1

Bayesian Resume Rating for sports math is explained in PDF https://www.jellyjuke.com/uploads/5/8/0/2/58022979/mathematical_explanation_of_the_bayesian_resume_rating_10-23-18.pdf

The formula used to calculate talent is: formula

However I am confused with the Pi(x) inside product set. It is stated that it depends on each game, however the probability ecuations found in the pdf only give constants (they contain x but it is eliminated after integration).

If I just calculate the product of these constant it yields unconsistent results.

I found a probability formula that depends on x, but it is the probability that the team have determinated talent, and don't depends on the game but the ligue.

What would be the correct formula Pi(x)?

NeDark
  • 117

1 Answers1

1

I'm the maker of this algorithm.

The formula for P(x) is on the first page of the pdf. In that context, using that formula from the first page is μA in the case of a win or μB in the case of a loss. So the formula you're talking about could be re-written as shown in the link below. I hope that helps.

To anyone else reading this, my email is [email protected] if you'd like to contact me directly.

https://i.stack.imgur.com/IbjW7.png

JellyJuke
  • 111
  • Thank you for your response. Unfortunatelly, I have more doubts. Why standard deviation calculation in page 7 includes parity of the league two times? It is strange for me that the equation has three addens. In the equation in the first page it only included two addens, the squares of the standard deviation of each team. – NeDark Jul 11 '20 at 11:13
  • On the first page, there are 2 terms because it's just an example to explain the Bienaymé formula. On page 7 there are 3 terms: one parity term for the first team, one parity term for the second team, and the uncertainty of one team's talent. The other team's talent is assumed at a certain value in this calculation, so that's why there's not a fourth term. – JellyJuke Jul 13 '20 at 19:26