Sun Runner: Zen vs HiLo

[Sun Runner]

Have you guys figured which one is better yet?

Let me know will ya'? I might want to change?.

[bjmagic]

> Have you guys figured which one is better
> yet?

Better is a relative term.

Take a look at the following from http://www.bjstats.com/bjre.asp

Depth: Good
Decks: 2
Rules: S17 DAS
Spread: 8

Hi-Lo Illustrious 18 & Fab 4 Score: 64.65
Hi-Lo Full Indexes, Heads Up Score: 76.90
Complete Zen Ill18 & Fab 4 Heads Up Score: 76.33
98 Zen Count Full Indexes Score: 75.12

Something must be amiss with the 98 Zen Count Full Indexes as there is now way that using Ill18 should beat full indexes.

[stainless steel rat]

> Have you guys figured which one is better
> yet?

> Let me know will ya'? I might want to
> change?.

>

troublemaker!



I am really convinced that for the game I have been using, DD, 66% pen, DAS, S17, LS, hilo is better. Not that I think it is better overall, or not that I think it is better for any other game. But I have run two sets of sims today, the most recent for 1 billion rounds for hilo and another billion for zen, and hilo was slightly better for this specific game.

I don't buy the "you did something wrong" any longer. I gave the exact CVCX settings I used, and verified that _everything_ was set identically for the two sims except for the strategy selected.

I think it is too easy to say "this is wrong". If it is wrong, _somebody_ ought to be able to point out _where_ it is wrong. As a computing professional, "this is wrong" doesn't cut it. I need more.

Even Don did exactly what he often accuses me of doing, testing the wrong thing. I've been using hi-lo _full_ indices vs zen _full_ indices. running a sim with I18 seems pointless although I am going to do it to see if it makes any difference...

[Myooligan]

. . . just kidding ;P

This whole saga is better than a brand new episode of MacGyver. Can't wait to find out what happens next!

[stainless steel rat]

> Better is a relative term.

> Take a look at the following from
> http://www.bjstats.com/bjre.asp

> Depth: Good
> Decks: 2
> Rules: S17 DAS
> Spread: 8

Someone has said that perhaps zen 98 is not the same as the original full zen count...

> Hi-Lo Illustrious 18 & Fab 4 Score:
> 64.65
> Hi-Lo Full Indexes, Heads Up Score: 76.90
> Complete Zen Ill18 & Fab 4 Heads Up
> Score: 76.33
> 98 Zen Count Full Indexes Score: 75.12

> Something must be amiss with the 98 Zen
> Count Full Indexes as there is now way that
> using Ill18 should beat full indexes.

[stainless steel rat]

> . . . just kidding ;P

> This whole saga is better than a brand new
> episode of MacGyver. Can't wait to find out
> what happens next!

Personally I'd take a new episode of Mac any day.

[Norm Wattenberger]

> Better is a relative term.

> Take a look at the following from
> http://www.bjstats.com/bjre.asp

> Depth: Good
> Decks: 2
> Rules: S17 DAS
> Spread: 8

> Hi-Lo Illustrious 18 & Fab 4 Score:
> 64.65
> Hi-Lo Full Indexes, Heads Up Score: 76.90
> Complete Zen Ill18 & Fab 4 Heads Up
> Score: 76.33
> 98 Zen Count Full Indexes Score: 75.12

> Something must be amiss with the 98 Zen
> Count Full Indexes as there is now way that
> using Ill18 should beat full indexes.

The '98 Zen Count is a 'lighter' version of Zen. It should not perform quite as well as the original Complete Zen. It also divides by quarter decks which may reduce the effect of bet optimization.

Serious Blackjack Software

[Parker]

I think I may have figured this out.

You have been using "Zen Complete" as defined in CVCX.

The game in question is a double deck game with Late Surrender.

I just took a look at the strategy tables in CVCX and "Complete Zen" has [b]no surrender indices![/i]

[bjmagic]

> The '98 Zen Count is a 'lighter' version of
> Zen. It should not perform quite as well as
> the original Complete Zen. It also divides
> by quarter decks which may reduce the effect
> of bet optimization.

So you're saying that the original zen with 18 indexes is better than the rounded 98 zen with over 100 indexes?

Are Auston's simulations detailed on pg 79 of Blackbelt in Blackjack 1998 wrong. I quote "It took Auston a billion hands ... to determine that the lite indexes underperfomed the regular indexes by about two-hundreths of one percent."

[Parker]

Again, looking at the CVCX stategy tables, I find that the "Complete Hi-Low" strategy tables list a total of 215 indices, including 30 different ones for surrender.

The "Complete Zen Count" tables list a total of 123 indices, and as previously mentioned, no surrender indices. Even the surrender basic strategy shown is not optimal, always surrendering 16 vs 9, 10 and A.

This differs with my Double Deck "Ultimate Basic Strategy Card." :-)

[Myooligan]

[Koolipto]

...I ran those exact same games on CVCX with 2D DAS (no surrender), 66% pen 1:8 spread but no forced bet ramp and I got the following cSCORES: Zen=70.34 and Hi-Lo=74.26.

Believe me that I am not trying to argue either side of this...just trying to figure out why the results are counterintuitive (or where I went wrong). The amount of effort expended by all in wading through the volumes of lengthy e-mails is ridiculous (and I am sorry to add to this). Am I missing something? I hope the explanation is obvious to someone.

The exact sim files used were:
Zen 1998 Full Indexes - 04/13/04 11:23
Hi-Lo Full Indexes - 04/04/03 06:23 [first]

I've recorded every setting in case anyone wants me to e-mail them. I will not mind being embarrased on this one.

> I think I may have figured this out.

> You have been using "Zen Complete"
> as defined in CVCX.

> The game in question is a double deck game
> with Late Surrender.

> I just took a look at the strategy tables in
> CVCX and "Complete Zen" has no
> surrender indices!

[bjmagic]

The Zen 1998 version is count per quarter deck. Did you use quarter deck resolution in CVCX for the indexes?