Any interest in a card counting system?

I have put together a top notch count that would be an excellent addition to BJA4 or a similar book. Not to give anything away I'll just say that the tag values are well known but I have put a new twist as to how to implement the system under almost any conditions. What I have come up with is simple, completely original, and the performance results kick butt.

This is a serious post. I have put a lot of time and energy into this system and do not wish to just give it away. Any advice on what my options are is greatly appreciated.

Without knowing the details of your 'new twist', I can say it doesn't sound that unusual to me. Wong made improvements to Hi-Lo. In 'Modern Blackjack', Norm introduced his 'felt' system, which uses the same tags as Revere Point Count. Most serious players make their own adjustments or refinements to whatever system they choose to play. I myself play hi-lo, but use a side count and a few other minor refinements and a more aggressive bet ramp than traditional hi-lo. Does that make it NOT hi-lo?

If you have truly found something so spectacular that it 'kicks butt' under almost any conditions, you would be more interested in playing it than selling it.

kewljason,

Why can't dcbclimb do both simultaneously? After all, it isn't as if I cannot play AOII because Bryce Carlson is still playing it... right?

Dog Hand

First of all, post some numbers on under what conditions and just how much it allegedly outperforms other well known systems and the methods you used to derive the numbers.

First of all I said the performance results kick butt, I never said it outperforms anything. I think it kicks butt because the results are on par with other popular yet more difficult systems.
There are 52 index numbers, splitting Tens is not recommended and will not be included in any of the simulations. One could obviously use fewer indices but more is better and 52 seemed like an appropriate number. I will use all 52 for the simulations. There are simple betting strategies for play all, 2 deck Wong-Out, and 6 or 8 deck back count. Play all and back count bet ramps for 6 & 8 decks are identical. I will use these in the simulations as well. One should be able to switch # of decks or rules mid-session with no problem.

CVCX 3 billion rounds
Play All Scores (some bets may not be optimal)
1 Deck H17 50% 3 players, 1-4 spread: 59.68
2 Decks H17 DAS 60% 3 players, 1-8 spread: 44.68
6 Decks S17 DAS LSr 4.5/6 4 players, 1-20 spread: 43.07
8 Decks S17 DAS LSr 6/8 4 players, 1-20 spread: 27.43

You have only posted half the information necessary for a comparison. Perhaps you should post results from the same sims using the the original count that you modified or even generic hi-lo for comparison sake.

"I myself play hi-lo, but use a side count and a few other minor refinements and a more aggressive bet ramp than traditional hi-lo. Does that make it NOT hi-lo?"

If you have a set of completely different indices that have never been published before then yes. There would be nothing stopping you from calling it "new". Norm mentions that he wanted to call his FELT count RPC-Lite but couldn't gain permission. My system is a more advanced version of UBZen-II (no it is not true counted). I may be mistaken but I don't think George C. owns the UBZen-II tag values. I do not know George C. but doubt he cares what I call my system. Having said that I'd be happy to call it whatever he wants. My work as well as Norm's is original, why wouldn't we call it "new".

I'll run the same simulations using the Zen count defined by Arnold Snyder in the 2005 3'rd edition of Blackbelt In Blackjack. I'll reference the appendix and use the same 52 indices which my system uses, no Ten splits. All Zen indices will be multiplied by 4 in order to use "full decks remaining" true count calculations. Also, for some reason I have the double 10 vs. Ten index of +1 crossed out and replaced with +2 so I will use that instead (probably a more risk averse index). I can't find anything in Snyder's book about truncating, flooring, or rounding true counts so I will truncate. Remaining decks estimation accuracy will be quarter deck for all sims. Quarter deck accuracy results were actually slightly better than exact card accuracy in single deck. Rounding vs. truncating produced practically identical results as well.

I will also simulate Norm Wattenberger's FELT count found in Modern Blackjack at qfit.com. I'll use the Full versions found in the appendix for the appropriate rules and number of decks (different indices for single deck and multi-deck). I believe Norm forgot the 1 deck indices for doubling 8 vs. 4 and 9 vs. 7 so I'm going to use the 6 deck indices here. FELT floors true count calculations rather than truncate but I will use the same deck estimation accuracy as Zen (1/4 deck). Although FELT-F has more than 52 indices most of them are for pair splits which make little difference. Basically this is a Fairly Accurate Unbiased Level Two (FAULT) count comparison.

I'm going to name my system SIM52 because it's simple and there are 52 indices excluding Ten splits. SIM actually stands for something but I can't give away my twist just yet. I'm hesitant to call the count my own because like I said the tag values are well known (UBZen-II and it is not true counted). However I do feel that the implementation of the count is original and completely different. That's why I'm calling it my system but not my count. SIM52 uses the same indices for 1, 2, 6, or 8 decks with 10 differences for S17/ H17. To remember both sets just make the index a decimal (the integer would be for S17 and the decimal for H17) keeping the total at 52.

The impetus of all of this was when I made the mistake of customizing a system for one particular game. When that game ran out I was unable to play until I re-learned new indices. SIM52 is IMHO much easier than both Zen and FELT. If you can add, subtract, and multiply simple numbers you can easily manage SIM52, there is no deck estimation or true counting. SIM52 also has simple bet ramps which happen to be identical for 6 and 8 decks; these will be used for the results below.

As stated earlier, I will consider sharing the system with a respected authority if they are interested or wish to verify. Last I'm going to re-run everything with 5 Billion rounds!

CVCX Play All Scores: Zen (52 indices) / FELT-F / SIM52

1 Deck H17 50%, 3 players, 1-4 spread:
60.77 / 57.67 / 59.95

2 Decks H17 DAS 60%, 3 players, 1-8 spread:
45.19 / 42.80 / 44.68

6 Decks S17 DAS LSr 4.5/6, 4 players, 1-20 spread:
44.35 / 42.97 / 43.49

8 Decks S17 DAS LSr 6/8, 4 players, 1-20 spread:
27.86 / 26.85 / 27.31

Sent you a private message.

they can really today only own the name for the count and perhaps the fame for inventing it. 30 years ago simulations to create counts were very difficult and few could do it. It was easy for Humble and Braun to develop Hi-Opt II and force people to pay good money for it. They were the ones who had access to one of the large mainframe computers needed to develop a count system. (There were no PC's yet). Today anyone can generate a count system and index numbers to go with it in an hour or two. Snyder had an article a few years ago called something like "Inventing the Best Count System" in which almost every conceivable set of tag values was tested for PE, BE, and IC.

Is anything under the sun (recently) really new? Wong High Low is not much different from Braun's High Low which can be tracked all the way back to Harvey Dubner in 1963.

What are you wanting for it