Set Pieces

[Pirate55]

So you are happier with our corners this season and think that we are getting better but we have scored less goals from corners this season than last.

Was the last sentence an attempt at trolling? Or do you have inside knowledge on what we work on in training?

No, I can see an improvement in the variation in our corners only recently ( apart from the last two games). This hasn't been the case all season.

If that improved variation continues, I can see us scoring more goals from corners - with hopefully one on Sunday.

 

[braineclipse]

Look at the last graph. The team that had the most corners (20) apparently scored zero goals.

The authors made no claim based on that one particular piece of information. Most likely it's a statistical outlier resulting from a small sample size. Read what they claimed in the text...

As far as I can tell your disputing the statistical analysis of people who know a lot more about statistical analysis than you do. Based on picking out a particular piece of information that doesn't make sense to you. This it's frequently referred to as anomaly hunting and it's quite common, but nowhere near a valid argument.

If you're going to disagree with the claims made in that article at least have the common courtesy to disagree with what they actually claim.

Not trying to sound patronizing here. But it seems like you have a rather limited understanding of statistics. Yet you assume fyi know better than the people who wrote that and to be able to point out their errors.

I just don't buy that there is a finite time on the training ground which doesn't enable us to work on corners as some argue.

There is infinite time available on the training ground?

Why not work more on all the other aspects of our game instead with all that extra time you claim is there? Why corners especially? On what basis is that the best choice for us.

Personally thinking back to the Palace game and many other games I'm quite happy that it seems like Eriksen has prioritized working on his long range shooting rather than his corner taking. It seems to be paying off.

Edit: putting is undeniably a vital skill in golf. Being very good at corners is demonstrably not a vital skill in football. I think your analogy breaks down at that point.

 

[parklane1]

I am waiting for someone to bring up about how poor we are at free kicks as well as corners. Some get over focused on both these ways of scoring goals and in all reality very few are successful. As for stats well we all know they can be misleading in most cases.

 

[Hemant]

The authors made no claim based on that one particular piece of information. Most likely it's a statistical outlier resulting from a small sample size. Read what they claimed in the text...

As far as I can tell your disputing the statistical analysis of people who know a lot more about statistical analysis than you do. Based on picking out a particular piece of information that doesn't make sense to you. This it's frequently referred to as anomaly hunting and it's quite common, but nowhere near a valid argument.

If you're going to disagree with the claims made in that article at least have the common courtesy to disagree with what they actually claim.

Not trying to sound patronizing here. But it seems like you have a rather limited understanding of statistics. Yet you assume fyi know better than the people who wrote that and to be able to point out their errors.



There is infinite time available on the training ground?

Why not work more on all the other aspects of our game instead with all that extra time you claim is there? Why corners especially? On what basis is that the best choice for us.

Personally thinking back to the Palace game and many other games I'm quite happy that it seems like Eriksen has prioritized working on his long range shooting rather than his corner taking. It seems to be paying off.

Edit: putting is undeniably a vital skill in golf. Being very good at corners is demonstrably not a vital skill in football. I think your analogy breaks down at that point.

Very interesting and technical discussion.

There are a lot of statistical techniques around and very easy to generate all manner of figures and inferences from them. In the medical world it is complicated enough (where cause and effect are better established) but in the sporting world there are other even more complicating factors, such variation due to human behaviour, including irrational behaviour, fashion changes, e.g. Attacking players lining up in a clump at corners then going to different positions, strength of selected players e.g. swansea can be vulnerable in the air if Llorente or Fer are not available but they can be more potent if they are. More important in this context is how players read the trajectory of the ball, and predict where they should go.

Statistical analysis is ok for mathematicians and sports scientists but sometimes you have to go with the evidence of your eye and adapt on the pitch for the peculiar circumstances of the game situation.

This is what allows all of us non-einsteins to play and comment.

 

[scaramanga]

A small improvement here can make a big difference in terms of results.

No it won't. This is a perfect example of how little you understand of the underlying numbers and how to analyse them.

A small improvement would do nothing, if we made a 100% increase in our conversion rate (a massive one) then we'd have scored 3 more goals across the season so far, a 4.3% increase in goals scored, or an extra goal every 8 or so matches. There's every chance that none of those goals would have lead to an extra point.

That's before we even consider the increased risk that taking long corners causes.

 

[scaramanga]

Very interesting and technical discussion.

There are a lot of statistical techniques around and very easy to generate all manner of figures and inferences from them. In the medical world it is complicated enough (where cause and effect are better established) but in the sporting world there are other even more complicating factors, such variation due to human behaviour, including irrational behaviour, fashion changes, e.g. Attacking players lining up in a clump at corners then going to different positions, strength of selected players e.g. swansea can be vulnerable in the air if Llorente or Fer are not available but they can be more potent if they are. More important in this context is how players read the trajectory of the ball, and predict where they should go.

Statistical analysis is ok for mathematicians and sports scientists but sometimes you have to go with the evidence of your eye and adapt on the pitch for the peculiar circumstances of the game situation.

This is what allows all of us non-einsteins to play and comment.

Our brains don't tell us what happened, they tell us what we think happened after what we see has gone through the filter of all our biases and preconceptions.

 

[Pirate55]

No it won't. This is a perfect example of how little you understand of the underlying numbers and how to analyse them.

A small improvement would do nothing, if we made a 100% increase in our conversion rate (a massive one) then we'd have scored 3 more goals across the season so far, a 4.3% increase in goals scored, or an extra goal every 8 or so matches. There's every chance that none of those goals would have lead to an extra point.

