Remove this ad

Lead

Jul 11 16 9:46 PM

Tags : :

Here are the economic results that deal with sport specific available population per game, this takes into account both population amount and wealth of each nation. Below is a list of all countries in the tournament as well as their soccer football specific available population in millions. This number determines the potential of people/resources contributing to the sport of soccer (directly or indirectly).

MATCHES WON BY THE COUNTRY WITH MORE SPORT SPECIFIC AVAILABLE POPULATION
GROUP STAGE: 14/25 (56%)
KNOCKOUT STAGE: 9/15 (60%)
TOTAL: 23/40 (58% victories for countries with more available population)

COUNTRIES (24)
Russia (80.9) 1st
Germany (53.3) 1st
France (42.6) 1st
Italy (39.5) 1st
England (34.5) 1st
Spain (31.3) 1st
Turkey (27.6) 1st
Ukraine (26.1) 1st
Poland (22.0) 1st
Romania (8.3) 1st
Belgium (7.3) 1st
Portugal (6.9) 1st
Sweden (6.2) 1st
Austria (5.7) 1st
Hungary (5.3) 1st
Slovakia (3.4) 1st
Czechia (3.2) 2nd
Switzerland (2.7) 2nd
Croatia (2.6) 1st
Northern Ireland (1.2) 1st
Wales (1.0) 2nd
Ireland (0.8) 3rd
Albania (0.7) 1st
Iceland (0.2) 1st
TOTAL: 413.3
AVERAGE: 17.2

Last Edited By: abramjones Jul 12 16 2:15 AM. Edited 1 time

Quote    Reply   
Remove this ad
Remove this ad
avatar

TheRoonBa

Posts: 5,442 Site Admin

#1 [url]

Jul 11 16 10:43 PM

Is the available population in millions or is it just an index figure (out of 100, or something else?).

I ask because only around 33% of England's male population would be between 15-40 years old (and thus available). And this would represent only around 9 million people. But there is a figure of 34.5 in your table (which, if in millions, is greater than the entire population of adult + non-adult males in England).


Or are females also included?  I wouldn't think they'd be relevant, as male and female football are generally separate.  And in any case, only 18 million males + females combined would be of football playing age (15-40) in England in total.

In any case, 

Last Edited By: TheRoonBa Jul 11 16 10:48 PM. Edited 1 time.

Quote    Reply   

#2 [url]

Jul 12 16 2:19 AM

TheRoonBa wrote:
Is the available population in millions or is it just an index figure (out of 100, or something else?).

I ask because only around 33% of England's male population would be between 15-40 years old (and thus available). And this would represent only around 9 million people. But there is a figure of 34.5 in your table (which, if in millions, is greater than the entire population of adult + non-adult males in England).

Or are females also included?  I wouldn't think they'd be relevant, as male and female football are generally separate.  And in any case, only 18 million males + females combined would be of football playing age (15-40) in England in total.

In any case, 
Good observation, and yes it is in millions (I have just fixed my OP to state that since you brought it up). Available population actually goes from 15 to 64 and includes females. The reason is because these people, though they can't play the sport, potentially economically contribute. So the amount of economically available females is very important (consider who is raising the child, and who driving the child to practice, who are going to watch the games and providing wealth for the sport etc... wealthy women and men, for that matter, have more ability to help). The reason why older people are important to (41-64) is because they still greatly contribute to the economic infrastructure of the sport (for instance, my father is a fanatical baseball, football (gridiron), and basketball fan as are many people his age). Certainly even older people can contribute also, but it is less likely as age increases because they don't generally don't work anymore or as much. It isn't a perfect calculation, it is just an estimate (enough for practical use in most cases). Age is also important to consider because in most impoverished countries there is a disproportionate amount of people under the age of 15, and this is a drain on the economic potential of a country.

Last Edited By: abramjones Jul 12 16 2:22 AM. Edited 1 time.

Quote    Reply   

#3 [url]

Jul 12 16 10:18 PM

abramjones wrote:
TOTAL: 23/40 (58% victories for countries with more available population)
58% is not the strongest correlation.  Mabel has a higher success rate at picking winners.

I think you should not count penalty shootout as victories - the convention for virtually all such comparisons/rankings is to count them as draws.

Quote    Reply   
avatar

TheRoonBa

Posts: 5,442 Site Admin

#4 [url]

Jul 12 16 11:39 PM

abramjones wrote:

Good observation, and yes it is in millions (I have just fixed my OP to state that since you brought it up). Ander the age of 15, and this is a drain on the economic potential of a country.

OK, and have you you used a multiplying factor for "poorer" countries? (For example, Turkey has around 55 million people from 15-64, but only 27 in your table - Russia has around 90 million, but only 80 in your table).

Quote    Reply   

#5 [url]

Jul 13 16 6:09 AM

nfm24 wrote:
abramjones wrote:
TOTAL: 23/40 (58% victories for countries with more available population)
58% is not the strongest correlation.  Mabel has a higher success rate at picking winners.

I think you should not count penalty shootout as victories - the convention for virtually all such comparisons/rankings is to count them as draws.
58% is actually moderately high (think of  a coin that has a 58% chance of going heads, that would greatly alter results), but keep in mind that the tournament average is 65% (this tournament happened to have more "economic upsets" than usual. For example, this year's Copa America had 72% victories for countries with more available populationAlso keep in mind that this doesn't account for extremes in size difference which would be even a higher rate (examples: countries the size of usa/china/russia/germany/italy/united kingdom/france/spain generally have an easier time beating in random sports countries the size of new iceland/angola/bolivia/bermuda than do countries the size of argentina/australia/turkey/poland and especially countries similar to the size of denmark/czechia/bolivia/egypt/tunisia). And of course, traditional prediction models should have higher accuracy or that would be really pathetic. and the experts should know which countries have better programs and how to calculate that into their predictions. wealth and population simply make it easier to establish better programs. wealth and population do not "know" what countries have better programs.

And I maybe shouldn't count those games, but this is just me dealing the shitty rules for scoring in football that I have already mentioned in a different post. so either way I go on this issue there are going to be issues with scoring in football, it is out of my hands. It also will not really effect the percentage on average in a negative way, it will simply provide more data to make the percentage slightly more accurate.

Last Edited By: abramjones Jul 13 16 6:38 AM. Edited 8 times.

Quote    Reply   

#6 [url]

Jul 13 16 6:10 AM

TheRoonBa wrote:
abramjones wrote:

Good observation, and yes it is in millions (I have just fixed my OP to state that since you brought it up). Ander the age of 15, and this is a drain on the economic potential of a country.

OK, and have you you used a multiplying factor for "poorer" countries? (For example, Turkey has around 55 million people from 15-64, but only 27 in your table - Russia has around 90 million, but only 80 in your table).

No I haven't. Available population counts all people living on the equivalent of at least $10 USD per day and in between ages of 15 and 64. It is not just about age.

Quote    Reply   
Remove this ad

#7 [url]

Jul 14 16 2:40 PM

> 58% is actually moderately high

I'd say 58% is not statistically significant and possibly within the error bars of the inputs to your comparison. But we are dealing with small data sets here so the idea of significance is a bit of a moot point anyway. In other words if a couple of marginal results had gone the other way (e.g. if Italy had beaten Germany on penalties), or if some extenuating circumstances were considered (e.g. if Italy had not fielded a B-team against Ireland) then the correlation would be even weaker. This wouldn't by itself mean that your method/idea was terrible, just that the effect you are looking to demonstrate is a weak one (in soccer). Which we know already.


