Friday, September 8, 2017

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.




The usual 10000 simulations, based on the current Elo ratings and including results up to and including 5 September 2017.

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

Peru, Sweden, Denmark, Tunisia and Senegal replace Chile, Republic of Ireland, Iceland, Congo DR and Burkina Faso.

AFC: Iran, Japan, Korea Republic, Saudi Arabia, Australia
CAF: Nigeria, Egypt, Tunisia, Côte d'Ivoire, Senegal
CONCACAF: Mexico, Costa Rica, USA
CONMEBOL: Brazil, Uruguay, Colombia, Argentina, Peru
UEFA: Belgium, Germany, Spain, England, Serbia, France, Portugal, Poland, Croatia, Italy, Switzerland, Sweden, Denmark

These would be the seeds: Russia, Brazil, Germany, Portugal, Argentina, Poland, Belgium and France.

Belgium and France replaced Chile and Colombia in the list of seeds.

100% - Korea Republic
100% - Brazil
100% - Saudi Arabia
100% - Mexico
100% - Belgium
100% - Japan
100% - Iran
100% - Costa Rica
99.99% - Germany
99.94% - Uruguay
99.8% - Spain
99.6% - England
96.29% - Serbia
96.01% - Colombia
95.04% - France
94.72% - Nigeria
93.13% - Portugal
91.84% - Poland
83% - Egypt
82.98% - Argentina
80.13% - Tunisia
78.54% - USA
75.91% - Italy
75.43% - Switzerland
66.34% - Côte d'Ivoire
59.54% - Peru
56.82% - Croatia
56.61% - Panama
50.46% - Sweden
49.84% - Senegal
40.91% - Chile
40.54% - Denmark
40.44% - Iceland
39.99% - Australia
36.67% - Burkina Faso
33.07% - Northern Ireland
31.05% - Slovakia
29.9% - Morocco
28.95% - Ukraine
27.34% - Wales
19.87% - Congo DR
16.76% - Turkey
13.23% - Uganda
12.47% - Honduras
12.39% - Syria
11.61% - Montenegro
11.38% - New Zealand
10.73% - Bosnia and Herzegovina
10.04% - Cape Verde Islands
8.81% - Greece
6.51% - Republic of Ireland
5.3% - Scotland
5.28% - Zambia
5.15% - Ecuador
4.09% - Paraguay
3.77% - Ghana
3.75% - Gabon
3.5% - Netherlands
3.45% - South Africa
0.97% - Slovenia
0.11% - Bulgaria
0.01% - Mali

Top 10 based on difference between these sims and those ran last time:

Peru 59.54 9.3 50.24
Serbia 96.29 54.46 41.83
Tunisia 80.13 40.35 39.78
Saudi Arabia 100 63.87 36.13
Denmark 40.54 13.43 27.11
Sweden 50.46 25.03 25.43
Nigeria 94.72 72.18 22.54
Korea Republic 100 80.84 19.16
Belgium 100 84.19 15.81
Ukraine 28.95 14.37 14.58

Bottom 10 based on difference between these sims and those ran last time:

Republic of Ireland 6.51 64.65 -58.14
Chile 40.91 91.9 -50.99
Uzbekistan 0 41.26 -41.26
Congo DR 19.87 58.21 -38.34
Australia 39.99 68.75 -28.76
South Africa 3.45 32.01 -28.56
Bosnia and Herzegovina 10.73 35.42 -24.69
Croatia 56.82 79.25 -22.43
Cameroon 0 21.22 -21.22
Greece 8.81 23.5 -14.69

Looking at the final draw seeds:

100% - Brazil
99.99% - Germany
93.13% - Portugal
82.94% - Argentina
75.44% - Poland
69.88% - Belgium
39.11% - France
37.56% - Spain
31.71% - Colombia
31.56% - Peru
16.83% - Wales
13.16% - Switzerland
8.67% - Chile
0.01% - England
0.01% - Northern Ireland

Top 5 based on difference between these sims and those ran last time:


Belgium 69.88 23.9 45.98
Peru 31.56 5.15 26.41
Spain 37.56 14.6 22.96
Portugal 93.13 72.02 21.11
France 39.11 28.11 11

Bottom 5 based on difference between these sims and those ran last time:


Chile 8.67 85.97 -77.3
Mexico 0 22.59 -22.59
Colombia 31.71 50.69 -18.98
Italy 0 6.1 -6.1
Argentina 82.94 88.8 -5.86

