Friday, June 11, 2010

World cup Domination

This article is from I figured some of you degenerates would want to put some cash down on the world cup.

I was given a gift the other day. And as soon as I opened it, I knew it was special.

It came from a reader, in an e-mail. His note was simple, along the lines of "Thought you might enjoy this."

He was right.

The single coolest trend in sports betting is the advancement in the use of algorithms. For the past decade the Strip has been filling up with reformed financiers and ex-accountants and actuaries who are applying all the math they spent their lives perfecting to predict who will win and by how much. Lately the bookmakers, like the guys doing in-running wagering at the M and the financiers in London who recently started a sports betting hedge fund, have been trying it too.

It makes sense, then, with the World Cup just two weeks away, that the brightest minds at the biggest bank on the planet would do their own quantitative study of who will win that golden FIFA trophy. What else are they going to do with all that knowledge? (Spare me any moaning about how these Richie Riches should be spending their time trying to make sure we don't suffer another global economic meltdown. If you say you don't blow off steam at work, you're lying.)

The report -- which is what my very kind degenerate reader had e-mailed me -- is officially called "A Quantitative Guide to the 2010 World Cup." It was put together by bankers at JP Morgan who are based overseas. I e-mailed Matthew Burgess, whose name is at the top of the list of authors, but he declined my request to be interviewed. And, while the analysts included a disclaimer on the title page that says "This report should be taken with a pinch of salt," it's a brilliant look at how to think about the tournament. And, as the authors also wrote, "an interesting exercise and ideal opportunity to lightheartedly explain Quantitative techniques."

To build their model, the bankers used data that was readily available to the public, no differently than what they do when determining a company's valuation. They based their results upon the following factors (Warning: If you don't have the stomach for lists and long division, skip to the end now to find out who will win the World Cup. I won't be offended):

The basic criteria

1. The official FIFA World Rankings (which take into account match results, match status, opponents strength, regional strength and results over the previous four years).
2. The latest odds for each team to win it all.
3. Betting trends on each team over the previous six months.
4. Results from previous World Cup tournaments and qualifying competitions.
5. Recent form from the previous 12 months.
6. Consistency in public sentiment for each team.
7. Team fundamentals.
8. Success ratio.

Now, if you looked at just the rankings (factor No. 1), your best bets are, in order: Brazil, Spain, Portugal, The Netherlands and Italy.

True, the good ol' USA isn't mentioned much in this post. But to keep up with the squad, keep checking the U.S. national team blog on Insider. U-S-A!

All of those are good choices. But any wiseguy will tell you that betting based on a single factor is the first step toward a losing bet. Games are complex, and so are the people who play them. The Morgan boys know this, which is why they take into account factor No. 2: the odds to win. This is a number established by the bookmakers but manipulated by the marketplace. The most talented teams always bubble to the top -- whether it's soccer, football or tiddlywinks -- but the public teams sneak in there, too. Reminder, the public teams are the ones everyone loves and roots for no matter what. Like the Steelers in football or England in soccer. When the report was published in mid-May, the favorites were: Spain (FIFA rank: 2), Brazil (1), England (8), Argentina (7) and Italy (5).

So now the bankers have two factors to build a formula around. But they took it further. They examined what they called "Price Trend Metrics." In their world, that reflects investors' opinions of a stock and whether or not it is trending up or down. Translated to soccer, that meant taking a look at how a team's probability to win has tracked over a six-month period. According to that metric, the biggest increases were for Slovenia, France, Ivory Coast, Greece and Uruguay.

Still, that wasn't enough info to build the quant formula they wanted to. Next on their list was establishing a number for each country's result expectations. This was done by assigning points for every first (50 points), second (25), third (15) and fourth (10) finish in the team's history. Using that method, the leaders are Brazil (340 points), Germany (305), Italy (275), Argentina (125) and France (115).

Getty ImagesCacau and Germany are close via quant models, but in the end, there's no cigar.