> maybe shouldn't count those games,.... will not really effect the percentage on average in a negative way,

Yes clearly it isn't a big effect because there aren't many penalty games (your new total is 21/37, just under 57%), the reason I made the point was more about "good form" in that you might follow established guidelines for such comparisons. When dealing with fairly weak correlations, we need to try and screen as much noise as possible and penalty wins are generally (not just by me) considered noise in such comparisons.

In addition to not counting "penalty victories" as victories, the converse is also relevant in that your numbers ignore all the matches which were draws between opponents of large population disparity (e.g. Iceland drawing with anybody). Perhaps you could instead consider a points system as e.g. that used by Kaizeler on another thread, giving points also for draws; given that teams are not always playing for a win, this seems fairer. Or using goal-difference.

Your comment about not liking the scoring system of football does again suggest to me that it would make a lot more sense to work with the sports where the effect is stronger. Indeed you might prefer chess, where the same opponents play each other dozens of times - so you really have a "fair" comparison, as opposed to football where it is hard to score despite dominance of possession say.

Quote    Reply   

#8 [url]

Jul 14 16 8:01 PM

nfm24 wrote:
> 58% is actually moderately high

I'd say 58% is not statistically significant and possibly within the error bars of the inputs to your comparison. But we are dealing with small data sets here so the idea of significance is a bit of a moot point anyway. In other words if a couple of marginal results had gone the other way (e.g. if Italy had beaten Germany on penalties), or if some extenuating circumstances were considered (e.g. if Italy had not fielded a B-team against Ireland) then the correlation would be even weaker. This wouldn't by itself mean that your method/idea was terrible, just that the effect you are looking to demonstrate is a weak one (in soccer). Which we know already.


> maybe shouldn't count those games,.... will not really effect the percentage on average in a negative way,

Yes clearly it isn't a big effect because there aren't many penalty games (your new total is 21/37, just under 57%), the reason I made the point was more about "good form" in that you might follow established guidelines for such comparisons. When dealing with fairly weak correlations, we need to try and screen as much noise as possible and penalty wins are generally (not just by me) considered noise in such comparisons.

In addition to not counting "penalty victories" as victories, the converse is also relevant in that your numbers ignore all the matches which were draws between opponents of large population disparity (e.g. Iceland drawing with anybody). Perhaps you could instead consider a points system as e.g. that used by Kaizeler on another thread, giving points also for draws; given that teams are not always playing for a win, this seems fairer. Or using goal-difference.

Your comment about not liking the scoring system of football does again suggest to me that it would make a lot more sense to work with the sports where the effect is stronger. Indeed you might prefer chess, where the same opponents play each other dozens of times - so you really have a "fair" comparison, as opposed to football where it is hard to score despite dominance of possession say.

Again, the average for each tournament is 65% which is not a weak correlation at all (and gets rid of the fluke possibility). 58% is also not weak. If it was consistently 50-55% then it could be considered weak. The closest sport this happens in seems to be ice hockey (of the major sports). 65% is a pretty large number. Ice hockey might take this down to about 63 to 64 once I'm finished, but without ice hockey it would be above 65%. This means 2/3 of games in major tournaments are won by countries with more wealth/population, this is a huge statistic. And if population disparity was taken into consideration it would be even higher. So far the lowest I've seen is 53%, it has not gone below 50% yet. The highest so far is 88% I believe. As I stated in my video I'm sure somewhere out there there is a large tournament or two that fall below 50%, but it would be a very rare event.

I agree with your second paragraph, I could get more specific and have more fine turned results, I have compiled these statistics to show just how obvious it is wealth and population effect games.


Last Edited By: abramjones Jul 14 16 8:03 PM. Edited 1 time.

Quote    Reply   

#9 [url]

Jul 15 16 9:07 PM

>> I'd say 58% is not statistically significant

> 58% is also not weak.

The results you've given here are not in themselves statistically significant (i.e. no correlation was demonstrated), due to the low 8% trend in tandem with the small sample size of a single tournament of 40 results.  Or to put it crudely, your comparison here has only 40 results: only three of these need to be anomalous for the error to be higher than the trend.

