Skip to content

Which country will win the World Cup?

June 14, 2018

socceroosOn that question, the understatement of the year has to go to CNBC: “Predicting the winner of a soccer World Cup isn’t an exact science” (UBS uses stock-picking tech to predict the winner of the 2018 soccer World Cup).

After reading about Goldman Sachs in the AFR using artificial intelligence to predict the World Cup winner, I wondered how many other people were using data analytics and modelling to pick the winner this year – and also how much store we should put on big data as a way to accurately forecast future outcomes.

It turns out there are a lot of pundits out there applying different technology approaches to the challenge, and willing to back themselves publicly with their predictions.

South African data analytics firm Principa is using “analytics and machine learning” to predict the results for every single 2018 FIFA World Cup match, but is clever enough to only commit to predictions round by round. Principa’s first round results are here.

A number of the big banks have applied “artificial intelligence, statistical modeling, portfolio theory, and economic analysis” to pick their winner (What country will win the 2018 World Cup? Here are big banks’ predictions).

Lloyds of London based its prediction on the Insurable value of the players, and experts from Germany’s Commerzbank analysed historical sports data over years, including home bias and the number of goals scored in previous World Cups, world rankings, and track record at previous tournaments, running 10,000 simulations on each game. Not surprisingly, Germany came out on top.

According to data company Gracenote, Brazil is the statistical favourite to win the World Cup in Russia, while Alteryx applied its analytics model to the 2014 World Cup to validate its 2018 predictions: “We took the 16-team 2014 round data and applied our model to it – we actually only got two results wrong” (Germany to win and England bow out to Brazil: Analytics firm predicts every single result in the World Cup 2018). They are much braver than Principa!

The Democracy Institute’s econometric soccer rankings “signal that Brazil, France, Spain, and Argentina are poised to under-achieve” while Uruguay, Switzerland, England, Sweden, Iceland, Denmark, Morocco, and Australia are likely “to exceed expectations.” (The freedom to win: World Cup 2018 economic predictor)

Team Twelve’s “Alpha Ball” system uses data derived from recent matches in 20 categories such as average attack success rates and attack and defence patterns, and EA Sports EA Sports used simulation created with the data from FIFA 18 to predict France as this year’s winner.

According to Sky News, bookmakers are backing Brazil while scientists using new machine learning (ML) techniques, including a method called the “random-forest approach”, have picked Germany. (How scientists are attempting to predict the World Cup 2018 winner)

Some other academics can’t split Brazil and Germany.

“Soccerbot”, developed by an applied mathematician who wrote Soccermatics, sounded promising, but I was put off finding out more when Australia (pictured above from a recent game in Canberra) was given odds of 500-1 to win, and I had to click through to a sports betting site to get more information. And weirdly, UK’s Telegraph wants us to apply our own rating of the key data factors – and to login to Facebook to “play”. I’m not prepared to do that.

While there are so many different analytical models out there that can be applied, what was interesting was that there were only four countries predicted to win across all the pundits reviewed: Brazil, France, Germany and Spain.

So I thought, why not use the work already done to make my own prediction? (‘Standing on the shoulders of giants’ and all that.)

I tabulated the predictions, and made some weighting decisions. All the banks had a weighting of ‘one’ each (can we really trust them?); as an insurance company, Lloyds is probably more careful with its data analysis (and had the impressive-sounding Centre for Economics and Business Research helping with the analysis as well as predicting the outcome of the 2014 World Cup), so scored a ‘two’; and Gracenote and Alteryx as generalist data firms were also given a ‘one’ each.

Although the Democracy Institute expects Australia to do well in the tournament and impressively “draws upon the data provided in the Heritage Foundation’s “2018 Index of Economic Freedom”, the vagueness of its prediction scored it a ‘one’.

I included data from the official FIFA world rankings, giving this a ‘two’ weighting; and EA Sports also gets a ‘two’, for past history, successfully predicting the World Cup winners of 2010 and 2014. Team Twelve also got a ‘two’ as a specialist sports data firm.

Weightings of ‘three’ were reserved for the bookmakers – who have the most to lose by getting it wrong – and “scientists” and “statisticians” who, as academics, are perhaps the most objective in the whole thing.

So what were my findings?

Applying the weighting, it’s Brazil to win the 2018 World Cup (34%) ahead of Germany (30%), with France and Spain equal in third place (18%).

Here’s the table of predictions.

Predicting Entity Winning Team Weighting
Principa Not brave enough to predict N/A
UBS Germany 1
Goldman Sachs Brazil 1
ING Spain 1
Nomura France/Spain 1
Lloyds of London France 2
“Bookmakers” Brazil 3
“Scientists” Spain 3
Democracy Institute Germany 1
Gracenote Brazil 1
FIFA Germany 2
Alteryx Germany 1
EA Sports France 2
Commerzbank Germany 1
Team Twelve Brazil 2
Soccerbot I’m not going to a betting site N/A
“Statisticians” Brazil/Germany 3
Weighting Total   25

 

Although I’ve predicted Brazil to win, you can’t count out Germany. Former England striker Gary Lineker might be right: “Football is a simple game. Twenty-two men chase a ball for 90 minutes and at the end, the Germans win.”

To be honest, I think the most accurate prediction will be a democratic one. More than 500,000 people worldwide have already predicted the winner in the FIFA World Cup Bracket Challenge. Hopefully they’ll announce the people’s choice before the World Cup starts tonight!

Whatever happens, the next month will be amazing.

 

No comments yet

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: