Us 23.7%
Tranmere 28.6%
So something like 20-30% will happen to a handful of the 92 per year, likewise some will go the other way. Using the same site last seasons best overperformers were
Arsenal again at 23.7%
Burnley 32%
Charlton 19.5%
Salford 28.5%
Under Manning our overperformance was 61.8%, so in the last two seasons with 184 teams only Burnley have come within 50% of what we were doing, and then only just. I'm not going to bother looking at other leagues or more seasons as I think that's enough to go on but I'd be staggered if there were more than a handful of teams, as in less than 1%, who get anywhere particularly close to 40% overachievement over the course of a season. To keep scoring at the rate we were was completely unsustainable and I'm sure I remember you lamenting our lack of striking options and mentioning how many goals came from whoever was playing left back. I don't think anyone was genuinely thinking Leigh and Brown were going to get 25 odd goals between them this year.
Meanwhile I notice you deliberately steered clear of the expected Goals Against (it's right there in your very own data source - in the next column!), whereby everyone in the promotion frame has outperformed their xGA except us? So our 'net overperformance' isn't quite as drastic as you suggest is it? Talk about misleading!
To be honest I couldn't be bothered to look into xGA which gives a value to the chances you concede, at this point it's basically down to your keeper and whether your opponents are clinical as to whether you meet it or not. As you correctly say we're the only promotion hunting team who has underperformed but if you take it back to when Manning was here we were also in the overperforming section. It's gone backwards since as teams have taken a couple of extra chances against us and also partly down to Cumming replacing Beadle and that being a bit of a step down. We were excellent in this metric under Manning so it's not a criticism, just pointing out he had the rub of the green a bit.
We've scored 12 'too many' goals is it? How about looking at individual games? 8 of our 'overperforming goals' came in the five matches in which we scored three goals under Manning - none of which, technically, were needed to win the game. So, xG is actually penalising us for having the gall to get a 3rd at Fleetwood or Stevenage.
I'm not sure you can look at individual games and I'm not taking the time to do it but the criteria is the same for everyone. Barnsley walloped Port Vale 7-0 on the opening day, should we take 6 of their total as well as they didn't need them? With any stats the longer term the more accurate result you get, no one is being penalised for scoring more goals.
For xG to improve vs real human goals, we really needed to stick to just the two goals in these games - and miss a couple of sitters, didn't we?
I'm really not sure what you mean here. The point I've made several times is that we were never going to continue scoring at the rate we were. If we were then brilliant, we've got some world class finishers on our hands but that's not the case. It's not trying to level things up, it's recognising when over a shorter term something is a bit of an anomaly that is likely to correct itself and that is what has happened. The alternative argument is that Manning is magic, and by virtue of him being on the touchline players take chances they wouldn't normally take.
And then maybe, at home and 3 up against Shrewsbury, what we really needed to do was try and boost our xG by, presumably, creating a raft of chances but deliberately missing them - ya know, get the old xG back on an even keel! Nothing like smashing a wasteful 6.98 on xG - that's four pretend goals we could have used later in the season after all.
Again this doesn't make any sense. To go back to Fleetwood you're point might be that we don't need to attack much as we've got a healthy lead which is absolutely fair enough. But what if we hadn't already scored three absolute screamers, where most days none of them go in? Do we create more chances? Probably, but are we then likely to concede more? Again probably, it works both ways.
Anyway, I'm just glad xG is finally getting the recognition it deserves on this forum.
I'm already excited to go 1-0 up at Shrewsbury and belt out a chorus of "1.13 - 0.86 to the Oxford Boys"...
As I've said it's not the be all and end all but it is a handy guide. Results are king but over the long term they go hand in hand with performances, we went and battered Pompey away a few weeks back and lost and were then crap at home against Cheltenham but won. The result was obviously better against Cheltenham but if you had the opportunity to pick a level of performance and guarantee you'd match it every game you pick Pompey hands down and over the course of the season it will yield results more often than not. We were good under Manning but if we had picked up 4/5 fewer points no one would have been saying how unlucky we were.
