wardos
wardos
  • Threads: 4
  • Posts: 12
Joined: May 22, 2010
June 14th, 2010 at 12:35:48 AM permalink
I want to work out the expected return for some split hands. Say you are permitted to split up to three times, so as to form four separate hands. The starting point should be what's the probability of having 2 hands, 3 hands or 4 hands to split, because from there you can work out expected returns.

Assuming an infinite deck, these are my results:
2 hands = 12/13 x 12/13 = 85.20710%
3 hands = 2x(1/13 x 12/13 x 12/13 x 12/13) = 12.10042%
4 hands = (1/13 x 1/13) + 2x(1/13 x 1/13 x 12/13) + 2x(1/13 x 1/13 x 12/13 x 12/13) = 2.69248%

Can any one confirm that's correct? For some reason from there I can't reconcile with the Wizard's Appendix 1 on infinite deck splitting expected returns:

I am only out by <0.1% but I want to get it spot on and understand the difference.

Thanks in advance, Wardos.
miplet
miplet
  • Threads: 5
  • Posts: 2142
Joined: Dec 1, 2009
June 14th, 2010 at 2:02:21 AM permalink
1 hand = 12/13
2 hands = 144/2197
3 hands = 3456/371293
4 hands = 769/371293
These match yours exacly if I multiply them by 13.
“Man Babes” #AxelFabulous
wardos
wardos
  • Threads: 4
  • Posts: 12
Joined: May 22, 2010
June 14th, 2010 at 5:14:26 PM permalink
Thanks Miplet - I didn't have 1 hand probability in there because I had already assumed you had a splitting hand once dealt. Either way, our numbers agree. I have a feeling I am not matching up with the Wizard Appendix 1 splitting expected returns because I have used unrounded numbers throughout. Like I said the difference was very small.

With thanks
Wardos
  • Jump to: