Euro 2020

[CrabCZ]

Bulgaria lack world class players at the moment and there aren't any breath-taking talents coming from below, but since Hubchev took care of the NT, he restore the fighting spirit in the team and we made some very good results at home. Our home record of the last WC qualifiers and the Nations League is 6-1-1, beating Holland, Sweden (for the first time since the 1960's), Norway and Cyprus. Whilst I still see the Czech Republic and especially England as favorites, I believe we still have realistic chances for the second place, if we manage to win our home games.

I can say the same thing about the Czech team, they've appointed a new coach in october, the air cleared, the players that are supposed to play actually play. They also beat the Dutch in the last euro qualifiers (as did everyone else). They did it twice and that team really was god awful, but it's past anyway. All I'm saying you need to judge it with the lens of the current reality + there are still gigantic differences even between teams that are not thought as the top ones.

[amirbachar]

Group F is the group of death, with 4 teams capable of qualifying and all teams lowering their chances of qualifying.
Euro 2020 Qualifying Simulations:
https://simtheworld.blogspot.com/2018/1 ... tions.html

If you have any question about a specific team or in general feel free to ask...

It looks great, thanks for sharing. I actually have a couple of questions regarding this methodology:
1) Are we sure that we can apply the method described on http://old.clubelo.com/Articles/Improvi ... model.html for clubs also to national teams?
2) When simulating the probability to qualify Before the draw, did you take into account all the 4 criteria regulating the draw? (host teams, Political clashes, winter venues and excessive travel?). They are not easy to implement in a simulation program, but they surely influence the probability of the the different nations to be drawn in the same group

These are two great questions.
I realize that national teams football may be different than club football, but since I haven't done the research, using the club research is a good compromise.
The only restriction taken into account is the political one. The other restriction have a small percentage anyway, since they only prohibit 3 countries in each group from lists of up to 12 teams each, so I didn't implemented them. Maybe in the future I will implement such restrictions as well for future draws.

[bugylibicska]

Best tweet from Wales after the draw:

@ffwtbol
Follow Follow @ffwtbol

Here is all you need to know about our draw for Euro 2020. The average price of a pint of beer. Croatia £1.80 ; Slovakia £1.51; Hungary £1.34; Azerbaijan 69p.

[seso]

Best tweet from Wales after the draw:

Here is all you need to know about our draw for Euro 2020. The average price of a pint of beer. Croatia £1.80 ; Slovakia £1.51; Hungary £1.34; Azerbaijan 69p.

Wales about £5.00, so the fans will be happy!

Greece very seldomly plays Italy so it will be fun to watch. The last official matches were in 1980/81. Even the last friendly match was in 2008. Also Bosnia and Herzegovina and Finland have been doing quite well lately. Let's see!

Very interesting the way UEFA has structured the 2018–19 UEFA Nations League and Euro 2020 qualifying play-offs; well-done to the person that thought of it! Let's see what the result will be!

[Michele]

While I'm very pleased with the draw, the match schedule worries me somewhat. We have a day off on match day 1 and our first match is then Switzerland away. That could easily be a loss. After that we have Ireland at home and we've seen how much we struggle to break them down. If we only have one point after two matches, we will have to play catch up all through the rest of the qualifiers with Ireland away in the final match. I would much prefer a schedule where we would likely be ahead of Ireland before we play them. We saw last fall what we can do to them if they are forced to score. On the other hand, they have held us scoreless 3 times in matches where we have chased the win in vain. I really hope that we can win the home game this time and be in a good position after the difficult opening matches.

[Ricardo]

@elkjiaer
Exactly!
Did Holland had an unlucky draw? No, because the goal is to qualify and end up in the top 2. And then the pot2 teams is not (very) important.
Somehow your luckometer is more in agreement with my stomach feeling as Dragonite's. I'm not a big favorite of ELO rankings, but in this case it does work.
Thanks

[Dragonite]

I don’t agree.
The “probability to advance” = “probability to be first placed” + “probability to be second placed”.
Saying that the strongest opponent doesn’t matter because you can still advance ALONGSIDE it and don’t have to finish ABOVE it is shortsighted.

