Leading up to the 2018 World Cup, there were many predictions made about which team would win the tournament. The predictions used various methods, such as analyzing bookmaker odds, using recent results and statistics, and player and team rankings.
There was no consensus for the predicted winner, though Germany and Brazil were mostly listed to finish first or second. The tournament was won by France. No prediction picked the winner correctly, and only half of them had France in the top four, and usually placed fourth behind Brazil, Germany and Spain.
The best prediction? It is hard to say, but the only one to have France not finishing fourth or worse was the prediction by Goldman Sachs, which used a combination of crunching data on team characteristics, individual players and recent team performance.
Table of 2018 Predictions
|Zeileis et al.||Brazil (16.6%)||Germany (15.8%)||Spain (12.5%)||France (12.1%)|
|Ekstrøm||Germany (25.4%)||Brazil (20.1%)||Belgium (~11%)||Spain (~10.5%)|
|Gracenote||Brazil (~22%)||Spain (~10%)||Germany (~8%)||Argentina (~8%)|
|Groll et al.||Spain (17.8%)||Germany (17.1%)||Brazil (12.3%)||France (11.2%)|
|Audran et al.||Germany (24.0%)||Brazil (19.8%)||Spain (16.1%)||England (8.5%)|
|Llaneras & Andrino||Brazil (17.9%)||Germany (15.6%)||Spain (15.3%)||Argentina (11.3%)|
|Goldman Sachs||Brazil (18.5%)||France (11.3%)||Germany (10.7%)||Portugal (9.4%)|
|FiveThirtyEight||Brazil (19%)||Spain (17%)||Germany (13%)||France (8%)|
|Gilch & Müller||Germany (26.0%)||Brazil (13.2%)||Spain (11.2%)||Argentina (9.2%)|
|Yuan||Brazil (16.2%)||Germany (14.3%)||Poland (11.7%)||France (8.0%)|
|Consensus||Brazil (42)||Germany (37)||Spain (21)||France (8)|
- Zeileis - Zeileis A, Leitner C, Hornik K (2018). "Probabilistic Forecasts for the 2018 FIFA World Cup Based on the Bookmaker Consensus Model", Working Paper 2018-09, Working Papers in Economics and Statistics, Research Platform Empirical and Experimental Economics, Universität Innsbruck. METHOD: based on a consensus model using quoted betting odds of 26 bookmakers. https://econpapers.repec.org/paper/innwpaper/2018-09.htm
- Ekstrøm - Claus Ekstrøm's from UCPH Biostatistics has created an algorithm using Elo ratings and Bet365 betting odds, then run thousands of tournament simulations. Results published in Danish: "Who will win the World Cup 2018 and who is best to preach it?" http://sandsynligvis.dk/2018/06/14/hvem-vinder-vm-2018-og-hvem-er-bedst-til-at-pr%C3%A6diktere-det/
- Gracenote - Gracenote Sports methodology is based on the Elo system. Every time two teams compete, one "wins" points from the other depending on the result and the following: 1) the winning margin, 2) which team is at "home," and 3) the relative strength of the two teams. The importance of the competition and match is also factored into the calculation.http://www.gracenote.com/sports/fifa-world-cup-predictions-2018/
- Groll et al. - Andreas Groll, Christophe Ley, Gunther Schauberger, Hans Van Eetvelde. Prediction of the FIFA World Cup 2018 – A random forest approach with an emphasis on estimated team ability parameters. Despite Spain being the most probable to win the final, after 100,000 simulations, the most probable tournament course was for Germany to beat Brazil in the final. To make sense of that you need to read their paper: https://arxiv.org/pdf/1806.03208.pdf
- Audran et al. - Audran, J., M. Bolliger, T. Kolb, J. Mariscal, and Q. Pilloud (2018): "Investing and football - Special edition: 2018 World Cup in Russia," Working paper, UBS. The researchers from Swiss bank UBS used a statistical model based on four factors that are supposed to indicate how well a team will be doing during the tournament: the Elo rating, the teams' performances in the qualifications preceding the World Cup, the teams' success in previous World Cup tournaments and a home advantage. https://www.ubs.com/content/dam/assets/wm/global/cio/doc/investing-in-emerging-markets-en.pdf
- Llaneras & Andrino - prediction by Kiko Llaneras and Borja Andrino as published on El Pais, based significantly on the Elo ranking, then simulatiing 10,000 alternative versions of the tournament.https://elpais.com/deportes/2018/06/04/actualidad/1528063475_409937.html
- Goldman Sachs - Using AI, crunching data on team characteristics, individual players and recent team performance. http://www.goldmansachs.com/our-thinking/pages/world-cup-2018/multimedia/report.pdf
- FiveThirtyEight - based on statistical models by Jay Boice and Nate Silver using ESPN's Soccer Power Index (SPI) ratings, a combination of each team's recent match results and the overall quality of its World Cup roster.https://projects.fivethirtyeight.com/2018-world-cup-predictions/
- Gilch & Müller - Lorenz A Gilch and Sebastian Müller. On Elo based prediction models for the FIFA Worldcup 2018. A prediction based on Poisson regression models that include the Elo points of the teams as covariates and incorporates differences of team-specific effects. https://arxiv.org/pdf/1806.01930.pdf
- Yuan - Andrew Yuan, a Brazilian software engineer, based his predictions on match results since 1930, ranking tables since 1993 and game location data. http://andrewyuan.github.io/FWC2018_prediction.html
- Elo Ratings - The World Football Elo Ratings are based on the Elo rating system, developed by Dr Arpad Elo. Ratings use the results of recent matches. The prediction lists the top four rated teams before the start of the tournament. https://www.eloratings.net/2018_World_Cup_start
- FIFA Rankings - official FIFA team ranking data from 7th June 2018https://www.fifa.com/fifa-world-ranking/ranking-table/men/index.html
- Consensus - my summary of all the predictions listed. I have given each of the top 4 predictions a score from 4 to 1 to calculate which team was more often predicted to win the tournament.
- Other World Cup Predictions
- Results Russia 2018
- Olympic Games medal predictions
- Winter Olympics Medal Predictions