KO Count Question

I am practicing the KO count and am confused on a particular point. Say I'm playing the KO Count in a 6 deck game with an initial running count of -20. At a running count of -4 you are supposed to begin raising your bets and stand on 16 vs 10. Even though on average you have an advantage on the running count of -4, it is quite possible to have a disadvantage at a running count of -4. Say that 4 decks have been played and the running count is -4. With an IRC of -20, this means that the average count after 4 decks should be a -4 (b/c the KO count is unbalanced by 4 points/deck). So, you truly do not have an advantage at this point even though you are raising your bet. If say 4.5 decks have been played then the count should be -2 but we're still raising our bet at the -4 count, so we actually have a disadvantage! So, maybe I'm missing something here but wouldn't it make sense to not raise your bet at a count of -4 if more than 4 decks have been played and for that matter, not raise your bet at a count up to a count of -2 if 4.5 decks have been played? In both cases you do not have an advantage but the book says to start raising your bet and stay on 16 vs 10 regardless of the number decks that have been played. I wonder how much this would improve your expectation if you do not raise your bet in these situations? Unfortunately, I don't know of any simulation software that can simulate the effect of only raising your bet on a -4 count if less than 4 decks have been played. Is my point valid or am I missing something here?

I was in Korea playing a game.

Some military boys came in, joined the table.

And without fail, one of the dudes would always raise his bet in neutral counts around the time 4 decks had been played.

I couldn`t resist...

"Gettin` knocked out over there?"

------------------------------------------------------------------
As for your observation ... very astute, but ...

Don`t worry about it, the sims say KO is a winner, even with these little quirks.

If you want to adjust your RC, you can do the following:

Add 4 in the first two decks (ie RC of 12 now equals 16)
don`t change the RC in the next two decks (Ie RC of 12 now equals 12)
Subtract 4 in the final two decks (RC of 12 now equals 8)

(one exception, at pivot, you are accurate no matter where you are in the shoe)

This creates a quasi-true count, but (in truth) I doubt you`d ever notice the difference in your win rate over the course of your life.

Best,
Syph

You're right. A running count of -4 after four decks of a six deck shoe have been played equals a true count of zero, and you still have the standard disadvantage.

Now if only two decks had been dealt, a -4 running count would equal a true count of +2. With three decks dealt out, -4 running equals +1.33 true. At all points across the spectrum however (and this is what unbalanced counts are about), a -4 running count will be, on average equal to a small positive true count. The error range is the small price you pay in trade for not having to convert to a true count based on the remaining decks.

BTW, count systems that are unbalanced by a half rank rather than a full rank (Red 7, KISS, UB Zen)tend to have a smaller error range in that regard.

Thanks for the responses. It always struck me as funny that I usually am raising my bets at the end of the shoe. After really understanding what an unbalanced count is all about it dawned on me that I in fact do not always have an advantage when I'm raising my bets. It's funny that the book never mentions this fact. It seems to me though that this would have a significant effect on your expectation and especially your deviation by raising your bets when in fact you have a disadvantage. I know the K-O count works but I wonder how much we're giving up when we are raising bets in an unfavorable TRUE count. It does seem pretty easy to remedy as explained by Syph but I don't want to go through the hassle of estimating the # of decks remaining and adjusting my running count if everyone thinks that it is not worth it. I guess we'll never know b/c I don't think this can be simulated with any simulators on the market.

Jeff,

You can simulate Syph's suggested KO adjustments by using a Multi-Strategy in CVSim. You'd define the "normal" KO playing and betting strategies for use on cards 105-208 (decks 3 & 4). For cards 1-104 (decks 1 & 2), you'd define the "minus 4" version of KO playing and betting strategies: thus, on the betting side you'd start to increase your bets at -8 (instead of the normal -4), while on the playing side you'd decrease all your index numbers by 4 (so you'd stand on 12 vs. 2 at 0 instead of the normal +4). Finally, for cards 209-312 (decks 5 & 6) you'd define the "plus 4" version of the playing and betting strategies: start increasing your bets at 0 (instead of -4) and increment all the normal index numbers by 4 (so now you'd wait for +8 before standing on 12 vs. 2).

I'll admit this would be somewhat cumbersome to do, but if it were easy, I'd have done it myself and presented you with a fait accompli ;-)

If you e-mail Norm at qfit.com, he might have a more elegant solution. He's always very quick and helpful when I e-mail him.

Hope this helps!

Dog Hand

I don't know the KO count in particular, but this would apply to all unbalanced counts. Given that you have this built-in disadvantage, what advantage do you gain by using an unbalanced count rather than a balanced one where you compute the true count? I mean, I guess it is simpler not having to divide by the number of decks, but even if your deck estimation skills are way off, I would assume that this is still more accurate than using an unbalanced count. Estimating the number of decks remaining down to the quarter deck or half deck in a shoe game might be difficult, but anyone should be able to be right within a reasonable margin of error (say, one deck).

