Doubling your money: Roulette vs. video poker

[Eduardo]

Okay, I have one for you.
 
Let's say you want the best odds of at least doubling your money (best chance as a percentage) in a single session without ever adding more money beyond what you start with. Let us say you start with $500.  You don't care about winning anything in between and you will not stop until you double to at least $1,000 or bust. If you win more, great, but that does not affect your odds of at least doubling (do not weight it that way).
 
Your choices are 100 play video poker or roulette.
 
The odds on roulette are simple as we all know, and that would go very fast one way or the other. So can vp ever beat those odds?
 
Let us assume for the sake of comparison that you are playing a machine with an expected payout of exactly 100% (which does not exist I think, it is usually a little higher or lower).  You can make it whatever game you want, DB, DDB, Deuces, etc. for the different payout amounts.
 
You can play one hand of max bet dollar 100 play for $500. But your odds of doubling are probably somewhat low on the first shot.  I have not looked them up but I know they are well below 50%.  Of course, if you do not double it you are likely to win <i>something</i> in all those 100 hands and would be able to place another bet with your remaining winnings and might eventually go on to double your starting amount, but the game would probably not end on a single draw.
 
So the question is,  if you take that $500, what would give you the best odds of at least doubling your starting amount in this high stakes situation, and would your odds at video poker ever be better than roulette?  I don't know how to simulate this in VPW or I would.
 
You can lay down all $500 at once on 100 play, split it in two on the first bet, or play it out in single line over a longer period. The question is what percentage of people will double their money before they bust. I don't know if the odds of doing that get better or worse with more hands in VP.
 
At first I thought it was simple but I think it gets complicated fast so I look forward to what the VP scientists here have to say.
 
I hope the question make sense. Thanks!

[jm002546]

   The practical problem of VP is not the right paytables.  You could never make a bet that would put you above 1000. If you had a table where every paying hand EV was almost zero except for 1 at 1.99999, then you would bet it all in one hand and approach 50% chance of doubling.  50%would be the optimal.  Roulette would be below 50.   Very complicated question for my simple mind and my answer is probably wrong.

[shadowman]

The best chance will always be a game where you can do no better than double your money as with even money bets in Roulette. In games where you can do better than double your money (like VP) the game must have more losing hands to make up for the bigger wins.

[Eduardo]

Hey that does make sense.
 
Is there a way to figure out what percentage it would actually be for the video poker scenarios?  I'm just curous how different it is, what percentage of people would double their money by continually going max bet on 100 play until they ran out of change or doubled? In such a high risk scenario it can't be high, but I'm curious about the percentage.
 
Say you put down all 500, win back 320.  You put that 320 back in, etc etc.  I know it would have to be simulated using software and then averaged many times over as the calculations would be ridiculous. Maybe that's not as easy to simulate as it should be though given the requirements I've put out.
 
It certainly makes sense though that the odds would be far lower for doubling considering if you were dealt a royal you could win... what, $400,000?!
 
I guess if we knew how to arrive at the percentage number we could also compare it to other games like craps with high multiplier payouts (snakeyes?) and easy-to-calculate odds. For instance, given $500 and a goal of x30, are you better off going 100 play or double 1's on the craps table?  Again, must be craps because of the chance to also win more at vp?
 
I do understand the huge difference between ordinary play and these high-risk big win scenarios and I would not suggest the latter unless you can accept the risk. I'm just curious what the actual numbers look like.
 
The numbers are so easy for a dice game or any clear 1-time payout. It is much more complicated when you are looking at multiple hands of vp and the many potential outcomes.

[shadowman]

OK, just send me that $400K check and I'll write the program.
Actually, I have a program that runs through simulated VP play. It would not be too difficult to change it to try this out. Right now I'm working on some GVP simulations ... but when I'm done with that I can give it a shot.

[jm002546]

  As long you never make a bet that puts you over 1000, and your last betputs you at 1000 or 0, you have a 50% chance.  It's obviously equally likelythat you will win 500 or lose 500.  Yes roulette pays 2 for 1, but not 1/2 the time.

[Eduardo]

Well, that doesn't work very well for VP, only the table games. And it is not quite correct either.  As stated in your example, in roulette you have below a 50% chance but your last bet also puts you at 1,000 or 0. If you were at a lower with a higher payout bet, the odds would be even worse of reaching 1,000 and higher of reaching zero.  So it isn't as simple as that.