Print and Go Back ESPN.com: World Cup 2010 [Print without images]

Thursday, December 3, 2009
Updated: December 7, 11:17 AM ET
Considering the 'impact' of the draw

By Nate Silver
Special to ESPN.com

There is arguably more on the line tomorrow than on any day of actual soccer competition. Based on the way Charlize Theron draws the pingpong balls in Cape Town, South Africa, the United States could have as much as a 63 percent chance of advancing to the knockout stages in next summer's World Cup -- or as little as a 25 percent chance. England could be anywhere from 94 percent to 69 percent.

The difference for some other teams may be even larger: Ivory Coast, for instance, has an 89 percent chance to advance in a group with Slovakia, New Zealand and Italy, but just a 48 percent chance against Brazil, the USA and Portugal.

Let's take a look at the 32 teams in the field in advance of tomorrow's draw, arranged as FIFA has placed them into four "pots." I'll explain how I came up with the percentages I teased above, and which teams you should be rooting for or against in tomorrow's sweepstakes.

You'll see several additional statistics here, above and beyond each team's SPI rating and ranking. The first two, OFF (for offense) and DEF (for defense), reflect the relative strength of each team's attack and defense. The higher the OFF rating and the lower the DEF rating, the better the team. These numbers are published regularly as part of the SPI and are explained at greater length in the FAQ.

But there's also a special statistic that I've created just for the purposes of the draw called "impact." Impact measures the effect a team will have on its group opponents' chances of advancing to the knockout stage. For example, Brazil's impact is a minus-6 percent, meaning that if a team happens to draw the Brazilians (as opposed to another, randomly selected team from Pot 1), its odds of making it to the knockout stage are reduced, on average, by about 6 percent. South Africa, on the other hand, is a plus-8 percent, giving an opponent significantly better odds of advancing.

The impact score is determined by a program I created that simulates the group stages of the World Cup based on each team's SPI ratings (the SPI was specifically designed with this purpose in mind). But a couple of additional factors are taken into consideration.

One of these factors is home-field advantage, which is very powerful in international soccer, the World Cup being no exception. The host country has won the World Cup one-third of the time since the tournament was first played in 1932, and it has reached at least the semifinals more than 70 percent of the time. South Africa is arguably the weakest host ever, but even relatively marginal teams that have hosted the tournament often have done remarkably well -- like South Korea in 1998, which reached the semifinals, or Sweden in 1958, which played in the final.

In addition to home-field advantage, we also account for home-continent advantage, treating it as 50 percent as valuable as home-field advantage itself. There is powerful evidence for this, too, according to both anecdotal and statistical evidence, and it may be especially important in Africa -- a continent to which the best teams in the world rarely travel.

Lastly, our program accounts for pace, as our research has found that teams that play a more defensively oriented game (and therefore have stronger DEF ratings) are at a slight advantage against elite competition relative to more offensively minded opponents.

So, back to South Africa for a moment: Certainly, any team that draws the South Africans is going to be thrilled to have done so. And on the road, that side is awful, having lost every away match it has played this year. But this tournament will be played in South Africa, and between home-field advantage and the team's slow pace, it is capable of pulling off an upset or two. Indeed, the South Africans have had some decent results here and there, going 1-1-1 in this summer's Confederations Cup (which was also hosted by South Africa) and beating three World Cup qualifiers -- Paraguay, Ghana and Cameroon -- in home friendlies played in 2008.

If there's another team in Pot 1 that is relatively beatable, it's probably the defending champion, Italy. Of the seven matches they played between the Confederations Cup and the European Championships, the Italians won just two. They are also a relatively old team, with several players on the wrong side of 30. Whether that experience proves to be an asset in South Africa or, as SPI seems to think is more likely, the Italians have trouble keeping pace with some of the younger teams in the tournament, we'll have to wait and see. Picking up even one point against the Azzurri is never easy, but compared with facing a Brazil or a Spain, an opponent should be happy to have them.

There's just as much at stake in Pot 2, where the diversity of the teams reaches its apex, both in terms of geography and quality of competition. The United States is the toughest team on paper, but considering how inconsistently the U.S. has played since losing striker Charlie Davies, it's anyone's guess which team will show up in South Africa. Mexico arguably suffers from big-fish-in-a-small-pond syndrome: It has tended to play very well against other North American teams, but hasn't had much to brag about against clubs from the rest of the world. The Mexicans' wide-open style of play is arguably a poor fit for the World Cup. Still, some other systems have them rated higher -- 15th in the FIFA rankings and eighth in ELO, for example -- and they'll certainly come into South Africa with a lot to prove.

The team to watch, however, may be Australia, which is tougher to score against and plays more cohesively as a unit than either the U.S. or Mexico. Although the Australians haven't had much opportunity to play in tournaments against teams from the rest of the world, they have started to play a more ambitious schedule of friendlies, with very impressive results on occasion -- including a trouncing of almost-qualified Ireland in a road friendly in August, plus a 2-0-1 record against the Netherlands in three friendlies during the past 18 months.

At the other end of the spectrum are New Zealand and North Korea, which are among the worst teams ever to qualify for the World Cup. Draw one of these sides, and your task is simplified greatly because you've essentially reduced your group to three teams competing for two slots. Hence, those teams' high impact ratings of plus-8 and plus-10 percent, respectively.

North Korea, because it plays a boring, defensively oriented brand of soccer, probably is more capable of annoying a superior club with a 0-0 or 1-1 draw. Nevertheless, this is a team that has lost matches to the likes of Lebanon, Thailand, Qatar and Tajikistan over the course of the past 24 months. As for New Zealand, any team that loses a qualification match to Fiji, as the Kiwis did last November, probably isn't much of a threat to advance.

Pot 3, in a nutshell, is why the group stages of South Africa are liable to be the most competitive of any in years -- and certainly since the World Cup expanded to 32 teams in 1998. Take a look at what we have here: First, there are five African sides that will benefit from home-continent advantage, at least two of which (Ivory Coast and Cameroon) would be fully capable of reaching the elimination stages even on neutral turf. Playing at "home," Ivory Coast and Cameroon will be as tough as many of the clubs from Pot 1, whereas Nigeria and Ghana should be roughly as strong as a solid, second-tier European squad like Denmark or Serbia.

Then we have three South American teams, two of which (Chile and Uruguay) are ranked in the SPI top 10; the third, Paraguay, isn't far behind and plays stifling defense. What this means, in essence, is that almost every group will have at least three competitive teams -- the team drawn from Pot 3, the team from Pot 1 and the European team from Pot 4. If the group also happens to draw Australia or the United States from Pot 2, there will be few breaks in the schedule at all. Yes, there are a couple of mathematically plausible "dream draws" that would warrant an exception to this rule (in this group, for instance, Algeria is a ways behind the others). But we should expect at least six or seven of the eight groups to be highly competitive down to the last day of competition, with all of the African teams supercharged by their home-continent advantage.

Coming at this from an American perspective, I have some trouble understanding how FIFA seems to make things up as it goes along, not deciding upon its seeding criteria until the last minute. Example: Until Wednesday morning, it was unclear which of the four teams from the group of England, the Netherlands, Portugal and France would be top seeds and placed into Pot 1 and which would wind up in Pot 4 instead. It does appear, however, that FIFA made the right decision, as England and the Dutch are rated marginally higher by SPI than are Portugal and France.

Still, Portugal and France are going to be very tough opponents, and whichever group draws one of them is liable to be referred to as the proverbial "Group of Death." Portugal, after struggling early in qualification, has played nearly flawless soccer in 2009, and will be even tougher once Cristiano Ronaldo returns from injury. France, on the other hand, hasn't played well of late -- although the talent and track record of success in the World Cup are there in abundance.

Serbia will also be very competitive. Its players have begun to take on key roles on elite club teams throughout Europe, and the Serbians beat out France for the automatic qualification slot from UEFA Group 7. The one disadvantage Serbia has is that, like most of the other Eastern European clubs, it tends to play a wide-open style that doesn't always translate to the World Cup.

The one weak link here is probably Slovakia, which has never before qualified for either the World Cup or the European Championship since Czechoslovakia split up in 1993. Yes, the Slovakians played excellent football during qualification, but they did so playing in a weak group, and some of their other results -- failing to qualify for Euro as the result of a 5-2 home loss to Wales, and losing friendlies to the likes of Cyprus and Iceland -- suggest a team that isn't quite ready for prime time. According to our impact ratings, the difference between drawing Portugal or France versus drawing Slovakia is as large as the difference between drawing the United States versus New Zealand, or Brazil versus South Africa.

Dream and nightmare draws

The program I developed can also be used to estimate the odds of each team advancing out of any given group -- something you'll see on display in my follow-up article once the draw takes place Friday. For the time being, let's take a look at dream and nightmare draws for a couple of teams that the ESPN audience should take a particular interest in: the United States and England.

The dream draw for the United States actually isn't all that terrific. The Americans get the host, South Africa -- which will be tougher than many people realize -- and the aforementioned Slovakia. Although Algeria is the weakest team from Pot 3, no group can have more than one African team, which means the United States would get a tough Paraguay team instead. So, no gimmes here, but the U.S. would have about a 63 percent chance of advancing.

By contrast, they could end up with this:

This would be the ultimate "Group of Death" -- the toughest possible configuration of teams in South Africa, according to SPI (although several other permutations would be nearly as competitive). Even the Brazilians would fail to advance from this group about once in five times, and the United States would have little chance.

Here, meanwhile is England's best possible draw:

This is one of the few exceptions to the rule we described above -- a group that would contain only one highly competitive club. England would likely breeze through this group, with about a 94 percent chance of advancing. Were it to fail to advance, England would take its place alongside the Chicago Cubs as the world's greatest chokers.

On the other hand, England could wind up this situation:

This is a very good English club, and even in this scenario it would advance about 69 percent of the time. But there would be no margin for error -- a loss or even a draw against the United States, for example, would probably do them in.

Now, let's take one last sweep through the field, this time focusing on a team's odds of advancing, assuming it gets an average draw from the other three pots, subject to the geographical restrictions that FIFA imposes. After the draw takes place Friday, we can compare these figures with the actual groups to see who got lucky and who was unlucky.

A few things worth pointing out here: First, look at the South Africans' number -- they are nearly even money to advance. This is because FIFA protects the host nation by placing it in the seeded pot, meaning that South Africa won't have to worry about teams such as Brazil, Spain and England until the elimination stages. A secondary advantage is that South Africa can't face another African team in the group stages, so it won't have to face Ivory Coast or Cameroon either. If the South Africans get a draw of, say, Japan, Slovenia and Paraguay, you'll probably hear some conspiracy theories about how things were rigged in their favor. But such a draw would be about par for the course given the way things have been structured. Indeed, South Africa would have a decent shot at advancing from such a group.

By contrast, the percentages are much lower in Pot 2, including that of the United States (42 percent). These teams really get the worst of all worlds because they're not only forced to play a match against both the Pot 1 and Pot 3 teams, but are also blocked from drawing the poor teams at the bottom of their own pot (North Korea and New Zealand, for example). The best they can hope for is to draw South Africa from Pot 1 or Slovakia from Pot 4 -- or both -- but odds are that they'll face a very tough group and had better get ready to play the best football of their lives.

Pot 3 is more straightforward, although there is one slight quirk. Because the five African nations in Pot 3 can't draw South Africa, that means one of the South American teams -- Uruguay, Paraguay or Chile -- will get the South Africans in Pot 1. Although these three teams are perfectly capable of advancing on their own merits, the one that does draw South Africa will be in a commanding position to go through. Also, almost all of the African teams carry high percentages. We're quite likely to see three of them among the 16 in the knockout stages, in fact, and four or even five advancing isn't out of the question.

Lastly, notice that France and Portugal are at a disadvantage from their placement in Pot 4, instead of among the seeded clubs. Were Portugal in Pot 1 instead, its advancement chances would rise from 62 to 70 percent; France's number would jump from 58 to 66 percent. This isn't necessarily fatal, especially if one of these teams is lucky enough to draw a South Africa or a New Zealand. But it does mean they're playing against a bit of a stacked deck.

Nate Silver is a renowned statistical analyst who was named one of "The World's 100 Most Influential People" by Time Magazine in 2009. He gained acclaim for outperforming the polls in the 2008 U.S. presidential elections and created baseball's popular predictive system, PECOTA.