I guess, my question is, WHY would you choose to go with an unbalanced count when there are simple balanced counts like hi-lo and hi-opt I to choose from? (or more powerful ones like hi-opt II if you dont mind a multi-level count)

I am vexed... wait, no, that's not it ;-)

Keith

I, and I assume many others have chose the KO Count over the HI-LO b/c this count has proven to be just as effective as the HI-LO without estimating the # of decks and dividing to get the TC. When you take into account human error (estimating errors/fatigue)then the KO comes out on top. This can be proven by any simulation. On the other hand, if I had already learned the HI-LO and could execute it flawlessly then I probably would have stuck to it. The KO just seemed to require less work to execute flawlessly than the HI-LO and if both resulted in the same result then why not. I do agree though that it's disturbing to know that you could be raising your bet at times when you don't know what the true count really is. But the sims don't lie. I decided to pass on the superior multi-level/multi-parameter counts with a ton of indexes due to the time factor in learning the system and confidence factor in playing the system flawlessly to get the extra advantage.

Unfortunately I do not have the casino verite sim software but thanks for the suggestion. This is not possible in the sim software I purchased. Although, anyone willing to simulate this and publish results would assure them a decent spot in blackjack heaven.

Vex no more ...

:)

For many games, and unbalanced count performs (for all intents and purposes) almost identically to their balanced counterparts.

If you go to bjstats.com (in the calculator section) you`ll find that KO actually outperforms Hi-Lo / I18 in many single and double deck games.

So if you are a perfectionist, eliminating any possible error resulting from TC calculations is a definite bonus. You can play like a machine, even if this handicaps your game slightly.

You can carry on a conversation, flirt with the cocktail waitress, harass the ploppies around you with your basic strategy card ... all the while a gentle hum of an RC plays in the back of your mind. Deck estimation? We don`t need any stinkin` deck estimation.

How much is this handicap? Well, the EV from TKO (true counted KO) is about 5-10% stronger than KO ... I`d hazard a guess that the little adjustment I suggest would only garner about half that.

(I have not simmed it, if anyone can ... it would be interesting to see)

One last bit, true counted unbalanced counts are generally a little more immune to TC errors than balanced counts, at least according to this paper:

http://www.bjfonline.com/Library/membr2.htm

One more link, for those who are interested in using KO, Norm has created "OK" a simpler form of KO, but one just as strong ... actually stronger with the improved betting ramps. All the deviations are reduced to two categories, check it out here:

http://www.qfit.com/OKStrategy.htm

And one final bit ...

Many find that learning different IRC`S and pivot points for KO to be a bit of a nusance. Here`s a nice solution for those who play a variety of games:

IRC= -3 x number of decks
Key = -1 (always)
Pivot = number of decks

(I think the key point fluctuates a little, depending on the game, in contrast to the original KO, but its close enough.)

Counting has never been so easy.

That all said, I prefer balanced counts as well.

Best,
Syph

(yah, i know ... i have to much spare time on my hands)

The first count I ever learned was the Hi-Opt I count. Then I learned the Uston APC count. Then I took some time off from playing. When I came back to the game the KO count had just been released and I was really dreading putting in the time to relearn Hi-Opt ot Uston APC (shudder...). I figured I would try the KO count. I quickly became convinced of its ease of use. I used it for a while.

A couple of years ago I had the opportunity to do some basic shuffle tracking with another player. He knew Hi-Lo (and it is much easier to shuffle track with a balenced count) so that is what we used. Then I had a unique opportunity to regularly do some heavy duty wonging where I was able to count down two tables at the same time. Counting two tables at the same time required a simple count with no side count and no true count conversion. KO worked beautifully.

The last couple of years my playing time has been sporadic. I will sometimes get to play intensely for a month or two and then not play again for 8 months or a year, and then only get to play for a quick weekend trip. With the sporadic play I have been using KO. For me KO is the easiest count to pick back up. It only takes a couple of days to get back in the swing of running down shoes and refreshing the play indecies. I always found the hardest skill to pick back up after a long layoff was accurately estimating the number of remaining decks. With other counts I would have to spend most of my refresh time on this skill. With an unbalenced count this isn't needed.

If I was going to start playing regularly again I would probably use Hi-Lo because doing TC conversions aren't that difficult once I get in the swing of it and Hi-Lo offers other opportunities such as shuffle tracking (if the opportunity presents itself). But knowing I am probably going to be playing very sporadically for the next few years I have been sticking woth KO. Not having to spend the time to relearn to correctly estimate the decks remaining is a huge bonus for me.

More importantly though, I don't think it really matters which count you choose. It is far more important to put in the time and effort to learn to expertly use a count than whichever specific count you choose. The differences in EV for all of the counts are very close. Just making an extra mistake or two a session will be a bigger difference than the difference between the counts themselves.

Just some thoughts.

Ok, so, KO is a powerful counting system. I don't actually know the system, but this just leaves me even more confused. Here's what I am thinking:

With KO, you are ignoring a certain piece of information, namely, the number of decks remaining. This is significant information, as it helps you get a more accurate assessment of the condition of the remaining deck/shoe.

So, it seems to me that you are giving away some power by ignoring the number of decks remaining. Now, if KO is just as powerful, or even more powerful than a count like hi-lo, then you must be doing something to counteract the fact that you are ignoring this important piece of information. There must be something else that makes it more powerful. Is it still a level 1 system? Do you have to keep track of more cards? Is there some other way in which it is more complicated than hi-lo? It doesn't make sense that you can just ignore a vital piece of information and gain power. There must be some sort of a trade-off somewhere. This is just my intuition, but it seems fairly logical to assume. Also, it seems like this would apply to any unbalanced system which is just as powerful as a similar balanced system.

So I guess my question is... I know that KO is simpler than hi-lo because you don't have to estimate the number of decks left. But.. what's the trade-off? In what way is KO more complicated than hi-lo?

Keith

The tags for KO are the same as Hi-Lo with the exception that you include the "7"s as +1. This increases the betting correlation from Hi-Lo`s .968, to about .977, and the playing efficiency from about .51 to about .54.

Can`t remember the insurance correlation numbers.

(I`m not sure if its still fashionable to rate counts on their BCs and PEs, but what the hell.)

Also, the placement of the pivot (equivalent to a +4 TC in Hi-Lo), gives an accurate point (using only the RC) to deviate from basic strategy and place your max bets out.

Neat, huh?

Basically its a slightly stronger count... which then gets slightly crippled with no TC conversion.

So it all susses out to be about the equivalent of Hi-Lo.

That all said, I`m not aware of too many serious players using it. Hi-Lo still seems to be the standard for many teams.

It would be curious to know if DD would allow any of his team to use it, and if not ... why that might be the case.

Best,
Syph

Hmmm... once you're going to count 7's, wouldn't it make more sense to "keep it balanced" by just counting deuces and sevens as half a point each? From what I remember from reading The World's Greatest Blackjack Book, the effect of removal for 2s and 7s is about half that of the other low cards. It seems (intuitively) that this should at least increase your betting correlation, and probably your playing efficiency too.

Actually now that I write this out, this sounds a lot like a counting system that I have read about before, although I can't remember which one. It doesn't really seem overly complicated though... counting by halves really isnt all that hard. If I were going to do this, could I still use the same basic idea that each point of TC is worth a 0.5% edge? And would the Illustrious 18 indices change? Would this end up working?

Keith

Here is an early post in 1997 by Ken Fuch (author of ko) regarding the benefits (or lack of) of converting the KO count to a true count.

I think.

Regarding the I18, you can keep the same values. There might be some slight differences, but I believe enough sims have been done on "rounded" indices at this point to not worry about,

Actually, if you double all the values, you`ve recreated the Revere Point Coint.

(2 to 7): 1,2,2,2,1
(Tens to Aces): -2

(BC= .984 PE=.55)

Expect about a 10% increase in SCORE over Hi-Lo.

Also, while we`re on the topic. Stanford Wong has a count that takes this one step further:

(2 to 7): .5,1,1,1.5,1,.5
(9 to A): -.5, -1, -1

(BC: .992, PE: .57)

But you might get a headache, many find it easier to double the values to create a level 3 count.

I think it`s generally considered to be one of the strongest counts for shoe games, but we`re talking very slight gains at this point.

Best,
Syph

(now I gotta get off these boards and get back to my poker game ... good cards everyone!)

Jeff,

First of all, what version of KO are you using? Rookie, Preferred, or something else? Are you using any index numbers other than those given in the Knocjout Blackjack book?

Second, what particular rule set and penetration would interest you?

Dog Hand

Jeff,

What betting scheme are you using? That'll also have an effect on the results, I believe.

Dog Hand

Jeff,

I too, follow the KO method. I do believe that when the count gets to +4 in a 6-deck game, is when they say to start raising your bet, and when it gets to +20, is when you stay with 16 against a 10.

I don't have the book in front of me, so I'm not quoting directly out of the book, but this is what I follow and I studied that book quite extensively.

If I'm wrong about what's stated in the book, let me know. Otherwise, you might give it a try..