That's before we even consider the increased risk that taking long corners causes.

And that is where you seem not to understand simple arithmetic. Even an extra goal every 8 matches, could lead to an extra 3 goals over the season ( to take your figure ) . Those three goals could have been scored against Leicester or Liverpool at home or Bournemouth or Sunderland away. Thus we would have converted draws into wins in three matches. 6 more points and we would be leading the PL by two points. Now do you understand?

 

[billyiddo]

That'd be great if you could choose wich corners you score goals from but alas I don't think that's possible, even if we were to train corners all day every day

 

[AuroRaman]

Our brains don't tell us what happened, they tell us what we think happened after what we see has gone through the filter of all our biases and preconceptions.

 

[scaramanga]

And that is where you seem not to understand simple arithmetic. Even an extra goal every 8 matches, could lead to an extra 3 goals over the season ( to take your figure ) . Those three goals could have been scored against Leicester or Liverpool at home or Bournemouth or Sunderland away. Thus we would have converted draws into wins in three matches. 6 more points and we would be leading the PL by two points. Now do you understand?

I understand, you're just wrong.

It's a false assumption that those goals would lead to points, mainly because we've won a high proportion of our matches. That makes it quite likely that these goals would come in matches where we already have 3 points.

This figure is then skewed even more by the fact that those goals are most likely to be scored in the matches where we get a lot of corners. We get lots of corners in the matches we dominate most and in the games we dominate most we tend to take away how many points?

Look, it get it. Stats are tough. It's not easy to understand how to properly apply numbers like this - that's why so many people who can't properly understand them think they're useless.

So you'll just have to trust me (or listen to any of the others here who know what they're talking about) when I tell you that we would have to massively increase our corner conversion rate to make a barely noticeable improvement in goals scored, that those goals scored are unlikely to have resulted in an increase in points and that doing so would have increased the likelihood of us conceding goals.

 

[Pirate55]

The authors made no claim based on that one particular piece of information. Most likely it's a statistical outlier resulting from a small sample size. Read what they claimed in the text...

As far as I can tell your disputing the statistical analysis of people who know a lot more about statistical analysis than you do. Based on picking out a particular piece of information that doesn't make sense to you. This it's frequently referred to as anomaly hunting and it's quite common, but nowhere near a valid argument.

If you're going to disagree with the claims made in that article at least have the common courtesy to disagree with what they actually claim.

Not trying to sound patronizing here. But it seems like you have a rather limited understanding of statistics. Yet you assume fyi know better than the people who wrote that and to be able to point out their errors.



There is infinite time available on the training ground?

Why not work more on all the other aspects of our game instead with all that extra time you claim is there? Why corners especially? On what basis is that the best choice for us.

Personally thinking back to the Palace game and many other games I'm quite happy that it seems like Eriksen has prioritized working on his long range shooting rather than his corner taking. It seems to be paying off.

Edit: putting is undeniably a vital skill in golf. Being very good at corners is demonstrably not a vital skill in football. I think your analogy breaks down at that point.

Before you cast meaningless assertions, as an Economist I read statistics as part of my undergraduate course at one of the leading Universities in the world. As Hemant said above, statistics only have limited usefulness in this context, due to definitions taken and multiple other variables . Anyone who has studied statistics should know to be wary of their applicability and always adopt a healthy scepticism - especially when they don't agree with other hypotheses.

Regarding your other point, while time for training is indeed not infinite, I should more accurately have stated that it is not finite either. How long do you think it would take to practice and implement say three or four differnt corner variations? We are talking about professional footballers here who have how many hours a day available?

 

[scaramanga]

as an Economist...... one of the leading Universities in the world.

Or

....statistics only have limited usefulness ........ Anyone who has studied statistics should know to be wary of their applicability and always adopt a healthy scepticism - especially when they don't agree with other hypotheses.

Pick one.

 

[milo]

as an Economist

 

[Pirate55]

And therein ends this conversation.

 

[Pirate55]

Pick one.

When you demonstrate you understand the difference between 'could' and 'would'.

Can't be bothered to discuss this with those that can't comprehend plain English.

 

[scaramanga]

When you demonstrate you understand the difference between 'could' and 'would'.

Can't be bothered to discuss this with those that can't comprehend plain English.

If we're working on "could", then why don't we suggest that working with Hugo more "could" lead to scoring direct from goal kicks?

 

[milo]

And therein ends this conversation.

I don't wish to be rude but you don't understand the basics of statistical analysis.

 

[Pirate55]

I don't wish to be rude but you don't understand the basics of statistical analysis.

And you don't seem to understand that football is not a science capable of analytical analysis. If it was, we would all be bored senseless. End of.

 

[90291Spur]

And you don't seem to understand that football is not a science capable of analytical analysis. If it was, we would all be bored senseless. End of.

This board is not capable of statistical analysis either but there is a high probability of verbal fisticuffs most of the time. One more sleep gents, until the day when, if we beat the goons, they cannot (statically) finish above us.

 

[milo]

And you don't seem to understand that football is not a science capable of analytical analysis. If it was, we would all be bored senseless. End of.

I rarely think of stats when I am watching a game but understanding how to use data helps me understand the game better.

The argument that you have failed to make for over a year now is exactly the kind of game situation that is well suited to statistical analysis.

You have been presented with plenty of large data sets that show that you are wrong. If you understood them you would have given up on this hopeless argument tonks ago.