For years, researchers and card counters alike, have used Griffin?s EoR?s
(SD Vegas Strip rules with NDAS), to estimate the betting correlations of
the different point count systems, available on the market. Even today,
if you ask the typical Hi-Lo player what his/her BC is, no doubts the answer is an invariable 0.97. The fact that almost all of them play exclusively the shoe-games won?t change a dime their appreciation of the Hi-Lo, as a ?simple and powerful count.? The KISS principle is in vogue, so, it?s hard to swim crosscurrent, admittedly.

Hi-Lo correlation and betting efficiencies.

Using TOB, page 44, EoR?s we get:

cc = .9682, thus rounding
cc = .97

If we want to know what our bc for multiple-deck play is, first at all, we?ll
need revamped eor?s for 6dks.

Base rules: 6dks, s17, das, spl3 and spa1 bse = -0.4041 


 

2	 0.3896 

3	 0.4471 

4	 0.6038 

5	 0.7665 

6	 0.4192 

7	 0.2699 

8	-0.0392 

9	-0.2151 

T	-0.5169 

J	-0.5169 

Q	-0.5169 

K	-0.5169 

A	-0.5741 

	 

ss	 2.9985 

cks	 0.0000



The table above is normalized and presented as one-deck values, for easiness
of comparison, obviously. If you want to know how its look in a six-deck
format, divide each entry by 311/51 and voil?.

E.g. the five would be =.7665/ (311/51) = 0.1257

Comparing this table with the one on page 44 (TOB), we note an increased volatily
(no surprise at all, with das in effect) as well as noticeable differences with the fives
and aces, among other minor ones. What?s important for us is to determine
the betting correlation of our point count. And so we get:

cc = .9620, thus rounding

cc = .96

Not a big deal the penalty for multi-deck, right? Maybe not, maybe yes.

It is my educated guess, that most of the players do have difficulties evaluating the betting efficiencies of their particular count systems. To swap terms like correlation coefficient and betting efficiency looks a common practice, among novices and week-end warriors.

The ?problem? with multi-deck (is there anything else, handy?) is not only that the strategic
gain diminishes compared with the hand-held variety, also the betting gain goes South. Level-one betting systems and mainly during the first decks of the deal, do suffer these effects in a more noticeable manner. Good news is that the efficiency pop-up, while our shoe is being depleted, and yes, as a limit it approaches the counting correlation. Evaluate yourselves.


Hi-Lo betting efficiencies for different penetrations. 


 

Dks	Cards 		Efficiency 

seen    left  

			 

0.5	286		0.8951 

1.0	260		0.9217 

1.5	234		0.9325 

2.0	208		0.9387 

2.5	182		0.9431 

3.0	156		0.9464 

3.5	130		0.9490 

4.0	104		0.9513 

4.5	 78		0.9534 

5.0	 52		0.9555



On average with the Hi-Lo point count you can count with a 0.52%
advantage for each positive true-count gained through counting. The main problem
you face as a user is the standard deviation. An average of 0.22% is
a good estimate. So when you have, let?s say TC = 2 exactly, actually you have:

((0.52 * 2) ? HA) +/- 0.22

For the above given set of rules, solving yields:

Something between 0.42% and 0.86% advantage over the house.

As you can see, it pays to be cautious with your bets. This is especially true during
the first two decks, where the Hi-Lo is more exposed. Betting the farm here doesn?t look a good idea, feel obliged to add. When in doubt, try to err on the conservative side.

A proven fact is that the vast majority of BJ players, pros included use the Hi-Lo point count. So, what? Nothing, the shoe can be beaten with its aid. No breaking-news at all!

Just don?t get crazy driven your ?Volkswagen?, pretending you?re inside a ?Mercedes-Benz.?

Until next,

Sincerely,

Zf