Results 1 to 6 of 6

Thread: lagavulin62: tko sim reflections/deck estimation concerns

  1. #1
    lagavulin62
    Guest

    lagavulin62: tko sim reflections/deck estimation concerns

    just got back from a 4 day road trip(no internet) and see the thread below generated good discussions so thanks for all the great posts. I did manage to use the sim/generator but I won't bore you with numbers and redundancy since cac's sims have already shown anything I discovered for myself. there is a point on deck estimation which may or may not be a problem so this is why i started a separate thread. more on that later.

    you may recall from my original thread how I rounded out a few indexes to make simulatiing easier, in particular, insurance was taken at the pivot instead of the hi/lo equivalent of TC-3. I ran a "greater accuracy" RA generation for TKO for the more common indexes and found very little difference in the way of gain but this has been stated on this board before. however there may be another use for TKO generated indexes and that is when you consider their convenience when setting up your betting ramps. consider the following indexes which show a difference from their hi/lo TC's. note I have converted tko tc's to their hi/lo equivalents for comparison.

    index hi/lo tko
    __________________________________________________ ____

    16v10 0 1
    16v9 5 7
    11vA 1 2
    9v2 1 2
    10v10 4 6
    __________________________________________________ ___

    since calculating TC for TKO in real-time is very difficult, common sense dictates creating betting ramps ahead of time for each deck level left. using whole deck precision and entering at TC 2 you would only need a total of 10 different
    ramps. 5 for each deck level(don't count 1 deck level since you will never see this) at TC 2 and 3. of course the pivot is a gimme. if you wanted to include a wong-out point thats another 5 for a total of 15 ramps. all of these follow a nice predictable pattern that is not difficult to learn. then conveniently, as you can see from above, the correct indexes don't kick in until tc 2 or higher, the exception of course being 16v10 which is marginal anyway.

    deck estimation
    here is the part which may or may not throw a monkey-wrench into this whole thing. consider deck estimation errors when using hi/lo. a running count of 10 with 3 decks left floors to a TC of 3. but suppose you actually have 2 decks left and so the true count is really 5. you error results in you underbetting. now consider TKO with a running count of 16(irc-0) with 4 decks left. this is a TC of 2. but here again, suppose you are late in your deck estimation and there are actually 3 decks left. this brings your TC to 1 floored. in this case your error results in overbetting. so it seems the question is now would this error rate be significant enough to seriously handicap TKO?

  2. #2
    koolipto
    Guest

    koolipto: Re: tko sim reflections/deck estimation concerns

    > you may recall from my original thread how I rounded
    > out a few indexes to make simulatiing easier, in
    > particular, insurance was taken at the pivot instead
    > of the hi/lo equivalent of TC-3. I ran a "greater
    > accuracy" RA generation for TKO for the more
    > common indexes and found very little difference in the
    > way of gain but this has been stated on this board
    > before. however there may be another use for TKO
    > generated indexes and that is when you consider their
    > convenience when setting up your betting ramps.
    > consider the following indexes which show a difference
    > from their hi/lo TC's. note I have converted tko tc's
    > to their hi/lo equivalents for comparison.

    > index hi/lo tko
    > __________________________________________________ ____

    > 16v10 0 1
    > 16v9 5 7
    > 11vA 1 2
    > 9v2 1 2
    > 10v10 4 6

    The TKO I18 indices are available in BJRM. Also, I believe that Cac posted these once (they were floored). Both those sources and my own sims (CVData) put 16vT at 0 for S17 DOA DAS. On 16v9 BJRM and Cac had 6; my own sims shaking out at 6 or 7. 11vA we are all three getting 1. 9v2 for all three is 1. 10vT Cac and BJRM have 3. I think mine came out at 4, but I play with 6 as an approximate RA number (I can't recall if I ran a RA index sim or if I just fudged).

    By the way, I grossly round my own indices without a lot of apparent harm to system performance based on sims I have done. Also, in case it makes you feel better, there are a surpisingly large number of indices where my own sims come out differently from Cac's and BJRM. Based on Roger Harris' earlier concept of EZ-TKO, I am toying with rounding all indices to 0, +2 and +4 (maybe +3 also). Using his system you wouldn't even have to true count.

    > since calculating TC for TKO in real-time is very
    > difficult, common sense dictates creating betting
    > ramps ahead of time for each deck level left. using
    > whole deck precision and entering at TC 2 you would
    > only need a total of 10 different
    > ramps. 5 for each deck level(don't count 1 deck level
    > since you will never see this) at TC 2 and 3. of
    > course the pivot is a gimme. if you wanted to include
    > a wong-out point thats another 5 for a total of 15
    > ramps. all of these follow a nice predictable pattern
    > that is not difficult to learn. then conveniently, as
    > you can see from above, the correct indexes don't kick
    > in until tc 2 or higher, the exception of course being
    > 16v10 which is marginal anyway.

    Bootlegger on BJ21 did something similar with his True Counted KO charts. I think EZ-TKO gets you what you need without having to memorize betting ramps.

    > deck estimation
    > here is the part which may or may not throw a
    > monkey-wrench into this whole thing. consider deck
    > estimation errors when using hi/lo. a running count of
    > 10 with 3 decks left floors to a TC of 3. but suppose
    > you actually have 2 decks left and so the true count
    > is really 5. you error results in you underbetting.
    > now consider TKO with a running count of 16(irc-0)
    > with 4 decks left. this is a TC of 2. but here again,
    > suppose you are late in your deck estimation and there
    > are actually 3 decks left. this brings your TC to 1
    > floored. in this case your error results in
    > overbetting. so it seems the question is now would
    > this error rate be significant enough to seriously
    > handicap TKO?

    My apologies if this is already obvious, but deck estimation errors for all systems are the most impactful the farther away you get from pivot. For Hi-Lo, estimation errors have the least impact the closer you are to the pivot of zero. Same is true for TKO at its pivot of 4. Personally, I am comfortable with the impact of estimation error in TKO especially since it is most accurate where you have the most at stake. Some time ago, there was a thread on TKO/A with a pivot of 8. I tooled around with that in a spreadsheet once and concluded that in addition to being difficult to do two running counts and True Count conversions, the system was way too sensitive to my mediocre deck estimation abilities.

    Rgds, Koo

  3. #3
    lagavulin62
    Guest

    lagavulin62: Re: tko sim reflections/deck estimation concerns

    > The TKO I18 indices are available in BJRM. Also, I
    > believe that Cac posted these once (they were
    > floored). Both those sources and my own sims (CVData)
    > put 16vT at 0 for S17 DOA DAS. On 16v9 BJRM and Cac
    > had 6; my own sims shaking out at 6 or 7. 11vA we are
    > all three getting 1. 9v2 for all three is 1. 10vT Cac
    > and BJRM have 3. I think mine came out at 4, but I
    > play with 6 as an approximate RA number (I can't
    > recall if I ran a RA index sim or if I just fudged).

    > By the way, I grossly round my own indices without a
    > lot of apparent harm to system performance based on
    > sims I have done. Also, in case it makes you feel
    > better, there are a surpisingly large number of
    > indices where my own sims come out differently from
    > Cac's and BJRM.

    I understand about that, and it seems indexes are dependent on which book or software you use but for the most part I have found them to be at least consistent and I round myself and don't notice much difference either. so much for exact indices.

    > Bootlegger on BJ21 did something similar with his True
    > Counted KO charts. I think EZ-TKO gets you what you
    > need without having to memorize betting ramps.

    this is where I obtained that formula

    > My apologies if this is already obvious, but deck
    > estimation errors for all systems are the most
    > impactful the farther away you get from pivot. For
    > Hi-Lo, estimation errors have the least impact the
    > closer you are to the pivot of zero. Same is true for
    > TKO at its pivot of 4.

    well sometimes your train of thought blinds you to the obvious. no need to apologize. if errors are going to be meaningful it may as well be when you have very little at stake.

  4. #4
    Roger Harris
    Guest

    Roger Harris: EZ-TKO Simplified Strategy

    Here is the simplified strategy I've come up with so far, although I'm still looking for improvements. When the running count is greater than the "current key" (TKO -2), then:

    Stand: 12v2 12v3 12v4 16v10 soft18vA
    Double: 8v5 8v6 9v2 11vA A3v4 A6v2 A7v2 A8v5 A8v6
    Split: 66v2 99vA

    Using the current key for insurance doesn't seem to work so well, so I'd recommend using the pivot for that, or even better, use the standard KO Preferred index, the running count at pivot-1. (That one exception isn't too hard to remember, since insurance is a special play anyway.)

    Using two groups of plays (like KO Preferred for multi-deck) at -2 and 0 produces better results, but of course they would be harder to learn and remember. I think the above simplified strategy makes EZ-TKO somewhat easier than KO Preferred, despite having to estimate the current key (which is very easy), but it gives better results.


  5. #5
    koolipto
    Guest

    koolipto: Re: EZ-TKO Simplified Strategy

    > Here is the simplified strategy I've come up with so
    > far, although I'm still looking for improvements. When
    > the running count is greater than the "current
    > key" (TKO -2), then:

    > Stand: 12v2 12v3 12v4 16v10 soft18vA
    > Double: 8v5 8v6 9v2 11vA A3v4 A6v2 A7v2 A8v5 A8v6
    > Split: 66v2 99vA

    > Using the current key for insurance doesn't seem to
    > work so well, so I'd recommend using the pivot for
    > that, or even better, use the standard KO Preferred
    > index, the running count at pivot-1. (That one
    > exception isn't too hard to remember, since insurance
    > is a special play anyway.)

    > Using two groups of plays (like KO Preferred for
    > multi-deck) at -2 and 0 produces better results, but
    > of course they would be harder to learn and remember.
    > I think the above simplified strategy makes EZ-TKO
    > somewhat easier than KO Preferred, despite having to
    > estimate the current key (which is very easy), but it
    > gives better results.

    Roger, thank you for posting your work. I'm curious if you have run any sims and as to how this compares to KO, reKO, TKO etc.

    Separately, I wonder if there is not a place here for a slightly more difficult, but more powerful strategy. My reasoning is that for the only slightly ambitious, it would be easy to learn more indices at perhaps more levels. Personally, memorizing indices before I get to the casino is a small price to pay for the simplicity of playing with the current RC and the pivot. I don't think I could go in to the casino without some indices at pivot, since it is such a "guaranteed" point. What do you think of adding pivot indices and -1 for insurance (the latter being a rare calculation)? I would think this would be close to full TKO (C22) for a wonger (and yet so easy).

    Would A2v4 work at the -2 current RC?

  6. #6
    YaleMoto
    Guest

    YaleMoto: Re: EZ-TKO Simplified Strategy

    > Separately, I wonder if there is not a place here for
    > a slightly more difficult, but more powerful strategy.
    > My reasoning is that for the only slightly ambitious,
    > it would be easy to learn more indices at perhaps more
    > levels. Personally, memorizing indices before I get to
    > the casino is a small price to pay for the simplicity
    > of playing with the current RC and the pivot. I don't
    > think I could go in to the casino without some indices
    > at pivot, since it is such a "guaranteed"
    > point.

    Try the Moto Unbalanced Halves Count which is a step up from KO. Counting Tags: 2:1; 3:1; 4:1; 5:1.5; 6:1; 7:.5; 8:0; 9:0; 10: -1; A: -1. (Similar to Uston SS, but pivot at TC +4 instead of +2 which I find more useful.) The betting efficiency is about the same as the other best betting efficiency systems - Halves and Uston SS: .99.

    Skew the key count for deck penetration to maintain the speed of the unbalanced system instead of true counting it: with a starting count of 0 in 2 decks for example, skew the
    key from 4 to 6 depending on pen. You can also skew the insurance count too. Learn to true count the system so you can make these adjustments to key and insurance.

    As Einstein said, make things as simple as possible, but no simpler. There has been too much emphasis on reducing the number of matrix numbers. I memorized the complete indices for 2 deck KO in less than 1 day. (I also have full indices for AOII memorized - which I play when fresh.) If you are really going to play blackjack for thousands of hands to get to NO, you can spend alittle time to memorize more index numbers.

    So you can add more indices, especially near pivot. That is why I like pivot at TC of +4. Use all of the indices
    published for KO, plus add 10 splits and A,8 v. 3-6. Others are possible. But remember that you lose accuracy away from pivot.

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

About Blackjack: The Forum

BJTF is an advantage player site based on the principles of comity. That is, civil and considerate behavior for the mutual benefit of all involved. The goal of advantage play is the legal extraction of funds from gaming establishments by gaining a mathematic advantage and developing the skills required to use that advantage. To maximize our success, it is important to understand that we are all on the same side. Personal conflicts simply get in the way of our goals.