Need help calculating risk.of ruin

.....

Quote: mccarthyj923

Hello, I am trying to calculate my expected.loss range for 5000 hands of $1 video poker playing 5 credits. AKA $25,000 coin in. Here is the issue. I want to adjust the strategy to where I treat going for a.royal flush as I do any other straight flush. This reduces my win rate but greatly reduces the variance. It Appears that in 9/5 poker even with the reduced amount of royal flushes my payout would be 96.8474 if a royal was $250. However a royal is really $4,000 so that .000852 return from royal flush would be multiplied by 16 to get .013632 when we add that to the rest of the return I get a payout of 98.125% instead of the typical 98.4498% from optimized play. This knocks the variance from 19.495 to 3.643
https://wizardofodds.com/games/video-poker/analyzer/

I am now.trying to find out with a 98.125% payout and a 3.643 variance what my 95% expected range would be on 5000 hands.

Sorry for this ridiculous post but I can not figure it out on my own.

"This knocks the variance from 19.495 to 3.643"

Really ??? perhaps I'm misunderstanding.

Standard Deviation = StdDev * (#hands ^ 0.5) * wager

Standard deviation is the square root of variance.

(3.64^0.5) * (5000^0.5) * $5 = $674



EV is .9845 - 1.00 = -1.55%.

EV for 25000 coin in is -1.55% * 25,000 = -$387.

Being within 3 standard deviations (about 96% of the time), that puts you at:

On the left (bad) side is -387 - (674*3)
On the good side, -387 + (674*3)

However, that isn't 100% perfect. Because while increasing the return (on RF) increases your return, it also increases the variance.

So same calculation, but replace the return (instead of the 98% return, plug in the 96% return). And it's somewhere in between those two.


If that makes any sense.


It isn't perfect, but yeah.

Quote: mccarthyj923

Hello, I am trying to calculate my expected.loss range for 5000 hands of $1 video poker playing 5 credits. <SNIP>

your post has nothing to do with a risk of ruin question in my opinion

how much are you trying to ruin?

your post wants to know a range of possible ending outcomes without risk of ruin
but a good question any ways
nice test

there are programs that can do this 4u
Video Poker for winners comes to mind
and Gamblers Odds (IF you can find it safely)

Sally

Quote: RS

Being within 3 standard deviations (about 96% of the time),<snip>

aint so, in my opinion
for a normal distribution 3SD covers abouts 99.73% of the outcomes

here is what normal could look like


here is something the op is after

Quote: RS

On the left (bad)<SNIP>

However, that isn't 100% perfect. <SNIP>

it is abouts 340% wrong, give or take, plus or minus, in my opinion
but I still could be wrong, give or take

This is easily calculated using a correct program for the task like Video Poker for Winners
then compare your answers to those you get and see (C)

in other words
ev and sd is useless for the original question
in my opinion of course

Quote: RS

If that makes any sense.

It isn't perfect, but yeah.

yes, I agree, not even close to perfect
no one is perfect
nice try I do say
Sally

Quote: AxelWolf

"This knocks the variance from 19.495 to 3.643"

Really ??? perhaps I'm misunderstanding.

really?
many do want to learn something at times, I think
please show your work

thank you
Sally

added: show your work for what makes you think you are misunderstanding, that is

Sorry I meant to say expected win loss for the bets. Would you guys agree that it is sensible to play this way so I can lower the required bankroll or is playing anything other than perfect strategy just throwing money away?

I'm sorry would this be $1 per unit or $5 per unit since it is a $5 total bet?

Quote: mccarthyj923

Here is the issue. I want to adjust the strategy to where I treat going for a.royal flush as I do any other straight flush.

I do not get this.
try again I say maybe with different words.

Quote: mccarthyj923

It Appears that in 9/5 poker <snip>

hold on now there and here

what exactly is 9/5 poker
in video poker?

please explain the game you are wanting to play

Quote: mccarthyj923

<snip> <snip>

Sorry for this ridiculous post but I can not figure it out on my own.

neither can I
computers do help

or even simulations that can be done online too
http://www.beatingbonuses.com/simulator_java.htm
(make sure they are large enough, not just 5 or 10 sessions each)

now
the distribution of 5000 hands played can be calculated once it is clear
what game you are playing and exactly what the pay table is and how you play each hand

for example
5000 hands of

9/6 JOB
prob of a loss
71.962339%

9/5 JOB
prob of a loss
79.812666%

8/5 JOB
prob of a loss
84.753248%

beware the trend!


yes, using a windows pc is easy for these type of problems (lots of software available)
and a used windows pc is very inexpensive, in my opinion,
for apple users

Sally

Game I would be playing is 9/5 jacks or better. When I say treating a royal flush like any other straight flush I am saying not to keep k 10 both of spades and instead give up the 10. The thought is that if I don't chase royal.flushes I will hit more of everything else and therefore have more predictability and can play longer sessions with a lesser bankroll

Quote: mccarthyj923

Game I would be playing is 9/5 jacks or better. When I say treating a royal flush like any other straight flush I am saying not to keep k 10 both of spades and instead give up the 10. The thought is that if I don't chase royal.flushes I will hit more of everything else and therefore have more predictability and can play longer sessions with a lesser bankroll

Adjustments like this won't make a lot of difference to the variance. The main reason why variance is lower when single-coining is because the Royal's payback is 250 units instead of 800 units. Variance_due_to_Royal = Royal_Probability*(Royal_Payout - Game_Return)^2. Heck, KT suited is not the correct basic strategy hold in 9/5 JoB either.

Max coin:
Variance_due_to_Royal = 1/40,168.71*(800-0.9845)^2 = 15.89

Min coin:
Variance_due_to_Royal = 1/50,393.88*(250-0.9722)^2 = 1.23

Let's say you followed min coin strategy while playing max coins...
Variance_due_to_Royal = 1/50,393.88*(800-0.9831)^2 = 12.67

The payouts are the biggest issue with the variance numbers...

Quote: mustangsally

really?
many do want to learn something at times, I think
please show your work

thank you
Sally

added: show your work for what makes you think you are misunderstanding, that is

I couldn't. Possibly show any work if I don't understand exactly what he is saying. Even if I did understand I couldn't show the work. I don't spend time trying to do fancy math and simulations. I if need to know something I can't figure out, and it's profitable (if its not profitable I don't want to know) to do so I find a way to get the awnser and more time finding and playing stuff. Math and simulations are great and nessisary upt to a point but only get you so far.

If he was asking what I thought he was asking.
IE if I go slightly less aggressive for Royals.
I couldn't imagine it was that big of a difference.
I was hoping when I said that he would explain more details.