Why KO doesn't work and how to modify it so it does.. inside!

KO must be modified if you're serious about playing
Blackjack with a positive expectation. The problem
with KO lies in the fact that your advantage is still
questionable when the running count's (RC) equal or
greater than the key count (KC).

The rules we're playing with:

- 8 deck shoe
- DOA
- DAS
- LS
- 75% penetration

Initial running count (IRC), KC, and pivot point (PP):

- IRC = 0
- KC = +22
- PP = +32

Using KO's unbalanced count, the final RC for one deck
will be +4. Multiply that by 8 decks and you'll have
+32. Give 75% penetration and the average final RC of
that 8 deck shoe will be +24.

Now imagine you're playing through a shoe that's 95%
complete. The RC's at +22. Do you raise your bet?
According to the book, yes. However, my answer is no.
On average, those last few hands dealt will have more
low cards than high cards so there's absolutely no
advantage. Basically you have to understand that the
RC will always progressively get higher and higher as
more cards get dealt out from the shoe.

My technique and strategy:

Never play through an 8 deck shoe because quite simply
it's hazardous to your bankroll. The only way to play
is by Wonging in at RC +22 when half of the shoe has
yet to be dealt. The average RC of a half completed 8
deck shoe with 75% penetration is +12. By Wonging in
at +22 you're definitely playing with an advantage.
Bail once the count dips below +22 or when the other
half of the shoe starts being dealt. The other half
of the shoe should only be played if the RC's at +34.
I realize the strategy could be tweeked and made more
efficient but that'd ruin it's simplicity.

More often than not, an RC equal or greater than the KC indicates a positive expectation and you should be betting higher. You are right, though. Occasionally the RC is high but the true count is 0 or even negative. If you worry about these less common circumstances and you want to take into account the number of decks left, you might as well just learn a balanced system. The whole point of KO is that you don't have to care about the number of decks left, and on average, this works just fine.

I am very interested in additional KO plays that are not listed in KO.
For example we know that at +2 in a six deck shoe against dealer 5 or 6 we should double on soft 19. (this is not listed in KO) I am trying to get additional advantages.
If anyone in here has any suggestions, I would sincerely appreciate it.

Regards,

Hollywood

I agree that learning a balanced system would be the best move. If I ever get dissatisfied with modified KO I'll definitely jump ship.

Regarding the "less common circumstances," they actually happen more times than you think. I remember playing the last few hands of an 8 deck shoe where the RC kept would not drop. It was at least +35 because at that point I just didn't bother to count. I kept waiting for the 10's and A's to come out but they never did.

I just don't see how KO Preferred can even come close to performing better than a balanced system. Like you said, tons of opportunities are missed early and late into the shoe.

That 95% of the shoe also takes into account the 75% penetration :)

"I just don't see how KO Preferred can even come close to performing better than a balanced system. Like you said, tons of opportunities are missed early and late into the shoe."

KO is inaccurate as already stated. But, computer simulations show it is just behind level one balanced counts like Hi-Lo. Counting the sevens adds value. However, HUMANS using balanced systems are inaccurate too! Having to "estimate" the number of decks left to play to convert to a true count. Plus the rounding involved in the conversion. Computers have no problem with this. The question is: Can "humans" count Hi-Lo with the accuracy necessary to beat KO? Maybe in single and double deck, but in shoe games, it may well be a toss up!

As quark mentioned, KO counts the seven, a fairly important card, which helps it gain on Hi-Lo.

Even more important, KO (or any count, for that matter) is most accurate at it's pivot point, regardless of penetration. And KO's pivot, at a Hi-Lo equivalent of +4 TC, is at a very useful place.

Here, many important strategy deviations take place, and at this count is when one should have a max bet (or close to it) out.

As a result, the two counts sim out very close.

Furthermore, a true-counted KO will outperform Hi-lo consistantly for the above reasons.

For an in-depth comparison of KO and Hi-Lo (and other counts as well) read Blackjack Attack by Don Schlessinger.

I believe that this is why KO seems to provoke strong reactions from counters. People have difficulty accepting that anything that simple can work as well as it does.

I think he meant 95% through the cards that are to be dealt. Am i right Chuck? If not, tell me where this game is!!

I don't know how to simulate myself, and perhaps it's about time i learn. But for a long time, i've been looking for simulation numbers of TKO compared to Hi-Lo (including the same deviations for both counts). Anyone out there can help me with these?

Posts 927-8 were messed up during the transition to the new format. Notwithstanding, you may straighten them out by pasting onto blank and then para-spacing after each
(break) sign. G

Knockout.

Thanks,

Regards,
Hollywood

has a link to his explanation. He puts the Color KO information in, where you can link to it from his regular email address. I can add also that if you go through his Color KO system, and are at the true counting system he has, and drop counting the 2s, you have approximate indexes averaged for the nearest even true count, for the Uston +/-.

T-hopper takes the KO numbers and moves upward to his level 2 unbalanced Bushito count. Then he has found the 2 sidecounts that take Bushito to a PE near .85 and a BE of .99! This is the limit for power/ease in a practical count.

Unbalanced Zen has some high PE figures by optimizing the card values for the process of having complex sets of pivot and other numbers in a shuffle. As a result its effective PE is a bit more than the 1/2 of the true count equivalent PE, that most unbalanced systems give you in running count mode.

JSTATII moves up in power almost as well as the Bushito, starting out as an unbalanced ten count.

Brett Harris' count systems are variations of the Wong halves count with the 9 counted as zero. Here he gains a bit of BE, and PE, but increases the insurance efficiency very much.

All of the multi-level unbalanced systems use the trick of having values that are level 3 and below, but where the true count equivalent of them is the same as a much higher level count. But more power is seldom as big a factor as your playing speed and accuracy.

Example 1: Most counts with a PE above .61 have little variation between their one deck and multi-deck indexes, except insurance. If you no longer need different tables to show any practical edge increase, you are not likely to be able to detect differences in playing results, between single parameter systems either. There will be differences, but few of them will be such that you will be able to note where one count would have given better results than another. The PE is a small section of your playing edge compard to your BE.

Example 2: Even hi-lo, with a BE=.96, and PE=.51, played at 240 hands per hour, will have a higher probability of player success per actual hour played, than a perfect play computer, at 200 hands per hour.

Example 3: Even with 180+ indexes accurately used, your plays are going to be the same as basic strategy about 70% of the time.

Example 4: The difference between perfect betting and hi-lo's BE is .04, or 4% of your betting gains. It takes 34,000 hands for this efficiency difference to excede one sigma of standard deviation between these systems.

Example 5: The differences in count power are miniscule compared to the gains you would get simply having a coupon or two every 200 hands of play. Suprisingly such coupons are most commonly given out for red chip and low black chip players; ie you are more likely to get flyers for bet $5 win $10 coupons, and bet $100, win $200 coupons, than you are bet $25, win $50.

No unbalanced count will allow you to shuffle track effectively however. No matter how they are used you get the equivalent of 2 pivots etc., as you add to the running count in the subdeck segment the also unbalanced offsets from previous shuffles that make up that subdeck.

The links below rate the relative performance of several systems in 1-8 decks with average or good penetration.

http://www.bjrnet.com/thop/intsystems.htm
http://www.bjrnet.com/thop/advsystems.htm