Foxxed Posted 7 September 2018 Posted 7 September 2018 2 hours ago, 80's fox said: I'm not sure you fully understand what expected goals is. This will explain it for you https://www.bbc.co.uk/sport/football/40699431 I've learnt something, I think.... So if your xG is low you've been in some good positions to score but fluffed them. If it's higher you're scoring some goals which are not obviously easy. So the xG shite is "Are you fluffing good chances"? And if you are fluffing good changes then the team should probably stop fluffing good chances. So when someone says "They've got a very low xG" I should just translate it to "Maybe they should buy some new strikers"? Educational!
Lesta Legend Posted 7 September 2018 Posted 7 September 2018 Thank God we'll get our top striker back soon, so kinda makes worrying about such a stat pointless, just yet....
lildave3 Posted 7 September 2018 Posted 7 September 2018 3 hours ago, Phube said: According to xG Burnley were over performing all season... but still managed to stay in 7th. I mean..they literally did over perform... 3 hours ago, Izzy Muzzett said: How on earth did we manage in the old days before we had to worry about expected goal ratio's? What a really weird thing to say. It’s not about managing it’s about collecting new information and improving the way we think about things. I love the analytical side of things and think it has a place in any sport. Not a coincidence that the most heavy analytics based clubs are thriving right now.
Ricey Posted 7 September 2018 Posted 7 September 2018 1 hour ago, Foxxed said: I've learnt something, I think.... So if your xG is low you've been in some good positions to score but fluffed them. If it's higher you're scoring some goals which are not obviously easy. So the xG shite is "Are you fluffing good chances"? And if you are fluffing good changes then the team should probably stop fluffing good chances. So when someone says "They've got a very low xG" I should just translate it to "Maybe they should buy some new strikers"? Educational! Not quite. It's not about fluffing chances, it's more an indication of quality of chances created. xG is basically how many goals a team would be expected to score based off the positions on the pitch their chances occurred in. The higher the xG the more and/or better chances have been created.
Wymsey Posted 7 September 2018 Posted 7 September 2018 Not sure how goals can be 'expected' really. For example, if the team is generating a lot of opportunities against a weaker team but their goalkeeper is preventing the ball from hitting the back of the net - not sure how you can measure which of those chances should have been goals. Statistics etc can help to evaluate performances and trends etc; but football is a lottery and the amount of goals, like how many yellow or red cards will be distributed during a game, can't be estimated.
RumbleFox Posted 7 September 2018 Posted 7 September 2018 Well reading Foxestalk's greatest minds trying to understand this is fun. Carry on please. X
Lambert09 Posted 7 September 2018 Posted 7 September 2018 2 hours ago, 80's fox said: Okay I get your point. However our xG against Wolves was 0.20, however you want to portray it, we was fortunate. For sure, we didn’t really deserve to win that day and nobody needed the xG to see it. But I was thinking more in general terms. I do think it’s a good measurement and great addition to the usual stats but I wont be worrying too much over a bad score on it. Nor getting too excited if we achieve a high score. But I do think the halfway stage will be a good indication of which teams performed well. I just care that we are scoring more than conceding, any other stat is just what if’s.
Izzy Posted 7 September 2018 Posted 7 September 2018 1 hour ago, lildave3 said: I love the analytical side of things and think it has a place in any sport. Not a coincidence that the most heavy analytics based clubs are thriving right now. I'm sure you're right mate. But as an aging dinosaur I struggle to keep up with all the changes and advancements in the modern game. Fair play to those who love and understand the analytical side of it, but I'm a bit thick and it all goes over my head really
MC Prussian Posted 7 September 2018 Posted 7 September 2018 Expected Goals is a rubbish invention. Period. It's a measure attempting to describe the indescribable. Can't see the use for it and won't ever buy into the hype. Football is turning into a sport invented by Americans obsessed with stats. What next? Expected fouls?
