A Program gets competitive CSGO results to rate CSGO teams.
The rating will give out probability of a team winning next match.
The probability can be used with Kelly Criterion to take advantage of house' odds.
** Progress **
Best result so far has quadratic loss of 0.2
Betting over 6 months would generate 300% profit.
The system seems to not be flexible enough in using the number of matches each team has played. There should be a way to use the difference in matches played between 2 teams to determined the k1 (first parameter) weight, etc.