I looked at this a while back, he had a good save/shot percentage that first year in the PL, although on xg he still saved slightly less than he 'should have'. Either way he was above the prem average. Every year since he has been at the bottom end. You're a stats man Aski... it's not a stretch to say his form relegated us. I'd never put that on one player (the DOF and Managers buy and pick them) , but take Struijk, Meslier and Bamford out of the prem team and replace them with just 'average', and we'd have been midtable under Bielsa that second year (wouldn't have needed to sack him). Alternatively, assuming we took them out that first marsh summer window, we'd have been midtable again... and probably still had Jesse. Choose your poison . FWIW Struijk is a confidence player and doing great now... whether he's mobile enough and confident enough to avoid being a rabbit in the headlights in the PL again? Dunno.
I’d agree with that regarding Meslier and arguably Pickford kept Everton in the PL two seasons running.
spot on. I'm reticent to quote Jamie Carragher, but the year we went down he said the teams with the worst 3 goalkeepers went down. I'm not sure whether he'd been checking the stats or just the eye test... but statistically that was pretty much correct. Pickford proves the point... saved more than he should statistically and his side stays up because of it. In absolute terms there were 16 goals between Pickford (saved 4 more than he should) and Meslier (12 less than he should). It's just stats, some of that is poor/great finishing, so obviously luck comes into it... but 16 goals unluckier?
Not sniping, I’m curious where do we get ‘saved less/more than he should’ from? I’m just thinking the stats of the worst teams will be relatively worse than the better teams, Can we make the stats single out one player? eg GK with useless defence in front of him will have useless stats? Or is there a way to use stats to judge just the GK?
We are reliant upon Xga (expected goals against). Thus with a **** defence where all the goals are scored from within say the 6 yard box, then Xga will show that the goal keeper had very little chance of not conceding, and most likely Xga will equal actual goals against or be even higher if the keeper is pulling off fantastic saves. However if the defence are limiting shots on goal, and teams are scoring 30 yard screamers against a keeper every week, then Xga would say that the keeper should be saving more than he concedes. Meslier's current Xga suggests that he should have conceded 7.8 goals this season, where as despite all the mistakes he is deemed to have made this season that has cost us goals, his stats say he is performing exactly as expected. Where as Swansea who have also only conceded 9 goals so far this season, have an XGA of 15.5, so stats say that Swansea should have conceded 6 more goals than they have, thus their keeper is performing better than expected. Luton who have conceded 20 goals this season so far, have a better Xga than Swansea at 14.5, so the Luton keeper has conceded 5 goals that on average you would expect to be saved
Is this a very statistically way of saying we tend to concede less when we don’t let teams have a shot at Mes
Specifically relating to Mes this season, then no, because all the stats are showing is that Mes has conceded the exact number of goals that statisticians expected him to have conceded. Given the goal at Sunderland, is definitely one that should have been saved, that also means Mes has saved one, that stats consider should have been almost a certain goal. According to the game stats we should have conceded against Sheff Wed. However your statement would be true for the previous 3 seasons. Leeds United Xga stats per season since Mes was first choice keeper
I keep my own stats 2020, compiling the data from about 10 different sites, because I always find that one site records a certain stat differently to the others. But https://fbref.com/en/comps/10/Championship-Stats and https://footballxg.com/xg-league-tables/ and https://www.fotmob.com/en-GB/leagues/48/table/championship?filter=xg are probably the ones I rely upon most when it comes to Xg and Xga. The first two sites haven't been updated yet to include this weekends games, where as the FotMob one has, and I just average out the stat between all 3 of those sites once they are all up to date Below is a copy of the FotMob Xg and Xga table, basically confirming that we should have scored 22 goals, conceded 8 and have gained 2 more points than we actually have
