Simulating the 2022 World Cup

In 2014, I simulated Brazil to win the World Cup. We saw how that went. I did that again in 2018. We saw how that went!

Well, here we are again. I didn't lose my previous code this time, so I've generated another composite ranking, this time comprising 5 models:

• FiveThirtyEight
• Ed Feng at The Power Rank
• Massey Ratings by Ken Massey
• Elo Ratings  (weighted lower)
• FIFA rankings (weighted lower, for good reason)

I then used this model's output, along with the parameters of 0.4 goals for home-field for Qatar (down from 0.5 previously), 1.62 as the standard deviation in margin of victory, and simulated the World Cup 10,000 times to get the following probabilities. It's nice to once again be able to include the color banding for the US:

Group	Country	MDS Rating	Avg Points	1st in Group	2nd in Group	Advance	Quarters	Semis	Final	Champion
A	Netherlands	2.02	5.98	62.02%	22.03%	84.05%	53.73%	31.01%	16.39%	9.08%
A	Ecuador	1.29	3.97	17.55%	35.50%	53.05%	23.01%	9.38%	3.12%	1.17%
A	Senegal	1.21	3.74	13.16%	26.13%	39.29%	16.68%	6.32%	2.13%	0.61%
A	Qatar	0.91	2.96	7.27%	16.34%	23.61%	10.02%	4.31%	1.57%	0.57%
B	England	1.89	5.50	53.67%	25.55%	79.22%	46.53%	25.22%	13.05%	6.22%
B	United States	1.45	4.32	22.79%	35.05%	57.84%	26.50%	11.30%	4.67%	1.75%
B	Wales	1.18	3.56	13.47%	22.25%	35.72%	13.93%	5.34%	1.68%	0.60%
B	Iran	1.06	3.23	10.07%	17.15%	27.22%	9.60%	3.72%	1.19%	0.40%
C	Argentina	2.02	5.86	59.73%	23.71%	83.44%	50.69%	30.29%	15.89%	8.81%
C	Mexico	1.38	4.17	19.80%	35.47%	55.27%	23.47%	10.69%	4.18%	1.45%
C	Poland	1.26	3.86	14.48%	26.39%	40.87%	16.45%	6.52%	2.35%	0.82%
C	Saudi Arabia	0.85	2.78	5.99%	14.43%	20.42%	6.38%	2.01%	0.64%	0.13%
D	France	1.94	5.56	52.35%	27.57%	79.92%	47.51%	27.09%	13.64%	7.17%
D	Denmark	1.76	5.05	31.88%	38.70%	70.58%	37.87%	20.31%	9.59%	4.44%
D	Australia	1.03	3.08	8.70%	18.45%	27.15%	9.82%	3.77%	1.06%	0.28%
D	Tunisia	1.00	2.96	7.07%	15.28%	22.35%	7.81%	2.72%	0.78%	0.18%
E	Spain	2.06	5.53	51.61%	27.91%	79.52%	48.43%	26.51%	15.33%	8.56%
E	Germany	1.91	5.15	32.70%	38.88%	71.58%	39.86%	19.91%	11.06%	5.57%
E	Japan	1.37	3.64	11.50%	22.55%	34.05%	14.44%	5.74%	2.39%	0.85%
E	Costa Rica	0.84	2.38	4.19%	10.66%	14.85%	4.38%	1.16%	0.33%	0.08%
F	Belgium	1.96	5.25	49.67%	25.41%	75.08%	41.86%	21.80%	12.33%	6.10%
F	Croatia	1.57	4.25	24.03%	32.26%	56.29%	24.73%	10.92%	4.85%	2.13%
F	Canada	1.35	3.60	14.12%	23.21%	37.33%	14.12%	5.30%	2.06%	0.70%
F	Morocco	1.29	3.50	12.18%	19.12%	31.30%	12.18%	4.23%	1.54%	0.59%
G	Brazil	2.42	6.35	67.98%	20.36%	88.34%	61.47%	40.16%	26.37%	17.57%
G	Switzerland	1.57	4.07	15.59%	39.61%	55.20%	25.77%	11.33%	5.35%	2.17%
G	Serbia	1.50	3.92	12.45%	28.65%	41.10%	18.80%	8.16%	3.37%	1.37%
G	Cameroon	0.91	2.37	3.98%	11.38%	15.36%	4.54%	1.36%	0.54%	0.12%
H	Portugal	1.99	5.91	58.26%	25.52%	83.78%	46.07%	25.47%	14.58%	7.41%
H	Uruguay	1.63	4.93	27.46%	41.75%	69.21%	30.33%	13.97%	6.53%	2.65%
H	South Korea	1.10	3.50	10.25%	21.71%	31.96%	9.82%	3.17%	1.19%	0.36%
H	Ghana	0.64	2.37	4.03%	11.02%	15.05%	3.20%	0.81%	0.25%	0.09%

• Group of Death: 
• Group G: Brazil/Switzerland/Serbia/Cameroon
• Narrowly beating out E and F, although F has the strongest weakest team, Morocco
• Top contenders (> 50% one of these teams wins):
• Brazil, 17.6%
• Netherlands, 9.1%
• Argentina, 8.8%
• Spain, 8.6%
• Portugal, 7.4%
• Odds a country outside of Europe/South America wins: 10%
• Top teams by my ratings that missed out:
• Italy, #7 in the world, better than 23 World Cup teams
• Colombia, #13, better than 21
• Peru, #21, better than 14
• Hungary, #25, better than 11
• Ukraine, #26, better than 11
• Honorable mention, worst team in the ranking set: Bhutan, "a Buddhist kingdom on the Himalayas’ eastern edge"

The above is maybe not the best omen, given my track record, if you're for Brazil (and I'm less optimistic than FiveThirtyEight, and recently it's been a good idea to fade their sports predictions). But this is a tournament being played out of season in winter, gained via bribery, accomplishing its goal of sportswashing - and we're all going to watch it anyway!