Creativity & Chance Conversion

Arsenal news and interviews
Discuss anything Arsenal related, players, tactics etc.

Re: Creativity & Chance Conversion

Postby theHotHead » Tue Sep 14, 2021 6:04 am

swipe right wrote:HH - it’s a bunch of BS and we all know it. It’s Jays way of distracting from the fact that he’s spent all summer biging up Arteta and it’s come undone in four games. That’s why my posts on the first page was show me the chances. Eight pages later, nothing.

Swipey, I saw every single chance, Jay is not incorrect, there were at least 30 chances, but its the quality of the chances. If you were able to put a filter on the chances, using an Excel spreadsheet and filter out all totally shit chances - chances that you would be taking the poss to claim were chances - you would end up with about 11 chances.

If you take those 11 chances, 3 of them were the result of pinball in the goal mouth, so not chances spread over a period of time but shots that took place in immediate succession. Those pinball shots skew the numbers slightly. If I said there were 5 shots all game but you found out all 5 shots came in a 10 second spell of penalty box mayhem, it paints a different picture than 5 shots coming across 80 mins of footie, for example.

Yes there were 30 shots, no, most of them were not worthy of being called chances.
User avatar
theHotHead
Poster Of The Month
Poster Of The Month
 
Posts: 12215
Joined: Mon Oct 17, 2016 12:44 am
Location: Norf Landon

Re: Creativity & Chance Conversion

Postby swipe right » Tue Sep 14, 2021 6:20 am

theHotHead wrote:
swipe right wrote:HH - it’s a bunch of BS and we all know it. It’s Jays way of distracting from the fact that he’s spent all summer biging up Arteta and it’s come undone in four games. That’s why my posts on the first page was show me the chances. Eight pages later, nothing.

Swipey, I saw every single chance, Jay is not incorrect, there were at least 30 chances, but its the quality of the chances. If you were able to put a filter on the chances, using an Excel spreadsheet and filter out all totally shit chances - chances that you would be taking the poss to claim were chances - you would end up with about 11 chances.

If you take those 11 chances, 3 of them were the result of pinball in the goal mouth, so not chances spread over a period of time but shots that took place in immediate succession. Those pinball shots skew the numbers slightly. If I said there were 5 shots all game but you found out all 5 shots came in a 10 second spell of penalty box mayhem, it paints a different picture than 5 shots coming across 80 mins of footie, for example.

Yes there were 30 shots, no, most of them were not worthy of being called chances.

Over 90 mins even a pub team would create what Jay terms as chances. This is all academic BS. When you see the game you see that we struggle to get the ball forward or penetrate the box. Our midfield is coached to push the ball wide and we are so slow in buildup that the wide players can’t get in behind the opposition defense to put in a decent cross. So we end up passing back to the center. It’s pathetic. God knows what Arteta is doing in training.
swipe right
David Rocastle
David Rocastle
 
Posts: 4371
Joined: Fri Dec 28, 2018 5:05 am

Re: Creativity & Chance Conversion

Postby jayramfootball » Tue Sep 14, 2021 7:38 am

theHotHead wrote:
jayramfootball wrote:
theHotHead wrote:But Jay, if we take the Norwich game in isolation, its not our finishing that prevented us from scoring more goals (ESR chance aside), it was the quality of the chances we created that was the problem.


It was a combination.
ESR chance, Auba chance and Pepe chance (blocked on the line) - you'd expect 1 goal from those three chances ( in fact the xG combined for all 3 was about 1). We missed them all.
Then all the other chances - based on league averages a goal would result about 2-3% of the time so you'd expect 35-40 of those types of efforts to yield a goal.
Then the actual goal, which obviously had a very high xG.
That is why the xG came out at 2.7.

Now, if you refer back to a previous post I have already stated that we're only 4 games in, so the numbers can not be reflective of individual players yet and even for the whole team we only have a small data set.
For example, that ESR chance - we can forgive him. It's one chance and can easily be missed. If over the season, however, he has 10 of those types of chance and misses ALL of them, we'll be saying (rightly) that he has finished poorly this season.

That is why this data will be updated each week. The picture becomes more and more solid as more data is added, but as of now, 4 games in ... we are creating lots of chances (6th in the league), but not very good ones (19th in the league on average) and our finishing is the worst in the league.

No Jay, I agree you could reasonably expect 1 goal from ESR/Pepe/Auba chances most definitely but I disagree that you would expect a goal from the other chances, they were poor low quality chances, they never looked like being goals.

I agree with PnG, of we continue to create such poor quality chances in high numbers, we will struggle to score any goals!!


Those other lower percentage chances lead to goals game after game, year after year. Lower percentage, yes, but goals nonetheless.

Already pointed out 6 goals scored from low percentage chances by just 3 teams.
If I looked at every team I bet I would find that the difference between the goals scored from low quality and high quality chances was not that big.
Last edited by jayramfootball on Tue Sep 14, 2021 8:19 am, edited 1 time in total.
Image
jayramfootball
Arsène Wenger
Arsène Wenger
 
Posts: 13684
Joined: Mon Sep 03, 2012 8:58 pm

Re: Creativity & Chance Conversion

Postby jayramfootball » Tue Sep 14, 2021 7:42 am

swipe right wrote:
theHotHead wrote:
swipe right wrote:HH - it’s a bunch of BS and we all know it. It’s Jays way of distracting from the fact that he’s spent all summer biging up Arteta and it’s come undone in four games. That’s why my posts on the first page was show me the chances. Eight pages later, nothing.

Swipey, I saw every single chance, Jay is not incorrect, there were at least 30 chances, but its the quality of the chances. If you were able to put a filter on the chances, using an Excel spreadsheet and filter out all totally shit chances - chances that you would be taking the poss to claim were chances - you would end up with about 11 chances.

If you take those 11 chances, 3 of them were the result of pinball in the goal mouth, so not chances spread over a period of time but shots that took place in immediate succession. Those pinball shots skew the numbers slightly. If I said there were 5 shots all game but you found out all 5 shots came in a 10 second spell of penalty box mayhem, it paints a different picture than 5 shots coming across 80 mins of footie, for example.

Yes there were 30 shots, no, most of them were not worthy of being called chances.

Over 90 mins even a pub team would create what Jay terms as chances. This is all academic BS. When you see the game you see that we struggle to get the ball forward or penetrate the box. Our midfield is coached to push the ball wide and we are so slow in buildup that the wide players can’t get in behind the opposition defense to put in a decent cross. So we end up passing back to the center. It’s pathetic. God knows what Arteta is doing in training.


All premier League teams create those types of chances every game. Most chances created are low quality by every team.

The data just keeps holding up to challenge.

Yes we don't create enough quality chances but we're also the worst team in the league at finishing so far this season.

I will be updating each week to see if we can improve both areas.
Image
jayramfootball
Arsène Wenger
Arsène Wenger
 
Posts: 13684
Joined: Mon Sep 03, 2012 8:58 pm

Re: Creativity & Chance Conversion

Postby Power n Glory » Tue Sep 14, 2021 9:51 am

jayramfootball wrote:
Power n Glory wrote:
jayramfootball wrote:
theHotHead wrote:81 mins, 24 chances, 7 of which were not absolutely fookin shit chances, 3 of which were rebounds, one was the actual goal and there were 2 decent half chances from Saka. But again I stress, not a single big chance from any of that lot, not a single chance that I would expect a goal to be scored from.

At this point I can say we were crap vs Norwich, Ben White had a good game, anyone criticising Lokonga's performance is blind, Odegaard was meh, AMN needs to give up ambitions of being a CM, for e very good thing he does, he will do 2 absolutely dreadful things. Tomi looked quality, Ramsdale looked good, Pepe was in and out, Saka was in and out, Auba was Auba - not a great game but you felt he would do damage given a chance.


There were 30 shots on goal - you'll just have to rewatch it ( :biggrin: ) as I am going to assume that all the major websites like Sky and BBC who also state there were 30 shots got it right).

4 good / decent chances listed above with video and time stamps.
26 speculative chances

You use the term 'expecting to score'. What does that mean???
You mean would NEVER expect to score? Would you NEVER expect ESR to have put his chance away, or Auba to NEVER score with his left foot with just the keeper to beat, or Pepe to NEVER score from 2 yards out instead of having his shot blocked???
Or do you mean - on percentages you would expect a miss more often than a goal?

Name one chance where the xG is not accurately reflected.

Your player assessment is similar to what I thought from being there.


Again, find a game from the weekend where a team has had a similar amount of shots on goals with a low xG ratings and scored? It's a rare thing.

Also, Norwich had 10 shots on goal with similar low 0.1 xG scores. Should they have scored a goal against us? The volume of shots argument doesn't make much sense


This data has already been provided for the whole season not just this weekend.
Example. The average quality of the chances created by Aston Villa this season is 0.09.
They have scored 5 goals.

3 of their 5 goals have xGs off 0.03, 0.06 and 0.09.
There are other teams in the table at the top of this thread.

Another example...
Liverpool have had a lot of lower quality chances like us. The average quality of their chances is less than 0.1.
They have scored with an xG chance of 0.03
There is another one of their goals with an xG of 0.14.

I can go on.

Obviously the accumulation of lower quality chances will eventually yield a goal. That's been football ever since it began. We've all heard the phrase 'keep shooting and one will eventually go in'
Some players are just better at converting harder chances than others.
At the moment our team, unlike Liverpool, Villa and others, have turned none of those difficult chances into goals

So yes, the volume of shots argument does make sense and can be tracked to show that lower quality chances do yield goals and some teams are better at converting them.

Whichever way you come at the data it pretty much always ends up reflecting reality.

As for Norwich, yes, if they had scored we could not have said they were lucky. They had enough chances to score, but you are wrong about their xG. It was 0.58, not 1.
So whether they scored or not was pretty 50/50 based on expectation from chances.
Again solid data.


Solid data but you're ignoring player quality and have a data analyst approach with some these quotes. Teams don't often score from multiple goals from these low xG chances, you have some examples below. It's usually the odd goal and not 3 or 4. But the beauty of football is that sometimes you get games where you get some high quality goals from low xG chances and that's a brilliant game to watch.

https://understat.com/match/16409

https://understat.com/match/16399

https://understat.com/match/16388

https://understat.com/match/16402

The other thing you have to address with all this data is what does that tell you about the team and manager? Have we got things right with our transfer strategy and not buying better attackers if we're not scoring enough of these low xG chances? Should we be playing this style if we're not converting chances but when Arteta first arrived we were taking less shots but creating higher quality chances and scoring 2 out of 3 attempts like this Liverpool game below? Look at the amount of shots Liverpool had in contrast?

https://understat.com/match/11993

Or look at this Norwich game where we beat them 4-0 in Arteta's first season. One wonder goal from Cedric but the rest were higher xG chances finished.

https://understat.com/match/11953

The data is solid but now you have to apply some of that to the team we have and playing to our strengths, not recreating a Man City lite.
User avatar
Power n Glory
Dennis Bergkamp
Dennis Bergkamp
 
Posts: 6315
Joined: Thu Nov 08, 2018 6:02 pm

Re: Creativity & Chance Conversion

Postby Özim » Tue Sep 14, 2021 10:26 am

Is t this data proving what we knew, that we struggle to create decent chances?

It also shows that shots can be deceptive as 3 shots can come from the 1 period of play for example and if you scored the first shot l the other 2 wouldn’t happen.

I also think different teams have different levels of goal threats (some teams have more goalscorers) and therefore it really depends who the chance falls to as well, a defender is much less likely to score then an attacker for example
Özim
Poster Of The Month
Poster Of The Month
 
Posts: 4596
Joined: Fri Aug 19, 2016 11:26 pm

Re: Creativity & Chance Conversion

Postby jayramfootball » Tue Sep 14, 2021 10:47 am

Power n Glory wrote:
jayramfootball wrote:
Power n Glory wrote:
jayramfootball wrote:
theHotHead wrote:81 mins, 24 chances, 7 of which were not absolutely fookin shit chances, 3 of which were rebounds, one was the actual goal and there were 2 decent half chances from Saka. But again I stress, not a single big chance from any of that lot, not a single chance that I would expect a goal to be scored from.

At this point I can say we were crap vs Norwich, Ben White had a good game, anyone criticising Lokonga's performance is blind, Odegaard was meh, AMN needs to give up ambitions of being a CM, for e very good thing he does, he will do 2 absolutely dreadful things. Tomi looked quality, Ramsdale looked good, Pepe was in and out, Saka was in and out, Auba was Auba - not a great game but you felt he would do damage given a chance.


There were 30 shots on goal - you'll just have to rewatch it ( :biggrin: ) as I am going to assume that all the major websites like Sky and BBC who also state there were 30 shots got it right).

4 good / decent chances listed above with video and time stamps.
26 speculative chances

You use the term 'expecting to score'. What does that mean???
You mean would NEVER expect to score? Would you NEVER expect ESR to have put his chance away, or Auba to NEVER score with his left foot with just the keeper to beat, or Pepe to NEVER score from 2 yards out instead of having his shot blocked???
Or do you mean - on percentages you would expect a miss more often than a goal?

Name one chance where the xG is not accurately reflected.

Your player assessment is similar to what I thought from being there.


Again, find a game from the weekend where a team has had a similar amount of shots on goals with a low xG ratings and scored? It's a rare thing.

Also, Norwich had 10 shots on goal with similar low 0.1 xG scores. Should they have scored a goal against us? The volume of shots argument doesn't make much sense


This data has already been provided for the whole season not just this weekend.
Example. The average quality of the chances created by Aston Villa this season is 0.09.
They have scored 5 goals.

3 of their 5 goals have xGs off 0.03, 0.06 and 0.09.
There are other teams in the table at the top of this thread.

Another example...
Liverpool have had a lot of lower quality chances like us. The average quality of their chances is less than 0.1.
They have scored with an xG chance of 0.03
There is another one of their goals with an xG of 0.14.

I can go on.

Obviously the accumulation of lower quality chances will eventually yield a goal. That's been football ever since it began. We've all heard the phrase 'keep shooting and one will eventually go in'
Some players are just better at converting harder chances than others.
At the moment our team, unlike Liverpool, Villa and others, have turned none of those difficult chances into goals

So yes, the volume of shots argument does make sense and can be tracked to show that lower quality chances do yield goals and some teams are better at converting them.

Whichever way you come at the data it pretty much always ends up reflecting reality.

As for Norwich, yes, if they had scored we could not have said they were lucky. They had enough chances to score, but you are wrong about their xG. It was 0.58, not 1.
So whether they scored or not was pretty 50/50 based on expectation from chances.
Again solid data.


Solid data but you're ignoring player quality and have a data analyst approach with some these quotes. Teams don't often score from multiple goals from these low xG chances, you have some examples below. It's usually the odd goal and not 3 or 4. But the beauty of football is that sometimes you get games where you get some high quality goals from low xG chances and that's a brilliant game to watch.

https://understat.com/match/16409

https://understat.com/match/16399

https://understat.com/match/16388

https://understat.com/match/16402

The other thing you have to address with all this data is what does that tell you about the team and manager? Have we got things right with our transfer strategy and not buying better attackers if we're not scoring enough of these low xG chances? Should we be playing this style if we're not converting chances but when Arteta first arrived we were taking less shots but creating higher quality chances and scoring 2 out of 3 attempts like this Liverpool game below? Look at the amount of shots Liverpool had in contrast?

https://understat.com/match/11993

Or look at this Norwich game where we beat them 4-0 in Arteta's first season. One wonder goal from Cedric but the rest were higher xG chances finished.

https://understat.com/match/11953

The data is solid but now you have to apply some of that to the team we have and playing to our strengths, not recreating a Man City lite.


I have already shown 6 goals scored by 3 teams this season from 'low quality chances', including 1 team where MOST of their goals have come from low quality chances.
I am going to provide a view of every PL goal scored this season and the associated quality of the chance.
I don't know what that will show yet, the data will just be the data.

IN terms of the where the team needs to go , it's pretty clear to me that we need to improve the cutting edge in terms of opening teams up to create better chances and also be more ruthless in front of goal.

Do we have the players - given what we had and what we've brought in? That remains to be seen in terms of build-up play that can more effectively break down defences. In terms of finishing - i.e. players that can finish better chances at a high rate, strike cleanly from distance and convert more of those half chances , I think we are short and probably need a couple of players.

That said, we're only 4 games in... the data needs to be measured over a longer period of time really. For example, Pepe has had a few half chances this season and missed them all, plus failed to convert his one good chance. He might bang in two goals from distance in the next game and then all of sudden he's back on track. It will be later in the season before we can conclude at the player level.
Last edited by jayramfootball on Tue Sep 14, 2021 10:57 am, edited 2 times in total.
Image
jayramfootball
Arsène Wenger
Arsène Wenger
 
Posts: 13684
Joined: Mon Sep 03, 2012 8:58 pm

Re: Creativity & Chance Conversion

Postby jayramfootball » Tue Sep 14, 2021 10:49 am

Özim wrote:Is t this data proving what we knew, that we struggle to create decent chances?

It also shows that shots can be deceptive as 3 shots can come from the 1 period of play for example and if you scored the first shot l the other 2 wouldn’t happen.

I also think different teams have different levels of goal threats (some teams have more goalscorers) and therefore it really depends who the chance falls to as well, a defender is much less likely to score then an attacker for example


Ironically yes.
The data shows merit to the argument that we struggle to create quality chances.
We are 19th in the league on that rating.

The summary I provided is still the most objective view to be read on GW.

Overall chances - OK
Quality chances - Poor
Finishing - Poor
Image
jayramfootball
Arsène Wenger
Arsène Wenger
 
Posts: 13684
Joined: Mon Sep 03, 2012 8:58 pm

Re: Creativity & Chance Conversion

Postby Power n Glory » Tue Sep 14, 2021 10:53 am

Özim wrote:Is t this data proving what we knew, that we struggle to create decent chances?

It also shows that shots can be deceptive as 3 shots can come from the 1 period of play for example and if you scored the first shot l the other 2 wouldn’t happen.


Yep. It;s a decent site worth checking for because those key passes and chances created figures sent around are deceptive at times. I've always maintained that and believe Jay may have found this site from a conversation we've had in the passed about chance creation.

Either way, I think there is still a misunderstanding on xG. Teams with that create a high volume of shots don't usually score from low xG chances. Even if it does happen, it's not sustainable. We can look at City's games and they usually get a lot of shots but they often score from quality chances created not just the low probability chances. If we want to play this high possession, volume style, then we will need to improve.

https://understat.com/match/16396

https://understat.com/match/16409

https://understat.com/match/16387
User avatar
Power n Glory
Dennis Bergkamp
Dennis Bergkamp
 
Posts: 6315
Joined: Thu Nov 08, 2018 6:02 pm

Re: Creativity & Chance Conversion

Postby jayramfootball » Tue Sep 14, 2021 11:00 am

Power n Glory wrote:
Özim wrote:Is t this data proving what we knew, that we struggle to create decent chances?

It also shows that shots can be deceptive as 3 shots can come from the 1 period of play for example and if you scored the first shot l the other 2 wouldn’t happen.


Yep. It;s a decent site worth checking for because those key passes and chances created figures sent around are deceptive at times. I've always maintained that and believe Jay may have found this site from a conversation we've had in the passed about chance creation.

Either way, I think there is still a misunderstanding on xG. Teams with that create a high volume of shots don't usually score from low xG chances. Even if it does happen, it's not sustainable. We can look at City's games and they usually get a lot of shots but they often score from quality chances created not just the low probability chances. If we want to play this high possession, volume style, then we will need to improve.

https://understat.com/match/16396

https://understat.com/match/16409

https://understat.com/match/16387


3 of Aston Villa's 5 goals this season have come from low probability shots.
2 of Liverpools 6 goals
At least 1 of Everton's goals

I am checking every team, I expect to see every team with goals from low probability shots a(except us and a couple of others) and also expect the majority of every team's chances to be heavily skewed to lower quality chances (not just us).
Image
jayramfootball
Arsène Wenger
Arsène Wenger
 
Posts: 13684
Joined: Mon Sep 03, 2012 8:58 pm

Re: Creativity & Chance Conversion

Postby theHotHead » Tue Sep 14, 2021 11:01 am

jayramfootball wrote:
theHotHead wrote:
jayramfootball wrote:
theHotHead wrote:But Jay, if we take the Norwich game in isolation, its not our finishing that prevented us from scoring more goals (ESR chance aside), it was the quality of the chances we created that was the problem.


It was a combination.
ESR chance, Auba chance and Pepe chance (blocked on the line) - you'd expect 1 goal from those three chances ( in fact the xG combined for all 3 was about 1). We missed them all.
Then all the other chances - based on league averages a goal would result about 2-3% of the time so you'd expect 35-40 of those types of efforts to yield a goal.
Then the actual goal, which obviously had a very high xG.
That is why the xG came out at 2.7.

