Halves I and Halves II? News to me.

KenSmith

KenSmith

Administrator
Staff member
#1
#1
I was flipping through my copy of N. Richard Werthamer's "Risk and Reward: The Science of Blackjack" last night and I came across a table of system tag values which included the following entries:

Tags for (Ace, 2, 3, 4, 5, 6, 7, 8, 9, Ten)
Halves I: -1.5, +1, +1, +1, +1.5, +1, +0.5, 0, -0.5, -1 (BC = 0.994)
Halves II: -1. +0.5, +1, +1, +1.5, +1, +0.5, 0, -0.5, -1 (BC = 0.992)

What he calls "Halves II" is Wong's Halves count from Professional Blackjack, which happens to be what I have used for years.
His Halves I count increases the value of the Ace and offsets it with the deuce.

I confirmed his Betting Correlation calculations using our own Efficiency Calculator:
(Just double the tags to use the calculator for Halves.)

Halves I
PE: 0.4939
BC: 0.9937
IC: 0.6667

Halves II (rhe normal Halves count)
PE: 0.5650
BC: 0.9919
IC: 0.7247

Just eyeballing these numbers, I'm sure with any index plays Wong's version is still going to be superior because of the better Playing Efficiency.

I still found it surprising that the BC of Wong's count could be improved by tweaking it. Has anyone heard of this count before?
 
QFIT

QFIT

Well-Known Member
#2
#2
Werthamer is one very sharp person, and it was nice of him to include my name in his acknowledgements, BUT the book is filled with so many serious errors that it is a quite dangerous read for anyone that does not understand the science well. His efforts were amazing considering his lack of reading on the subject. But, he was new to the subject and made critical errors. R Michael Canjar pointed out some of the more dangerous in Optimal Play: Mathematical Studies of Games and Gambling edited by Stewart N. Ethier and William R. Eadington. (Ethier is a frigin genius on probability theory.)

Sorry, just realized I didn’t really answer the question. Theoretic value based on PE, BC, and IC is an estimate. An amazingly good estimate. But, certainly not good enough to calculate advantage to three decimals. Which is to say that the strategies probably have about the same efficacy, depending on the particular indexes that you use. But, I personally don’t like strategies where the ten and ace are counted differently, as I believe they are more difficult to count quickly. Most people count most cards in pairs. It is far easier, in my mind, to count pairs of cards if “high” cards have the same tag values.
 
Last edited by a moderator:
iCountNTrack

iCountNTrack

Well-Known Member
#3
#3
KenSmith said:
I was flipping through my copy of N. Richard Werthamer's "Risk and Reward: The Science of Blackjack" last night and I came across a table of system tag values which included the following entries:

Tags for (Ace, 2, 3, 4, 5, 6, 7, 8, 9, Ten)
Halves I: -1.5, +1, +1, +1, +1.5, +1, +0.5, 0, -0.5, -1 (BC = 0.994)
Halves II: -1. +0.5, +1, +1, +1.5, +1, +0.5, 0, -0.5, -1 (BC = 0.992)

What he calls "Halves II" is Wong's Halves count from Professional Blackjack, which happens to be what I have used for years.
His Halves I count increases the value of the Ace and offsets it with the deuce.

I confirmed his Betting Correlation calculations using our own Efficiency Calculator:
(Just double the tags to use the calculator for Halves.)

Halves I
PE: 0.4939
BC: 0.9937
IC: 0.6667

Halves II (rhe normal Halves count)
PE: 0.5650
BC: 0.9919
IC: 0.7247

Just eyeballing these numbers, I'm sure with any index plays Wong's version is still going to be superior because of the better Playing Efficiency.

I still found it surprising that the BC of Wong's count could be improved by tweaking it. Has anyone heard of this count before?
Click to expand...
As Norm said, i think BC and PE are more like guidelines and sometimes inaccurate. I cant really say much about BC because i didnt do much work, however BC is calculated using effects of removals of each card denomination, however the effects of removals are dynamic quantities and depend on the composition of the shoe. Typically the values for full shoe are used.

for PE, KC and I worked on sims of perfect play using composition dependent CA as implemented by his program. Running the sims is computationally expensive and i had to use up 4 different computers with high end CPUs (each computer would run a sim of 5 million rounds, sum up and average the results). The results are not published yet. But i can give two illustrative examples.

For 1 deck S17 DAS 75% flat betting (to ensure that all the gain is from playing)
Using High Opt II (full indices) ev is 0.882%
Using Perfect Play (quasi-perfect) ev is nearly doubled standing at 1.531%

If we take the ratio of ev_HO2/ev_PP=0.576 vs PE using calculator=0.667
My guess is the difference will be larger as the penetration is deeper because strategy variations will be more important.
Anyway what I am trying to get to, is that one should be wary when comparing counting systems using BC, PE and the most accurate comparison would be using a sim
 
jack.jackson

jack.jackson

Well-Known Member
#4
#4
KenSmith said:
I was flipping through my copy of N. Richard Werthamer's "Risk and Reward: The Science of Blackjack" last night and I came across a table of system tag values which included the following entries:

Tags for (Ace, 2, 3, 4, 5, 6, 7, 8, 9, Ten)
Halves I: -1.5, +1, +1, +1, +1.5, +1, +0.5, 0, -0.5, -1 (BC = 0.994)
Halves II: -1. +0.5, +1, +1, +1.5, +1, +0.5, 0, -0.5, -1 (BC = 0.992)

What he calls "Halves II" is Wong's Halves count from Professional Blackjack, which happens to be what I have used for years.
His Halves I count increases the value of the Ace and offsets it with the deuce.



I confirmed his Betting Correlation calculations using our own Efficiency Calculator:
(Just double the tags to use the calculator for Halves.)

Halves I
PE: 0.4939
BC: 0.9937
IC: 0.6667

Halves II (rhe normal Halves count)
PE: 0.5650
BC: 0.9919
IC: 0.7247

Just eyeballing these numbers, I'm sure with any index plays Wong's version is still going to be superior because of the better Playing Efficiency.

I still found it surprising that the BC of Wong's count could be improved by tweaking it. Has anyone heard of this count before?
Click to expand...
I really like this count, and actually ran some numbers on it awhile back. http://www.blackjackinfo.com/bb/showthread.php?t=13385.

For one, is was fairly simple to count(sorry qfit) since the X,2 X,3 X,4 & X,6 would off-set eachother. And second of all, it seemed to be an ideal count for wongin-in situations b/c of the 5 and A tags.

Also, a Side Count of the Ace and Tens could be used converting it into 12223210-1-3 for stiff and ins hands, once you became efficient at the main count.
 
NAP

NAP

Well-Known Member
#5
#5
jack said:
And second of all, it seemed to be an ideal count for wongin-in situations b/c of the 5 and A tags.
Click to expand...
I'm not following this. Can you please explain further? Do you mean that you only do an A-5 count while backcounting?
 
jack.jackson

jack.jackson

Well-Known Member
#6
#6
NAP said:
I'm not following this. Can you please explain further? Do you mean that you only do an A-5 count while backcounting?
Click to expand...
Certain counts have better efficiency for particular hands. And since we wong into postivie counts this count would seem almost ideal, because for "one" example would be when we split Tens at high(er) counts. Putting more emphasis on the Aces and Fives would prove better accuracy when splitting Tens because the A would give us 21 while the 5 would give the dealer 11 or 21 when facing a vs 5 or 6.

2223210-1-2-3
 
Last edited:
KenSmith

KenSmith

Administrator
Staff member
#10
#10
Zero errors is an unattainable goal, with any system. You WILL make mistakes, and yes, you will make more mistakes with a complicated system. Realizing that is essential.

If I were learning to count today, I wouldn't choose a complex count. I would use either Hi-Lo or KO, and leave it at that.
 
H

HsiaoDi

Well-Known Member
#11
#11
KenSmith said:
Zero errors is an unattainable goal, with any system. You WILL make mistakes, and yes, you will make more mistakes with a complicated system. Realizing that is essential.

If I were learning to count today, I wouldn't choose a complex count. I would use either Hi-Lo or KO, and leave it at that.
Click to expand...
For shoe games I assume...?
 
KenSmith

KenSmith

Administrator
Staff member
#12
#12
HsiaoDi said:
For shoe games I assume...?
Click to expand...
Even for double deck games I would recommend just a simple count. If you are concerned in the small decrease in hourly EV, just play a few extra minutes per session. :)

Single deck play is different enough that I'm willing to concede that the more powerful counts are probably worth it. But there are very few locales where good single deck still exists for counting.
 
R

ringlejames

Banned
#13
#13
KenSmith said:
Halves II (rhe normal Halves count)
PE: 0.5650
BC: 0.9919
IC: 0.7247

Just eyeballing these numbers, I'm sure with any index plays Wong's version is still going to be superior because of the better Playing Efficiency.

I still found it surprising that the BC of Wong's count could be improved by tweaking it. Has anyone heard of this count before?
Click to expand...
Why not just double each tag to get the same PE, BC and IC? That way you would not be

If your going to use a multi level system, would it not make more sense to play with wholes rather than halves for simplicity?

That is what I did when I learned this count.

The halves were very frustrating and I already had experience with a level three system so learning the halves seemed like it would be a hindrance rather than an advantage.

Anyways I have always thought that halves were annoying.

Just wanted to get posted up on this website. Im surprised U dud bit find it sooner.
 
You must log in or register to reply here.
Top