Hi-Lo plus perfect insurance

[Freightman]

Freightman’s Pro Tip
When Rule of 9 produces a strike percentage of EXACTLY 33.33333%, consider hand quality prior to making insurance decision :)

I note the negative rating - why am I not surprised

[DSchles]

Well, one of the reasons might be that I pointed out an error in your previous post and, instead of acknowledging it, you simply deflected and attempted to change the topic to promote your "hand quality" concept. I guess the point is: some people never learn and some people grow tired of it.

Don

[Freightman]

Well, one of the reasons might be that I pointed out an error in your previous post and, instead of acknowledging it, you simply deflected and attempted to change the topic to promote your "hand quality" concept. I guess the point is: some people never learn and some people grow tired of it.

Don

The main point is that Novella’s can be written to substantiate minutiae - but why bother. The secondary point, which you missed, is my never ending fan base, unwilling to acknowledge good ideas and dribbling to pounce on pretty much everything else.

Now, what deflection - this ground has been covered a thousand times. The non enunciated Novella, if written, would have concentrated on the word “Perfect” - that being voluminous commentary surrounding the taking or non taking of insurance either below or above strike point. The phrase “hand quality” need not be raised as the concept of insurance below strike point has been commented on many times by credible posters.

[iCountNTrack]

Ion Saliu has a big fan base too.

[Cacarulo]

The idea is interesting as it brings together a system with a very high PE (including perfect insurance) with one that has a very high BC.
We'll see if the result beats HO2/A or not. The simulation is on its way.

Sincerely,
Cac

TEN count analysis using HALVES for betting purposes

The purpose of this study was to determine how much could be improved on
my previous analysis in which I used exclusively HALVES to play and bet except
for insurance in which I used the TEN count.

The SCORE to beat is then:

Halves/Ten:

1-12: 24.99
1-16: 29.14

I did the following study based on four types of TEN counts:

a) 0 0 0 0 0 0 0 0 0 -1 (unbalanced)
b) 1 1 1 1 1 1 1 1 1 -2 (unbalanced)
c) 1 1 1 1 1 1 1 1 1 -1 (unbalanced)
d) 4 4 4 4 4 4 4 4 4 -9 (balanced)

Although all systems have the same efficiency and the same correlation for insurance (1.0),
in the first, since only tens are counted, there is not a good dispersion in the indices.
That effect is not found in the other three. That's why the study took me longer than expected.
In the first case there was no improvement as you will see.

a)

1-12: 24.99
1-16: 29.14

b)

1-12: 25.31
1-16: 29.46

c)

1-12: 25.31
1-16: 29.46

d)

1-12: 25.36
1-16: 29.51

Still not enough to overcome HO2/A.

Sincerely,
Cac

[Gramazeka]

This is strange. What are we missing? We have all indicators of BC and IC better. Its shoes game ?

[Cacarulo]

This is strange. What are we missing? We have all indicators of BC and IC better. Its shoes game ?

The heart of the matter is in the PE. HO2 has better shoe efficiency than the TEN count.
TEN count is only better in IC (1,000) against HO2's 0,9100.
Halves has a higher BC than HO2/A but not that much: 0,99 vs. 0,98.
A good PE combined with a good BC achieve a better SCORE.
Now, what if instead of using the TEN count combined with Halves, we were to use HO2 combined with Halves?
This is sure to outperform HO2/A.

Sincerely,
Cac

[Gramazeka]

HO2 combined with Halves?

Well then already RAPC 71 + HO 2 ))

p.s.

I sometimes feel like your simulations overestimate the contribution of PE. I have a bunch of tools to double check, but I really don't like doing simulations.

[Cacarulo]

Well then already RAPC 71 ))

p.s.

I sometimes feel like your simulations overestimate the contribution of PE. I have a bunch of tools to double check, but I really don't like doing simulations.

No, RAPC 71 is a level 4. What I say is to use the HO2 indices for playing but for betting to use Halves.

Sincerely,
Cac

[Gramazeka]

RAPC 71 is a level 4. What I say is to use the HO2 indices for playing but for betting to use Halves.

This is compatible on a computer, but not in real life. But if we are using a computer, then I would replace Halves with RAPC 71.

[Cacarulo]

This is compatible on a computer, but not in real life. But if we are using a computer, then I would replace Halves with RAPC 71.

Ahh, I understand now. I thought you wanted to replace the HO2/Halves combo with RAPC 71.

Cac

[iCountNTrack]

Well then already RAPC 71 + HO 2 ))

p.s.

I sometimes feel like your simulations overestimate the contribution of PE. I have a bunch of tools to double check, but I really don't like doing simulations.

Actually a simulation is not a theoretical model, so it will not overestimate or underestimate anything and with a good RNG and large enough number of rounds will give you the "true answer"

Like i always, PE and BC as detailed in Theory of Blackjack were good and handy concepts 40 years ago. But now we can run 10 billion rounds in less than one minute, so running simulations is the only authoritative way to compare the performance of counting systems

[DSchles]

Actually a simulation is not a theoretical model, so it will not overestimate or underestimate anything and with a good RNG and large enough number of rounds will give you the "true answer"

Like i always, PE and BC as detailed in Theory of Blackjack were good and handy concepts 40 years ago. But now we can run 10 billion rounds in less than one minute, so running simulations is the only authoritative way to compare the performance of counting systems

I agree. "Old-timers" like to invoke BC, PE, and IC as the be-all, end-all manner of comparing strength of systems, but it is truly impossible to weight these three components accurately, given that every time we change the rules of the game, the pen, and, especially the bet spread, the contribution to the overall gain varies accordingly. So, trying to paint the worth of a system by invoking these three metrics alone, while not altogether invalid, certainly is not the most efficient, nor reliable, way of doing it today.

Don