Praying Mantis: UBZ II or TKO?

[Cacarulo]

> If you double the values of Halves tags you
> should use 2 as the correct TC divisor.
> Otherwise you would made Halves somehow
> imprecise. E.g. the insurance index = 7.
> Probably here lies the problem. If you use
> Halves the way I count it, then the 1 is the
> correct divisor obviously. I don't think we
> have here an apple to apple comparison
> because that way Halves is at an initial
> slightly disadvantage with the Zen count.

If for Halves you use 2 as the TC divisor then you'll have to use 2 for Zen as well (Apple-to-Apple).
Maybe all these discrepancies have to do with the inclusion of LS. Probably CVCX sims used different indices (published) or the number of rounds were not enough.
I've generated indices (without LS) using CVdata and SBA and both are the same. This means that my sims should be correct. In any case, I would like to hear Norm's comments.

Sincerely,
Cac

[Zenfighter]

Rules: 6dks, das, spl3 and spa1.

EV max indices, precision 3.5 sd Max n of pairs at index 200 million.

 

 

Insurance	3 

16 vs T	        0 

15 vs T	        4 

16 vs 9	        5 

12 vs 6	       -1 

12 vs 5	       -2 

12 vs 4	        0 

12 vs 3	        2 

12 vs 2	        4 

13 vs 2        -1 

13 vs 3	       -2 

9 vs 2	        1 

9 vs 7	        4 

11 vs A	        1 

10 s A	        4 

10 vs T	        4 

8 vs 6	        2 

8 vs 5	        4 

A8 vs 6	        1 

A8 vs 5	        1 

T,T vs 6	5 

T,T vs 5	5 

 

I don?t agree that Zen can out SCORE Halves provided 1-12 and or 1 ? 16 spread in any given standard shoe,

(4.5/6 and/or 5/6) if the Halves player uses the above printed indices. How can an ace-reckoned two level count (Zen) beat another three level one? (Halves) Can?t be true.

Sincerely

Zenfighter

[Don Schlesinger]

> I don?t agree that Zen can out SCORE Halves
> provided 1-12 and or 1 ? 16 spread in any
> given standard shoe,
>
> (4.5/6 and/or 5/6) if the Halves player uses
> the above printed indices. How can an
> ace-reckoned two level count (Zen) beat
> another three level one? (Halves) Can?t be
> true.

I agree. Halves was meant to be true counted by whole decks, using the half-indices that Wong created. If you double all the indices, then you true count by half-decks remaining.

Use Zen however Snyder said you should.

It can't possibly outperform Halves in a play-all 1-12 shoe game. No way.

Don

[Cacarulo]

> Rules: 6dks, das, spl3 and spa1.
> EV max indices, precision 3.5 sd Max n of
> pairs at index 200 million.
>
> Insurance 3
> 16 vs T 0
> 15 vs T 4
> 16 vs 9 5
> 12 vs 6 -1
> 12 vs 5 -2
> 12 vs 4 0
> 12 vs 3 2
> 12 vs 2 4
> 13 vs 2 -1
> 13 vs 3 -2
> 9 vs 2 1
> 9 vs 7 4
> 11 vs A 1
> 10 s A 4
> 10 vs T 4
> 8 vs 6 2
> 8 vs 5 4
> A8 vs 6 1
> A8 vs 5 1
> T,T vs 6 5
> T,T vs 5 5

So now you would have to use half-deck estimation for TC purposes, right? If I were to run a FAIR comparison I would need to do the same for Zen.
In my sims I used full-deck indices and full-deck estimations. Maybe something is reverted when more precision is used but I don't know. We have to run sims for this. You agree that we can't compare a system estimated to half decks to another estimated to full decks, don't you?

Sincerely,
Cac

[Zenfighter]

For the same rules as above and with the same degree of precision, here you have a full set of Zen indices adjusted to half ?deck estimation.

 

 

Insurance	3 

16 vs T	        0 

15 vs T	        3 

16 vs 9  	4 

12 vs 6	       -1 

12  vs 5       -1 

12 vs 4	        0 

12 vs 3	        1 

12 vs 2	        3 

13 vs 2	       -1 

13 vs 3	       -2 

9 vs 2	        1 

9 vs 7	        3 

11 vs A	        1 

10 vs A	        3 

10 vs T	        3 

8 vs 6	        2 

8 vs 5	        3 

A8 vs 6	        1 

A8 vs 5	        1 

T,T vs 6	4 

T,T vs 5	4 

	 

 

 

You have to use them because your simulator do not support half-integer math. Am I right?

Otherwise were would be the problem to match full deck vs. full deck? If this is the case then you?re forced to use these Zen?s ones adjusted for half.

Good Luck! :-)

Zenfighter

[Cacarulo]

> For the same rules as above and with the
> same degree of precision, here you have a
> full set of Zen indices adjusted to half
> ?deck estimation.
>
> Insurance 3
> 16 vs T 0
> 15 vs T 3
> 16 vs 9 4
> 12 vs 6 -1
> 12 vs 5 -1
> 12 vs 4 0
> 12 vs 3 1
> 12 vs 2 3
> 13 vs 2 -1
> 13 vs 3 -2
> 9 vs 2 1
> 9 vs 7 3
> 11 vs A 1
> 10 vs A 3
> 10 vs T 3
> 8 vs 6 2
> 8 vs 5 3
> A8 vs 6 1
> A8 vs 5 1
> T,T vs 6 4
> T,T vs 5 4
>
> You have to use them because your simulator
> do not support half-integer math. Am I
> right?

My simulator does support half-deck estimation so when I get some time I will run 2 more sims but with your indices. You could also run the same sims to double-check what I eventually get.

Sincerely,
Cac

[Zenfighter]

I think we have to wait for Norm for a double check.

Meanwhile somehow surprised by Zen beating Halves in shoe games (provided a 1 to 12/16 spread) and after reading you calmly I?ve extracted the following based exclusively in Richard Reid?s OSR formula who evaluate potential gains based exclusively in linear estimates of Betting correlations, Playing efficiency and Insurance correlations. Here are my results:

 

Overall System Ratings Comparison 

 

Count			1-12		1-16 

 

Halves			97.99		98.18 

 

Zen			96.65		96.58 

 

The higher IC of the Zen count doesn?t act with enough power to fight the monster BC of Halves.

Pure math, I know, no cut card placement, so pure linear estimates in gain. Moral? Let?s wait for confirmation.

Hope this helps, anyway.

Zenfighter

[Don Schlesinger]

I thought I provided the CVCX comparsion above, for I18. Adding four more indices can't possibly make any major difference to completely reverse the magnitude of the SCOREs.

There simply is no logic that I can think of behind Zen's outperforming Halves.

Don

[davinci]

> I thought I provided the CVCX comparsion
> above, for I18. Adding four more indices
> can't possibly make any major difference to
> completely reverse the magnitude of the
> SCOREs.

> There simply is no logic that I can think of
> behind Zen's outperforming Halves.

> Don
Back at original question. I changed about 6 mos. ago from KO to UBZ2. I found the level 2 counting easier than true, and I had little trouble adjusting to new values.