Remove this ad

Lead

Jul 12 14 10:36 PM

Tags : :

Just wondering if and when the rankings are going to be put back up.
Quote    Reply   
Remove this ad
Remove this ad
avatar

TheRoonBa

Posts: 5,503 Site Admin

#1 [url]

Jul 14 14 9:18 AM

I wonder that every day. :-D

When I can find the will power to fix the ranking program. :-D

Meanwhile, anybody else who wants to create a ranking program, feel free to do so. I am far from being an expert on ranking systems, computer programming or mathematics.

Quote    Reply   
avatar

TheRoonBa

Posts: 5,503 Site Admin

#3 [url]

Jul 15 14 11:03 AM

pieter wrote:
we still have FIFA rankings....Cute

True.  I think everybody should just give up and leave FIFA to rank teams alone, because they are the best smiley: pimp

Quote    Reply   

#5 [url]

Jul 15 14 1:08 PM

TheRoonBa wrote:
pieter wrote:
we still have FIFA rankings....Cute

True.  I think everybody should just give up and leave FIFA to rank teams alone, because they are the best smiley: pimp

They didn't do too bad last time... Though this be madness, yet there is method in't.

Quote    Reply   

#6 [url]

Jul 15 14 4:28 PM

Indeed, now that the 10.18g WPL has been disbanded due to office refurbishment, the reigning champions of the World Cup of Rankings will be unable to defend the title and FIFA may be the favourites for the next edition...

Quote    Reply   
avatar

TheRoonBa

Posts: 5,503 Site Admin

#8 [url]

Jul 15 14 8:20 PM

pieter wrote:
strange: life goes on without rankings....

Life also goes on without league tables, cup competitions, football, sport.  They serve a useful purpose, otherwise they would not exist.

It is natural for teams to seek competition and to judge themselves against other teams, otherwise the World Cup would not have evolved.  Rankings are the ONLY way for teams who do not qualify for World Cups to see where they stand relative to other teams worldwide.  And they are an important indicator of how the smaller teams are developing.

However, I do realise you are the king of the anti-ranking brigade, so I shall not try to convince you of their worth smiley: laugh

Last Edited By: TheRoonBa Jul 15 14 8:50 PM. Edited 1 time.

Quote    Reply   
avatar

TheRoonBa

Posts: 5,503 Site Admin

#9 [url]

Jul 15 14 8:56 PM

Kaizeler wrote:
TheRoonBa wrote:
pieter wrote:
we still have FIFA rankings....Cute

True.  I think everybody should just give up and leave FIFA to rank teams alone, because they are the best smiley: pimp

They didn't do too bad last time... Though this be madness, yet there is method in't.

Don't be fooled by this. That was mainly due to luck and a not-very-good method to compare the rankings.  In a longer-term prediction, including teams from lower down the ranking, they would fare much worse.

Quote    Reply   

#10 [url]

Jul 15 14 9:28 PM

If only there was some way of proving it ;-)

Quote    Reply   
Remove this ad
avatar

TheRoonBa

Posts: 5,503 Site Admin

#11 [url]

Jul 17 14 2:24 PM

nfm24 wrote:
If only there was some way of proving it ;-)

By golly, nfm, if only there was. smiley: roll


(For some reason, images of the 80s TV programme "Gentle Ben" entered my mind at this point, and I decided to shut off my brain for the time being).

Quote    Reply   

#12 [url]

Jul 17 14 4:57 PM

Suggestion : use FIFA ranking to predict the outcome of every international match, and bet $10 on each result. When you have either won or lost $1,000, declare FIFA's ranking to be brilliant / awful (delete as desired).

Quote    Reply   

#13 [url]

Jul 17 14 10:18 PM

nfm24 wrote:
If only there was some way of proving it ;-)

I got bored today so there is one now smiley: tongue


I pulled up all of the monthly FIFA rankings since the 2006 methodology change and used them to 'retroactively predict' the results of international football matches since then, with similar rules to the World Cup of Rankings: only matches that did not end in a draw count, and whichever team is on top in the ranking is the predicted winner. If at least one team is unranked (or both teams are level), then no prediction is made.