Next, they evaluated form. For all soccer bettors, recent form is the most important factor they consider when making a bet. But during the World Cup, it's hard to quantify what that is in the early rounds of the tourney. The Morgan guys determined this by taking the average FIFA World Ranking points over the previous 12-month period. But they only wanted to include World Cup qualifiers, so their data period ended Dec. 31, 2009.

In that category, the leaders were Netherlands (1,121 points), Spain (1,039), Brazil (894), Portugal (859) and Chile (811).

There's more. When bankers examine a company, they look at its fundamentals, meaning whether it has delivered a healthy return on investment over the long haul. Consistency is important in any kind of gambling, whether it's on the stock market or the sports market. To examine the fundamentals of the teams in the World Cup, the Morgan boys created a formula that accounted for the probability of each team winning from a variety of sources offering odds. Their formula was: average probability of winning divided by max probability of winning minus minimum probability of winning. Here, the leaders were:

1. Brazil
2. England
3. Spain
4. Ivory Coast
5. Argentina

We continue. Every wiseguy loves to know a team's strength of schedule. The bankers were no different. They created a "Success Ratio" based on a team's winning percentage over the previous 12 months and its opponents' average strength of schedule. According to that formula, the best teams are:

1. Netherlands
2. Spain
3. Chile
4. Brazil
5. England

That's a lot of data to consider. But finance guys are used to working out messy problems. This is where the quantitative analysis comes in. When they are analyzing a company via a quant model, their formula is made up of four categories: Valuation Metrics, Market & Analyst Sentiment, Company Fundamentals and Price Trend.

Getty ImagesKaka and Brazil are favorites of many, but not these mathematical models.

Each of the eight statistical sets outlined above was put into one of these four categories, then each category was weighted to make up a different proportion of the formula, with some being more important than others, and each piece of data within the category was weighted, too. It looked like this:

Valuation Metrics (40 percent)
• Odds (50 percent)
• FIFA Ranking (50 percent)

Company Fundamentals (15 percent)
• Success Ratio (50 percent)
• Consistency (50 percent)

Market and Analyst Sentiment (15 percent)
• Result Expectations (33 percent)
• Recent Form (67 percent)

Price Trend (30 percent)
• Trend in probability to win (50 percent)
• FIFA Ranking Trend (50 percent)

Once Morgan's quant squad put all this data into its superconducting spreadsheets, it spit out a model score. That number indicated, based on all the information available, what the strongest teams in the World Cup are. This is no different than a sharp bettor in Vegas making a power rating for college hoops team. They take into account factors such as defensive and offensive efficiency and returning starters to get a score for each team. The difference in scores between two teams that are playing becomes what the gamblers think the final score of the game should be. According to Morgan, the strongest World Cup teams are:

1. Brazil (Model score of 1.68)
2. Spain (1.53)
3. England (0.91)
4. Netherlands (0.63)
5. Argentina (0.48)

But the model score formula did not account for the possibility of a draw, so another formula designed to calculate a team's Penalty Shoot-out metric was created. This one was achieved by taking the number of goals scored divided by the number of games played and multiplied by 50 percent. From that was subtracted the number of goals conceded divided by the number of games played times 50 percent.

Getty ImagesIn the end, it appears England may hoist one of the world's most coveted trophies this summer.

In this formula, England was far and away the leader, as seen below:

1. England (1.65)
2. Spain (1.29)
3. Netherlands (1.13)
4. Germany (1.09)
5. Ivory Coast (0.78)

Finally, the Morgan team applied their model scores and Penalty Shoot-out metrics to every game. The team with the higher score advanced. Here's how it shook out.

In the quarterfinals, Netherlands beat Brazil, England beat France, Slovenia beat Argentina and Spain beat Italy.

In the semis, England defeated the Netherlands and Spain toppled Slovenia.

And in a thrilling shootout in the finals, England -- with the higher Penalty Shoot-out metric -- took down top-ranked Spain.

So there you have it. England will win the World Cup.

Consider this my gift to you.

No comments: