Creativity & Chance Conversion

[DiamondGooner]

No they can’t, hence the reason no one that watches these matches sees these chances.

Stats = fake reality

No one in their right minds relies on stats over their eyes.

The stats state that we don’t struggle offensively, isn’t strange how everyone that watches thinks we do then? Hmmmmm

Exactly.

Some bell end inputs this data so its not perfect or reliable.

We live in an age of information but if you dig a little deeper you come to realise someone no different to me or you collects this data and its subject to human opinion just like books are written by someone.

Inputs which data?

All data.

Even if the data is captured by a machine, that machine was programmed to capture certain items of data by a human.

Or do you think machines programme or design themselves?

[jayramfootball]

No they can’t, hence the reason no one that watches these matches sees these chances.

Stats = fake reality

No one in their right minds relies on stats over their eyes.

The stats state that we don’t struggle offensively, isn’t strange how everyone that watches thinks we do then? Hmmmmm

Exactly.

Some bell end inputs this data so its not perfect or reliable.

We live in an age of information but if you dig a little deeper you come to realise someone no different to me or you collects this data and its subject to human opinion just like books are written by someone.

Inputs which data?

All data.

Even if the data is captured by a machine, that machine was programmed to capture certain items of data by a human.

Or do you think machines programme or design themselves?

Well a significant portion of the data that is presented is in fact determined by machine learning.
The inputs are the raw data which are actual events and situations on the pitch.
I am not sure why a person would knowingly input incorrect data - if they did their machine learning model would fall apart and they would not be able to market it to sports franchises who use this data.
There could be data input mistakes of course, as in any model but then the outputs are so transparent they can be checked easily - which is part of any model building QC and testing.

Take the example I gave - Auba's left-foot shot at an angle of about 35 degrees when he was put through - the xG number suggests he should score that about 1 in 8 times. Not unreasonable at all. Every single chance can be assessed and compared to similar situations to make sure it is not under or overrepresented.

No, the fact is that the data is going to be more representative than an agenda based or subjective opinion.
Hence this data will be updated after every game.

Also the actual shots and goals are clearly going to be largely error-free. They are about as black and white a statistic as you can get.
The Ai is only relevant to the chance quality.

[Hybrid47]

Have to say this is going to be one of the most interesting threads over the season.

The quality of our chance creation has been something that has deteriorated over the years.

The fact that Auba could miss an opportunity but knew he'd always get another is something we took for granted. Now, a good chance for us is an anomaly.

That also can go hand in hand with the football iq of players we have had in the past that we haven't had for many years.

It will definitely be interesting to see how our conversion rate changes with a settled first 11.

I mean for now we just need to build some momentum and realistically from our first 3 games we were only going to come away with 3 points 4 if we were lucky. In the grand scheme of things games against Chelsea and City were basically free hits.

[Power n Glory]

Majority of the shots are 0.1 for xG. Those are poor chances with a low probability of scoring. We controlled the game, dominated and had the volume of shots, but only 6 shots on target, a lot of blocked shots and the best chance we missed fell to Pepe. We still have a long way to go when it comes to our creativity.

[jayramfootball]

Majority of the shots are 0.1 for xG. Those are poor chances with a low probability of scoring. We controlled the game, dominated and had the volume of shots, but only 6 shots on target, a lot of blocked shots and the best chance we missed fell to Pepe. We still have a long way to go when it comes to our creativity.

That is what the data says. yes, except you missed our poor finishing.
Lots of chances created, not enough good ones (we rank 10th in the league for that), and poor finishing meaning we scored 1 instead of the 4-5 expected goals we should have from the chances created so far this season. Even those 0.1 chances are converted at, well, 10% on average and we're not converting any of them. Never mind the fewer better chances we are missing.

I reckon we will rise in team xG over the coming weeks because right now we're 10th best in the league having played Chelsea and Man City .. we'll be up to 6th or so in a few games is my prediction - but we'll see. Whether we convert more of our chances I am less confident about - our shooting is abysmal.

[Power n Glory]

Majority of the shots are 0.1 for xG. Those are poor chances with a low probability of scoring. We controlled the game, dominated and had the volume of shots, but only 6 shots on target, a lot of blocked shots and the best chance we missed fell to Pepe. We still have a long way to go when it comes to our creativity.

That is what the data says. yes, except you missed our poor finishing.
Lots of chances created, not enough good ones (we rank 10th in the league for that), and poor finishing meaning we scored 1 instead of the 4-5 expected goals we should have from the chances created so far this season. Even those 0.1 chances are converted at, well, 10% on average and we're not converting any of them. Never mind the fewer better chances we are missing.

I reckon we will rise in team xG over the coming weeks because right now we're 10th best in the league having played Chelsea and Man City .. we'll be up to 6th or so in a few games in my prediction - but we'll see.

https://understat.com/league/EPL

Look through the weekend fixtures and highlight the goals scored from a 0.1 xG score or less? It's rare but look it up. Those are wonder goals of individual brilliance that make the goal. Not the pass.

If such a low xG score is a good chance, that would also mean Norwich had some good chances to win but were let down by poor finishing. That's far from the truth. Every shot isn't a good shot. But again, open to anyone, look up the 0.1 xG chances that were scored, watch Match of the Day, or just go to Arsenal Player and rewatch the game to see the chances we created.

[jayramfootball]

Majority of the shots are 0.1 for xG. Those are poor chances with a low probability of scoring. We controlled the game, dominated and had the volume of shots, but only 6 shots on target, a lot of blocked shots and the best chance we missed fell to Pepe. We still have a long way to go when it comes to our creativity.

That is what the data says. yes, except you missed our poor finishing.
Lots of chances created, not enough good ones (we rank 10th in the league for that), and poor finishing meaning we scored 1 instead of the 4-5 expected goals we should have from the chances created so far this season. Even those 0.1 chances are converted at, well, 10% on average and we're not converting any of them. Never mind the fewer better chances we are missing.

I reckon we will rise in team xG over the coming weeks because right now we're 10th best in the league having played Chelsea and Man City .. we'll be up to 6th or so in a few games in my prediction - but we'll see.

https://understat.com/league/EPL

Look through the weekend fixtures and highlight the goals scored from a 0.1 xG score or less? It's rare but look it up. Those are wonder goals of individual brilliance that make the goal. Not the pass.

If such a low xG score is a good chance, that would also mean Norwich had some good chances to win but were let down by poor finishing. That's far from the truth. Every shot isn't a good shot. But again, open to anyone, look up the 0.1 xG chances that were scored, watch Match of the Day, or just go to Arsenal Player and rewatch the game to see the chances we created.

No one said a 0.1. xG was a good chance - it is a chance that has about a 10% chance of being scored.
Stop trying to muddy the waters.

The data is the data - we had 30 shots that should have yielded 2-3 goals against Norwich.

[VCC]

The goal probably counted as 3 chances

[jayramfootball]

The goal probably counted as 3 chances

No because even those chances can be missed - a defender might get a lunging block in for example.
The goal was awarded an xG of 0.67.

Our total xG for the game was about 2.7

[Power n Glory]

Majority of the shots are 0.1 for xG. Those are poor chances with a low probability of scoring. We controlled the game, dominated and had the volume of shots, but only 6 shots on target, a lot of blocked shots and the best chance we missed fell to Pepe. We still have a long way to go when it comes to our creativity.

That is what the data says. yes, except you missed our poor finishing.
Lots of chances created, not enough good ones (we rank 10th in the league for that), and poor finishing meaning we scored 1 instead of the 4-5 expected goals we should have from the chances created so far this season. Even those 0.1 chances are converted at, well, 10% on average and we're not converting any of them. Never mind the fewer better chances we are missing.

I reckon we will rise in team xG over the coming weeks because right now we're 10th best in the league having played Chelsea and Man City .. we'll be up to 6th or so in a few games in my prediction - but we'll see.

https://understat.com/league/EPL

Look through the weekend fixtures and highlight the goals scored from a 0.1 xG score or less? It's rare but look it up. Those are wonder goals of individual brilliance that make the goal. Not the pass.

If such a low xG score is a good chance, that would also mean Norwich had some good chances to win but were let down by poor finishing. That's far from the truth. Every shot isn't a good shot. But again, open to anyone, look up the 0.1 xG chances that were scored, watch Match of the Day, or just go to Arsenal Player and rewatch the game to see the chances we created.

No one said a 0.1. xG was a good chance - it is a chance that has about a 10% chance of being scored.
Stop trying to muddy the waters.

The data is the data - we had 30 shots that should have yielded 2-3 goals against Norwich.

Not trying to muddy the waters, just helping people understand xG. The data is the data I haven't seen where you get this idea that 30 shots should have yielded 2-3 goals. It doesn't work like that if you're creating low xG chances.

The one clear cut chance we should have scored but didn't was the 56 minute one from Pepe. If you watch that back, that was a 0.56 chance and nothing to do with bad finishing, just excellent defending from William to block it.

But again, anyone in fact, look up a 0.1 xG or less goal in match of the day or rewatch our game to get an idea of how xG works. That 3rd Lukaku goal is a 0.1 xG chance. The Fernandes goal for Utd is 0.03. Ronald's goals were a 0. 89 and 0.30. Newcastle's goal was a 0.26. Rewatch the goals for context.

[jayramfootball]

Majority of the shots are 0.1 for xG. Those are poor chances with a low probability of scoring. We controlled the game, dominated and had the volume of shots, but only 6 shots on target, a lot of blocked shots and the best chance we missed fell to Pepe. We still have a long way to go when it comes to our creativity.

That is what the data says. yes, except you missed our poor finishing.
Lots of chances created, not enough good ones (we rank 10th in the league for that), and poor finishing meaning we scored 1 instead of the 4-5 expected goals we should have from the chances created so far this season. Even those 0.1 chances are converted at, well, 10% on average and we're not converting any of them. Never mind the fewer better chances we are missing.

I reckon we will rise in team xG over the coming weeks because right now we're 10th best in the league having played Chelsea and Man City .. we'll be up to 6th or so in a few games in my prediction - but we'll see.

https://understat.com/league/EPL

Look through the weekend fixtures and highlight the goals scored from a 0.1 xG score or less? It's rare but look it up. Those are wonder goals of individual brilliance that make the goal. Not the pass.

If such a low xG score is a good chance, that would also mean Norwich had some good chances to win but were let down by poor finishing. That's far from the truth. Every shot isn't a good shot. But again, open to anyone, look up the 0.1 xG chances that were scored, watch Match of the Day, or just go to Arsenal Player and rewatch the game to see the chances we created.

No one said a 0.1. xG was a good chance - it is a chance that has about a 10% chance of being scored.
Stop trying to muddy the waters.

The data is the data - we had 30 shots that should have yielded 2-3 goals against Norwich.

Not trying to muddy the waters, just helping people understand xG. The data is the data I haven't seen where you get this idea that 30 shots should have yielded 2-3 goals. It doesn't work like that if you're creating low xG chances.

The one clear cut chance we should have scored but didn't was the 56 minute one from Pepe. If you watch that back, that was a 0.56 chance and nothing to do with bad finishing, just excellent defending from William to block it.

But again, anyone in fact, look up a 0.1 xG or less goal in match of the day or rewatch our game to get an idea of how xG works. That 3rd Lukaku goal is a 0.1 xG chance. The Fernandes goal for Utd is 0.03. Ronald's goals were a 0. 89 and 0.30. Newcastle's goal was a 0.26. Rewatch the goals for context.

Our team xG for the game was around 2.7.
If you have 10 shots all of 0.1 xG you expect to score 1 goal.. i.e 9 will not go in, but 1 on average willl. That is the whole point of xG rating per attempted shot. The better the chance the more likely it is on average to be scored.


That is the overall cumulative effect of all the percentage chances of scoring from every shot we had in the Norwich game on a running total of xG for the game for the team. We scored 1 , so we were well below expectations based on the chances we had.

