I was reading "Blackbelt in Blackjack" when I thought of a novel approach for a count and was wondering if anybody has done any research on this yet:
Keep a "fractions count" for single-deck games where the numerator is the number of high cards left to be dealt and the denominator is the the number of low cards to be dealt. Intuitively, this would eliminate the need to true-count while increasing accuracy. Thoughts and comments?
> I was reading "Blackbelt in
> Blackjack" when I thought of a novel
> approach for a count and was wondering if
> anybody has done any research on this yet:
> Keep a "fractions count" for
> single-deck games where the numerator is the
> number of high cards left to be dealt and
> the denominator is the the number of low
> cards to be dealt. Intuitively, this would
> eliminate the need to true-count while
> increasing accuracy. Thoughts and comments?
Sort of like an "others to tens" ratio, huh?? I'm sorry, but you're 42 years late!! This was Thorp's original "Tens Count," enunciated in Beat the Dealer, in 1962. Not a bad idea, though! :-)
Don
LOL! Not a bad idea indeed (albeit a little tardy) ;-) It seems the perfect alternative for anybody who would like to use a balanced count but doesn't want to use the TC conversion. When the fraction is less than 1, it's the same as a negative count. Any fraction greater than 1 should be rounded to the nearest .5. It would be nice to see more work done on this as well as an level 1 unbalanced count with an ace side-count.
Don, I was also wondering if you did any research on an off-shoot of clumping (please hear me out on this first). Here was my thought: In Ken Uston's "Million Dollar BJ" he mentions a hand that he played where the count was astronomical, so he decided to split and resplit 10's against the dealer's 6 upcard. Even though the count dropped, it was still high enough for him to expect the dealer to bust his hand, but alas, the dealer pulled a 21. With a certain count, let's say a TC of 10 with 1 deck left to play, the expected % of the time you would expect to pull a 10 or Ace (hi-lo count) would be roughly 40%. If he received 2 10's and splits them and receives 2 more 10's, all of the sudden, not only has his TC gone down, shouldn't the expectation of the next card being a high card go down even steeper than the TC (in dependent trials, the formula for there to be 5 high cards in a row is 40%^5, right?).
The question is: does it make sense to split 10's so many times and NOT expect a low card to come for the dealer when you receive a bunch of high and zero-cards even if the count is still high? Or another example is, if you have a 5-card 13 vs. a dealer's 10, what are the chances of you drawing a bust-card even with a low count?
All this BJ reading and playing has really gotten me thinking. I just want to know if I'm on the right track. :-)
> Sort of like an "others to tens"
> ratio, huh?? I'm sorry, but you're 42 years
> late!! This was Thorp's original "Tens
> Count," enunciated in Beat the Dealer,
> in 1962. Not a bad idea, though! :-)
> Don
> When the fraction is less than
> 1, it's the same as a negative count.
Well, not quite, in Thorp's version. He counted 36 non-tens ("others") to 16 tens, a starting ratio of 2.25 to 1. But, of course, other methods are possible.
> Don, I was also wondering if you did any
> research on an off-shoot of clumping (please
> hear me out on this first).
My blood pressure just spiked 20 points, but I'll cut you a little slack -- just this once! :-)
> With a certain
> count, let's say a TC of 10 with 1 deck left
> to play, the expected % of the time you
> would expect to pull a 10 or Ace (hi-lo
> count) would be roughly 40%. If he received
> 2 10's and splits them and receives 2 more
> 10's, all of the sudden, not only has his TC
> gone down, shouldn't the expectation of the
> next card being a high card go down even
> steeper than the TC (in dependent trials,
> the formula for there to be 5 high cards in
> a row is 40%^5, right?).
Right so far. But, you make it sound that, when we split 10s v. a dealer's 5 or 6, only getting another 10 is "good." What's wrong with getting aces, 9s, or 8s? In fact, with the TC as high as you mention, the delaer is near 50% probability of breaking with those upcards. So, no matter what you hit your 10s with, you're probably a slight favorite to win any hand.
> The question is: does it make sense to split
> 10's so many times and NOT expect a low card
> to come for the dealer when you receive a
> bunch of high and zero-cards even if the
> count is still high?
Short answer: Yes, it makes sense.
> Or another example is,
> if you have a 5-card 13 vs. a dealer's 10,
> what are the chances of you drawing a
> bust-card even with a low count?
Not high enough to stand!
> All this BJ reading and playing has really
> gotten me thinking. I just want to know if
> I'm on the right track. :-)
No, you're not. Forget you ever knew the word "clumping" existed.
Don
I will probably test the fractions count to see if there's something worthwhile to be gained from that.
As for the other thing, it's forgotten. :-)
I'll be sure to send you a bottle of aspirin for the migraine I'm sure to have started. ;-)
> Well, not quite, in Thorp's version. He
> counted 36 non-tens ("others") to
> 16 tens, a starting ratio of 2.25 to 1. But,
> of course, other methods are possible.
> My blood pressure just spiked 20 points, but
> I'll cut you a little slack -- just this
> once! :-)
> Right so far. But, you make it sound that,
> when we split 10s v. a dealer's 5 or 6, only
> getting another 10 is "good."
> What's wrong with getting aces, 9s, or 8s?
> In fact, with the TC as high as you mention,
> the delaer is near 50% probability of
> breaking with those upcards. So, no matter
> what you hit your 10s with, you're probably
> a slight favorite to win any hand.
> Short answer: Yes, it makes sense.
> Not high enough to stand!
> No, you're not. Forget you ever knew the
> word "clumping" existed.
> Don
Posting Permissions
-*-
Bookmarks