Predictions II: Making the 2nd Round and Quarterfinals

Four years ago, we were living in England as the tournament was getting underway. Excitement and expectations were high; as others and I have noted before, probably unrealistically. But I also vividly recall my friend David’s state of mind. Although he now lives in England, David grew up in Spain, and growing up in Spain clearly shaped his expectations about World Cup success (or failure, as it may be) the way growing up in German shaped mine. In particular, I remember his wonderment about the English optimism as well as his own almost complete certainty that Spain was not going to go anywhere in the tournament.

And yet, every once in a while, I detected a glimmer of hope in David’s eyes and the occasional hopeful comment about the Spanish team. So, the other day I had an email from him that he was feeling increasingly excited about Spain’s chances this year, but this statement was immediately followed by the question: Is this just “irrational exuberance?”

So, let’s look at Spain’s prospects (and everyone else’s) to see how irrational David’s optimism is – statistically speaking of course.

Predicting World Cup Success, Take I: Recent Form and Home Continent

Following up on the post about Measuring World Cup Success, I thought I’d take a look at what might predict a successful 2010 World Cup. And along the way, I thought I’d try to make my friend Dan’s life easier when picking winners in his World Cup brackets (or at least have something to talk about around the water cooler - do those things still exist?).

Recall that I earlier measured “success” in three ways: matches won, points won, and advancing in the tournament. Since matches won and points won are similar (more wins = more points), they are highly correlated to the tune of .98. So, for starters, let’s take a look at the odds of advancing in the tournament and winning matches. And, to make it easier and more tractable, let’s look at World Cups since the inception of the 32-team format – that is, since 1998.

The current format of the tournament prescribes that 50% of all teams play in the second round. It also prescribes that the maximum number of games played is 7 (3 in the first round, and then another four up to the final). Since 1998, the distribution of wins is quite uneven, with about a third of all teams failing to win a single match, while the World Cup was won by teams with 6 (Italy, 2006) or 7 wins (France, 1998; Brazil, 2002):

A number of academic articles have been written about success in international soccer generally. But only a modest amount of research has been done on explaining World Cup outcomes specifically (there are about a handful of studies). Instead, the majority of analyses have focused on explaining FIFA rankings, which are taken to be indicators of international soccer success. (A good example of this kind of work is the recent paper by Leeds and Leeds in theJournal of Sports Economics.)

Measuring World Cup "Success"

My friend Dan is an executive with a Fortune 500 company. In his job, he travels globally and works with people from all walks of life and all around the world. Of course, wherever he goes these days, one of the few things almost everyone is interested in talking about is the World Cup. Since Dan recently was in charge of the NCAA basketball tournament office pool and did such a bang-up job with it, his colleagues (naturally) thought it’d be neat if he also set up a World Cup betting pool. Trouble is, Dan knows nothing – ok, next to nothing for the purposes of betting on the World Cup – about soccer (aside from what he gleaned from watching the Champions League final with me and the kids the other day). So other than knowing that the Brazilian defensive line is pretty good and Lahm is fast for a little guy, not that much ...

So I was trying to be helpful – which, in itself, is curious, because I know nothing about office betting pools. Turns out, it’s not all that obvious how one ought to organize a World Cup betting pool. For those of you out there who have given this any thought, you know what I mean. For those of you who haven’t, I found this blog pretty instructive:

http://www.officepoolguy.com

But just in case you were thinking of doing this without spending too much time contemplating the various advantages and disadvantages of different systems, I found the following website very helpful:

http://web.me.com/andyjc/WorldCup/Welcome_to_the_World_Cup_2010.html

It (very helpfully I thought!) comes with an excel sheet already set up to keep track of everything. A nifty alternative is the ESPN bracket predictor, but you have to create an account first:

http://games.espn.go.com/bpredictor/en-us/frontpage?

Add to this (these slightly hilarious, but also useful) tips or a quick look at a predictions website, even those not dipped in the holy water of soccer knowitall-ness will be ok. Couldn’t be easier, right?!

Well, Dan’s question got me thinking and race for the computer to see if, say, knowing the FIFA rankings and a little bit of soccer history – and I mean a tiny bit of it – helps to explain outcomes in the World Cup.

