Click here to Skip to main content
15,881,746 members
Please Sign up or sign in to vote.
1.00/5 (2 votes)
See more:
I have a boatload of stats for a team, and I want to determine the probability that they beat their next opponent. There's a wealth of statistical information about each team, INCLUDING statistics for results versus common opponents. Obviously, any previous head-to-head results should be weighed the most heavily. Next in line would be results versus common opponents, lastly, general match results.

I have no idea where to begin, other than I know I need to come up with the "probability" of a team A beating team B. I'm assuming that's on a percentage scale, so it team A is predicted to beat team B 23.7% of the time, they are most-likely going to lose. But for example, I'd then pick a number between 1 and 1000, and if that number is greater than or equal to 237, then they would lost, etc...

Any examples, or at least a starter that says "take stat a, multiply by games played, divided by blah, add coefficient...." you can see my math knowledge, which is close to nothing...
Posted
Comments
[no name] 17-Mar-14 9:38am    
If your math knowledge is zero how did you get this job. Seriously its like asking "how do I fly a 747" and then saying I know nothing about flying. Your notion that there is a cookbook answer to this question is naïve. What do you think statisticians spend their time learning?
andegre 17-Mar-14 9:50am    
Obviously I was exagerating on the "I know nothing" part. I used to know some, but haven't used probablity since college. I'm looking at getting more of a kick start and/or some guidance to start... Obviously what I do will be nothing compared to what statisticians are able/are doing...
[no name] 17-Mar-14 9:52am    
Thank you for clearing that up. Cpallini has posted a very interesting response.

1 solution

You could start form the "ELO rating system" page at Wikipedia[^].
 
Share this answer
 
Comments
andegre 17-Mar-14 10:34am    
Wow, never thought of that! (And I'm a Chess player on Chess.com everyday) Fantastic idea, even though this doesn't even factor in the statistics, which is interesting...but may be a better solution.

This content, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)



CodeProject, 20 Bay Street, 11th Floor Toronto, Ontario, Canada M5J 2N8 +1 (416) 849-8900