Question about KISS with Ace Side Count

Recently I began making a switch from KISS III to UBZII for double deck games, the reason being that I want a higher playing efficiency and insurance correlation. when I play DD.

Stats for KISS III : .98 .56 .78
Stats for UBZII : .97 .62 .84

Then I began to look a little closer at QFIT's page on playing strategies and I noticed that KISS II had this:

KISS II : .90 .62 .87

So it has the same PE and a higher IC than the UBZII and the only difference is that you don't count the Aces in KISS II. So theoretically could I use KISS III but just side count the Aces and use the aces for deciding my bet and take them out for index and insurance calls to achieve the higher PE and IC?

Would this make my stats look like: .98 .62 .87?

Or am I completely misinterpreting what I'm looking at? (most definitely not impossible :laugh

It would work, except for the fact that KISS III also includes 7's in its count, which means that you have to side count those too.

Oops, didn't notice that 7's aren't counted in KISS II. Still an interesting idea.


Would I need to side count them both separately, or in the same side count?

Just off the top of my head... no math analysis...

Why not side count Aces and add +1 to your bet count for every extra Ace remaining. Play the indices the same. Your only gain would be in betting efficiency.

my guess it would be something like .96 .62 .87

If you're going to be spending a high proportion of your playing time at DD, I believe your best move is to go to a balanced count that semi-neutralizes the Ace, such as Zen or Mentor -- and forget the side counts.

In double deck play, you generally must "play all", practically speaking. TC's of +3 and -3 are fairly common. Since you'll be playing most of those -2, -3 and -4 TC's, you want/need to kick in your negative index plays. But all unbalanced counts are very poor at playing negative indices.

That's why comparing published PE's of unbalanced counts with balanced counts is "apples-to-oranges". That's because the unbalanced link between the RC and TC dissolves at negative counts! Their Insurance correlation however, holds up pretty well -- as does their betting correlation because unbalanced systems are calibrated in the direction of rising counts -- and away from falling counts.

Case in point: UBZ has a 62% published playing efficiency. That would be the case if you true counted it. But used right out of the box, it plays the hands considerably poorer than that! The 62% is just a calculation of what that card tag configuration is capable of if utilized fully. This PE shortcoming is true of all unbalanced counts.

With the regular Zen, or with Mentor, you will be playing your hands with a 62% or 63% PE -- and the semi-neutralized Ace will make your Insurance correlation pretty good too. As for me, the sims I've run show regular Zen and Mentor to perform well above UBZ in double deck games. Cross-check please??

I didn't know that Fred. Thank you for the info. Maybe I'll have to finally make the switch to a balanced then.

I have charted the results of 20,000 double-deck sims based on the DD rules currently found in the US at Modern Blackjack Page 192. However, these sims use the Sweet 16 indexes and will not see the full effects of playing efficiency. The problem with comparing strategies with full indexes is that they are all published with different index sets.

Norm, I'm assuming that your sims with Hi-Opt I, Hi-Opt II and AOII were using a side count of Aces???????

Also, I'd say it's probably reasonable to assume that systems with a higher PE would move up in ranking if 40 indices were played by all systems in the double deck comparison analysis.
It's not all that rare that you'll find yourself hitting 14 vs. 2, and 13 vs. 3, and 13 vs. 4 -- and not doubling with 10 vs. 9, and 10 vs. 8, and A/2 vs. 5, and A/4 vs. 4, and A/7 vs. 3 -- then going ahead and doubling with 8 vs. 5, and 8 vs. 6, and 9 vs 2, and 9 vs. 7, and 10 vs. 10, and A/8 vs. 4, 5 or 6 -- then splitting 9/9 vs. 7, and occasionally even 9/9 vs. A. It makes the game more fun too.

Is this still true with KISS III if you added 2 to your running count for every deck played for playing decisions when you went into negative counts? Would it be able to handle the PE better?

Edit: or would you alter the index number for the variation? Like how 16 V 10 is 15 for 6 Deck, but after 4 decks in the neutral count should be 17. You wouldn't stand with a neutral count.

So would the index plays be based off 3 decks in? If so would you subract 2 early on and add2 after the 3 deck mark for every deck played? I'm
probably thinking about this too hard considering I'm about to swith counts anyway.

With unbalanced counts, handling the negative index plays becomes more work than using a balanced count. For example, with 12 vs. 5 and with 13 vs. 3, you should hit if the TC is anywhere below -2.0

Now, if you're just a half deck into a double deck game using KISS, a neutral RC at that point would be "18". With a deck-and-a-half left, you would hit either hand at all RC's below "15" (2 x 1.5 remaining decks = 3).
But if you were a deck-and-a-quarter into it approaching the shuffle, a neutral RC would now be "19.5". With just three quarters of a deck left, you'd hit any RC below "18". That's how you'd get a true 56% PE out of KISS III. You'd need to do the same thing with UBZII to get the full 62%.

More extreme negative index plays require larger adjustments -- such as not doubling 11 vs. 10, and it does come up. That carries a TC of -4.0.
Picture it. There's three-quarters of a deck in the discard tray and the RC is "13". You've got 11 vs. 10. What do you do?

Well, a neutral RC here would be "18.5". With 1.25 decks remaining, 5 points below neutral would be -4 true. You're 5.5 below neutral (an actual TC of -4.4), so you should just hit it!

If you play double deck and only estimate to the nearest half deck, you'll make some mistakes. In the previous example when three-quarters of a deck was in the tray, if you called that a half deck, your computed "hit" RC with an expected 1.5 decks remaining would be "11", and you'd double. How much difference will those kinds of mistakes make? I dunno. But you can see how much adjusting goes into trying to play all your negative indices right with an unbalanced count.

It's pretty easy to do all the "-1 TC" plays, and just let it go at that. Simply note what a neutral count would be at this point, and hit it if the RC is below that by more than the remaining number of decks. Pretty straightforward! This would include hands like 9 vs. 3, and 12 vs. 6 S17 only, and 13 vs. 2, and A/2 vs. 5, and A/4 vs. 4, and A/8 vs. 6 H17 only, and 4/4 vs. 5 DAS only, and 3/3 vs. 2 DAS only.

Beyond that, you have to perceive "2 x the remaining decks" and 3 x the remaining decks", etc. In the end, it's an individual choice.