How I calculated Molly's video poker loss rebate

This may be elementary for some folks here, but it also may be useful for some others. I thought I should share it anyways. It's how I calculated the loss rebate stopping points for Molly's lose-20-get-20 promotion:

https://blog.vidpoke.com/2022/08/how-i-calculated-mollys-video-poker.html

You have a line that says assuming you are playing perfectly. Is the perfect strategy the same for the game when you are playing on a loss rebate versus just playing the game normally for cash?

This is the funniest sh*t in the world.

Filius Bruce: "Finally, Bob Dancer was careful in not revealing the exact expected value of the promotion at Molly's. I've learned from the master and won't do it here either."


But you guys will say the when, where, how, and give every other detail about stuff. ˂˂˂ That's just blanking hilarious.

I don't care about the Molly's promotion, since I haven't been there in ages. I do care about revealing certain strategies that can be applied to much more lucrative promotions.

I guess there's a bright side, it only took 3 decades for him to realize the value of using double-up in key situation.

p.s. My guess is that you guys don't know the exact value because that would require knowing what an infinite double-up situation would yield/how many doubles someone is gonna do(Not sure if there's a limit), and what their starting denomination is. But let me take a stab, it's worth close to full value -the EV on the machine.

Quote: randomperson

You have a line that says assuming you are playing perfectly. Is the perfect strategy the same for the game when you are playing on a loss rebate versus just playing the game normally for cash?
link to original post

I can't see why not. You could possibly change your strategy if you take into account tipping on jackpots.

Quote: randomperson

You have a line that says assuming you are playing perfectly. Is the perfect strategy the same for the game when you are playing on a loss rebate versus just playing the game normally for cash?
link to original post

Using any non-zero value for “Nothing” hands would absolutely change optimal strategy. Determining the value to use, however, is not so simple. At least in this promotion, the largest available bet size is relatively close to the rebate amount as compared with other such promotions.

​My local casino has 9/6 Jacks or Better and 9/6 Bonus Deluxe (which returns 99.64%). On certain days the casino offers 25% loss rebates on up to $1,200 in total losses (good for up to $300 in free play).

I've lost the maximum ($1,200) a couple times and cashed out with profits of a few hundred bucks other times. But I've never had any particular number in mind that would trigger me to stop playing.

After reading the article linked above, it occurs to me that perhaps I should always be playing for a royal flush type gain of $4,000 or more and also be willing to endure more $1,200 max losses.

From a practical standpoint, I can't imagine ever NOT cashing out after hitting a royal even if an algorithm determined it was +EV to keep playing until I catch another one or see my bankroll dip to -$1,200. I'm not going to sit in front of a machine for 24 hours straight because of math. But it would be nice to know what the cash-out number is for optimal play.

Quote: randomperson

You have a line that says assuming you are playing perfectly. Is the perfect strategy the same for the game when you are playing on a loss rebate versus just playing the game normally for cash?
link to original post

The short answer is I don't know. The expected value of the promotion depends on the variance and the expected value of the game itself: the higher the variance, the higher the value of the promotion, but also (to a lesser degree) the higher the expected value of the game, the higher the expected value of the promotion.

If there are strategy changes that increase the value of the promotion, they must be increasing the variance and decreasing the expected value of the game. But I have no clue about how to find them except for changing the strategy here and there randomly and running the loss rebate again to see if its EV changes. My guess is if there are such changes, their effect in this specific promotion would be at most a couple of cents, no more.

Quote: FiliusBruce

Quote: randomperson

You have a line that says assuming you are playing perfectly. Is the perfect strategy the same for the game when you are playing on a loss rebate versus just playing the game normally for cash?
link to original post

The short answer is I don't know. The expected value of the promotion depends on the variance and the expected value of the game itself: the higher the variance, the higher the value of the promotion, but also (to a lesser degree) the higher the expected value of the game, the higher the expected value of the promotion.

If there are strategy changes that increase the value of the promotion, they must be increasing the variance and decreasing the expected value of the game. But I have no clue about how to find them except for changing the strategy here and there randomly and running the loss rebate again to see if its EV changes. My guess is if there are such changes, their effect in this specific promotion would be at most a couple of cents, no more.
link to original post

We agree on the theoretical point about how this could change the strategy. The problem with saying it probably doesn’t matter is that we don’t really know until we try.

The solution technique you described in the article would just as well solve the more complex problem, with more computational steps, maybe too many.

It might not change the overall ev of the promotion but there have to be certain hands that flip. Imagine a 3 to royal and 4 to flush situation that would already be a penalty card spot in normal vp without the rebate. The ev starts so similar that it almost has to flip from the flush to the royal draw in the promotion. You could as a first step isolate a hand like that and see how the ev change within the promotion versus outside. Maybe it doesn’t change the ev of the promotion much because the hands don’t occur often enough but it could be a big percentage of the return of the hand when it does occur.

Fundamentally, you would go for more royals in tournament strategy, and play much more aggressively if you had to hit a high score in a fixed time. My guess would be somewhere between optimal cash strategy and tournament strategy, but more on the side of optimal strategy.

(This post is regarding loss rebates. It’s an interesting idea that tourney play would be similar; however, I believe the method, if it exists, would be different.)

Royal holds wouldn’t be the only type affected. The values of all non-winning holds would increase, some more than others depending on the odds of landing a busted hand for each. These would include holding 4 or fewer to a royal, straight flush, flush, or straight; low pairs and any singletons; and, of course, holding nothing. We may find that new holds become viable, such as 3 to a flush or straight, 2 to a straight flush, or a single 10, among others. As all of these holds are typically accounted for “behind the scenes” by VP Analyzers, it shouldn’t be too hard to come up with a method to alter someone’s homemade version (Filius…?). I suggest homemade, as I don’t believe there is a “commercial” version that allows altering the payout for “Nothing” hands and/or using decimal payouts. While WoO may be an exception, as Tring found that the URL could be manipulated, it believe that it may not work on all devices. Besides, simulations may be necessary to complete the iterations outlined below.

Rough Draft of Method for Creating an Optimal VP Strategy for Loss Rebates:
(Please feel free to tear this apart in order to improve it!)

1. Determine the expected number of hands to reach either stopping point. (Initial value = Rebate amount / Bet amount. Subsequent values may need to be determined via simulation.)
2. Take the reciprocal of the figure from (1.). (This is the new value for “Nothing” hands.)
3. Analyze the current game using all of the normal payouts, substituting the figure from (2.) for “Nothing” hands.
4. Determine the upper and lower stopping points.
5. Repeat steps (1.) thru (4.) until there is negligible change in the expected number of hands.
6. Create a VP strategy to accommodate the final version of analysis from (3.) and utilize the final stopping points from (4.).

Molly's used to have a lose $100 get $200 in free play.