Thank you Don, Norm, and Rick! (nt)
Based on your post, I added the range of results to my output. This exposed a bug in my simulation. Here are the revised results with the range of results.
Cut Mean SD Min Max
0 15.83 4.17 5 44
13 15.72 4.18 4 49
26 15.51 4.21 4 46
39 15.22 4.27 4 46
52 14.85 4.38 2 44
65 14.41 4.53 0 45
78 13.93 4.69 -3 44
91 13.41 4.87 -5 45
104 12.81 5.07 -9 43
"I'm a fan of interesting theoretical questions, so thanks for answering this, Don and Panama Rick."
Don't forget Norm!
"Just to make sure I'm understanding correctly: are you saying that, on average, there is a two-deck portion of every eight-deck shoe that, if counted separately, would yield a final count of +15?"
Yes.
"If that's the case, I'm surprised."
Well, I wasn't. :-)
"I would have expected the average worst two-deck portion to be substantially lower than that,"
Not as an AVERAGE.
"while recognizing that the absolute worst two-deck portion would be substantially higher (theoretically it could get all the way up to 104, though I'm assuming the odds against that happening are pretty astronomical)."
Rick shows 44 or so to be just about the ultimate.
"In any case, your intuition certainly does seem to have been incredibly accurate, Don."
Bad habit of mine. :-)
"Can you explain how you came to your extremely-well-educated guess?"
Maybe. I've been at this for 33 years, and my intuition re things of this nature is rather sharp. I just tried to call upon all the play I've done and came up with my best estimate. Frankly, I didn't expect it to be THAT precise! :-)
Don
I frequent a number of automobile forums. Unfortunately, they are populated predominantly by kids (ok, young adults) that never even make an effort to spell, capitalize, or punctuate in their posts. You make the effort and it makes your posts clear and articulate. As someone you couldn't spell himself out of a wet paper bag, I not only don't see most simple spelling mistakes, but have a lot of understanding for the occasional mistakes of others.
I doubt that the profitability of shuffle tracking can be established theoretically.It's largely dependant on shuffle procedures and dealer specifics like plug-in-habits and grab sizes.
Even if a plug-in doesn't divide a slug it will for most regions of the discards displace it and make a successful tracking virtually impossible.
Francis Salmon
In theorie, the shuffle ought to be random. Tracking benefits become possible only by the fact they are not.
There are other examples where we can exploit dealer imperfections. I just mention payout errors and hole-carding.
Sure, card counting is based on sound theory but the real world makes enhancements possible.
Francis Salmon
