Thanks. I understand that. But variance isn't about cost. Unless you only care about negative BR swings and not variance I am still not getting it.
My thinking is your free play has an EV. Your array of actual results expected will define a SD for that array of possible results around that EV, and variance would be the square of that SD. Unless you get the same return for the free play every single time, it has variance associated with it. But it will never result in a BR downswing. You have to bet actual money to see that.
I assume Don just chose to negate the variance associated with the free play for some good reason. I am just wondering what that reason was. Was it the lack of a possibility of BR downswing? The amount of actual play required to get the free play? Or something else? Or any combination of these factors?
I'm sorry. You're right. I started thinking about the bonus as actual money that was given to you, because its slot value is just about $100, but that isn't the case. If it's free play for, say, the slots, then it simply adds to your e.v. and the play certainly is subject to variance of its own. The distinction is that this play does not come from your original bankroll, which isn't any further at risk for playing through the $100 bonus, so I'm not quite sure what to do about that in the ROR equation.
Suppose you made 100 $100 bets, with their associated e.v. and variance. Your $10,000 br was completely cycled through one time. Now, someone comes along as says, "Have one more bet; it's on the house." For ROR purposes, this has to help! You can't lose any more of your own bankroll than you already have for the first 100 bets, but you might make some more. So, you have positive e.v., which helps your ROR. If you count the e.v. from the bet AND the variance from the bet as if they came from your own money, then you'd have to calculate the ROR as if your original bank were $10,100 and use the e.v. and variance from 101 plays. But I'm not convinced that that's the right way to go about it.
Hmm.
Don
I was thinking the same thing. Clearly RoR wouldn't be higher. I am not sure how to calculate how much lower, if any, RoR would become. It is tricky enough calculating ROR when betting your BR on two different games that use real money. How do you factor in funny money play? I guess my solution would be to calculate RoR without the free play and just know it would be less. I am used to doing that were software can give the info to make all your decisions but can't sim the results for your complete decision strategy. I do the closest sim I can and know my results should outperform the sim. It is frustrating but it is the best I can do.
Getting 1% CB on 99.73% game makes the play a 0.73% advantage. Variance is ~26.
I guess if you want to get nit-picky and it's freeplay, then you're getting 0.9973% back, added to 99.73% gets you to 100.7273% or a 0.7273% advantage. But even more nit-picky, the game isn't a 99.73% return, it's 99.7283% return....and you're not going to be playing perfectly to begin with. Not that any of that really matters at the end of the day.
IOW: Just calculate it as if it were a 0.7% advantage and go from there. No need to get nit-picky on the crap that really doesn't matter.
For what it's worth, you need about 4 royals worth of BR to play at full kelly.
EDIT: Wait are you getting both 1% in freeplay and $100 cash per 500 hands? If so, kinda need to know the denomination. If you're playing 25c, then that $100 is gonna be huge. If you're playing $25, that $100 is a paltry amount.
"Everyone wants to be rich, but nobody wants to work for it." -Ryan Howard [The Office]
This is Mike's answer:
"The risk of ruin doing this once is obviously 0%, so I assume you’re asking about the ROR is the player does this promotion over and over a very large number of times.
"I agree with your estimation of just adding $100 to the bankroll and basing the variance on 101 plays. In fact, it may be exactly right. Then I would put an advantage of 1/101, a bankroll of 100 units, and a standard deviation of about 8 into whatever the appropriate formula is and see what you get. That is about as deep as I can take this on short notice."
Don
What are you talking about? When we factor RoR for card counting, that is based on infinite play. Maybe i am playing a game that has no restrictions on how long i can play. Maybe i have been playing said game for years and have been given free play for years that comes out to adding 1% to my edge. Maybe im just trying to figure out bankrolls needed to play this game for friends who are more limited in roll than i am.
Good for you, I am happy for you, but what you are talking about is prior history. Pertaining to current situations enjoy it while you can because you cannot guarantee what you see today will be there tomorrow. Good plays tend to be eliminated more often than a shitty one, as there is little incentive to get rid of any machine that has a higher house edge, not for the player. Free play can disappear at any time.
Last edited by BoSox; 01-03-2019 at 06:30 PM.
ROR is calculated on infinite play unless otherwise stipulated, but that's not the question here. The only thing æffecting your ROR would be the chance that the casino rescinds on the freeplay (or whatever 'bonuses' you're getting). But if they announce tomorrow there's no more bonuses and you don't play anymore (duh), then that's not going to æffect your ROR.
"Everyone wants to be rich, but nobody wants to work for it." -Ryan Howard [The Office]
I am not sure how you got and "a" and "e" combination, but I assume it was intentional because you weren't sure which to use. "Affect" is used for verb forms, and "effect" is used for noun forms. Don, the grammar police chief, taught me that one. I always did it backwards until he taught me the right way. I am not sure what to do with a gerund (verb form that functions as a noun), I guess it would be affecting.
And don't forget that there are more uses of each word that further complicate the story: "effect" can be a verb, also, as in, I want to effect a change in the system, which means I want to bring it about. And "affect" can be a noun (emphasis on the first syllable), when it means a strong feeling or emotion.
Don
Bookmarks