That's all I'm saying really - I don't quite understand the point of considering one tournament on its own, it's a bit like saying: I've done a much bigger comparison and here is one more data-point for it, let's discuss that one by itself.  


> I have compiled these statistics to show just how obvious it is wealth and population effect games.

To look at one tournament separately doesn't really illustrate anything mathematically, it's only with much larger sample size that you can really say something solid about the trends.  [I'm not sure whether me spelling this out is patronizing or not, apologies if so]
Conversely, once you have established an overall large-set trend, that doesn't necessarily say anything about the size of the effect in individual tournaments (or matches) which can be viewed as random variables from some distribution the parameters of which may be measurable by your large-set statistics.


> Again, the average for each tournament is 65% which is not a weak correlation at all

If you mean that 65% is the average over all tournaments of all sports (?) then I agree, because that comparison involves many thousands of results, albeit taken from possibly rather disparate distributions per sport and/or era, so the sample size will be enough to imply statistical significance (although it might be wise to estimate the margin of error).


> This means 2/3 of games in major tournaments are won by countries with more wealth/population, this is a huge statistic.

It's not quite that, it means that out of the games *which were not draws*, 65% were won by the team with more [insert bespoke population/wealth combo here].  To ignore draws is misleading, particularly in a low scoring game.  Doing this would have almost no effect on, say, rugby or basketball, but draws are a big deal in football.  

Quote    Reply   

#10 [url]

Jul 16 16 2:59 PM

nfm24 wrote:
>> I'd say 58% is not statistically significant

> 58% is also not weak.

The results you've given here are not in themselves statistically significant (i.e. no correlation was demonstrated), due to the low 8% trend in tandem with the small sample size of a single tournament of 40 results.  Or to put it crudely, your comparison here has only 40 results: only three of these need to be anomalous for the error to be higher than the trend.

That's all I'm saying really - I don't quite understand the point of considering one tournament on its own, it's a bit like saying: I've done a much bigger comparison and here is one more data-point for it, let's discuss that one by itself.  


> I have compiled these statistics to show just how obvious it is wealth and population effect games.

To look at one tournament separately doesn't really illustrate anything mathematically, it's only with much larger sample size that you can really say something solid about the trends.  [I'm not sure whether me spelling this out is patronizing or not, apologies if so]
Conversely, once you have established an overall large-set trend, that doesn't necessarily say anything about the size of the effect in individual tournaments (or matches) which can be viewed as random variables from some distribution the parameters of which may be measurable by your large-set statistics.


> Again, the average for each tournament is 65% which is not a weak correlation at all

If you mean that 65% is the average over all tournaments of all sports (?) then I agree, because that comparison involves many thousands of results, albeit taken from possibly rather disparate distributions per sport and/or era, so the sample size will be enough to imply statistical significance (although it might be wise to estimate the margin of error).


> This means 2/3 of games in major tournaments are won by countries with more wealth/population, this is a huge statistic.

It's not quite that, it means that out of the games *which were not draws*, 65% were won by the team with more [insert bespoke population/wealth combo here].  To ignore draws is misleading, particularly in a low scoring game.  Doing this would have almost no effect on, say, rugby or basketball, but draws are a big deal in football.  

Yeah that's what I'm saying, maybe I wasn't clear. A single tournament alone means nothing of course, but putting many together show that wealth/population are a big factor. In addition I have the other spreadsheets which blatantly show this fact in a more mathematically assured manner. The reason I am showing the factor of wealth/population using tournaments in this manner is to show the same concept from the spreadsheets from a different angle.

Last Edited By: abramjones Jul 16 16 3:02 PM. Edited 2 times.

Quote    Reply   

#11 [url]

Jul 17 16 1:36 PM

OK. I suppose one advantage of looking at individual tournaments is to get an idea of the variance of the size of the effect.

Quote    Reply   
Remove this ad
Add Reply

Quick Reply

bbcode help