AFC 5th place play-off

69.63% - Australia
30.37% - Syria

AFC-CONCACAF play-off

52.38% - AFC
47.62% - CONCACAF

By team:

39.99% - Australia
23.19% - Panama
15.53% - USA
12.39% - Syria
8.9% - Honduras

OFC-CONMEBOL play-off

88.62% - CONMEBOL
11.38% - OFC

By team:

31.6% - Chile
28.79% - Peru
16.51% - Argentina
11.38% - New Zealand
4.62% - Ecuador
3.67% - Colombia
3.15% - Paraguay
0.28% - Uruguay

UEFA play-offs

Worst second:

26.88% - Wales
21.16% - Greece
17.44% - Bosnia and Herzegovina
11.46% - Republic of Ireland
8.02% - Slovakia
2.64% - Croatia
2.22% - Iceland
1.61% - Montenegro
1.57% - Turkey
1.35% - Serbia
1.19% - Sweden
1.18% - Denmark
1.05% - Northern Ireland
0.89% - Scotland
0.73% - Ukraine
0.24% - Slovenia
0.16% - Bulgaria
0.11% - Netherlands
0.05% - Italy
0.04% - Austria
0.01% - Cyprus

Seeded:

92.64% - Italy
67.97% - Switzerland
42.09% - Wales
41.68% - Northern Ireland
40.33% - Denmark
30.8% - Portugal
22.11% - Slovakia
15.07% - France
12.7% - Poland
9.92% - Turkey
7.71% - Croatia
5.53% - Sweden
5.44% - Ukraine
1.84% - Bosnia and Herzegovina
1.37% - Spain
1.11% - Scotland
0.75% - Netherlands
0.54% - Iceland
0.21% - England
0.17% - Bulgaria
0.01% - Montenegro
0.01% - Germany
0% - Greece
0% - Republic of Ireland
0% - Serbia
0% - Slovenia

Seeded if reaching the play-offs

100% - Wales
100% - Portugal
100% - Germany
99.28% - Spain
98.75% - France
98.22% - Switzerland
98.07% - Poland
93.98% - Italy
70.83% - Bulgaria
65.26% - Denmark
64.17% - Turkey
42.13% - Northern Ireland
35.1% - Ukraine
33.1% - Croatia
31.34% - England
30.63% - Slovakia
12.61% - Netherlands
7.61% - Scotland
7.17% - Sweden
6.63% - Bosnia and Herzegovina
1.4% - Iceland
0.04% - Montenegro
0% - Greece
0% - Republic of Ireland
0% - Serbia
0% - Slovenia


About me:

Christian, husband, father x 3, programmer, Romanian. Started the blog in March 2007. Quit in April 2018. You can find me on LinkedIn.

37 comments:

  1. thanks Edgar, great work as always

    ReplyDelete
  2. Thank you very much. So we are probably having Pot 1 with the seeds, and 3 Pots: AFC+CONCACAF (8 teams, if repeating the way used for the 2014 draw), CAF+3 CONMEBOL non seeds (URU, COL, PER - 8 teams in total) and UEFA non seeds (8 teams). There won't be 2 CONMEBOL teams in a group, and no more than 2 UEFA teams in a group (so 6 groups with 2 UEFA teams, and 2 groups with 1 UEFA team and 1 CONMEBOL seed).

    It seems that it won't be so hard for BRA-ARG (hardest possibility: BRA/ARG-Spain-Mexico-Nigeria), but it could be very hard for some UEFA teams (hardest possibility: Germany-Spain-Uruguay-Mexico).

    If Chile finaly is out of the WC, it would be terrible for them. 2 months ago they were almost qualified and with almost a seed assured, and the 1/8 final almost assured (with an easy group at the WC, beeing seeded).

    As I am spanish, if finally we are not a seed (it looks like we're not gonna be seeded), I would like to have the easiest possible group, maybe Poland-Spain-Senegal-Saudi Arabia as it's not a good idea to face the host country.

    Thanks!

    Juan

    ReplyDelete
    Replies
    1. Senegal as an easiest possible group? Are you mad? They are the most quality packed African team since Nigeria 1998 and you know how that one ended.

      Delete
    2. Przemek, he thinks about our Polish team

      Delete
  3. Great reply from you Juan as well. As a Serb my only prayer is that Heidi Klum doesn't come near drawing balls or that FIFA doesn't create extra pot just for Serbia :)

    ReplyDelete
  4. Does Uruguay have really no chance of being seeded in the final draw???

