The case against Ryan Allsop

[Eric the Half a Bee]

Our resident messageboard comic, DML, recently commented that Allsop's goals conceded record was good (last season). I thought it was time to check that out.

Last season Allsop conceded 53 goals. Of all goalkeepers who played virtually a full season, that's the 6th worst record when you look at goals conceded per game or per 90 minutes.

Some claim that Allsop didn't make many saves, but he faced 142 shots on target and saved 64% of them. That's the 7th worst of all goalkeepers in the Championship last season and the worst of all who played virtually every game. His clean sheet percentage was 4th worst of all regular keepers in the Championship last season.

I'm not the greatest fan of xG stats but there is one interesting one regarding Allsop - post-shot expected goals minus goals allowed. In other words, how likely a goalkeeper is to save a shot with goals scored subtracted. A positive score means a goalkeeper has saved more shots that might have been expected to have gone in, a negative score means the opposite - they've let in more goals that should have been saved. Allsop, for the season, stands at -7.4, the third worst in the Championship. In other words, he's cost us 7.4 goals that normally would have been saved overall.

Some other stats regarding Allsop. There were 8 occasions where he made 4 or more saves in a game - we won 2 points. He made 7 saves at QPR but we lost 3-0. He made 5 at West Brom in a 0-0. He made 5 on two occasions against Millwall yet we lost both without scoring. In total, we won 3 and drew 4 of 23 games where he made 2 or more saves. I guess such is not surprising for a side finishing 21st in games where teams had plenty of shots on target against us. In all bar one of those 23 games he had a saving percentage of at least 50%.

There were 20 games where he made 1 or no saves at all. We won 9, drew 6 and lost 5. There were 12 games where he saved less than 2/3rds of all shots on target, drawing 3, losing 9. Of course, a lower save percentage could easily mean a team was just clinical on the day, like the 3-2 defeat at home to the Jacks, where they had 4 attempts on target and scored 3 - their second was saved by Allsop but the rebound found the net.

All in all, though, they're a poor set of statistics for a goalkeeper at this level.

[2b2bdoo]

The sky guy on the red button loved stats and would always mention how Allsops stats were the worst in the division. I don’t think he made a huge amount of mistakes but simply doesn’t save enough.

The in the know twitter people were going on for ages saying we weren’t signing a keeper this summer. I kept saying it’s obvious we would sign a keeper and if we didnt there would be huge question marks over bullut because it’s so obvious, they kept saying we aren’t looking for a keeper…. We signed a keeper.

[Rjk]

to add to this, we can introduce the concept of post-shot Xg.
yes I can almost hear eyes rolling at that already, but bear with me.

so regular Xg is usually from where a shot is taken, how often does it result in a goal from there (+ some other contributory factors)

post shot xg just looks at the trajectory of the ball after it has been hit, so for a shot on target, how fast it is going, how much swerve , and where in the goal of is going to hit, how many shots like that typically get saved and how many are typically goals.

over the last 12 months, Ryan allsops post shot xg minus the number of goals he has conceded is significantly negative i.e. he has let in more than the average keeper in the population they are looking at: -0.3 goals per game on average.

that puts him in the 6th percentile of goalkeepers in similar levels for this metric. (i.e. 94 percent of the other keepers were better than him)

his save % was 62.1% of shots saved, which is the 4th percentile
and he stops 3.2% of crosses which is the 8th percentile

[Des Parrot]

In my opinion his reactions are slow and he’s poor with shoots close. Not agile enough for me.

[Carl Dale's Mole]

The problem is, in order to be a good shot stopper, you need to be good at stopping shots.

[the other bob wilson]

In my opinion his reactions are slow and he’s poor with shoots close. Not agile enough for me.

I think that’s it - I still say Allsop didn’t make any howlers of the type Runnarson made against Man City, but there were no end of goals where I asked “could the keeper have done better there?”. I tend to be less quick to blame keepers than many, because it seems to me that they are expected to maintain far higher standards than outfield players, but there were too many of those could the keeper have done better moments with Allsop. I wasn’t expecting us to sign a keeper this so called summer, but, as I mentioned in another thread, I’m glad we have.

[olderblue]

Allsop is a league one player at best, like most of the dross signed by Morison.
He just isn’t good enough for the Championship

[WJ99mobile]

Allsop is a league one player at best, like most of the dross signed by Morison.
He just isn’t good enough for the Championship

So why was he so good for Derby

I don’t get it, he’s looked useless since being here but clearly had a bit of pedigree

[Eric the Half a Bee]

So why was he so good for Derby

I don’t get it, he’s looked useless since being here but clearly had a bit of pedigree

He wasn't so good for Derby. His stats were slightly better than they were last season but that was in a side that would have comfortably avoided relegation without all the points deductions.

Here's one for everyone - Allsop's save percentage and "expected saves" stats last season were better than Alex Smithies and Dillon Phillips the season before.