A team aiming for the top has a better chance to advance than a team that isn’t even aiming for the top and is already putting itself in a secondary role.


By the way, We Global Football is already estimating the 24 Euro 2020 participants.
This is their projection:



I don’t know if these guys’ projections are good or not, I only started paying more attention to them a few months ago because of the 2019 Africa Cup.

[Vickzq]

I don’t know if these guys’ projections are good or not, I only started paying more attention to them a few months ago because of the 2019 Africa Cup.

I can't say if they used the correct pattern for 'who goes to UNL play-offs' (especially according to their own prediction of group winners, etc), but it looks quite reasonalbe for a 24-team EURO.

I wondered who might have a 'theoretical disadvantage' (or advantage) after the match schedule was fixed (looking at early qualifier stage). Potentially, because they don't play and can't influence the other teams' results (and might feel like they need to catch-up)... or because they play 'away'... I see this:

A) Czech Republic has to start away against England... and then 'pause' at second match-day. Montenegro plays away vs Bulgaria, and then has to host England. This could leave both with that feeling of 'we need to catch-up' right after two rounds. Atleast Kosovo starts at home.

B) Portugal plays at home vs Ukraine and Serbia in a row right at the beginning - looks more like a clear advantage for them. Ukraine also has to 'pause' during second round, so they might feel left behind more than anybody else, when Serbian can catch up away vs Luxemburg?

C) Good for the dutch to play Belarus and Germany at home; might leave them with 6 pts already, when Germany potentially could still have zero (not playing on first day). Northern Ireland could also get 6 pts playing twice at home first, vs Estonia and Belarus - but then again, Germany should still be strong enough to avoid stumbling.
But it might be more 'entertaining' because of the match schedule.

D) Republic of Ireland can play away Gibraltar first, then Georgia at home - might put them into a favorable position at the beginning. While Switzerland starts away in Georgia, but atleast can face Denmark at home... Danes (pausing on first day) might feel the 'we need to catch-up now' right after two match days.

E) Croatia is set up to a good start, with Azerbaijan at home and then Hungary away. Slovakia will have to score at home vs Hungary first, then go to Wales (they will start their qualifiers on second match day).

F) The first two match days seem going in favor of Spain, playing Norway at home and Malta away. It will already split up possible opposition during these two rounds: Sweden plays Romania at home first, then Norway away... while Romania will host Faroe Islands.

G) If we assume, that this is one of the most balanced groups overall... Israel might have a small advantage - they play Slovenia and Austria at home first. Poland has to face Austria first away, then Latvia at home - while Macedonia hosting Latvia and then playing Slovenia away might have good chance for a decent start.

H) Iceland starts away vs Andorra, but has to visit France after this - while the French start at home to host Moldova. Either Albania or Turkey might get a good head-start, facing each other first in Albania... Turkey hosting Moldova on second day, Albania visiting Andorra.

I) Belgium facing first Russia at home and Cyprus away has a decent setup to start qualifiers. Kazakhstan hosts Scotland first, and then Russia. Visiting San Marino on second day Scotland could make some points to be in favorable position right at the beginning.

J) Italy hosts Finland and then Liechtenstein; while Bosnia and Herzegovina hosts first Armenia and the Greece... when Finland has to visit Armenia on second day.

The door to an 'agreable start' is wide open for England, Croatia, Belgium, Portugal, France, Spain and Italy ... but also teams like Israel or Scotland...

[Ricardo]

I don’t agree.
The “probability to advance” = “probability to be first placed” + “probability to be second placed”.
Saying that the strongest opponent doesn’t matter because you can still advance ALONGSIDE it and don’t have to finish ABOVE it is shortsighted.

A team aiming for the top has a better chance to advance than a team that isn’t even aiming for the top and is already putting itself in a secondary role.

yes, but calculate it yourself, you will see that the chance of qualifying for Holland is better with Germany and Northern Ireland as with Sweden and Turkey? Apparently most of us feel that intuitively, but can't get the numbers behind it.
For this the numbers shouldn't be hard to find.

[Lorric]

I don’t agree.
The “probability to advance” = “probability to be first placed” + “probability to be second placed”.
Saying that the strongest opponent doesn’t matter because you can still advance ALONGSIDE it and don’t have to finish ABOVE it is shortsighted.

A team aiming for the top has a better chance to advance than a team that isn’t even aiming for the top and is already putting itself in a secondary role.

yes, but calculate it yourself, you will see that the chance of qualifying for Holland is better with Germany and Northern Ireland as with Sweden and Turkey? Apparently most of us feel that intuitively, but can't get the numbers behind it.
For this the numbers shouldn't be hard to find.

There was a model posted further back in the thread. It had % chance to qualify both before the draw and after it. Despite being put together, both Germany and Netherlands had their chances improved. NI though took easily the biggest hit to their chances out of everyone.

[HibeeJibee]

What are probabilities of Germany & Netherlands, Denmark & Eire and/or Finland & Greece entering playoffs?

If so they could meet 5 times in 1 qualifying campaign, having met in NL + qualifiers.

[Vickzq]

What are probabilities of Germany & Netherlands, Denmark & Eire and/or Finland & Greece entering playoffs?

Netherlands, Denmark and Finland have 'play-offs' guaranteed IF they don't qualify during group qualifier stage. So their probability to enter play-offs should be 100% minus the predicted (has to be published somewhere) probability for them to qualify directly (and not needing the play-offs).
For Germany, Ireland and Greece it should be their probability 'to fail' (100% minus predicted chances) multiplied with the probabilities of every team (except 2 others, to make a 4-team play-off) ranked 'ahead of them', which has to qualify to leave the play-off spot to them.
I will leave out Croatia & Poland... this is - honestly - rather random... but let's make this approximation. I added them in another try, and it didn't change much in context of probability.

Well, might be an error somewhere - but it sounds right just now
I will take % from this simulation of qualifiers - although it seems they modified it later after draw:
http://simtheworld.blogspot.com/2017/10 ... tions.html

- Netherlands 14.7% (100%-85.3%)
- Switzerland 84.4%
- Portugal 96.6%
- England 91.7%
- Belgium 92.1%
- France 95.4%
- Spain 97.7%
- Italy 92.8%
- Germany 0.6% (100%-99.4%)

I end up with ~0.05% probability to see Germany and the Netherlands again together participating both in UNL play-offs.
Does this seem half-way reasonable, or is there a major bug?

[Witkop1983]

I guess the probability of Germany and Netherlands meeting again in the Euro 2020 group stage is much higher than them meeting in the playoffs
(and then maybe in the knockout too to make it 6 times?)

Edit: oops...... they are both hosting group stages so they cant be together. So maybe knockout then?

Although teams from group C and F only meet in the semi's or final unless one of them finishes third in their group

[amirbachar]

What are probabilities of Germany & Netherlands, Denmark & Eire and/or Finland & Greece entering playoffs?

Netherlands, Denmark and Finland have 'play-offs' guaranteed IF they don't qualify during group qualifier stage. So their probability to enter play-offs should be 100% minus the predicted (has to be published somewhere) probability for them to qualify directly (and not needing the play-offs).
For Germany, Ireland and Greece it should be their probability 'to fail' (100% minus predicted chances) multiplied with the probabilities of every team (except 2 others, to make a 4-team play-off) ranked 'ahead of them', which has to qualify to leave the play-off spot to them.
I will leave out Croatia & Poland... this is - honestly - rather random... but let's make this approximation. I added them in another try, and it didn't change much in context of probability.

Well, might be an error somewhere - but it sounds right just now
I will take % from this simulation of qualifiers - although it seems they modified it later after draw:
http://simtheworld.blogspot.com/2017/10 ... tions.html

- Netherlands 14.7% (100%-85.3%)
- Switzerland 84.4%
- Portugal 96.6%
- England 91.7%
- Belgium 92.1%
- France 95.4%
- Spain 97.7%
- Italy 92.8%
- Germany 0.6% (100%-99.4%)

I end up with ~0.05% probability to see Germany and the Netherlands again together participating both in UNL play-offs.
Does this seem half-way reasonable, or is there a major bug?