To give more, the chance ESR had when he got a shot away from just around the edge of the box was an xG of 0.11.. i.e. an 11% chance of going in based on the norm. It's a chance but more often that not it will be missed but if he had 9 of those chances, 1 would be likely to go in on average.

[Power n Glory]

The goal probably counted as 3 chances

We had the volume. The final xG score is an accumulation of all the shots tallied up to give us the 2.69 score. All the 0.1, 0.14, 0.04 shot at up if you dominate with shots. We did dominate but just need to create better chances.

[jayramfootball]

The goal probably counted as 3 chances

We had the volume. The final xG score is an accumulation of all the shots tallied up to give us the 2.69 score. All the 0.1, 0.14, 0.04 shot at up if you dominate with shots. We did dominate but just need to create better chances.

Almost correct.
Yes, this season the quality of our chances created ranks 19th in the league, the volume of chances ranks 6th.
Even though the quality is ranked low, we have created a lot and over the season would have been expected to score 4-5 goals but have scored 1.
That is the worst ratio in the league of expected goals to actual goals - i.e. our shooting is shit as well as the quality (on average) of the chances we are creating.

[Power n Glory]

Majority of the shots are 0.1 for xG. Those are poor chances with a low probability of scoring. We controlled the game, dominated and had the volume of shots, but only 6 shots on target, a lot of blocked shots and the best chance we missed fell to Pepe. We still have a long way to go when it comes to our creativity.

That is what the data says. yes, except you missed our poor finishing.
Lots of chances created, not enough good ones (we rank 10th in the league for that), and poor finishing meaning we scored 1 instead of the 4-5 expected goals we should have from the chances created so far this season. Even those 0.1 chances are converted at, well, 10% on average and we're not converting any of them. Never mind the fewer better chances we are missing.

I reckon we will rise in team xG over the coming weeks because right now we're 10th best in the league having played Chelsea and Man City .. we'll be up to 6th or so in a few games in my prediction - but we'll see.

https://understat.com/league/EPL

Look through the weekend fixtures and highlight the goals scored from a 0.1 xG score or less? It's rare but look it up. Those are wonder goals of individual brilliance that make the goal. Not the pass.

If such a low xG score is a good chance, that would also mean Norwich had some good chances to win but were let down by poor finishing. That's far from the truth. Every shot isn't a good shot. But again, open to anyone, look up the 0.1 xG chances that were scored, watch Match of the Day, or just go to Arsenal Player and rewatch the game to see the chances we created.

No one said a 0.1. xG was a good chance - it is a chance that has about a 10% chance of being scored.
Stop trying to muddy the waters.

The data is the data - we had 30 shots that should have yielded 2-3 goals against Norwich.

Not trying to muddy the waters, just helping people understand xG. The data is the data I haven't seen where you get this idea that 30 shots should have yielded 2-3 goals. It doesn't work like that if you're creating low xG chances.

The one clear cut chance we should have scored but didn't was the 56 minute one from Pepe. If you watch that back, that was a 0.56 chance and nothing to do with bad finishing, just excellent defending from William to block it.

But again, anyone in fact, look up a 0.1 xG or less goal in match of the day or rewatch our game to get an idea of how xG works. That 3rd Lukaku goal is a 0.1 xG chance. The Fernandes goal for Utd is 0.03. Ronald's goals were a 0. 89 and 0.30. Newcastle's goal was a 0.26. Rewatch the goals for context.

Our team cG for the game was around 2.7.
If you have 10 shots all of 0.1 xG you expect to score 1 goal.. i.e 9 will not go in, but 1 on average willl. That is the whole point of xG rating per attempted shot. The better the chance the more likely it is on average to be scored.


That is the overall cumulative effect of all the percentage chances of scoring from every shot we had in the Norwich game on a running total of xG for the game for the team. We scored 1 , so we were well below expectations based on the chances we had.

To give more, the chance ESR had when he got a shot away from just around the edge of the box was an xG of 0.11.. i.e. an 11% chance of going in based on the norm. It's a chance but more often that not it will be missed but if he had 9 of those chances, 1 would be likely to go in on average.

Again, I challenge you to find a game where 0.1 xG chance has been scored or look again the low 0.1 chances missed and show me where it's an example of bad finishing.

[jayramfootball]

Majority of the shots are 0.1 for xG. Those are poor chances with a low probability of scoring. We controlled the game, dominated and had the volume of shots, but only 6 shots on target, a lot of blocked shots and the best chance we missed fell to Pepe. We still have a long way to go when it comes to our creativity.

That is what the data says. yes, except you missed our poor finishing.
Lots of chances created, not enough good ones (we rank 10th in the league for that), and poor finishing meaning we scored 1 instead of the 4-5 expected goals we should have from the chances created so far this season. Even those 0.1 chances are converted at, well, 10% on average and we're not converting any of them. Never mind the fewer better chances we are missing.

I reckon we will rise in team xG over the coming weeks because right now we're 10th best in the league having played Chelsea and Man City .. we'll be up to 6th or so in a few games in my prediction - but we'll see.

https://understat.com/league/EPL

Look through the weekend fixtures and highlight the goals scored from a 0.1 xG score or less? It's rare but look it up. Those are wonder goals of individual brilliance that make the goal. Not the pass.

If such a low xG score is a good chance, that would also mean Norwich had some good chances to win but were let down by poor finishing. That's far from the truth. Every shot isn't a good shot. But again, open to anyone, look up the 0.1 xG chances that were scored, watch Match of the Day, or just go to Arsenal Player and rewatch the game to see the chances we created.

No one said a 0.1. xG was a good chance - it is a chance that has about a 10% chance of being scored.
Stop trying to muddy the waters.

The data is the data - we had 30 shots that should have yielded 2-3 goals against Norwich.

Not trying to muddy the waters, just helping people understand xG. The data is the data I haven't seen where you get this idea that 30 shots should have yielded 2-3 goals. It doesn't work like that if you're creating low xG chances.

The one clear cut chance we should have scored but didn't was the 56 minute one from Pepe. If you watch that back, that was a 0.56 chance and nothing to do with bad finishing, just excellent defending from William to block it.

But again, anyone in fact, look up a 0.1 xG or less goal in match of the day or rewatch our game to get an idea of how xG works. That 3rd Lukaku goal is a 0.1 xG chance. The Fernandes goal for Utd is 0.03. Ronald's goals were a 0. 89 and 0.30. Newcastle's goal was a 0.26. Rewatch the goals for context.

Our team cG for the game was around 2.7.
If you have 10 shots all of 0.1 xG you expect to score 1 goal.. i.e 9 will not go in, but 1 on average willl. That is the whole point of xG rating per attempted shot. The better the chance the more likely it is on average to be scored.


That is the overall cumulative effect of all the percentage chances of scoring from every shot we had in the Norwich game on a running total of xG for the game for the team. We scored 1 , so we were well below expectations based on the chances we had.

To give more, the chance ESR had when he got a shot away from just around the edge of the box was an xG of 0.11.. i.e. an 11% chance of going in based on the norm. It's a chance but more often that not it will be missed but if he had 9 of those chances, 1 would be likely to go in on average.

Again, I challenge you to find a game where 0.1 xG chance has been scored or look again the low 0.1 chances missed and show me where it's an example of bad finishing.

Fella - any long shot from about 25 yards is going to be a goal scored from a shot that had a 0.1xG - or below (expected 10% of those types of shots going in).
Any single shot from distance (or a difficult circumstance) is not a bad finish if it misses, but if you keep on shooting from distance and score NONE after 15, 20,30,40+ attempts, then you are a shit finisher compared to your peers.

We , this season have had 59 shots with an expected xG of 4.46 goals.. i.e. our average per shot xG is 0.076 . Instead of scoring 4-5 goals, we've scored 1.. so our shooting is pants. The table at the top of this thread shows every team and their expected goals vs their actual goals. Some are performing better than the expectation, others like us, are not.

Here is the current table again


Now one thing i will say is that after 4 games the ranking means far less than it will after 15-20 games because a higher number of shots will make the averages per team more stable - hence this will be updated after every game.