I know how many hate anything statistical, but Stoke currently is under-performing compared to xG more than any other club. They are not doing great, but it may be down partially to bad luck. I say stick with him and let things sort out. PS- Note that we are most over-performing, after Swansea.
I asked this on another thread but never got a reply.......please can someone explain what "expected goals" means. I see it on MOTD & it completely baffles me.
Expected goals is the chance of scoring. Its out of 1. For example, striker right in front of an open goal, xG will basically be 1. Striker at a dodgy angle 15 yards away, xG will be 0.1
It's calculated by analysing thousands of shots and goals to see the likelyhood of scoring. A striker with a lower xG to goals scored means they are over performing. Pukki last year had an xG lower than his actual tally, meaning he scored more than he should have.
Simon Church in front of open goal, 10 yards out ... xG = 0.01
I know how many hate anything statistical, but Stoke currently is under-performing compared to xG more than any other club. They are not doing great, but it may be down partially to bad luck. I say stick with him and let things sort out. PS- Note that we are most over-performing, after Swansea.
I asked this on another thread but never got a reply.......please can someone explain what "expected goals" means. I see it on MOTD & it completely baffles me.
Like I posted in that other thread, here’s a video that explains it neatly.
I know how many hate anything statistical, but Stoke currently is under-performing compared to xG more than any other club. They are not doing great, but it may be down partially to bad luck. I say stick with him and let things sort out. PS- Note that we are most over-performing, after Swansea.
I asked this on another thread but never got a reply.......please can someone explain what "expected goals" means. I see it on MOTD & it completely baffles me.
This video explains it really well. Discusses the pros and cons. But what statisticians have figured out is that "genius" is not what fans thought it was. Data show that Messi, for example, is not a "great finisher" but is actually just good at shooting from optimal positions and that he gets there more than your average striker. xG shifts what genius... is. Clubs are using xG to maximize their chances at scoring. This is why there are less shots from outside the box now... xG shows it is a poor strategy, thus more clubs are training players to get into the box, before shooting. I doubt there is one club right now with any level of success that does not use xG regularly. Including ours.
Contrary to popular belief, it is not a load of meaningless twaddle.
Tifo Football is my favorite YouTube football site. A must watch. Since you know about it, do you have other sites like it that I should be checking out, Callum?
I know how many hate anything statistical, but Stoke currently is under-performing compared to xG more than any other club. They are not doing great, but it may be down partially to bad luck. I say stick with him and let things sort out. PS- Note that we are most over-performing, after Swansea.
Its in the presentation, not just stats. Eg. Expected goals is a load of bullshit, doesn't mean anything at all. All that matters is the actual result regardless how many goals were expected. One of the most stupid stats I have ever read
I know how many hate anything statistical, but Stoke currently is under-performing compared to xG more than any other club. They are not doing great, but it may be down partially to bad luck. I say stick with him and let things sort out. PS- Note that we are most over-performing, after Swansea.
Its in the presentation, not just stats. Eg. Expected goals is a load of bullshit, doesn't mean anything at all. All that matters is the actual result regardless how many goals were expected. One of the most stupid stats I have ever read
Utterly idiotic non point of view. Pukki is the case study that proves XG can work and maximise resources. Norwich signed him because he had a higher conversion rate XG for certain positions in the box. Norwich created a proliferation of those sort of chances. Pukki's goalscoring record wasn't always great, quite good at Brondby, Norwich sign him for free. He then starts converting the chances Norwich make and is prolific. Seems like you'd sign forwards and defenders without any understanding of the structure they'll play in, or any care to adapt your understanding.
Brentford turned themselves from a poor to average League One outfit into a good Championship side basically by using stuff like xG a lot. Other clubs are catching on. I'm sure we use something similar - why sign Lockyer, for example, from a team that let in ten more goals than us last season?
Brentford turned themselves from a poor to average League One outfit into a good Championship side basically by using stuff like xG a lot. Other clubs are catching on. I'm sure we use something similar - why sign Lockyer, for example, from a team that let in ten more goals than us last season?
Don't know, maybe Gallen and Bowyer watched Lockyer play football.
Whenever we sign a player Gallen and Bowyer talk about his strengths and style of play. They have never IIRC mentioned his XG stats.
That's not to say that they don't use them but they are an aid to analysis not the wonderdrug that some make them out to be.
Does anyone think Brentford just looked at Konsa's stats or that Norwich never watched Pukki play?
Charlton have a stats board in the dressing room, I'll see if I can find my photo of it.
If you hit a worldie into the top corner then its a goal, what's the expectation of that?
Impossible to judge. I dont care how many goals you have watched historically, you cant rely on that alone.
I dont think any of the models on Xg take into account each individual player across their whole career and THEN apply the objective historical data. Believe it is mapped on a judgment of goals scored but then applied universally to all future goals..doesnt take into account the subjective player, its a Bot player used as an average.
But if you are Claus Jensen you find chipping the keeper easier than a one on one. If you are Carl Leaburn your head is more a weapon than your foot.
This Stoke example is nice to prove the importance of the context any statistical analysis.
Parachute payments, high earners, large squad, Loads of good football potential but get beat every week and get relegated to League 1
Xg will tell you thats unlikely but context will tell you thats 1/3 teams that drop out of PL.
Im a big fan of data but football doesnt lend itself to objectivity in any way shape or form.
Whats the Xg on a penno to win the World Cup? Dont think you can find the answer just by looking at all the shoot outs in the past - its only on that day, in that moment.
If you hit a worldie into the top corner then its a goal, what's the expectation of that?
Impossible to judge. I dont care how many goals you have watched historically, you cant rely on that alone.
I dont think any of the models on Xg take into account each individual player across their whole career and THEN apply the objective historical data. Believe it is mapped on a judgment of goals scored but then applied universally to all future goals..doesnt take into account the subjective player, its a Bot player used as an average.
But if you are Claus Jensen you find chipping the keeper easier than a one on one. If you are Carl Leaburn your head is more a weapon than your foot.
This Stoke example is nice to prove the importance of the context any statistical analysis.
Parachute payments, high earners, large squad, Loads of good football potential but get beat every week and get relegated to League 1
Xg will tell you thats unlikely but context will tell you thats 1/3 teams that drop out of PL.
Im a big fan of data but football doesnt lend itself to objectivity in any way shape or form.
Whats the Xg on a penno to win the World Cup? Dont think you can find the answer just by looking at all the shoot outs in the past - its only on that day, in that moment.
You just proved the usefulness of xG without realising it...
Only by using the “average” as a baseline can you use xG to effectively measure individual player performance.
No idea how/if we use expected goals, but under Bowyer we rarely shoot from long range or swing in hopeful crosses from deep positions. We also look to dribble or pass our way into the box and from there we put in low crosses/passes and win penalties.
It's partly about playing attractive football but also those type of chances being more effective.
A few of our players have talked about Bowyer encouraging them to keep getting in the box as things will fall for them. They don't have to be the best finishers to score the type of chances we create.
You can see we have this way of playing in the goals we scored last season:
Brentford turned themselves from a poor to average League One outfit into a good Championship side basically by using stuff like xG a lot. Other clubs are catching on. I'm sure we use something similar - why sign Lockyer, for example, from a team that let in ten more goals than us last season?
Don't know, maybe Gallen and Bowyer watched Lockyer play football.
Whenever we sign a player Gallen and Bowyer talk about his strengths and style of play. They have never IIRC mentioned his XG stats.
That's not to say that they don't use them but they are an aid to analysis not the wonderdrug that some make them out to be.
Does anyone think Brentford just looked at Konsa's stats or that Norwich never watched Pukki play?
Charlton have a stats board in the dressing room, I'll see if I can find my photo of it.
I hope Barry Fry still uses his Rothmans Yearbook. Whatever Peterborugh do they're good at unearthing Non-League and League strikers: What people miss with Peterborough is they invest and are willing to lose a fair bit on signing multiple strikers, with many failures along the way. There should always be multiple evaluations going on for a player and squad assembly.
Every method should evolve, be re-interpreted, looked at again when it reaches heavy adoption by others, but most importantly combined with individuals experience and knowledge. I don't think any of us respect or want to go back to Pardew's addiction to signing strikers, without a strategy on how they'd play or how the team could supply them. Fortunately we have a management team that have a structure, have experience and talent in both recruiting and coaching for that structure. It would be a bit pointless if Pukki was identified by Norwich, but say like Piotor Parzysek physically couldn't meet the challenges of the Championship right away: Even more so in a small squad.
I'm not sure I see enough cohesion, especially defensively, at Stoke in the few times I've seen them. Having said that it's madness to sack him in the first few months of the season after he's assembled his squad. Give him to early November unless he does a Pardew.
"Nathan Jones will go down as the most abysmal appointment in Stoke’s history"
"Nathan Jones is starting to remind me of that bit in Father Ted where they try to tap a little dent out of the car and end up with a crumpled wreck."
"Nathan Jones, you’ve been here too long"
"Nathan jones said ‘now is the time to judge me’ he’s give us:
3 wins in 28, 1 point out of a possible 18, -9 goal difference. Bottom of the league. He’s got go!!!"
"cant even look at today with any positive thoughts, what formation is that, we're bottom of the league and with the squad we have he plays that I was in shock before the game. He's building the gallows himself!"
Comments
Contrary to popular belief, it is not a load of meaningless twaddle.
Players would have been on a high when their winning penalty was scored.
Its in the presentation, not just stats. Eg. Expected goals is a load of bullshit, doesn't mean anything at all. All that matters is the actual result regardless how many goals were expected. One of the most stupid stats I have ever read
Whenever we sign a player Gallen and Bowyer talk about his strengths and style of play. They have never IIRC mentioned his XG stats.
That's not to say that they don't use them but they are an aid to analysis not the wonderdrug that some make them out to be.
Does anyone think Brentford just looked at Konsa's stats or that Norwich never watched Pukki play?
Charlton have a stats board in the dressing room, I'll see if I can find my photo of it.
If you hit a worldie into the top corner then its a goal, what's the expectation of that?
Impossible to judge. I dont care how many goals you have watched historically, you cant rely on that alone.
I dont think any of the models on Xg take into account each individual player across their whole career and THEN apply the objective historical data. Believe it is mapped on a judgment of goals scored but then applied universally to all future goals..doesnt take into account the subjective player, its a Bot player used as an average.
But if you are Claus Jensen you find chipping the keeper easier than a one on one. If you are Carl Leaburn your head is more a weapon than your foot.
This Stoke example is nice to prove the importance of the context any statistical analysis.
Parachute payments, high earners, large squad,
Loads of good football potential but get beat every week and get relegated to League 1
Xg will tell you thats unlikely but context will tell you thats 1/3 teams that drop out of PL.
Im a big fan of data but football doesnt lend itself to objectivity in any way shape or form.
Whats the Xg on a penno to win the World Cup? Dont think you can find the answer just by looking at all the shoot outs in the past - its only on that day, in that moment.
Only by using the “average” as a baseline can you use xG to effectively measure individual player performance.
It's partly about playing attractive football but also those type of chances being more effective. A few of our players have talked about Bowyer encouraging them to keep getting in the box as things will fall for them. They don't have to be the best finishers to score the type of chances we create.
You can see we have this way of playing in the goals we scored last season:
Every method should evolve, be re-interpreted, looked at again when it reaches heavy adoption by others, but most importantly combined with individuals experience and knowledge. I don't think any of us respect or want to go back to Pardew's addiction to signing strikers, without a strategy on how they'd play or how the team could supply them. Fortunately we have a management team that have a structure, have experience and talent in both recruiting and coaching for that structure. It would be a bit pointless if Pukki was identified by Norwich, but say like Piotor Parzysek physically couldn't meet the challenges of the Championship right away: Even more so in a small squad.
I'm not sure I see enough cohesion, especially defensively, at Stoke in the few times I've seen them. Having said that it's madness to sack him in the first few months of the season after he's assembled his squad. Give him to early November unless he does a Pardew.
Do they give him more time or make a change over the international break?
"Nathan Jones will go down as the most abysmal appointment in Stoke’s history"
"Nathan Jones is starting to remind me of that bit in Father Ted where they try to tap a little dent out of the car and end up with a crumpled wreck."
"Nathan Jones, you’ve been here too long"
"Nathan jones said ‘now is the time to judge me’ he’s give us: 3 wins in 28, 1 point out of a possible 18, -9 goal difference. Bottom of the league. He’s got go!!!"
"cant even look at today with any positive thoughts, what formation is that, we're bottom of the league and with the squad we have he plays that I was in shock before the game. He's building the gallows himself!"
Oh yeaaa... Effectively three weeks down the line and Huddersfield are no closer to appointing someone