A prediction (made on Monday, 28 June) verified:
Although—and this has nothing to do with the issue at hand—Argentina’s goaltender Sergio Romero stinks, and that’s putting it nicely. Every shot on goal Mexico made—and there just weren’t enoughâ€”flummoxed the poor man. Klose, Podolski, and the rest of the German side are going to eat him alive Saturday.
It’s nice to be right once in a while. This was that while: taking into account Romero’s performance, it was easy to guess that Argentina would never have a chance against Germany.
But can we work our statistical magic for the rest of the tournament, and answer such questions as: Is the World Cup getting harder to win? Is the game becoming more rambunctious? And what will be the likely outcome?
The first Cup was held in 1930: there were only 13 teams that participated, and over the course of 18 games, 80 goals were made. Only one man was sent off (two yellows or one red card) in all of the 18 matches. By 2006, there were 32 teams in the tournament, with 64 games: 147 goals were scores and 28 men were sent to the showers early to curse their fate.
Here’s a picture, through time, of the goals per game scored (note that there were Cups while the world was engaged in its global folly):
Are you able to answer this question? What is the probability that the number of goals per game was lower in Cups played after 1958, than in those Cups played before that date? The probability, given the data before our eyes, is 1: were are certain that there were less goals per games since 1962.
In fact, we are certain that there is a downward trend in the GPG (goals per game); we are certain, because that is what happened. Now, the only question is: will this trend continue? This can be answered with a model, which is the dark line over-plotted on the data. The black dot over 2010, and the red lines above and below it, indicate the guess of how the 2010 tournament will finish.
So far—by my count, which might be off—there have been 133 goals in 60 games, for 2.2 GPG (I did not use any 2010 data in the modeled predictions). Unless something remarkable happens—and with Germany scoring like it has been, it’s possible—then the prediction of 2.38 GPG (from 1.83 to 3.10), looks to be on target.
Conclusion 1 Scoring is becoming more difficult in World Cup games.
Here’s a picture of the average number of men sent off per game, or SOPG:
Quite a shocking upward trend! This could mean that men are meaner now than they used to be, or that the referees are more card happy, or a combination of the two. Either way, we are close to the point where there is a 50% chance of a man going off in each game.
The prediction for 2010 is 0.44 SOPG, (from 0.27 to 0.61). So far (again by my count, which is possibly in error), there have been 16 men who have had to walk in disgrace through the tunnel before their teammates. That’s 0.27 SOPG already, which is on the low end of our prediction. And even if, for example, 2 more men are red carded over the four final games, then the SOPG will only grow to 0.28. That means that this year’s cup will likely have fewer altercations than the last one.
Conclusion 2 More are being sent off with red cards at World Cup games.
How about the goal differential of the final? Here’s a quick table, then a picture.
This table shows the winning numbers of goals crossed with the losing number of goals. The most common results are 2 to 1, 3 to 1, and 4 to 2. That last would be my guess for this final, which I predict will feature the Dutch and the Deutsche. There have been two ties at the end of play—0 to 0, and 1 to 1—which were decided on (yuck) penalty kicks.
Now a shot of the winning goal differential (WGT), calculated by subtracting the losing number of goals from the winning number of goals: zeros indicate a tie in regulation play.
Up until the 1980s (or so), the WGT bounced around from 1 to 3, but after that started the ties. This, don’t forget, is also the era of lower number of goals per game scored, so this result fits in with the trend.
The model predicts a 30% chance of a tie for this final. It also says that there is a 47% chance of the winner besting the loser by just one goal: a two-goal lead has a 20% chance; a three-goal lead is unlikely at 4%. (These percentages are rounded and don’t quite add to 100%.)
If there is a tie, that means penalty kicks will decide the winner.
Conclusion 3 More World Cup finals will be decided by penalty kicks.
What are your predictions?
Raw statistics for this article were culled from this website.