There is a major bug... The events are not independent, for example for both Germany and The Netherlands not to qualify means not only Northern Ireland pass both of them, but also another team from that group, which is very very unlikely.
I will post in the evening the probabilities for the simulations results for a 5th meeting.
I'll post here the PO semi finals and finals probability as well.

[amirbachar]

Playoffs probabilities:

Code: Select all

Team	Total PO	Q-PO	PO-Final	PO-semis
Switzerland	12.96%	6.85%	4.12%	1.99%
Portugal	9.75%	5.52%	2.78%	1.45%
Netherlands	3.29%	1.91%	0.92%	0.47%
England	3.03%	1.66%	0.93%	0.45%
Belgium	2.05%	1.32%	0.50%	0.23%
France	0.36%	0.24%	0.09%	0.04%
Spain	7.59%	4.80%	1.88%	0.91%
Italy	9.18%	4.22%	2.87%	2.10%
Croatia	10.50%	5.44%	3.03%	2.03%
Poland	23.30%	6.96%	7.90%	8.44%
Germany	6.03%	3.02%	1.68%	1.32%
Iceland	58.71%	12.02%	18.94%	27.76%
Bosnia and Herzegovina	25.95%	9.54%	9.06%	7.36%
Ukraine	51.31%	17.55%	17.84%	15.93%
Denmark	16.67%	7.84%	4.68%	4.15%
Sweden	36.72%	15.06%	10.31%	11.36%
Russia	43.30%	11.93%	12.46%	18.92%
Austria	20.32%	5.67%	5.58%	9.07%
Wales	49.21%	13.22%	13.21%	22.78%
Czech Republic	42.93%	9.47%	10.71%	22.75%
Slovakia	53.43%	12.66%	13.18%	27.59%
Turkey	48.85%	8.09%	10.67%	30.09%
Republic of Ireland	71.49%	12.05%	13.88%	45.56%
Northern Ireland	80.55%	8.22%	13.22%	59.12%
Scotland	56.09%	21.27%	18.63%	16.20%
Norway	85.78%	27.07%	27.56%	31.16%
Serbia	39.25%	16.28%	9.79%	13.19%
Finland	82.51%	14.61%	18.77%	49.13%
Bulgaria	75.41%	12.19%	16.13%	47.10%
Israel	77.33%	7.69%	12.99%	56.65%
Hungary	63.19%	7.94%	10.20%	45.05%
Romania	30.06%	7.31%	5.02%	17.73%
Greece	2.33%	0.36%	0.42%	1.55%
Albania	0.47%	0.04%	0.07%	0.36%
Montenegro	0.08%	0.01%	0.01%	0.06%
Cyprus	0.01%	0.00%	0.00%	0.01%
Estonia	0.00%	0.00%	0.00%	0.00%
Slovenia	0.00%	0.00%	0.00%	0.00%
Lithuania	0.00%	0.00%	0.00%	0.00%
Georgia	95.68%	33.63%	29.11%	32.94%
Macedonia	90.03%	27.74%	30.22%	32.06%
Kosovo	94.30%	16.99%	19.97%	57.35%
Belarus	97.86%	19.93%	17.66%	60.27%
Luxembourg	20.46%	1.47%	2.73%	16.26%
Armenia	1.60%	0.23%	0.30%	1.06%
Azerbaijan	0.08%	0.01%	0.01%	0.06%
Kazakhstan	0.00%	0.00%	0.00%	0.00%
Moldova	0.00%	0.00%	0.00%	0.00%
Gibraltar	0.00%	0.00%	0.00%	0.00%
Faroe Islands	0.00%	0.00%	0.00%	0.00%
Latvia	0.00%	0.00%	0.00%	0.00%
Liechtenstein	0.00%	0.00%	0.00%	0.00%
Andorra	0.00%	0.00%	0.00%	0.00%
Malta	0.00%	0.00%	0.00%	0.00%
San Marino	0.00%	0.00%	0.00%	0.00%