Remove this ad

Lead

Jul 12 16 2:30 AM

Tags : :

Well, since you guys generally dislike my ideas, I have come up with a new one. And you're the first I'm announcing it to (: I'm not sure if I will always have the time to maintain them, but I'm at least going to get a few months done. Here is how they work: If a country plays a country with more wealth/population and wins they will receive a win, the losing country will receive a loss. If a country defeats another country that is moderately smaller than them they will receive a win, but the losing country will not receive a loss. This is a primitive idea, but I think it may be pretty accurate, let's see how it goes. I will use most men's team sports.

some possible problems i see with this idea:
1. too few matches in some sports, and unbalanced because some sports are played less than others
2. some sports are harder to organize, especially those with more players on the field, because they require more resources and skill. i have no weight system for sports that require more resources
3. the common problem of teams that often play harder teams suffering unfairly with winning percentage that elo rankings try to solve, but i don't think it will be as big of an issue as with regular ranking because wealth/population is taken into consideration. some regions certainly have more efficient sports culture though, so it could still be an issue
4. typical mathematical problem when comparing and using available populations because of the disparity of the numbers in many cases
5. very small countries have few opponents that qualify to give them a loss

edit: i have decided to try elo rankings based on matches played by countries of relatively the same population/wealth instead.

Last Edited By: abramjones Jul 24 16 12:22 AM. Edited 8 times

Quote    Reply   
Remove this ad
Remove this ad

#1 [url]

Jul 12 16 10:23 PM

abramjones wrote:
Well, since you guys generally dislike my ideas, I have come up with a new one. And you're the first I'm announcing it to (: I
Why don't you do an Elo ranking but with "available population" as a handicap?  I assumed that was what you were doing all along?

Quote    Reply   

#2 [url]

Jul 13 16 5:15 AM

nfm24 wrote:
abramjones wrote:
Well, since you guys generally dislike my ideas, I have come up with a new one. And you're the first I'm announcing it to (: I
Why don't you do an Elo ranking but with "available population" as a handicap?  I assumed that was what you were doing all along?
No, I never did that unfortunately. When I first started with all of this i used the elo rankings here and then tried to adjust them by available population, but it's too problematic. i'm not really sure there is a way to do this, if so it's over my head mathematically (i no longer have this spreadsheet posted). All of the other spreadsheets  I have done are simply showing how larger and richer countries generally perform better. I think an elo ranking system that takes wealth/population into consideration is a great idea, i'm just not sure how to go about it. I don't even have any experience in setting up a regular elo ranking system.

Last Edited By: abramjones Jul 15 16 8:02 AM. Edited 4 times.

Quote    Reply   

#3 [url]

Jul 13 16 8:40 AM

I have had a look at that sort of relationship some time ago, but only looking at total population, so not considering any economic variables. Uruguay came out on top having historically been on a par (in terms of football level) with their much larger neighbours Brazil and Argentina.

Of course, generally speaking there is a clear trend between population and results; the relationship is just perhaps flatter than one would expect:
image

As we can see in the leftermost circle in the plot, even teams with a relative population size below 0.05 (i.e. teams that were outmatched by their opposition, in terms of population, by a factor greater than 19) still got about 30% of the available points in their games.

The effect on goal difference:
image

Quote    Reply   

#4 [url]

Jul 13 16 9:16 PM

Kaizeler wrote:
I have had a look at that sort of relationship some time ago, but only looking at total population, so not considering any economic variables. Uruguay came out on top having historically been on a par (in terms of football level) with their much larger neighbours Brazil and Argentina.

Of course, generally speaking there is a clear trend between population and results; the relationship is just perhaps flatter than one would expect:

Excellent statistics and discussion. I am glad that some people (even announcers on occasion) talk about the factor of population. And yes, differences from population alone are going to be pretty flat in general if wealth isn't considered... because there are many impoverished countries with very high populations. population alone will be a small but noticeable factor as you've shown, wealth and population combined is a very significant factor. I would like to see more people who talk about the consequences of population difference also mention wealth.

Last Edited By: abramjones Jul 13 16 9:28 PM. Edited 2 times.

Quote    Reply   

#5 [url]

Jul 14 16 1:31 PM

Looks like a nice inverse error-function fit to the second graph.

Without wishing to repeat discussion elsewhere, there is also the issue that many of the most populous countries are not football specialists, e.g. 4 or 5 of the top 10 : China, India, USA, Indonesia, Brazil, Pakistan, Nigeria, Bangladesh, Russia, Japan. This is another input to "available population". It also adds noise to the previous graphs - rather tedious but maybe possible to filter this by imposing a sort of "football development" criteria (ignore beginners), or more conveniently perhaps just look only at European results after 1950. Looking at Europe-only would also to some extent filter out the effects of large geographical and cultural disparity on the other continents.

Quote    Reply   
Add Reply

Quick Reply

bbcode help