Izzy Posted 7 September 2018 Posted 7 September 2018 1 minute ago, MC Prussian said: Expected Goals is a rubbish invention. Period. It's a measure attempting to describe the indescribable. Can't see the use for it and won't ever buy into the hype. Football is turning into a sport invented by Americans obsessed with stats. What next? Expected fouls? Amen brother
Dan Posted 7 September 2018 Posted 7 September 2018 5 hours ago, Lambert09 said: No I get it. But my point still stands. As an example, if a game is 0-0 but one team is happy to take a point they may not be trying to create goal scoring opportunities and will happily try and absorb some pressure which would allow the other side more attempts to create goal scoring opportunities. Which they would be scored on. But if that side conceded in the first 10 minutes, the rest of the game they would have a different approach and would be attempting to create chances, whilst the other team now with a lead might be more defensive and happy to try and absorb pressure. How many times have we been against a side that doesn’t try to score and is happy to just limit us but as soon as we get one suddenly spring to action and start attempting to actually score? My point being a team facing united and Liverpool and having a 2 goal lead against a side like wolves, is bound to create less chances. Against wolves we didn’t need another goal so you can understand why we wouldn’t be creating the goal scoring chances you get good scores for in this measurement. Yes we rode our luck at times but this stat doesn’t tell the full story for teams abilities. Good example was us against Wolves really. We scored twice from two quite lucky scenarios so didn't really need to go all out attacking against them when we had a 2-0 lead.
Stadt Posted 7 September 2018 Posted 7 September 2018 The data has came from hundreds of thousands of shots and is used by practically every top club in Europe but no, you, the foxestalker is much more knowledgeable about football analytics than people that have careers in it
KingsX Posted 8 September 2018 Posted 8 September 2018 3 hours ago, Lambert09 said: I just care that we are scoring more than conceding, any other stat is just what if’s. 7 hours ago, Izzy Muzzett said: Jeez my head hurts. Sorry, but what a load of old pony 6 hours ago, foxes_rule1978 said: Aren’t these the statistics the spurs fans worry about while we just get on with actual figures 4 hours ago, Kopfkino said: It's such a terrible stat Goals are obviously the number that counts. But would anyone here really argue they are the only number that shows which players and teams played effectively, independent of chance? Quantity and quality of chances are a huge deal -- the meat of discussion on here whether xG is cited or not. Clubs that learned to use advanced metrics ahead of others, Including Leicester City Football Club, were able to recruit better and reaped the rewards. So many articles two years ago were crediting the Walsh and Pearson team as early and effective adopters, partly explaining how such talent was assembled under everyone else’s noses. If as a fan, you don’t care to pay attention to these measures, fine. It’s your choice, and we should all enjoy the footy in our own way. But this passive-aggressive bitching about their very existence, is a poor look. If your club’s staff felt the same way, they would still be in the Championship.
Guest Kopfkino Posted 8 September 2018 Posted 8 September 2018 14 hours ago, KingsX said: Goals are obviously the number that counts. But would anyone here really argue they are the only number that shows which players and teams played effectively, independent of chance? Quantity and quality of chances are a huge deal -- the meat of discussion on here whether xG is cited or not. Clubs that learned to use advanced metrics ahead of others, Including Leicester City Football Club, were able to recruit better and reaped the rewards. So many articles two years ago were crediting the Walsh and Pearson team as early and effective adopters, partly explaining how such talent was assembled under everyone else’s noses. If as a fan, you don’t care to pay attention to these measures, fine. It’s your choice, and we should all enjoy the footy in our own way. But this passive-aggressive bitching about their very existence, is a poor look. If your club’s staff felt the same way, they would still be in the Championship. Don't dispute any of that. It's more the fact most people use it completely incorrectly that makes it daft, ie for individual matches or even using it for just 4 games. For clubs using it as part of a set of data fine, for people using it not exclusively but with context fine, for people doing real statistical analysis fine. There's too many variants based on different assumptions and some daft r-squared's for us mere mortals to make use of it as a statistic to be used alone. That's why I think it's terrible.
treer Posted 8 September 2018 Posted 8 September 2018 its all garbage at the end of the day someone has to decide when a chance is a a chance. Many differing opinions on what constitutes a chance, so there can never be a definitive xg. As it is the ball goes in the onion bag, the ref decides if it should count, and bingo, you got a goal.
