Out of thread, I know. But there is more about KO.

Hello Orange Cnty KO and distinguished readers. Sorry I'm a little late posting some comments about this, but I also have a job to attend, still unretired. Besides I was thinking about two things:

1) I don�t like the relation between KC and #Decks being not linear. It just bother me, just doesn't fit, given that all the other objects involved are linear.
2) The 2.75 appearing in your formulas.

Now I�ve come to a customized table -for me, at least- for determine KC, the magic number on which presence I�m gonna start raisin my bets... So help me Bernoully if I�m wrong!

Obviously the 2.75 you mention in your kind post comes from three assumptions:

I) There is a linear function between KC and #Decks
II) With 0 decks all parameters equal zero (Which in my last post I overviewed)
III) 22 is a correct value to KC for 8 decks.

With this in mind, the linear equation that links KC and #Decks is

You�ll see that speakin of money this is a nontrivial change.

WELL BELOW (LOL!) is a table wich contains: #Decks, KC values given by you, KC calculated, KC rounded. (Table went to the bottom due to some bug in the HTML here or some dwarf. I already reported it to webmaster). Please move the cursor down or press PgDown twice. TY

Of course, I know there are no 3, 5 or 7 deckers games. But there is a major change for the one deck situation, which somehow explains some strange behavior I�ve been noticing in my hours in front of my PC practicin KO. On the other hand, I chosen to round to the next integer -higher than you propose- for safety in the one and 6 decker. What do you people think about it?

Greetings from M�xico City.

You make a good point - it seems like there should be a linear relationship between the key count and # of decks.

However, it's not a true linear relationship due to penetration considerations. Single deck penetration averages 50%, double deck 60%, and shoes (4+ decks) 70%.

In KO Blackjack, the authors stipulate 65% for single deck, and 75% for 2d, 4d, 6d, 8d. While these numbers do not reflect realistic playing conditions, we should assume their key counts are based on these numbers.

If you're more comfortable with rounding, you could use (KC = 2.7 * #Decks) with an adjustment for single deck of 65%/75%. This would give you the same key counts the authors use.

I would not use key counts that are different than those stipulated by the authors. Using a higher key count will only decrease your EV.

The key counts for 3d and 7d are obviously correct - they fall exactly in the middle of the range. The key count for 5d is probably 13, but 14 is a more conservative number. BTW, there are 5d games - I played one in Amsterdam, Holland and there's one in Regina, Canada.

Hi again Orange County KO. So far we�ve had these possible 'straight line' relationships between KC and #D. I�ll make a brief summary:

A)
KC = 3*#D - 1
KC = 3*#D - 2
KC = 22(D - 1)/7

B) KC = 2.75*#D

Now there is a last one. Since the book is for players, not for boring mathematicians, maybe there was not much space for science methodology -this I assume from some words you wrote-. We can only assume how the values given by Dr. Vancura and Mr. Fuchs were obtained. They gave the values for each card and made a computer-aid simmulation for -maybe- thousands or even millions of hands. They annalized those results and applied some math and somehow validated the values. So, in strict sense, this data must be treated as experimental data. When we deal with this kind of information, the right approach is to try to find some correlation -if any- between the numbers and find the parameters.

Below is a link to a table and a graph. The graphic came first: is the one that pictures out the experimental data. The resulting equation is given by:

The table contains the original data and those numbers resulting after apply the equation -obtained by linear regresion- to get the unknown values. In some entries these results are different from others presented before, they seem to be lower than the others, WHICH IS COMFORTING!!!!

But, finnaly you may say: Hey Praomx man!!!... �What is the point of all this argument?

First of all, in my trip to Vegas I know that in the first hour or two I�ll be very busy learning the body language needed to hit, stand, double or split, and yet keeping the count, so I don�t think would be wise to play a one decker. Everyone holding the cards could be too much for a rookie. Besides, unfortunately for me there are no 'mouse and buttons' at the table. So I�ll seat on 2, 4 or 6 deckers and see how it goes. For this number or decks KC has proven to be quite stable thru almost all the equations presented.

It is the SD-game that I�m concerned about. If the book�s number is correct, no problem. But suppose for a moment that KC=2 is wrong, that should instead be 3. Then you�re going to war 'too soon' and with the wrong weapon and your EV is on danger. I�ve seen in other forums that some experts say KO is not considered very trustful for one deck... Maybe this is one reason. Maybe not.

Then again, it�s been my pleasure.

Good Cards and Greetings from M�xico City.