Is there a regional bias in the elo rating ?

An interesting question popped up in the discussions around my World Cup simulations and more specific the somewhat lower than expected predicted performance of Senegal compared to home nation Qatar in group A. 

Amir commented:

"Elo indeed has a regional problem - since there are not many matches between teams from different confederations, it takes a long time for ELO to adjust to the actual average per confederation.

If the difference are getting smaller (which is something worth testing), it means that Elo will indeed under-estimate CAF."

Apart from the difficulty to determine the "actual average per confederation", I personally have the impression that elo ratings are not regionally biased, because there isn't anything in the elo calculation that's influenced by the team's confederation. But maybe, indeed the number of intraconfederational versus interconfederational matches a team playes could be the exogenous factor. I decided to give it a go and research this.

Uncertainties about FIFA's new calculation method

Example: let's assume that in an elo-system team A has 1200 points as 'the value before a match' and team B has 1500 points as 'the value before a match' when they play a friendly match (not in a FIFA International Match Calendar window) against each other with A as the home team. A wins 1:0.

FIFA Ranking: January 2018 probable results

Here are the probable results used for the January 2018 probable ranking. They are generated using the Elo Ratings.

Before posting comments like: "Hey my team actually beat the other team! It's a mistake!" please read the previous paragraph again and look when these results were posted (Yes, before the matches actually took place!) ;)

2018 World Cup draw: What I would have liked to see

I think I've mentioned this before, but I'm getting old and I'm allowed to say things over and over. Getting plenty of practice with my kids :)

Long time ago - about 8 or 9 years ago perhaps - my blog didn't have a custom domain and its name was "Uninteresting things" (of course, in Romanian). This is one of those posts, so feel free to skip it.

2018 World Cup simulations (13 October 2017)

We're almost there. Only 9 spots left - 6 through play-offs and the other 3 in Africa.

The pots for the UEFA play-offs draw (seeded using the October 2017 FIFA ranking):

2018 World Cup simulations (6 October 2017)

There will be no updates until Friday. Plenty of websites will publish the seeds for both the final draw and the UEFA play-offs, so there's no need to worry :)

6 October means all matches played up to and including 5 October. So no matches later than this.

Taking into account the actual group distribution, these are the teams most likely to make it to Russia:

2019 AFC Asian Cup 3rd round sims (6 October 2017)

The usual 10000 sims after the third match day (5 September 2017). Kyrgyz Republic bounce back and replace Myanmar in the list of teams most likely to qualify.

FIFA to seed ALL the teams using October ranking

It finally happened. I must admit I wasn't expecting this, but it's good news. Especially since this will increase the number of FAs looking to improve their ranking :) Perhaps it will finally get FIFA to change the ranking algorithm.

I think the African teams could complain as the October ranking won't include their final group stage matches - scheduled for November, but that's life.

2018 World Cup simulations (8 September 2017)

It seems while I was climbing to 1750m in the mountains, all sort of things happened. One of the most amusing was this call to arms from one of the blog readers. Ed was kind enough to answer in detail and here's the usual update of the sims.

I've added info regarding the AFC 5th place play-offs, the intercontinental play-offs and the UEFA play-offs (worst 2nd place, seeded / unseeded). Scroll to the end.

2019 AFC Asian Cup 3rd round sims (14 July 2017)

The usual 10000 sims after the second match day (13 June 2017). Great improvement for Myanmar and Philippines. On the other hand, Kyrgyzstan, Tajikistan and Singapore stumbled and lost momentum.

2017 CONCACAF Gold Cup simulations

The Gold Cup kicks off today.

The usual 10000 simulations.

2018 World Cup simulations (23 June 2017)

The usual 10000 simulations, based on the current Elo ratings and including results up to and including 22 June 2017.

Previous simulations (31 March 2017).

2019 AFC Asian Cup 3rd round sims (21 April 2017)

The usual 10000 sims after the first match day. Great improvement for India.

2018 World Cup simulations (31 March 2017)

The usual 10000 simulations, based on the current Elo ratings and including results up to 29 March 2017.

With the 2017 release dates announced, I've used the October 2017 ranking for the UEFA play-off seeding and, of course, for the final draw seeding.

2019 AFC Asian Cup 3rd round sims

10000 simulations - before and after the draw.

First, remember the pots:

2017 Confederation Cup simulations

By request.

The usual 10000 simulations.

First the average points in the group stage.

2018 World Cup simulations (18 November 2016)

The usual 10000 simulations, based on the current Elo ratings and including results up to 15 November 2016.

The play-offs are also simulated, with a slight twist: I used Elo ratings to seed the UEFA play-off draw.

FIFA Ranking: November 2016 probable results

Here are the probable results used for the November 2016 probable ranking. They are generated using the Elo Ratings.

Before posting comments like: "Hey my team actually beat the other team! It's a mistake!" please read the previous paragraph again and look when these results were posted (Yes, before the matches actually took place!) ;)

2018 World Cup simulations (14 October 2016)

The usual 10000 simulations, based on the current Elo ratings and including results up to 13 October 2016.

The play-offs are also simulated, with a slight twist: I used Elo ratings to seed the UEFA play-off draw.

2018 World Cup simulations

Finally found some time and energy today to write the code for the Russia 2018 simulations.

They are based on the current Elo ratings and include results up to 6 October 2016.