RumbleFox Posted 17 December 2018 Posted 17 December 2018 I know some people think "it is all garbage" but I think they slightly miss the point of statistics. Of course they are not the be-all and end-all. Of course there are outliers but they are an interesting tool to look at and to say they have no use at all is going too far. With that in mind, I was just looking at the stats so far this season.... All teams have currently played 17 games (almost half way through the season) and we currently sit in 12th position. Our "expected goals for" are 20.96. This puts us 13th in the table in terms of how many goals our team has been "expected to score". For a team of our investment I can understand why some people are not happy with this and we do seem to not be creating quite enough goal scoring opportunities (as most of you will know I am a "Puel In" poster and think he still needs time but I am fully able to be objective and look at things from both sides). Our "expected goals against" are 19.01. This puts us 5th in the table in terms of how many goals our team has been "expected to concede". So in terms of a defensive unit we do seem to be quite solid. The question is can we keep this defensive strength whilst improving going forwards? In my honest opinion? I have no idea. So if we had scored all of our "expected goals" and conceded all of those "expected" to be scored against us where would we be? It seems we would be 8th.... So what does all this rambling mean? With all the recent talk (and 17 million threads) regarding Puel being sacked I just thought it might be interesting to have a more objective look. As I said, I think we should stick but I would never want to be accused of being blinkered. Of course stats mean nothing on their own but it does seem that according to goal chances created and conceded we "are about where we expect to be". In fact, if anything we "deserve" to be slighlty higher. This backed up my gut feeling as we seem to have conceded a few worldies and missed quite a few sitters. I think this shows that no matter where you stand on Puel he certainly is not "sh1t/terrible/the worst ever blah blah" or any of the other hyperbolic statements we see on here. If anything, he is "average". Now the question is should we put up with "average"? It depends where you stand but despite being behind Puel I can see why some want him out. My own opinion is that to be on course for a mid table season with the second youngest squad in the league, a change in style and with the tragic off-field events is probably acceptable and the hope would be we will improve. Sorry for long boring post I just hate all the non-constructive, non backed up, emotionally fuelled "opinions" on here. X
RumbleFox Posted 17 December 2018 Posted 17 December 2018 On a side note, just a quick shout out to Ghezzal. He has received some stick on here but when watching him my own personal "eye test" seemed to tell me he has performed pretty well. If we look at "Expected Assists" he is second behind Maddison.... This is despite having played fewer minutes than anyone else in the top ten on Leicester's list of "expected assist" providers. If this were extrapolated to "expected assists per 90 minutes" he sits top of the pile by quite a decent margin........ Of course this does not paint a full picture. He has usually been brought on when we are chasing games and so probably more likely to create chances. It is also difficult to maintain that level over 90 minutes. It does, however, make an interesting read and coupled with his one only real appearance "as a number 10" where he played very well I just thought it was a wee bit of good news amongst the doom and gloom on here. Given our inability to create enough goal scoring opportunities, could he be an option for a starting place again once fully fit? X
coolhandfox Posted 17 December 2018 Posted 17 December 2018 24 minutes ago, RumbleFox said: I know some people think "it is all garbage" but I think they slightly miss the point of statistics. Of course they are not the be-all and end-all. Of course there are outliers but they are an interesting tool to look at and to say they have no use at all is going too far. With that in mind, I was just looking at the stats so far this season.... All teams have currently played 17 games (almost half way through the season) and we currently sit in 12th position. Our "expected goals for" are 20.96. This puts us 13th in the table in terms of how many goals our team has been "expected to score". For a team of our investment I can understand why some people are not happy with this and we do seem to not be creating quite enough goal scoring opportunities (as most of you will know I am a "Puel In" poster and think he still needs time but I am fully able to be objective and look at things from both sides). Our "expected goals against" are 19.01. This puts us 5th in the table in terms of how many goals our team has been "expected to concede". So in terms of a defensive unit we do seem to be quite solid. The question is can we keep this defensive strength whilst improving going forwards? In my honest opinion? I have no idea. So if we had scored all of our "expected goals" and conceded all of those "expected" to be scored against us where would we be? It seems we would be 8th.... So what does all this rambling mean? With all the recent talk (and 17 million threads) regarding Puel being sacked I just thought it might be interesting to have a more objective look. As I said, I think we should stick but I would never want to be accused of being blinkered. Of course stats mean nothing on their own but it does seem that according to goal chances created and conceded we "are about where we expect to be". In fact, if anything we "deserve" to be slighlty higher. This backed up my gut feeling as we seem to have conceded a few worldies and missed quite a few sitters. I think this shows that no matter where you stand on Puel he certainly is not "sh1t/terrible/the worst ever blah blah" or any of the other hyperbolic statements we see on here. If anything, he is "average". Now the question is should we put up with "average"? It depends where you stand but despite being behind Puel I can see why some want him out. My own opinion is that to be on course for a mid table season with the second youngest squad in the league, a change in style and with the tragic off-field events is probably acceptable and the hope would be we will improve. Sorry for long boring post I just hate all the non-constructive, non backed up, emotionally fuelled "opinions" on here. X I think it show we have improved defensively, which is down to some what to good coaching, improving players and recruiting better players in Evans and Riccardo and maybe being a little less gun ho. We need to invest in quality players in wide areas, someone like a Anderson would help on end, one quality paler could make a massive difference
Duncan_K Posted 17 December 2018 Posted 17 December 2018 I don't mind a bit of stats analysis, but i always want to scream at these expected goals stats "GAME SITUATION, GAME SITUATION, GAME SITUATION". I don't believe we have a great defence at the moment and certainly not top 5 in the Prem, because far too often we concede the first goal, as we all know; 'Goals change games', and then we are up against it.
RumbleFox Posted 17 December 2018 Posted 17 December 2018 16 minutes ago, Duncan_K said: I don't mind a bit of stats analysis, but i always want to scream at these expected goals stats "GAME SITUATION, GAME SITUATION, GAME SITUATION". I don't believe we have a great defence at the moment and certainly not top 5 in the Prem, because far too often we concede the first goal, as we all know; 'Goals change games', and then we are up against it. Yep, fully agree but still worthy of looking at. (Not aiming this at you) but the anti-stat folk annoy me as they are so black and white. No one who likes stats claims they prove everything but they are certainly a useful tool and and interesting matrix to at least act as a springboard for a more informed discussion. As you say, we do concede first quite a lot which changes the complexion of the game. Still, I do believe we are about where our performances merit (mid table, somewhere between 8th and 12th). A bit of luck we could be 7th and a bit less luck we could be near the bottom. I honestly don't know quite where this team can go but it is certainly not all doom and gloom. X
Raw Dykes Posted 17 December 2018 Posted 17 December 2018 On 08/09/2018 at 16:33, treer said: its all garbage at the end of the day someone has to decide when a chance is a a chance. Many differing opinions on what constitutes a chance, so there can never be a definitive xg. As it is the ball goes in the onion bag, the ref decides if it should count, and bingo, you got a goal. I don't think it's garbage at all. It's an objective measure of how good a chance is, depending on a number of factors, including angle to goalmouth, distance from goal, number of defenders in the way, etc. I'm sure you'd agree that a 2 yard tap-in into an open net has a better chance of going in than a desperate long shot from the halfway line with 11 men behind the ball. There are reasons for this, and we know it's true, not just think it's probably the case. It's nothing to do with opinions. That's the point of it.
st albans fox Posted 17 December 2018 Posted 17 December 2018 One thing to add here is that the xg and stats said that Burnley were way over achieving last season and they had a long way to fall when things came back to ‘normal’. I expect you will find that out stats and cg when we won the league said we had a fair way to fall the follow season Also, note that spurs are over achieving this season - it may catch up with them over the remaining months or it may be next campaign ...
treer Posted 17 December 2018 Posted 17 December 2018 I suppose you could take your Xg preditions to the bookies, and use them against the odds offered, soon enough find out if the xg predictions were any good, or a waste of time
lildave3 Posted 17 December 2018 Posted 17 December 2018 10 minutes ago, treer said: I suppose you could take your Xg preditions to the bookies, and use them against the odds offered, soon enough find out if the xg predictions were any good, or a waste of time Bookies take these things into account when pricing games up.
RumbleFox Posted 17 December 2018 Posted 17 December 2018 2 hours ago, st albans fox said: One thing to add here is that the xg and stats said that Burnley were way over achieving last season and they had a long way to fall when things came back to ‘normal’. I expect you will find that out stats and cg when we won the league said we had a fair way to fall the follow season Also, note that spurs are over achieving this season - it may catch up with them over the remaining months or it may be next campaign ... Good points. Of course teams can make the stats look foolish, our title win is a perfect example but, in general, given enough time xg does seem a decent tool. It has flaws, especially as a predictive tool but those teams with higher xg and lower xga do tend to do well over a longer number of games. X
Recommended Posts
Archived
This topic is now archived and is closed to further replies.