With all that being said....if this was a school assignment and I had to try to come up with an accurate EV for jai-alai....I definitely would use singles games not doubles. And focus on one player. So betting Goiko to win in singles as an example I can easily track his historical singles win percentage. Then I can also compute the advantage/disadvantage of each post position using historical data across all singles games of jai-alai. Assuming post 1 helps Goiko as much as it seems to help in general...then I am on my way to calculating his probability of win. Ideally come up with a way to adjust for other players in the game, maybe limit that to adjusting for other players with high or above average win percentages. In other words....what is the Lopez or Aritz or Irastorza affect? That gets tricky based on post position for each if they are in the game. Then the payout factor I think you would have to go to the archives and get the averages for Goiko winning singles in post 1, post 2, etc. Use that mean value for each winning payoff by post position. Then multiply times our "calculated" win probability. And I'm fairly certain we would not see a positive EV, due to the fronton's 20-25% takeout.
