vorosmccracken.com

The “triumphant” return of The Knack of the baseball world

vorosmccracken.com header image 2

World Cup 2010: Group Stages, Two Down One to Go

June 21st, 2010 · 4 Comments

Every group has now finished the first two rounds of games, and now comes the final game of the group stages, and for half the teams their last game of the tournament.

Let’s get all the updated numbers out of the way first. The National Team Rankings have now been updated. An important note for these is that the numbers are capable of moving more at this time than any other point in a four year cycle. Not only do these games count more than any other in the system, but they are also the only inter confederation matches with a high importance rating (besides the final qualifiers). So not only can the ratings move for the teams playing, but teams not in the tournament can also move some depending on how well (or poorly) the teams from their region do. The only major change of note is that the North Koreans have now slipped behind New Zealand as the lowest rated side in the tournament.

These rankings can be found here.

The table that folks would probably find most interesting is the one which details how often each team went out at each stage. For example the French at this point went out at the group stage 9535 times out of 10,000. This table can be found here. Finally a data dump of the results of every one of the 10,000 simulated tournaments including the results of every match can be found here.

Happy Trails

The following teams failed to win a single tournament: North Korea, Cameroon, Honduras, South Africa, New Zealand, Ivory Coast, Algeria and Slovakia. North Korea and Cameroon are officially eliminated and will only have pride on the line in their final games. Of the rest of the teams, the Ivory Coast are now the longest shots to advance as that is a near impossible goal differential to make up. Of those teams, the one with the best chances of advancing are Slovakia at 16.5%, but they’re not a very strong team and they’re unlikely to go much further even if they somehow escape into the next round.

Group A – Prisoners’ Dilemma Revisited

Last year I discussed the famed Game Theory example called the Prisoners’ Dilemma and applied it to a football match situation. Well unfortunately we now have another example in Group A in the World Cup tomorrow. Mexico is faced with a situation where it can either collaborate with their opponent thereby ensuring a good result for them (advancement) but giving up a chance for an even better result (first place and a likely avoidance of Argentina). Uruguay is less of a textbook example as a draw or a win both ensure the best possible result for them (winning the group). But the potential for a loss does bring a little risk as they could fail to advance should France unscrew themselves against the hosts.

So what should each team do? The idea behind the Prisoners’ Dilemma is that despite the fact that collaboration would bring about the best combined results for the prisoners, inevitably the only logical choice each prisoner has is to work against the other because their can be no guarantee as to the trustworthiness of the other. Mexico and particularly Uruguay find themselves in this situation as well. Theoretically if Mexico offered to not try and score and just let the game play out to a 0-0 draw, you’d think Uruguay would gladly take it. However were Mexico to suddenly charge down the field and stick one in the net in the 88th minute, Uruguay would now be at risk for being out of the tournament? Can they trust their opponents that much? They certainly can’t take out a contract on the matter.

So while everyone may be expecting a very cagy, very boring match tomorrow between the two, don’t be surprised if both teams decide to make some attempts on goal (though each will likely try and stay very secure at the back).

For their part, the French are engaging in a self-immolation rarely seen at this level of professional athletics. If you thought France has looked bad on the field, they have looked far worse off it. They actually still have a small chance because of the above, but they certainly haven’t been acting like it and with the epidemic of stupidity besieging the team’s camp right now, you’d have to put their actual chances of going through as much smaller than the above simulations.

The hosts are pretty much done for, they have goal differential problems, talent problems, and need a really fortunate result in both games to have a chance. It almost certainly isn’t going to happen, though theoretically you could envision them dropping a 7 or 8 spot on the French if things really start to spiral out of control on the other sideline. Though there’s little I’d put past the French team at this point, that still seems unlikely.

Advancement chances: Uruguay = 98.59%, Mexico = 96.24%, France = 4.65%, South Africa = 0.52%.

Group B

While not the strongest group, it isn’t nearly as easy as Argentina’s made it look so far. A close but relatively secure 1-0 win over Nigeria (that could have been several more goals) and then a 4-1 dismantling of a not awful South Korean side. They haven’t officially qualified, but they only missed advancing 7 times out of 10,000 tries and that probably over estimates the chances. They’re going through, probably in first. Everyone says that an underdog has to go out and challenge the favorites before a game, but South Korea may very well have been advised to try and keep the margin of their defeat down to a goal. Now they have goal differential issues. That said, I think they are the favorites to advance from here at about a 58% clip. It’s unlikely Argentina is going to do the Greeks any favors even if they rest some starters and the Greeks will go in needing a result to go through, and maybe needing a win to do it. The Nigerians need the Argentines to beat the Greeks, and then all they need to do is beat South Korea to advance, even though they have yet to record a point in the tournament. I think South Korea is probably the slightly stronger team, and so you wouldn’t think Nigeria’s chances would be substantially more than 20% here.

Advancement chances: Argentina = 99.93%, South Korea = 57.68%, Greece = 22.65%, Nigeria = 19.74%.

Group C

Having notched the only win in the group so far, Slovenia currently sits atop the group with 4 points. They obviously would love to see a draw in the other game, as that would put them through even if England beat them by a 100 goals. However they are still ranked third in terms of their chances of advancement, even though those chances are up considerably since the start of the tournament. The US will feel they are in fairly good shape as a win automatically sees them through, and even a draw wouldn’t necessarily put them out as they currently have a goals scored advantage on the English. They are playing what most feel are the weakest team in the group (and only one of two teams who have yet to score in the tournament) and if they could somehow shore up their problems in their backline, should have enough firepower in attack to go through. The English once again are facing recriminations back home after a very troubling scoreless draw against Algeria. They will come out looking to drop the hammer on Slovenia, and they too go through automatically with a win. Algeria is still alive, but they have to win, and may need some goal differential help as well should the English beat Slovenia.

Advancement chances: USA = 70.09%, England = 64.72%, Slovenia = 53.39%, Algeria = 11.80%.

Group D

The Germans saw a red card make this first round become more interesting than they thought it was going to be after the first game, and now need to be very careful with Ghana. A draw almost certainly puts them through (Australia would need to beat Serbia by four goals just to bring the other tie breakers into play), and a win likely wins them the group. Serbia bounced back after a disappointing loss in the first game, and now are well-positioned to still advance with a win getting it done and even a draw maybe doing the trick. The Aussies are starting to roo (get it) that 4-0 loss in the opening game as they are in big trouble. Still, a win against Serbia and an upset win for Ghana against the Germans and they are through. All other advancement scenarios for them are probably not worth discussing. As for Ghana, they are now Africa’s best hope to avoid getting completely swept out of the tournament in the first round. The next best chance for another African team is Nigeria and three of the other four are basically out of it. A win or a draw will do it. A loss and they need Australia to beat Serbia, but not by much.

Advancement chances: Germany = 74.96%, Serbia = 60.44%, Ghana = 57.60%, Australia = 7.00%.

Group E

The Dutch are through and Cameroon is out. The Dutch could still theoretically cough up first place, but with a fair amount of chaos in Group F, first place doesn’t look like it will necessarily be any advantage. Cameroon will have to find what it is they want to play for, though their Coach would be well advised to start the youngest players available to him in an early start for preparation for the next cycle. As for the Danes and the Japanese, the knockout stages start a game early. The Japanese have the goal differential battle won, so a win or a draw puts them through. The Danes are likely the stronger team (but not by a ton) but generally in situations like this, the team who advances with a draw will be the favorite to go through. That appears to be the case here as well.

Advancement chances: Netherlands = 100.00%, Japan = 55.26%, Denmark = 44.74%, Cameroon = 0.00%.

Group F

Paraguay has the only win in the group, and with four points a multiple goal win and the Kiwis due up next, they are in a very advantageous position right now. They were excellent for most of South American qualifying and South America has been brilliant thus far in the Cup. The teams who advance in Group E may prefer to play Paraguay over the Italians, but they’re easy pickings for no one in this tournament. A group that was supposed to be little more than a formality for the Italians has become the sort of thing the Italians like to inflict upon themselves: a self-inflicted complicated situation. Even a draw against Slovakia has a decent chance of putting them through, but I’m not sure they want to rely on Paraguayan generosity in beating the Kiwis. They probably should put their game faces on now rather than waiting for the knockout stage like they normally do, and end the drama by sinking Slovakia and getting ready for whatever comes next. Slovakia obviously must win and then a Paraguay win or draw would put them through. The Kiwis could advance with a draw theoretically, if the Italians also draw, and I think at this point they have to be absolutely thrilled with where they sit going into the final game. A win and they improbably would be through to the knockout stages.

Advancement chances: Paraguay = 94.19%, Italy = 77.87%, Slovakia = 16.50%, New Zealand = 11.44%.

Group G

The “group of death” has become the group of bizarre as lopsided scorelines, strange refereeing decisions and invisible phones. The Brazilians are through, the Portuguese are probably through, the North Koreans are out and the Ivory Coast is probably out. The Portuguese decided that they were going to take the goal differential tie-breaker seriously and sent seven past the North Korean keeper putting the Ivory Coast in a situation where they must do the same just to have a shot. Brazil and Portugal will play for first place, but since Spain might not take first place in their group, there’s a chance first place will prove to be a Pyrrhic victory anyway. Portugal will probably try and get some sort of result just to put the Ivory Coast away once and for all, and a draw would probably not bother the Brazilians either.

Advancement chances: Brazil = 100.00%, Portugal = 99.87%, Ivory Coast = 0.13%, North Korea = 0.00%.

Group H

Spain won this group 54.95% of the time in these simulations. The two goal victory over Honduras was very nice for them. Chile has won both of their first two games by identical 1-0 scorelines and head into the final game against the Spanish in good, but not great shape. For the Swiss, the ecstasy of the upset win over Spain turned to agony after a 1-0 loss to Chile in a game where a result would have been huge for them. Honduras is basically finished, but in truth they really haven’t played that badly. Whatever chances they have are predicated on winning by at least two goals against the Swiss and having the Chileans do the same against Spain. Not likely.

Advancement chances: Chile = 80.18%, Spain = 77.14%, Switzerland = 42.14%, Honduras = 0.54%.

Handicapping the Big Prize

The Brazillians continue to hold steady at up around the 25% mark for winning the whole enchilada (which coincidentally I had for dinner tonight). The Dutch have passed the Spanish (now 3rd favorite) for second favorite while the Argentines have passed the English (now 5th favorite) as fourth favorite. Rounding out the top ten in order are Portugal, Germany, Uruguay, Mexico and Paraguay; all with chances from roughly two and a half to six and a half percent. My American team remains around the 1% mark though it looks like it’s nudged up ever so slightly. The only pre-tournament favorite to be completely wiped out are of course the French who only won 0.12% of the time, and that’s likely 0.12% higher than what their actual chances are at this point.

It really starts to get fun from here now.

Tags: Soccer!! · South Africa 2010 · Uncategorized

4 responses so far ↓

  • 1 Daniel // Jun 22, 2010 at 3:05 am

    South America is having a brilliant participation, while Europe (and specially) Africa are a big disappointment. Maybe next time this big mistake of giving Africa and South America the same number of spots will be corrected… but I don’t believe so :(

  • 2 Rasmus // Jun 23, 2010 at 5:51 am

    Hey, Vörös, one question.
    Do you update the national team rankings throughout your simulations?
    Let’s use one real simulation you had there as an example: #11. Winner is Germany in the final vs. Greece. But Greece played 0-0 vs. Argentina in the last game of Group B, drew Uruguay 1-1 and won on penalties in the Round of 16, beat the US 2-0 in the Quarterfinals, Brazil 2-1 in the Semifinals. Would those results be incorporated into new national team rankings before simulating the final vs. Germany (which they lost 0-2)? Or did the simulation use the old rankings for the final?

    Of course it would be more calculation work to trigger recalculations of off-def ratings during the simulations (which you’re doing via VBA, I guess?), but it would definitely be more realistic.

    Thanks in advance,
    Rasmus

  • 3 DSMok1 // Jun 23, 2010 at 8:09 am

    Rasmus: that would actually be less accurate, because the ratings we have now are our best guesses for how good the team actually is.

  • 4 Voros // Jun 23, 2010 at 8:38 am

    DSMok1 is correct. Whatever random results the machine spits out won’t effect the ratings because effectively nothing has changed in terms of our best guess as to the quality of the team.

    I do see how on a more wide view how such a system might make more sense than you’d think at first glance. If you were to theorize that part of the randomness inherent in the results is from the assumed uncertainty of the accuracy of the various rankings, it would make sense for the good results and bad results to cluster for teams more than we’d expect from chance (IE, taking into account that sometimes the ranking might be a little high and sometimes it might be a little low).

    Nevertheless to set up such a sim, a better solution is to tinker with the rankings beforehand giving them a little variability before each sim. This achieves the same result, is a lot easier on my coding and is probably a more accurate model for a real life situation (because in this instance, an underrated team could nevertheless get unlucky and go three and out).

    It’s an interesting idea, but a whole lot of testing would need to be done before I could implement it, if it would work at all.

Leave a Comment