For a given team of 6 individuals, calculate the probability that this particular configuration of 6 will defeat a different, known, team based on how well each of the chosen 12 individuals have performed in past matches within different teams.
Breakdown of rules and clarifications:
- 2 teams compete against each other.
- For each competition, team configurations differ.
- 6 individuals are randomly assigned to each team from a large pool of N individuals.
- Each "match" will result in either a win, draw, or loss.
- Rows of raw data display: the individuals within both teams AND the match result
what is the best way to go about solving this problem? Initially, I thought about using a modified Elo system for each individual. Is this the correct path, and what else could be used instead? I would love to read some papers on the subject.
