Friday, December 7, 2007

Week 14 - Parlay Math Clarification Request

I like your analytic intent, here. Let me post some questions and comments, and see if things can get a little clearer...

1. What do the numbers under the single-game bet indicate? The other numbers in your chart seem to indicate winnings per dollar bet (with a given win %rate). They are all less than $1. Then under the single-games you show betting 14 teams yields $814..... this obviously cannot be $814 per dollar bet, so what is this number, and all the others in that column?

2. 52.88% is the win % you would need in order to break even on single game betting, not parlay betting, correct?

3. What assumptions are you making in parlay winning bets? If you bet 4 (overlapping) parlays per week, as my 9-team strategy suggests, and you miss 4 games (56% win rate), you could either
a) lose all 4 parlays, if the losers are all in different bets = loss of $4
b) lose 3 of 4 parlays = gain of $2
c) lose 2 of 4 parlays = gain of $10

Depending on how you weight the likelihood of those outcomes, your expected returns will be drastically different. In addition, who's to say your win rate happens uniformly? Take this as an example, both assume a 50% win rate:

i) Week 1 I win all four parlays for a gain of $24, and in Week 2 I lose all parlays. Win rate = 50%, and average winning is $12/week.
ii) Week 1 and week 2 I split the parlays, winning twice each week. This gives me $10/week as an average.

My point is I would rather win (and lose) in chunks than constantly split. And how do you determine at what rate each occurs?

4) I'd like to see a graph showing $/bet expected returns on the y-axis and winning Percentage on the x-axis, then run the numbers for each type of bet. This could be genious if we ever hit our stride and have consistent winning percentages.... That way we would know how we should be betting based on our performance.

No comments: