Lining Up The Chances To Be Champion

By: chris | December 4th, 2007

Some people have way too much time on their hands. This one must, and for that we’re thankful. Ken Roberts, a software engineer from the States of America United, has written a computer program which generates the probability that each team wins the championship of each league. He’s done it for all the major American sports, as well as a number of the larger Euro football leagues. Which begs the question: Why even play the games?

For all intents and purposes, this software means absolutely squadoosh. But it’s pretty cool and fun to peruse, especially if you spent your college years staring at a bunch of books with numbers and formulas and such – or at least supposed to be staring at them. Here is the explanation of the ‘how’ directly from the source itself:

I wrote a computer program that randomly generates results for each remaining match. Then it tallies each team’s position and repeats, millions of times. When the simulation “plays” a match it assumes each team has an equal chance of winning, with no regard for record, injuries, matchups, or streaks. To help flush out each teams highest and lowest possible seeds, I force them to win or lose all their remaining matches for 1000 of the simulation runs. Here is an interactive chart that gives you a feel for how playing the season 10 million times sheds light on the possible finishing scenarios.

Onto the good stuff – the percentage of times during the simulation each team won their respective league.

Serie A:

Inter 42%
Roma 22%
Juventus 11%

La Liga:

Real 36%
Barcelona 16%
Villareal 14%
Atletico 12%

Prem:

Arsenal 43%
Liverpool 13%
ManYoo 13%
Chelsea 11%

Bundesliga:

Bayern Munchen 39%
Werder Bremen 29%
Hamburger SV 14%

Ligue 1:

Lyon 55%
Nancy 22%

Some Related Stories:

Category: Bundesliga, English Premier League, French Ligue 1, Serie A, Spain La Liga, World Football

« Sam Allardyce = Scrooge | Home | Malouda: Training Is “Terrifying”. Wah. Wah. »

The Offside RSS Feeds

Mat

Well Juve has the same apparel as referees in many sports too...what a coincidence :)
Ian

I don't know - I think there's a very even race in Scotland right now, with Celtic missing their fullbacks and with their star keeper going in for knee surgery ... anyway, it's an interesting system. I like Roma a lot more than Inter, so I hope your boys can prove this system wrong too.
chris

Ian, it doesn't seem like it would work, but those odds do seem pretty legitimate when looking at the future, current standings aside. I'd still like to see more details on how it works at a deeper level.

Inara, then you obviously don't watch much Serie A, or check the standings. Or the papers where even Juve players admit that Roma and Inter are better. Don't let history cloud your judgment.

* - Btw, these results I believe are before ManYoo's win last night.
Ian

Well, the thing is, with a system like this, whoever is in the lead now is most likely to win, regardless of form, injuries, etc. Celtic has a huge advantage over Rangers this way, when their one-point advantage is, in reality, pretty small. It gets updated each week with new results, so it's always going to look accurate.
Ian

It seems like treating all teams like they are equal makes this system pretty much useless for prediction of actual odds. Putting in a variable that makes it slightly more likely for a home team to win or draw rather than lose, and more likely for Manchester United to win than Derby on any given day would make it a little more robust. I would happily bet my house that Arsenal will win more matches than Derby, but in this system, they each have the equal chance to win every game. Still, neat toy. The fact that Arsenal can apparently lose seven matches out of 24 and still have a 100% chance of finishing first (because all the other top teams are handicapped by being considered the equals of relegation teams). We'll see about that one.