    ReplyDelete
    Replies
    1. Daniel, Uruguay can reach 1100 points in October with two wins over Venezuela away and Bolivia at home. I think they are perfectly capable of that.
      But with 1100 points they will be 16th in the predicted October ranking, so all teams between them and 7th spot with 1250 points (Wales, France, Spain, Colombia, Chile, Switzerland, England, Italy and Peru) should not qualify and/or become victim of severe points-loss in October. Not very realistic and I can understand why Edgar's simulations give Uruguay no chance whatsoever to be seeded.

      Delete
  5. Hi Ed. How much ranking points will Sweden have with WW compared to WD against LUX-NED (in that order)

    ReplyDelete
    Replies
    1. Thank you Ed! Which site can you use to calculate ranking points?

      Delete
    2. You're welcome.

      Apart from FIFA's own prognosis tool I'm not aware of any other site where one can interactively make points calculations for scheduled matches. But beware, FIFA's tool presents only the scheduled matches between the publishing date of the current ranking and the publishing date of next month's ranking, so you can only look one month ahead. And the tool has a proven flaw: you can't input a PSO win or loss for a match (like the second leg of a 2-legged play-off or a knock-out match in a tournament) and research the effects of such an outcome.
      To be able to look forward until the end of each year and have complete control over the calculation I have made my own software (and match database) to calculate ranking prognosis automatically.

      Delete
    3. Hi Ed, when can we expect October ranking preview? I'm especially interested in Poland chances of being seeded in WC2018 draw if they qualify. Thanks in advance.

      Delete
  6. Hi Ed, how much ranking points will the Netherlands have with two wins against Belarus and Sweden?

    ReplyDelete
    Replies
    1. Exactly what elo predicts: 931 points and around spot 23 in October.
      Stupid elo-ranking knows nothing about the poor form of the Dutch team this year :)

      Delete
  7. Is there any word on how UEFA will be organising their qualifying competition for 2022? With 55 teams due to take part, the usual 9 groups presumably won't be possible as it would mean one group having 7 teams and thus needing extra matchdays.
    The 2 obvious options are 11 groups of 5 OR 10 groups (5x5 & 5x6) but both options would mean several runners-up not even making the play-offs if the Europe quota is again 13 teams.

    Or perhaps Europe will finally come to its senses and have pre-qualifiers for the smaller nations like Asia, Africa & Concacaf, and having a leaner second stage?

    ReplyDelete
    Replies
    1. 10 groups (5x5 & 5x6) and only winners will qualify. next 3 teams will qualify via stupid uefa national league

      Delete
    2. I see. Is that official? The 10 group set-up makes sense, but I'd be interested to see how they get their 3 other qualifiers. Will it be from a play-off series involving the highest ranked Nations League teams who haven't already qualified?

      Delete
    3. This is not official. That was Platini's idea. The way of qualifying via UEFA National League is the same as for EURO 2020 but play-off only for leagues A, B and C.

      Delete
    4. Sending a League C team to the World Cup as one of Europe's "best" 13 is a bat-shit crazy idea.
      So yes, it must have been Platini!

      Delete
    5. I don't see how UEFA would choose to send teams to the World Cup via Nations League. I'm fairly certain NL is used only to avoid friendlies and qualify teams to the UEFA championship. I can't even imagine a European qualification where Portugal and Italy are out and Finland plays Georgia for a spot in the World Cup.

      Almost certainly they will rank the 10 second place teams, eliminate the bottom four, and let the other six play-off for the last three spots.

      Delete
    6. I reckon that 11 groups of 5 and the 8 best runners up playing off for 2 spots over 2 rounds of playoffs would work reasonably well

      Delete
  8. Hi Ed, when can we expect October ranking preview? I'm especially interested in Poland chances of being seeded in WC2018 draw if they qualify. Thanks in advance.

    ReplyDelete
  9. With the news coming out that all 32 teams will be seeded by the FIFA rankings, I'm curious as to what chances the US has to avoid Pot 4, should they qualify. I'm also curious if beating both Panama and Trinidad and Tobago gives them any chance of moving into Pot 2. Thank you in advance and keep up the good work.

    ReplyDelete
    Replies
    1. Good news. It's about time FIFA properly seeded all 4 pots rather than basing Pots 2-4 on sheer geography. There are still means of keeping teams from the same confederation apart, and it's a shame it has taken them so long to figure this out.
      This may reduce the chances of a "group of death" but it makes perfect sense from the point of view of sporting integrity.

      Delete
  10. Yeah, finally a good decision by FIFA.
    This will most probably be the pot distribution if Edgar's prediction becomes reality:

    pot1:
    host Russia (64, UEFA)
    Brazil (1, CONMEBOL)
    Germany (2, UEFA)
    Argentina (3, CONMEBOL)
    Portugal (4, UEFA)
    Belgium (5, UEFA)
    Poland (6, UEFA)
    France(8, UEFA)

    pot 2:
    Spain (9, UEFA)
    Colombia (10, CONMEBOL)
    Switzerland (12, UEFA)
    England (13, UEFA)
    Italy (14, UEFA)
    Peru (15, CONMEBOL)
    Mexico (16, CONCACAF)
    Uruguay (17, CONMEBOL)

    pot 3:
    Denmark (19, UEFA)
    Croatia (20, UEFA)
    Costa Rica (21, CONCACAF)
    USA (25, CONCACAF)
    Sweden (27, UEFA)
    Iran (28, AFC)
    Senegal (29, CAF)
    Egypt (31, CAF)

    pot 4:
    Tunisia (33, CAF)
    Serbia (34, UEFA)
    Japan (35, AFC)
    Australia (37, AFC)
    Nigeria (40, CAF)
    Korea Republic (51, AFC)
    Côte d'Ivoire (60, CAF)
    Saudi Arabia (62, AFC)

    ReplyDelete
    Replies
    1. That's interesting. Pot 2 arguably isn't much weaker than Pot 1. Serbia would probably be the Pot 4 team to avoid if they are indeed in there.
      Best "group of death" from that scenario: Brazil, Spain, Croatia, Côte d'Ivoire?
      Arguably, the weakest group could be: Russia, Switzerland, USA, Saudi Arabia?

      Delete
    2. Or maybe Germany, Colombia, Croatia, Nigeria as the "Group Of Death", with Poland, Peru, Iran, Tunisia as the "Group Of Life".
      It's still all very hypothetical, of course. I still see Chile pipping Peru to make the play-off with NZ, and Europe's qualifiers will depend much on how their play-off draw turns out.

      Delete
    3. The chance of a top European team (Spain, England, Switzerland, Italy) to end up with Argentina or Brazil seems very high, but I'd rather see this system than the geographical pots.

      Delete
    4. England & Switzerland aren't "top" European teams and probably no stronger than the European teams projected to be in Pots 3 & 4.
      I'm pretty sure Brazil or Argentina would be happy to get either of those.
      But you're right in that their options are limited due to the likely presence of 3 Conmebol teams in Pot 2.

      Delete
    5. Hello Ed,
      May you please provide us the source at the FIFA website saying that all the pots will be based on the FIFA ranking ? What I understand is that only the seeded teams will be based on FIFA October ranking , and the rest 3 pots will be based on geographical location. ( i.e.( pot A for Russia and the top 7 seeded teams) + ( pot for non seeded European teams ) + ( pot for African and non seeded South American teams ) + (pot for Asian and concacaf teams ))

      Delete
    6. There's a link at the beginning of this post.

      Delete
    7. This comment has been removed by the author.

      Delete
    8. If the ranking in all pots will be considered, the possibility of deadlocks during the draw is very very high . In order to keep the geographical zones while choosing from the pots , there will be many skips and it might lead to a deadlock in the end. If they made a program to avoid the deadlocks , the draw will seem to be directed and it will lose its credibility. Infantino just wants to change the system without being much aware of the consequences, he wants to change for the change . If FIFA is looking for fairness , they wouldn't have done the shameful African pots seeds in the African qualifiers and the special ranking back in June 2016. I lost trust in FIFA since that date .

      Delete
  11. Ed, who would be 7th team of October ranking? Chile? So what are France's chances for pot 1 in the case of Chile qualifies?

    ReplyDelete
  12. Hi Edgar, the Rep of Ireland are only a 6% chance of making the finals.
    But what are their chances of making it to the playoff's??
    Thanks

    ReplyDelete
  13. Ger, Republic of Ireland has 17% probability to make it to the play-offs and will then always be unseeded. Or they could become worst runner-up with a chance of 12%. Their chance of top spot in the group is, after the disastrous results in September, almost non-existent: 1.4%.

    ReplyDelete