England's Prospects in 2010

Growing up in Germany, I remember the run up to every World Cup (or even World Cup qualifying campaign) as an arduous, painful, self-critical anticipation of imminent failure that would finally end the country’s extraordinary run of lucky World Cup campaigns up until that year. The “Miracle of Bern” (“Das Wunder von Bern”) became so deeply ingrained in the national psyche, it was about more than just soccer and in fact became the founding legend of the new, democratic Germany as a modern nation.

Fast forward a few years to last fall. I was reading about England’s qualifying campaign and prospects for the upcoming World Cup.

Let me say, before I say anything else, that I think Fabio Capello has done an excellent job getting England ready for the World Cup. Really, I do. He seems to have inspired a different spirit in a squad that, in terms of quality, isn’t all that different from the 2006 squad, but now they seem to believe they actually can win.

So, perhaps I shouldn’t have been surprised a few months ago when I read that Capello thought England would and should make the finals.

First I thought Mr. Capello was getting a little carried away, then that he had a lot of chutzpah that he was trying to impart to his team, or perhaps that he, as an expert, simply knows better. After all, he does have a great squad and anyone with a good team should be optimistic about an upcoming competition.

But then, he did it again in March when he declared that, “The semi-finals is a minimum."

At that point, instead of commending his chutzpah or thinking he just got carried away, my reaction was a bit of an “uh oh, here we go again ...”. I had this sudden flashback to the summer of 2006, when we happened to be living in England. I remember, all too well, the wall to wall news coverage of Wayne Rooney’s foot, but more importantly the palpable excitement about the upcoming tournament. Surely this was the year (40 years after Wembley - a nice round number) that England was finally going to win the tournament and the moment its extraordinary run of bad luck was going to end. What a difference in perspective to the years of my youth!

Turns out, 2006 wasn't the year. We all know what happened. But as soccer economists/writers Simon Kuper and Stefan Szymanski have pointed out with great wit and detail in their highly readable – ok, hilarious - Chapter 2 of Soccernomics on “Why England Loses and Others Win” this happens with great regularity. Clearly, Capello seemed to have "gone native" and internalized English attitudes about the World Cup.

Does Size Matter in the World Cup?

Some people (I won’t name them, but you know who you are) have asked me how it can be a World Cup if China, Russia, and India aren’t in the tournament while nations like New Zealand (pop. 4.37 mio) and Slovenia (pop. 2.06 mio) are. (These are the same folks who used to pay amused inattention to the World Cup before the U.S. started qualifying regularly every four years.)

But their questions do raise the question of whether the “importance” or “size” of a country matters for international soccer success. While it’s sometimes hard to explain to the non-soccer audience that the World Cup is not the Olympics and that, therefore, not everyone participates in the "finals", it is true that fans of World Cup statistics and international soccer have long suspected that the size of countries matters for success.

U.S. World Cup History

As a fan of U.S. soccer, the country’s consistent ability to qualify for this year’s World Cup and World Cups going back 20 years (to 1990) is good and bad news. Qualifying is great; clearly, the U.S. men have become consistent performers on the world stage. The U.S. are currently ranked no. 14 in the FIFA/Coca Cola World Ranking. Ok, not Top 10, but ahead of our neighbors to the South and North (Mexico and Canada).

At the same time, qualifying for the greatest show on grass and not winning is tough to take for any fan. So, I catch myself rooting for the U.S., but also knowing that the football gods are probably against us. So, what exactly are the odds, historically speaking, of the U.S. winning matches at the World Cup?

World Cup Goal Production

One month to go! With the World Cup upon us, I thought I’d look at some data on goals scored during past World Cups.

To judge performance, it's nice to start with a baseline. In all 18 World Cup tournaments held to date (1930-2006), teams have scored a total of 2,061 goals. That’s almost 115 goals per World Cup (114.5 to be exact).

Here’s what the data show about frequently goals were scored and who scored them:
Most Goals Scored By A Team: 27. By Hungary in the 1954 World Cup held in Switzerland. The 27 goals scored by Hungary that year made West Germany’s win even more miraculous. To see how truly unusual scoring 27 goals in a World Cup is, take a look at this: