Need a Wall Street Expert to review this document for me for World Cup

mathdotcom · 05-20-10, 12:04 PM

In layman's terms they find a bunch of variables that they think influence the probability of winning the world cup. The seeding matters, recent performance of these teams matter, and so on. Then they look at old data and see what the pattern is between the probability of winning and these variables. So if they find that Brazil's probability of winning decreases by 0.05 for every degree of latitude above the equator the game is played [and do this for all the other variables], they come up with a prediction given the current facts we know (latitude of South America, etc.).

Then they take their estimated probabilities and compare them to the market probabilities [implied by the lines] to find the best value [difference between estimated probability and price].

Justin7 · 05-20-10, 12:17 PM

It's at least an interesting read.

Pavy, did you see how they try to predict the winner of shoot-outs? The article suggests that the average goal differential is the best predictor for a shoot-out... What do you think of that?

mathdotcom · 05-20-10, 12:26 PM

Surely the goalies' past records in shootouts is a better predictor? I imagine the kickers don't matter much.

Karayilan9 · 05-20-10, 12:41 PM

Originally posted by mathdotcom

Surely the goalies' past records in shootouts is a better predictor? I imagine the kickers don't matter much.

Big tournament penalty shoot outs have the lowest scoring rates, the technical ability of the player can come second to his mental state in a high pressure penalty shoot out. Some countries have done better in penalty shoot outs than others, England and Holland are notorious for losing penalty shoot outs, Germany have a habit of winning them.

Jordet, G., Hartman, E., Visscher, C. and Lemmink, K. A. P. M. (2006) Kicks from the penalty mark in soccer: The roles of stress, skill, and fatigue for kick outcomes. Journal of Sports Sciences, 1-9, Preview article.
This paper is an essential starting place for an analysis for the impact of penalty shoot-outs in competitive international football as it reports some empirical findings on events in penalty shoot-outs in the World Cup (WC), European Championships (EC) and the Copa America (CA). The results are fascinating; here is a glimpse. The percentage success rate in the World Cup is 71.2% compared to 82.7% (CA) and 84.6% (EC), possibly reflecting the greater importance and consequent pressure of the world stage. The success rate of each penalty kick changes throughout the competition:

First kick 86.6%
Second kick 81.7%
Third kick 79.3%
Fourth kick 72.5%
Fifth kick 80%
‘Sudden death’ kicks 64.3%

These results highlight the increasing pressure as the competition progresses and may also highlight the ‘best player should go first’ fallacy. The idea of ‘getting off to a good start’ by putting the best penalty taker first appears wrong as there is least pressure on this kick.

pavyracer · 05-20-10, 01:46 PM

Originally posted by mathdotcom

In layman's terms they find a bunch of variables that they think influence the probability of winning the world cup. The seeding matters, recent performance of these teams matter, and so on. Then they look at old data and see what the pattern is between the probability of winning and these variables. So if they find that Brazil's probability of winning decreases by 0.05 for every degree of latitude above the equator the game is played [and do this for all the other variables], they come up with a prediction given the current facts we know (latitude of South America, etc.).

Then they take their estimated probabilities and compare them to the market probabilities [implied by the lines] to find the best value [difference between estimated probability and price].

Thanks mathy. I have printed the document as a reference for the WC and will study over the weekend.

Originally posted by Justin7

It's at least an interesting read.

Pavy, did you see how they try to predict the winner of shoot-outs? The article suggests that the average goal differential is the best predictor for a shoot-out... What do you think of that?

The formula they use is a good one. They are basically saying that a team who scores a lot of goals and concedes a few goals has a lesser chance of going to a penalty shoot-out and losing or winning the game there(50/50). I don't think it says England has a better chance of winning a penalty shoot-out. It says it has a better chance of not going to one and winning the game without extra-time.

This is important when you are looking at the odds of 90 min to win and odds to advance (which includes extra time and penalty shoot-out if needed).

mathdotcom · 05-20-10, 02:19 PM

Originally posted by mathdotcom

In layman's terms they find a bunch of variables that they think influence the probability of winning the world cup. The seeding matters, recent performance of these teams matter, and so on. Then they look at old data and see what the pattern is between the probability of winning and these variables. So if they find that Brazil's probability of winning decreases by 0.05 for every degree of latitude above the equator the game is played [and do this for all the other variables], they come up with a prediction given the current facts we know (latitude of South America, etc.).

Then they take their estimated probabilities and compare them to the market probabilities [implied by the lines] to find the best value [difference between estimated probability and price].

Sorry that should be South Africa. You're taking the information about the current event [World Cup 2010] and matching it with the relationships uncovered over observing previous events (ie. that the 0.05 latitude relationship holds). The problem is that the world cup is a fairly distinct event, friendlies probably don't mean that much most of the time, old relationships discovered over previous world cups may no longer hold for some reason. That's why guys doing this in hockey may ignore pre-lockout data because of different rules that were in place at the time and so on.

So these things aren't useless but should always be read with a grain of salt.

Dark Horse · 05-20-10, 02:48 PM

Originally posted by Karayilan9

Some countries have done better in penalty shoot outs than others, England and Holland are notorious for losing penalty shoot outs, Germany have a habit of winning them.

True, but these things can change. Holland for instance used to take the approach that they couldn't train pk's, because they could never imitate the live conditions in the stadium and the pressure at the time. After losing one too many pk series they changed that philosophy. Since then they've been winning penalty shootouts.