
98%: Eidetic Counting
I've just learned from my exgirlfriend that she is somewhat eidetic. As such, she can actually memorize and instantly recall groups of cards. For example, if I ran 20 cards by her, she could instantly tell me how many of each cards (although not in order, though that can be taught). In other words, in any given situation, she would know the exact composition of the remainder of the deck(s) after each card is dealt. It may sound absurd, but it's true, I've tested her and I'm extremely jealous.
I was considering teaching her to count but that seems pretty weak considering her freakish, Rainmanesque ability. Are there any systems which take into account exact deck composition (at least for a single deck scenario) or is there anything that attempts to approximate such a thing more closely than traditional counting (a regular count with a bunch of sidecounts perhaps)? At any given point in the game, she'll know exactly what's left in the deck, the only problem is then figuring out what the correct play is based on that knowledge.
I've perused some of my blackjack texts, including the card removal section of 'The Theory of Blackjack' and the prospect of developing some sort of strategy for this seems daunting, to say the least. Though I do have a question about this section of the book. Are the removal effects additive? To find the advantage of hitting vs. standing, could you just add together the removal effects of all the cards which had been played already?
I realize this is pretty out there, but any help would be greatly appreciated.

ET Fan: As far as I know ...
... ToB is the best source for information on this. Did you read Chapter 5 on multiparameter counts and ultimate human capability?
Yes, the strategy EORs are additive, but they should also be scaled for penetration. Work through the examples on p. 86, and if it still isn't clear, give a holler.
Did this person play cards a LOT as a child? I have been called eidetic, but not in cards  on the piano.
Maybe the lady should work on ace sequencing. Sorry to hear she's your exgirlfriend. ;)
ETF
> I've just learned from my exgirlfriend
> that she is somewhat eidetic. As such, she
> can actually memorize and instantly recall
> groups of cards. For example, if I ran 20
> cards by her, she could instantly tell me
> how many of each cards (although not in
> order, though that can be taught). In other
> words, in any given situation, she would
> know the exact composition of the remainder
> of the deck(s) after each card is dealt. It
> may sound absurd, but it's true, I've tested
> her and I'm extremely jealous.
> I was considering teaching her to count but
> that seems pretty weak considering her
> freakish, Rainmanesque ability. Are there
> any systems which take into account exact
> deck composition (at least for a single deck
> scenario) or is there anything that attempts
> to approximate such a thing more closely
> than traditional counting (a regular count
> with a bunch of sidecounts perhaps)? At any
> given point in the game, she'll know exactly
> what's left in the deck, the only problem is
> then figuring out what the correct play is
> based on that knowledge.
> I've perused some of my blackjack texts,
> including the card removal section of 'The
> Theory of Blackjack' and the prospect of
> developing some sort of strategy for this
> seems daunting, to say the least. Though I
> do have a question about this section of the
> book. Are the removal effects additive? To
> find the advantage of hitting vs. standing,
> could you just add together the removal
> effects of all the cards which had been
> played already?
> I realize this is pretty out there, but any
> help would be greatly appreciated.

98%: Re: As far as I know ...
Thanks for the tips. I'll see what I can come up with.
> Did this person play cards a LOT as a
> child? I have been called eidetic, but not
> in cards  on the piano.
That I don't know. Her ability is not limited just to cards (she's a particularly troublesome opponent in spades, hearts and seven card stud), but to groups of things in general, particularly numbers, but possibly other mundane objects. For example, you could take her to a used car lot and, assuming she knows the identities of all the cars there, she could tell you later all of the cars that were present. For some peculiar reason, her freakishness extends to other phenomema as well. One example of such a thing would be her ability to solve differential equations. If I see a coupled, multivariable linear diffeq, I usually find the determinant of the matrix of coefficients and then use this knowledge to make appropriate variable substitutions and then go on to solve the equations. She, somehow, just does it in her head nearly instantaneously.
> Maybe the lady should work on ace
> sequencing. Sorry to hear she's your ex
> girlfriend. ;)
Yeah, ace sequencing is probably next on the agenda. As for the ex part...she's also a good cook and a very nice person with a ridiculously high paying job so, in other words, I am the dumbest man on earth. But, hey, that's life and at least we're still friends.

Kyle Sever: Out of curiosity...
What is her background and what does she do (if you don't mind)?

98%: Re: Out of curiosity...

Claymore22: Re: Eidetic Counting
Though he doesn't offer any advice or knowledge specific to blackjack, you might be interested in reading some of the books by Oliver Sacks (one of the consultants to the makers of the movie, "Rainman"). In his work, he discusses eidetic ability, especially in the context of braininjured and autistic patients; nonetheless, the case histories are fascinating. By gaining a better understanding of how eidetic memory works, you might be able to apply the lessons to blackjack.

98%: Re: As far as I know ...
Alright, I've taken your suggestion and reviewed ToB yet again. The card removal section all makes sense, though it still doesn't give me quite the ammunition I would need to take on devising a multiparameter card counting system which is essentially a 13 card sidecount. If her abilities extended to doing large amounts of arithmetic (basically adding and subtracting all of the pertinent removal values at the decision point) this would be a piece of cake but, unfortunately, she's not actually Rainman.
It seems that with each added sidecount, the calculation of indices for a count + sidecounts system becomes more and more intractable. Griffin does mention the Einstein and Gordon counts 'fortified' with sidecounts of several cards. Have such strategies been developed and published? The playing efficiency of the Einstein + A,7,8,9,2 count clocks in at .891 and Gordon + A,6,7,8,9 performed even better, coming in at .922. Griffin surmises this as being the reasonable upper bound of human capability. I think this is well within the reach of my...student, considering that, in the span of about one hour, her skill at blackjack (as far as counting cards goes anyway) went from virtually none at all to surpassing my own, and I have been doing this for about two years. I TOLD her basic strategy ONCE and she knows it cold. Furthermore, she was able to learn the entire set of Zen indices after perusing the tables in 'Blackbelt in Blackjack' for an hour or so. Given her ability to memorize tables of numbers and her ability to keep track of every card dealt out in a game of blackjack, I think the PE maximizing approach is the way to go. For normal people like me a "simple" count has to do, but I don't think she should be subject to such restrictions.
Considering that Griffin mentions that he used a complicated multiparameter (of his own invention, I believe) count in a series 5000 hands as a sort of test of PE, I can only assume that such beasts do exist somewhere. I'd just like to know where to find them. It seems the Key Card Concept mentioned by Sklansky in 'Getting the Best of It' is the same sort of thing. Unfortunately, he does not go into all that much detail. Does anyone know if he or anyone expanded on this idea anywhere?

oneshot: Re: Out of curiosity...
98%, why the hell have you not married this treasure???

ET Fan: Re: As far as I know ...
> Alright, I've taken your suggestion and
> reviewed ToB yet again. The card removal
> section all makes sense, though it still
> doesn't give me quite the ammunition I would
> need to take on devising a multiparameter
> card counting system
Correct. I thought that was what you were considering. It goes without saying this is a major (months long) project. The ammunition is all there in ToB. I know it's there, because I once wrote a simulator that played perfectly, according to Griffin's tables. I'm not saying the ammo is all in one chapter. And obviously it takes some talent to load and fire accurately.
> which is essentially a 13 card sidecount.
Not necessarily.
> If her abilities extended
> to doing large amounts of arithmetic
> (basically adding and subtracting all of the
> pertinent removal values at the decision
> point) this would be a piece of cake but,
> unfortunately, she's not actually Rainman.
> It seems that with each added sidecount,
> the calculation of indices for a count +
> sidecounts system becomes more and more
> intractable. Griffin does mention the
> Einstein and Gordon counts 'fortified' with
> sidecounts of several cards. Have such
> strategies been developed and published?
I don't think so. Don would know better than I.
> The
> playing efficiency of the Einstein +
> A,7,8,9,2 count clocks in at .891 and Gordon
> + A,6,7,8,9 performed even better, coming in
> at .922. Griffin surmises this as being the
> reasonable upper bound of human capability.
> I think this is well within the reach of
> my...student, considering that, in the span
> of about one hour, her skill at blackjack
> (as far as counting cards goes anyway) went
> from virtually none at all to surpassing my
> own, and I have been doing this for about
> two years. I TOLD her basic strategy ONCE
> and she knows it cold. Furthermore, she was
> able to learn the entire set of Zen indices
> after perusing the tables in 'Blackbelt in
> Blackjack' for an hour or so. Given her
> ability to memorize tables of numbers and
> her ability to keep track of every card
> dealt out in a game of blackjack, I think
> the PE maximizing approach is the way to go.
But even with perfect PE, the PE contribution is low in a shoe game. I think I'd go for the ace sequencing if she were MY girl, you lucky DOG!
Or how about gin rummy? There are gin houses and tournaments in New York and Chicago with $5,000  $25,000 prizes. Memorizing the discards, including the order they go down, is important in gin. Luck still plays a major role, but far less than blackjack.
> For normal people like me a
> "simple" count has to do, but I
> don't think she should be subject to such
> restrictions.
> Considering that Griffin mentions that he
> used a complicated multiparameter (of his
> own invention, I believe) count in a series
> 5000 hands as a sort of test of PE, I can
> only assume that such beasts do exist
> somewhere. I'd just like to know where to
> find them.
I'd assume they exist on paper and disk, but not in the public domain.
> It seems the Key Card Concept
> mentioned by Sklansky in 'Getting the Best
> of It' is the same sort of thing.
> Unfortunately, he does not go into all that
> much detail. Does anyone know if he or
> anyone expanded on this idea anywhere?
Here's a possible lead. A friend of DD' (frequent bj21 poster) uses a two parameter system (A,2 verses 3,4,5,6,T, I think). Sounded very scientific. They probably wouldn't mind sharing some tricks. Best I can do.
ETF

98%: Ugh
So what you're saying is that I'm going to have to ask questions about this on bj21?! I must say that sounds unpleasant given the yahoo:intelligent poster ratio.
Posting Permissions
 You may not post new threads
 You may not post replies
 You may not post attachments
 You may not edit your posts

Forum Rules
Bookmarks