高盛-世界杯与经济学-全球宏观经济研究-20180611-49页_1mb

2018 World Cup and Economics Summary

Core Content

This document is the sixth edition of a book by Goldman Sachs on the World Cup and Economics, focusing on the 2018 World Cup hosted by Russia. It combines economic analysis with football predictions, using a statistical model that incorporates both team-level and player-level data. The model is based on machine learning techniques, including random forest, Bayesian ridge regression, Lasso regression, and gradient boosted machines, and it forecasts match outcomes by simulating 1 million possible tournament paths.

Main Predictions

• Brazil is expected to win the 2018 World Cup, defeating Germany in the final on July 15th.
• France has a higher probability of winning the tournament than Germany, but faces Brazil in the semi-finals, which could be a challenge.
• Germany is predicted to beat England in the quarter-finals on July 7th.
• Spain and Argentina are expected to underperform, losing to France and Portugal in the quarter-finals, respectively.
• Russia is forecasted to fail to progress past the group stage, despite hosting the tournament. It is ranked 66th in the FIFA rankings, the lowest ever, compared to its 13th position in 2010.

Key Insights

• Team-level results are the most significant factor in predicting success, contributing about 40% to the model's explanatory power.
• Player-level characteristics add another 25% of explanatory power, highlighting the importance of individual talent.
• Recent momentum (wins/losses ratio over the past ten matches), goals scored, and goals conceded are also important variables.
• Brazil is the strongest team, with the highest Elo rating, talented players, and a good win/lose ratio.
• France and Germany are in a close race for second place, but France's draw is more challenging.
• Spain's chances are diminished due to a tough draw and poor recent performance.
• Argentina's performance is affected by recent poor results, despite having a high Elo rating.

Statistical Model

• The model combines 200,000 probability trees and 1 million simulations to produce forecasts.
• The Elo rating system is used to measure team strength, originally developed for chess.
• The model was five times more accurate than the previous Poisson-based regression.
• The inclusion of friendships in the data slightly reduced model accuracy, possibly due to experimental tactics or weaker teams being fielded.

Interview with Carlo Ancelotti

• Ancelotti believes Brazil and Spain are the two teams most likely to win the World Cup.
• He highlights France and Germany as strong contenders, and Argentina, with Messi, as a potential surprise.
• He suggests that Belgium and Croatia are underrated and could surprise people.
• Ancelotti notes that the World Cup is less prone to surprises than the European Championship.
• He also mentions that while China and USA are improving, it is unlikely they will win in his lifetime.
• Managing a national team is different from managing a club, as it requires a focused preparation in a short period.

Russia's Performance

• Russia was the host nation and did not have to qualify, but its preparations were less than ideal.
• It is drawn with Saudi Arabia, Uruguay, and Egypt in a group of emerging market economies.
• Uruguay is the clear favorite in the group, with a FIFA ranking of 17.
• Stanislav Cherchesov is the coach, who has a history of success in Austria and the domestic league.
• The Russian team has a weak defense, with several key defenders missing due to injuries.
• Alexander Kokorin is injured, increasing pressure on Fyodor Smolov to perform.
• The midfield is uncertain, with Alexandre Golovin, Roman Zobnin, Daler Kuzyaev, and Alan Dzagoev likely to compete for positions.

Additional Sections

• A World Cup dream team was selected by clients.
• Group stage results are provided for several groups, showing the performance of teams.
• Exhibits are included to illustrate the model's results and the importance of team and player characteristics.
• Disclosures are noted, emphasizing the uncertainty of the forecasts and the stochastic nature of football.

Conclusion

The document highlights the economic and footballing insights from the 2018 World Cup, emphasizing the importance of data and modeling in predicting outcomes. It also discusses the challenges faced by Russia, both in terms of its footballing performance and economic conditions, drawing parallels between the two. The publication serves as a guide and companion for the tournament, offering interesting forecasts and analytical insights.