Now, if you refer back to a previous post I have already stated that we're only 4 games in, so the numbers can not be reflective of individual players yet and even for the whole team we only have a small data set.
For example, that ESR chance - we can forgive him. It's one chance and can easily be missed. If over the season, however, he has 10 of those types of chance and misses ALL of them, we'll be saying (rightly) that he has finished poorly this season.

That is why this data will be updated each week. The picture becomes more and more solid as more data is added, but as of now, 4 games in ... we are creating lots of chances (6th in the league), but not very good ones (19th in the league on average) and our finishing is the worst in the league.

No Jay, I agree you could reasonably expect 1 goal from ESR/Pepe/Auba chances most definitely but I disagree that you would expect a goal from the other chances, they were poor low quality chances, they never looked like being goals.

I agree with PnG, of we continue to create such poor quality chances in high numbers, we will struggle to score any goals!!


Those other lower percentage chances lead to goals game after game, year after year. Lower percentage, yes, but goals nonetheless.

Already pointed out 6 goals scored from low percentage chances by just 3 teams.
If I looked at every team I bet I would find that the difference between the goals scored from low quality and high quality chances was not that big.

No they don't Jay, they typically never amount to a goal, all you have done is show us some exceptions to the rule, you can always find an exception but, for the most part, for the vast majority of the so-called chances we created against Norwich, they don't amount to goals.

Lets do the analysis PNG has asked for, lets look at the percentage of 0.1xG chances that actually end up being goals. In fact, surely we can take this a step further, can't we look at each team's xG and match up actual goals to expected and see on how many occasions a team scored more than their expected goals. That will give us an idea of how many unexpected goals have been scored, right?!

My guess is there will be hardly any instances of that if any.
User avatar
theHotHead
Poster Of The Month
Poster Of The Month
 
Posts: 12215
Joined: Mon Oct 17, 2016 12:44 am
Location: Norf Landon

Re: Creativity & Chance Conversion

Postby Power n Glory » Tue Sep 14, 2021 11:14 am

jayramfootball wrote:
Power n Glory wrote:
Özim wrote:Is t this data proving what we knew, that we struggle to create decent chances?

It also shows that shots can be deceptive as 3 shots can come from the 1 period of play for example and if you scored the first shot l the other 2 wouldn’t happen.


Yep. It;s a decent site worth checking for because those key passes and chances created figures sent around are deceptive at times. I've always maintained that and believe Jay may have found this site from a conversation we've had in the passed about chance creation.

Either way, I think there is still a misunderstanding on xG. Teams with that create a high volume of shots don't usually score from low xG chances. Even if it does happen, it's not sustainable. We can look at City's games and they usually get a lot of shots but they often score from quality chances created not just the low probability chances. If we want to play this high possession, volume style, then we will need to improve.

https://understat.com/match/16396

https://understat.com/match/16409

https://understat.com/match/16387


3 of Aston Villa's 5 goals this season have come from low probability shots.
2 of Liverpools 6 goals
At least 1 of Everton's goals

I am checking every team, I expect to see every team with goals from low probability shots a(except us and a couple of others) and also expect the majority of every team's chances to be heavily skewed to lower quality chances (not just us).


Ok. Let's go with it.

So what does that tell you about Arsenal? You're not addressing the elephant in the room. Link it with all you have said about the manager, players and transfer strategy. If the data is the data, everything the manager is doing in regards to strategy and transfers can't be right, can it? How does Arteta fix this issue going forward if it's a case of poor finishing? Is it back to waiting for another transfer window?
User avatar
Power n Glory
Dennis Bergkamp
Dennis Bergkamp
 
Posts: 6315
Joined: Thu Nov 08, 2018 6:02 pm

Re: Creativity & Chance Conversion

Postby Power n Glory » Tue Sep 14, 2021 11:18 am

theHotHead wrote:
jayramfootball wrote:
theHotHead wrote:
jayramfootball wrote:
theHotHead wrote:But Jay, if we take the Norwich game in isolation, its not our finishing that prevented us from scoring more goals (ESR chance aside), it was the quality of the chances we created that was the problem.


It was a combination.
ESR chance, Auba chance and Pepe chance (blocked on the line) - you'd expect 1 goal from those three chances ( in fact the xG combined for all 3 was about 1). We missed them all.
Then all the other chances - based on league averages a goal would result about 2-3% of the time so you'd expect 35-40 of those types of efforts to yield a goal.
Then the actual goal, which obviously had a very high xG.
That is why the xG came out at 2.7.

Now, if you refer back to a previous post I have already stated that we're only 4 games in, so the numbers can not be reflective of individual players yet and even for the whole team we only have a small data set.
For example, that ESR chance - we can forgive him. It's one chance and can easily be missed. If over the season, however, he has 10 of those types of chance and misses ALL of them, we'll be saying (rightly) that he has finished poorly this season.

That is why this data will be updated each week. The picture becomes more and more solid as more data is added, but as of now, 4 games in ... we are creating lots of chances (6th in the league), but not very good ones (19th in the league on average) and our finishing is the worst in the league.

No Jay, I agree you could reasonably expect 1 goal from ESR/Pepe/Auba chances most definitely but I disagree that you would expect a goal from the other chances, they were poor low quality chances, they never looked like being goals.

I agree with PnG, of we continue to create such poor quality chances in high numbers, we will struggle to score any goals!!


Those other lower percentage chances lead to goals game after game, year after year. Lower percentage, yes, but goals nonetheless.

Already pointed out 6 goals scored from low percentage chances by just 3 teams.
If I looked at every team I bet I would find that the difference between the goals scored from low quality and high quality chances was not that big.

No they don't Jay, they typically never amount to a goal, all you have done is show us some exceptions to the rule, you can always find an exception but, for the most part, for the vast majority of the so-called chances we created against Norwich, they don't amount to goals.

Lets do the analysis PNG has asked for, lets look at the percentage of 0.1xG chances that actually end up being goals. In fact, surely we can take this a step further, can't we look at each team's xG and match up actual goals to expected and see on how many occasions a team scored more than their expected goals. That will give us an idea of how many unexpected goals have been scored, right?!

My guess is there will be hardly any instances of that if any.


Yes and if we look at it that way, shouldn't we be conceding more goals to those low xG chances teams create against us?
User avatar
Power n Glory
Dennis Bergkamp
Dennis Bergkamp
 
Posts: 6315
Joined: Thu Nov 08, 2018 6:02 pm

Re: Creativity & Chance Conversion

Postby jayramfootball » Tue Sep 14, 2021 11:21 am

theHotHead wrote:
jayramfootball wrote:
theHotHead wrote:
jayramfootball wrote:
theHotHead wrote:But Jay, if we take the Norwich game in isolation, its not our finishing that prevented us from scoring more goals (ESR chance aside), it was the quality of the chances we created that was the problem.


It was a combination.
ESR chance, Auba chance and Pepe chance (blocked on the line) - you'd expect 1 goal from those three chances ( in fact the xG combined for all 3 was about 1). We missed them all.
Then all the other chances - based on league averages a goal would result about 2-3% of the time so you'd expect 35-40 of those types of efforts to yield a goal.
Then the actual goal, which obviously had a very high xG.
That is why the xG came out at 2.7.

Now, if you refer back to a previous post I have already stated that we're only 4 games in, so the numbers can not be reflective of individual players yet and even for the whole team we only have a small data set.
For example, that ESR chance - we can forgive him. It's one chance and can easily be missed. If over the season, however, he has 10 of those types of chance and misses ALL of them, we'll be saying (rightly) that he has finished poorly this season.

That is why this data will be updated each week. The picture becomes more and more solid as more data is added, but as of now, 4 games in ... we are creating lots of chances (6th in the league), but not very good ones (19th in the league on average) and our finishing is the worst in the league.

No Jay, I agree you could reasonably expect 1 goal from ESR/Pepe/Auba chances most definitely but I disagree that you would expect a goal from the other chances, they were poor low quality chances, they never looked like being goals.

I agree with PnG, of we continue to create such poor quality chances in high numbers, we will struggle to score any goals!!


Those other lower percentage chances lead to goals game after game, year after year. Lower percentage, yes, but goals nonetheless.

Already pointed out 6 goals scored from low percentage chances by just 3 teams.
If I looked at every team I bet I would find that the difference between the goals scored from low quality and high quality chances was not that big.

No they don't Jay, they typically never amount to a goal, all you have done is show us some exceptions to the rule, you can always find an exception but, for the most part, for the vast majority of the so-called chances we created against Norwich, they don't amount to goals.

Lets do the analysis PNG has asked for, lets look at the percentage of 0.1xG chances that actually end up being goals. In fact, surely we can take this a step further, can't we look at each team's xG and match up actual goals to expected and see on how many occasions a team scored more than their expected goals. That will give us an idea of how many unexpected goals have been scored, right?!

My guess is there will be hardly any instances of that if any.


I am doing just that, so lets start with Man Utd

Image

11 goals , 4 of which have been scored with an xG of <0.1 AND nearly 70% of all their chances being <0.1 quality.

I am doing this for every team and I suspect it's going to show a lot of goals scored from these low-quality chances.
Image
jayramfootball
Arsène Wenger
Arsène Wenger
 
Posts: 13684
Joined: Mon Sep 03, 2012 8:58 pm

Re: Creativity & Chance Conversion

Postby Power n Glory » Tue Sep 14, 2021 11:27 am

jayramfootball wrote:
theHotHead wrote:
jayramfootball wrote:
theHotHead wrote:
jayramfootball wrote:
theHotHead wrote:But Jay, if we take the Norwich game in isolation, its not our finishing that prevented us from scoring more goals (ESR chance aside), it was the quality of the chances we created that was the problem.


It was a combination.
ESR chance, Auba chance and Pepe chance (blocked on the line) - you'd expect 1 goal from those three chances ( in fact the xG combined for all 3 was about 1). We missed them all.
Then all the other chances - based on league averages a goal would result about 2-3% of the time so you'd expect 35-40 of those types of efforts to yield a goal.
Then the actual goal, which obviously had a very high xG.
That is why the xG came out at 2.7.

Now, if you refer back to a previous post I have already stated that we're only 4 games in, so the numbers can not be reflective of individual players yet and even for the whole team we only have a small data set.
For example, that ESR chance - we can forgive him. It's one chance and can easily be missed. If over the season, however, he has 10 of those types of chance and misses ALL of them, we'll be saying (rightly) that he has finished poorly this season.

That is why this data will be updated each week. The picture becomes more and more solid as more data is added, but as of now, 4 games in ... we are creating lots of chances (6th in the league), but not very good ones (19th in the league on average) and our finishing is the worst in the league.

No Jay, I agree you could reasonably expect 1 goal from ESR/Pepe/Auba chances most definitely but I disagree that you would expect a goal from the other chances, they were poor low quality chances, they never looked like being goals.

I agree with PnG, of we continue to create such poor quality chances in high numbers, we will struggle to score any goals!!


Those other lower percentage chances lead to goals game after game, year after year. Lower percentage, yes, but goals nonetheless.

Already pointed out 6 goals scored from low percentage chances by just 3 teams.
If I looked at every team I bet I would find that the difference between the goals scored from low quality and high quality chances was not that big.

No they don't Jay, they typically never amount to a goal, all you have done is show us some exceptions to the rule, you can always find an exception but, for the most part, for the vast majority of the so-called chances we created against Norwich, they don't amount to goals.

Lets do the analysis PNG has asked for, lets look at the percentage of 0.1xG chances that actually end up being goals. In fact, surely we can take this a step further, can't we look at each team's xG and match up actual goals to expected and see on how many occasions a team scored more than their expected goals. That will give us an idea of how many unexpected goals have been scored, right?!

My guess is there will be hardly any instances of that if any.


I am doing just that, so lets start with Man Utd

Image

11 goals , 4 of which have been scored with an xG of <0.1 AND nearly 70% of all their chances being <0.1 quality.

I am doing this for every team and I suspect it's going to show a lot of goals scored from these low-quality chances.


Man Utd have not score 4 goals in one game from low xG chances. That’s the difference. You're looking at one Norwich game for us and saying we should have beaten them by 3 or 4 goals from mostly low xG chances.
User avatar
Power n Glory
Dennis Bergkamp
Dennis Bergkamp
 
Posts: 6315
Joined: Thu Nov 08, 2018 6:02 pm

PreviousNext

Return to Arsenal Talk

Who is online

Users browsing this forum: Anciunask, FnniElupe and 3 guests