[Eric the Half a Bee]

to add to this, we can introduce the concept of post-shot Xg.
yes I can almost hear eyes rolling at that already, but bear with me.

so regular Xg is usually from where a shot is taken, how often does it result in a goal from there (+ some other contributory factors)

post shot xg just looks at the trajectory of the ball after it has been hit, so for a shot on target, how fast it is going, how much swerve , and where in the goal of is going to hit, how many shots like that typically get saved and how many are typically goals.

over the last 12 months, Ryan allsops post shot xg minus the number of goals he has conceded is significantly negative i.e. he has let in more than the average keeper in the population they are looking at: -0.3 goals per game on average.

that puts him in the 6th percentile of goalkeepers in similar levels for this metric. (i.e. 94 percent of the other keepers were better than him)

his save % was 62.1% of shots saved, which is the 4th percentile
and he stops 3.2% of crosses which is the 8th percentile

Alnwick got a rating of 1 for post shot xG against QPR. He saved 1, conceded 2 but was expected to have only conceded 1. The first goal he had no chance with but is xG claiming that he should have saved the second?

[the other bob wilson]

Alnwick got a rating of 1 for post shot xG against QPR. He saved 1, conceded 2 but was expected to have only conceded 1. The first goal he had no chance with but is xG claiming that he should have saved the second?

What did I say about people expecting higher standards from keepers than they do from outfield players? Alnwick’s let in six in three games, but I wouldn’t blame him for any of them.

[Rjk]

Alnwick got a rating of 1 for post shot xG against QPR. He saved 1, conceded 2 but was expected to have only conceded 1. The first goal he had no chance with but is xG claiming that he should have saved the second?

yes, the post shot xg is suggesting he should have done better for the second, which seemed harsh, I'll have to go and look at it again, maybe it was more centrally in the goal than it first seemed.

as with regular Xg, you aren't really supposed to place too much weight on individual shot xg , even individual game because small inconsistencies in the model can throw up spurious results, but over a slightly longer time frame (I've seen 9 games mentioned) these even out to give you a more accurate picture

completely ignoring that and going back to the QPR game, Begovic faced 2 post shot xg and conceded 1, and Alnwick faced 1 and let 2 in

[dml1954]

Our resident messageboard comic, DML, recently commented that Allsop's goals conceded record was good (last season). I thought it was time to check that out.

Last season Allsop conceded 53 goals. Of all goalkeepers who played virtually a full season, that's the 6th worst record when you look at goals conceded per game or per 90 minutes.

Some claim that Allsop didn't make many saves, but he faced 142 shots on target and saved 64% of them. That's the 7th worst of all goalkeepers in the Championship last season and the worst of all who played virtually every game. His clean sheet percentage was 4th worst of all regular keepers in the Championship last season.

I'm not the greatest fan of xG stats but there is one interesting one regarding Allsop - post-shot expected goals minus goals allowed. In other words, how likely a goalkeeper is to save a shot with goals scored subtracted. A positive score means a goalkeeper has saved more shots that might have been expected to have gone in, a negative score means the opposite - they've let in more goals that should have been saved. Allsop, for the season, stands at -7.4, the third worst in the Championship. In other words, he's cost us 7.4 goals that normally would have been saved overall.

Some other stats regarding Allsop. There were 8 occasions where he made 4 or more saves in a game - we won 2 points. He made 7 saves at QPR but we lost 3-0. He made 5 at West Brom in a 0-0. He made 5 on two occasions against Millwall yet we lost both without scoring. In total, we won 3 and drew 4 of 23 games where he made 2 or more saves. I guess such is not surprising for a side finishing 21st in games where teams had plenty of shots on target against us. In all bar one of those 23 games he had a saving percentage of at least 50%.

There were 20 games where he made 1 or no saves at all. We won 9, drew 6 and lost 5. There were 12 games where he saved less than 2/3rds of all shots on target, drawing 3, losing 9. Of course, a lower save percentage could easily mean a team was just clinical on the day, like the 3-2 defeat at home to the Jacks, where they had 4 attempts on target and scored 3 - their second was saved by Allsop but the rebound found the net.

All in all, though, they're a poor set of statistics for a goalkeeper at this level.

Utter rubbish. Statistics can be manipulated to prove anything you like, to fit in with your campaign to get rid of Alsop and all this xG stuff is a load of guff. The fact that I look at is that we conceded 58 goals last season, which for a struggling team that finished in 21st place is excellent. Plus there were ten other teams that finished with the same or worse goals against record, including Swansea who finished in 10th place. Even Middlesborough, who finished 4th, shipped 56 goals. Alsop made a number of key saves, including a couple of penalties. The reason we nearly got relegated was therefore nothing to do with the defence or goalkeeper, it was because we only scored 41 goals, the second worst in the league. There is a strong argument therefore that if it wasn't for Alsop and the defence, we probably would have got relegated.

Anyway, now we have another keeper, who isn't a ‘shoe in’ for the team and we shall see what happens with Alsop. You can now shift your attention elsewhere in the squad and choose your next player to continually criticise and want rid of - enjoy.

