Been ages since I've done this stuff, so I'm a little rusty.
A relative asked me to go over some video poker stuff for him and I don't want to advise him incorrectly.

basically, he plays 5 coins
the expected value of the play is 5.036162986 coins for a win of 0.036162986 coins
sigma is 27.7125824726266
variance is 767.987227

I remembered all that with no problem

Next, I wanted to come up with a bankroll recommendation.
Risk of Ruin is 1 in 1000 or 0.1%

I did -(sigma^2/2xbar)*ln(RoR)
or -767.987227/(2*0.036162986) * ln(0.001)
= about $73,350

am I right here?


He also wanted some projections for outcomes at 2,000,000 hands
pretty sure central limit theorem applies

for that, I did Xbar+-z*sigma*sqrt(n)
so, 2000000*0.036162986 +- z*27.7125824726266*sqrt(2000000)
or 72325.97 +- z*39191.51
z = 1.96 for 95% CIs
z about 2.57 for 99% CIs

yes?

ty for any help... it's much appreciated as I don't want to lead the person astray

Skew is problematic here...google Evgeny Sorokin.

very interesting.... I was just thinking since the central limit theorem applied, I could use the standard risk or ruin stuff

thanks... I'll put it together this way as well