also matchplay coupons are worth more on longshots. for example if you had a $10 matchplay coupon that you could play on one number in roulette. your EV would be ($20)(35)(1/38) - $10(37/38) which is better than an even money bet such as red or black, $20(18/38) - $10(20/38). so the higher the variance the better.

to make the most from this promotion i should play high denomination games with very high variance to make it similar to betting one number on roulette and as close to a matchplay coupon as possible.

so my question is, which videopoker/slot games have the most variance?

Quote:rudeboyoi

so my question is, which videopoker/slot games have the most variance?

Considering just video poker, 10-5 Royal Aces Bonus Poker has a standard deviation of 13.58. You can just square that to get the variance, but I think standard deviation is the better measurement of volatility. However, that game is hard to find. I've only seen it once in Mesquite in 2006.

Among games you have a realistic chance to find, I'm going to go with Triple Double Bonus. The 9-7 pay table has a standard deviation of 9.91.

You might also try a 2-3 coin dollar machine with a nice payout, like Wheel of Fortune or something.

Quote:WizardConsidering just video poker, 10-5 Royal Aces Bonus Poker has a standard deviation of 13.58. You can just square that to get the variance, but I think standard deviation is the better measurement of volatility. However, that game is hard to find. I've only seen it once in Mesquite in 2006.

Among games you have a realistic chance to find, I'm going to go with Triple Double Bonus. The 9-7 pay table has a standard deviation of 9.91.

He'd be hard-pressed to locate that paytable in anything over $2 and he'd be lucky if the Mirage or South Point still had those.

How much play did you give them beforehand to get this offer?

Quote:rudeboyoialso matchplay coupons are worth more on longshots. for example if you had a $10 matchplay coupon that you could play on one number in roulette. your EV would be ($20)(35)(1/38) - $10(37/38) which is better than an even money bet such as red or black, $20(18/38) - $10(20/38). so the higher the variance the better.

to make the most from this promotion i should play high denomination games with very high variance to make it similar to betting one number on roulette and as close to a matchplay coupon as possible.

At first look, the idea that a game with a greater variance would increase the value of match play seemed counterintuitive - wouldn't cutting the cost in half without changing the payouts result strictly in doubling the expected return (added:) regardless of the game?

Then, I saw the example of roulette in the quote above, and I contemplated. Perhaps a higher variance does prove useful; after all, it is easy to see the difference in the EV of the two roulette examples. And I fail to see fault with either the calculations or the underlying assumptions. So, then I pondered the Wizard's suggestions for high variance VP games, and thought, "What about Super Aces Bonus, or Loose Deuces, or even Five Aces DDB?" Granted you cannot always find full pay, but if variance is at a premium, maybe a slightly inferior paytable will do. Why? If you are looking for a high variance game, it's because you want a "decent" chance at a jackpot-like payout. But at what cost? How low can the return of the game be, and how high does the variance have to be to cover? In other words, how can I combine the two into one metric to evaluate and compare the available paytables or other games? (If you want a high variance, and return may not be priority one, then is video keno an option?

So, in case you have not notices, I have this driving need to quantify things in order to qualify an assertion. (Can anyone show me figures to prove that higher variance give higher EV during match play?) In order to formulate some VP paytables, I took a closer look at our Roulette example as a simplified model. While I do agree that picking a number in Roulette has a higher variance than picking red or black, I have a tendency to disagree that variance is the only driving force in the change in EV. I believe that the probability of loss to be a greater factor.

More to come...

~camapl

Quote:camaplI believe that the probability of loss to be a greater factor.

If we analyze the Roulette example we see that the expected returns (using "for 1" figures instead of "to 1") for the respective bets would be as follows:

Single Number: (36 * 1/38) - (0.5 * 37/38) = (36 * 2.6316%) - (0.5 * 97.3684%) = 143.4211%

Half the Field: (2 * 18/38) - (0.5 * 20/38) = (2 * 47.3684%) - (0.5 * 53.6316%) = 121.0526%

Note the result of subtracting the base returns from the figures above and multiplying the difference by two.

Single Number: (143.4211% - 94.7368%) * 2 = 97.3684%

Half the Field: (121.0526% - 94.7368%) * 2 = 52.6316%

The figures above are the respective probabilities of loss of the two Roulette games. What does this mean? If we divide the probability of a loss in half and add it to the base return of the game, we get the return of the game during match play.

Return (match play) = Return (base) + 1/2 * Probability of Loss

Could this be right? If we consider how the coupon works, then perhaps it is. Due to the nature of the coupon, or the partial loss coverage, there is actually a premium for losing - namely, one half a unit. If we play VP or some other game where a decision affects the outcome, we may need to alter the strategy to account for the value of losing! Seem strange? I know. You may blame my driving need to analyze these things.

As always, have fun - it makes winning big even better!

~camapl

Quote:camapl(Can anyone show me figures to prove that higher variance give higher EV during match play?)~camapl

Actually, variance does not affect EV. The two concepts are entirely separate.

Example: Game A: There are 99 ping-pong balls in a bowl, numbered consecutively from 1 to 99. If you draw a number 50 or higher, I pay you even money. Since you have 50 ways to win but only 49 ways to lose, the EV of your bet is +1% (rounded down). Game B: Same bowl of balls, but this time you have to draw the "mystery number" to win. However, if you do, I will pay you 99-1. This is the same +1% EV (approximately), but in the case of Game B, your variance is MUCH higher. However, after a gazillion trials in both games, your expected result would be the same.