I then tallied up the percentage of correct 'predictions' and compared that with what other systems would deliver:
Random: basically a coin-flip model
Home: always go for the home team; in matches played on a neutral venue, there is a 50-50 chance of picking either team
FIFA: using the monthly rankings that had last been published
Elo: using each team's rating immediately prior to the match
Cubic: a rating system I developed, strongly influenced by the Elo model. It also means ratings are calculated/updated immediately after each match.

image

Unsurprisingly, the first two, 'lazy' models aren't really that good. A random choice ensures a 50% success rate in the long-run; picking the home team captures the home advantage effect but nothing else. FIFA did do worse than Elo with this larger sample, as you would expect, but they were not that inept:
image

And similarly, the 'randomness' of friendly matches makes them more difficult to reasonably predict. The same chart, considering only competitive matches:
image

Last Edited By: Kaizeler Jul 17 14 10:24 PM. Edited 2 times.

Quote    Reply   
avatar

TheRoonBa

Posts: 5,503 Site Admin

#14 [url]

Jul 18 14 12:30 AM

An interesting (and more accurate) version would involve the correlation between margin of victory and difference in rating points. Saying Team A is Number 100 and Team B is number 200 doesn't tell you anything about the difference in skill (and therefore, margin of expected victory) between these 2 teams.

Quote    Reply   

#15 [url]

Jul 18 14 3:52 AM

TheRoonBa wrote:
I wonder that every day. :-D

When I can find the will power to fix the ranking program. :-D


Haha okay. Do you keep an archive of rankings from the past? I searched the way back machine and most pages did not archive.

Quote    Reply   

#16 [url]

Jul 18 14 1:01 PM

You could also add "bookmakers' favourite" as another ranking system, which I expect would score better than all of the above.

It may be worth ignoring matches where there is a large difference in ranking. Rewarding the systems for predicting that Germany will beat San Marino is unnecessary (and indeed penalizes the "random" and "home" systems). More weighting could be given to successful prediction of matches between teams which are close together in the rankings, as this is a stronger test of the system.

It would also be interesting to see the situation with draws incorporated, but that makes it a lot more complicated and winning margin would also need to be considered then, as Mark said.



Teams ranked close together could be expected to draw, provided home advantage was not too dominant.

Quote    Reply   

#17 [url]

Jul 19 14 5:46 PM

TheRoonBa wrote:
An interesting (and more accurate) version would involve the correlation between margin of victory and difference in rating points. Saying Team A is Number 100 and Team B is number 200 doesn't tell you anything about the difference in skill (and therefore, margin of expected victory) between these 2 teams.

And as it turns out the difference in FIFA points is also not that strong of a predictor of margin of victory (in all matches, including draws, there was a correlation of 0.256, compared with Elo's 0.481 and 0.506 for my system). A team ranked at #100 will tend to beat both a team ranked at #150 and one ranked at #200, but there will not be a large difference between those two results.

This point is perhaps better illustrated by plotting the relationship between points- and goal-difference across ranking systems. I grouped matches together depending on the teams' points difference (rounding to the nearest 50 points, so one group includes all matches where teams were separated by less than 25 points; the next group includes matches where teams were separated by 25-75 points, and so on...) and evaluated the actual results in each group.

In the charts below, the solid black line represents the average goal difference within a group. For example, in the FIFA chart, in matches where both teams were very close (separated by less than 25 points, follow the vertical line from the 0 in the x-axis), the favourites scored on average 0.09 more goals than the underdogs. In matches between teams separated by around 200 points (175-225) the average goal difference increased to 0.88, and so on. We can notice that on FIFA's chart this line becomes quite flat, which means that a wider points difference will not necessarily return a wider goal (skill) difference. In the other models the trend is clearer.

As a dispersion analysis (to measure how much amplitude there is within each group of matches), the solid black line is surrounded by dashed lines. These are approximations to quartiles.* Let's say that, on the FIFA rankings, there were 600 matches where teams were separated by less than 25 points. I take the 300 "worst" results (from the favourite's perspective), average those goal differences and plot that average on the yellow line. I repeat the process for the 300 "best" results, average those goal differences and plot that average on the blue line.
The red/green dots come from a similar approach (averaging the goal difference in the best/worst 10% of matches).

Ideally you would want the distance (height) between dots on the same column to be as small as possible. This would mean that the points difference between teams is an accurate reflection of their skill difference (i.e. you wouldn't see many large wins in matches between teams ranked closed together, or close results in teams ranked far apart).

A final word on scales. The x-axis ends on different values for different rankings because each one operates on a different scale (e.g. Germany and San Marino are separated by ~1700 points on FIFA's ranking, ~1300 on Elo's and ~900 on Cubic). Effectively the x-axis stops at the point where fewer than 10 matches have been played within a group (e.g. on FIFA's case, 16 matches have been played between teams with a points difference around 1200 (between 1175 and 1225), but only 6 matches have been played between teams with a points difference around 1250 (between 1225 and 1275), so the chart will stop at 1200).


*The mathematically-minded will recognize that as goals are a discrete variable with a small amplitude, using the strict definition of quartiles would be of limited use when comparing groups.

image

image

image

Quote    Reply   

#18 [url]

Jul 19 14 6:00 PM

nfm24 wrote:
You could also add "bookmakers' favourite" as another ranking system, which I expect would score better than all of the above.

It would indeed be great if I could find a large collection of odds for these matches. For club matches at least there seems to be enough data around.

nfm24 wrote:
It may be worth ignoring matches where there is a large difference in ranking. Rewarding the systems for predicting that Germany will beat San Marino is unnecessary (and indeed penalizes the "random" and "home" systems). More weighting could be given to successful prediction of matches between teams which are close together in the rankings, as this is a stronger test of the system.

It would also be interesting to see the situation with draws incorporated, but that makes it a lot more complicated and winning margin would also need to be considered then, as Mark said. 

Teams ranked close together could be expected to draw, provided home advantage was not too dominant.

The point was mostly to compare these three 'non-lazy' systems, so while rewarding systems for correctly predicting Germany beating San Marino is unnecessary if all of them get it right, I wanted to evaluate all matches in case one system got it terribly wrong - for instance, new entrants to the FIFA ranking (Montenegro, South Sudan...) will start with 0 points, which is not necessarily accurate.

Of course ideally I would evaluate the degree to which the prediction was correct. Say that Team A beats Team B by 4-0. A ranking system that had those teams further apart should be rewarded more than one that had them very close together. The problem in determining this is that the FIFA ranking does not really translate to a "win expectancy" formula like Elo's. This makes draws more difficult to analyse (when would two teams be expected to draw? Given a certain points difference, or points ratio, or number of ranking places between them?)

Last Edited By: Kaizeler Jul 19 14 6:02 PM. Edited 1 time.

Quote    Reply   

#19 [url]

Jul 19 14 8:50 PM

Kaizeler wrote:
This makes draws more difficult to analyse (when would two teams be expected to draw? Given a certain points difference, or points ratio, or number of ranking places between them?)

Provided an "expected winning margin" can be defined (possibly including home advantage) an "expected draw" would be interpreted as a winning margin of below 0.5 goals, I suppose.   Though I only relabelled the problem...

Perhaps misleadingly, FIFA seems to do slightly better (as an average) in the first column (closest teams) than the other systems, although with a larger variance.


Quote    Reply   
avatar

TheRoonBa

Posts: 5,503 Site Admin

#20 [url]

Jul 6 16 2:08 PM

Rankings are coming back - I know not when, but they are...

With a view to their return, I am trying to bring in some teams who have played matches, but not against national teams. This is possible if, for example, another team on the rankings has played against the same non-national teams as the team who is not ranked.

As an example, I've just managed to give Herm a ranking based on the fact they played 2 teams that Sark also played, around about the same time (Saint Martins and Captains FC). It would appear that Herm are around 4 goals worse than Sark.

I'm currently trying to give Saint Helena a ranking - I am basing this on their match with HMS Montrose (4-3 win). The HMS Montrose also played Falkland Islands on this trip, on 27 January 2012, but I've not been able to find the score yet. Finding this score could enable me to give St. Helena a ranking (albeit slightly shaky!).

For Christmas Island and Cocos Islands, I am looking to somehow get them ranking points based on Christmas Island's matches against the Malaysian team UiTM. However, I have not yet been able to establish a link between UiTM and any already-ranked team.

Tristan da Cunha beat HMS Clyde 3-2 in 2011, and HMS Clyde beat South Georgia 2-1 in 2016. From this, we can make an unofficial ranking of these 2 islands, but so far, I haven't found any matches by HMS Clyde against any already-ranked teams, so TDC and SG remain "unrankable".

Quote    Reply   
Remove this ad
Add Reply

Quick Reply

bbcode help