[Pedro de la Rosa]

Alnwick got a rating of 1 for post shot xG against QPR. He saved 1, conceded 2 but was expected to have only conceded 1. The first goal he had no chance with but is xG claiming that he should have saved the second?

XG is a lot of tosh, then.

I remember Kieffer Moore scored a tap in from a yard out and the xG was 0.67 .

Alnwick also made a great save from Dykes’ header. I’m surprised we are signing a keeper after Alnwick’s start with our need for a centre half so great.

[dml1954]

to add to this, we can introduce the concept of post-shot Xg.
yes I can almost hear eyes rolling at that already, but bear with me.

so regular Xg is usually from where a shot is taken, how often does it result in a goal from there (+ some other contributory factors)

post shot xg just looks at the trajectory of the ball after it has been hit, so for a shot on target, how fast it is going, how much swerve , and where in the goal of is going to hit, how many shots like that typically get saved and how many are typically goals.

over the last 12 months, Ryan allsops post shot xg minus the number of goals he has conceded is significantly negative i.e. he has let in more than the average keeper in the population they are looking at: -0.3 goals per game on average.

that puts him in the 6th percentile of goalkeepers in similar levels for this metric. (i.e. 94 percent of the other keepers were better than him)

his save % was 62.1% of shots saved, which is the 4th percentile
and he stops 3.2% of crosses which is the 8th percentile

Honestly, this is just gobble de gook and means absolutely nothing. A golden example of how mathematicians and scientists shouldn't get involved in a sport where there are so many variables.

[Rjk]

Honestly, this is just gobble de gook and means absolutely nothing. A golden example of how mathematicians and scientists shouldn't get involved in a sport where there are so many variables.

it's fine if you do t understand it, or find it interesting, but every club is looking at this kind of thing, some spending millions on it

[Rjk]

XG is a lot of tosh, then.

I remember Kieffer Moore scored a tap in from a yard out and the xG was 0.67 .

Alnwick also made a great save from Dykes’ header. I’m surprised we are signing a keeper after Alnwick’s start with our need for a centre half so great.

Dykes doesn't seem to have registered a headed shot on target in the game according to the opta stats

[Rjk]

Dykes doesn't seem to have registered a headed shot on target in the game according to the opta stats

guessing you mean the Steve cook header.
that did have a pretty low xg, which presumably means that headers are very rarely scored from that position, if he'd let that in then there would have been big questions as it was basically straight at him.
what he did do well is get good distance on the save to stop a follow up shot coming in (I think it went for a corner) which probably isn't reflected in the stats

[The Lone Gunman]

Anyway, now we have another keeper, who isn't a ‘shoe in’ for the team and we shall see what happens with Alsop.

His name is Allsop, but you’re right - we’ll see what happens to him in due course.

Out of interest, why do you believe City have decided to bring in an another keeper? The club went with Allsop and Alnwick last season. Why do you reckon they’ve gone out and got a third first team keeper this time around?

I believe Bulut and his staff have assessed Allsop and Alnwick and quickly concluded that one (or maybe both) of them is not good enough.

[PontBlue]

Honestly, this is just gobble de gook and means absolutely nothing. A golden example of how mathematicians and scientists shouldn't get involved in a sport where there are so many variables.

Brentford supporters would probably disagree with you. I'd like to be in their position rather than hoping we won't be in another relegation scrap to avoid dropping into the third tier.

[John Buchanan]

Honestly, this is just gobble de gook and means absolutely nothing. A golden example of how mathematicians and scientists shouldn't get involved in a sport where there are so many variables.

I 1+1 to the power of nd that comment.

[Eric the Half a Bee]

Utter rubbish. Statistics can be manipulated to prove anything you like, to fit in with your campaign to get rid of Alsop and all this xG stuff is a load of guff. The fact that I look at is that we conceded 58 goals last season, which for a struggling team that finished in 21st place is excellent. Plus there were ten other teams that finished with the same or worse goals against record, including Swansea who finished in 10th place. Even Middlesborough, who finished 4th, shipped 56 goals. Alsop made a number of key saves, including a couple of penalties. The reason we nearly got relegated was therefore nothing to do with the defence or goalkeeper, it was because we only scored 41 goals, the second worst in the league. There is a strong argument therefore that if it wasn't for Alsop and the defence, we probably would have got relegated.

Anyway, now we have another keeper, who isn't a ‘shoe in’ for the team and we shall see what happens with Alsop. You can now shift your attention elsewhere in the squad and choose your next player to continually criticise and want rid of - enjoy.

You know that thing when you're the only one with a particular opinion, but everyone else is suggesting the opposite? You should be used to it on here by now.

[Eric the Half a Bee]

Another interesting stat about our defence last season. Only Millwall, West Brom, Stoke, Burnley, Sheff Utd and Luton had less on target attempts on their goal last season. That suggests we were decent defensively yet we conceded more than any of them.

[City123]

He's shit. Case closed.

[Dorcus]

Jak Alnwick, case closed.