MJ: Flooring Question

[MJ]

If one decides to floor the TC for Hi-Lo, wouldn't the net effect ultimately be putting less $$$ on the table as opposed to somebody who rounds the TC to the nearest integer? Let me provide an example of what I mean:

If the TC = +1.9 and you decide to floor, then the TC = +1.0 and you place a minimum wager. However, if you round to the nearest integer then the TC = +2.0 and you INCREASE your wager. This simple example is applicable anytime the TC = +.5-.99, +1.5-1.99, +2.5-2.99,...etc. So,if you floor the TC it appears as though you will put less action on the table vs the player who rounds to the nearest integer. I reckon this is due to the increased frequency of lower TC's that comes about as a result of flooring.

Furthermore, it also appears that one who rounds will place wagers that are closer to the proper kelly bet. With a TC = +1.9, do you think it's better to place a wager consistent with a TC = 2.0 or 1.0?

-MJ

[Norm Wattenberger]

recalculate your optimal bets accordingly. The point is you should calc your betting ramp using the TC frequencies and advantages that exist for the method of TC calculation used. Change the method from rounding to flooring, and all the numbers change.

Serious Blackjack Software

[MJ]

Any chance you will make a canned SIM available for CVCX(Hi-Lo) that uses flooring for the TC as well as the indices from BJA3?

-MJ

> recalculate your optimal bets accordingly.
> The point is you should calc your betting
> ramp using the TC frequencies and advantages
> that exist for the method of TC calculation
> used. Change the method from rounding to
> flooring, and all the numbers change.

> Serious Blackjack Software

[Norm Wattenberger]

> Any chance you will make a canned SIM
> available for CVCX(Hi-Lo) that uses flooring
> for the TC as well as the indices from BJA3?

I'll probably get around to it. Of course users can do it themselves.

[Phinitum]

On a similar theme, quite often the betting ramp comes out with a 2 unit bet at a floored/truncated TC of 1. That being based on a range where the weighted (I assume) average IBA is just over .3% but the IBA at a TC of exactly 1 is presumably about 0%. In practice I ramp up when the TC just exceeds 1, but in CVData I don't have that choice. It would be helpful if CVData supported that option to see if there is any value to the refinement.

>Any chance you will make a canned SIM available for
>> I'll probably get around to it. Of course
>> users can do it themselves.

[Norm Wattenberger]

Have to think about that. It might improve cover betting.

[Don Schlesinger]

> Any chance you will make a canned SIM
> available for CVCX(Hi-Lo) that uses flooring
> for the TC as well as the indices from BJA3?

All of the charts in Chapter 10 of BJA3 were created that way. That would be a start, no?

Don

[Brick]

> Have to think about that. It might improve
> cover betting.

Phinitum is probaly more concerned with the effects of ev and risk of ruin. I thought CV data allows these specified features.

For what ever reason,if I chose to find optimal bets at a TC of 1.5 can I do this by simulations using CV data?

[Norm Wattenberger]

> Phinitum is probaly more concerned with the
> effects of ev and risk of ruin. I thought CV
> data allows these specified features.

Yes, but not a small fractions of a count.

> For what ever reason,if I chose to find
> optimal bets at a TC of 1.5 can I do this by
> simulations using CV data?

No, CVCX.

[Brick]

that was fast,what a great answer!

[Phinitum]

Thanks for checking. I doubt I'd do a good job of using fractional TC's for betting so it wouldn't be realistic for me to sim that, but I do manage to distinguish 'equals' from 'greater than'. And yes, cover considerations do often eliminate my use of the information, but I still like to know when I'm betting with an edge and when I'm just taking an even gamble.

> Yes, but not a small fractions of a count.
> No, CVCX.

[Norm Wattenberger]

> but I still like to know
> when I'm betting with an edge and when I'm
> just taking an even gamble.

Yes. But remember, card counting is a compromise strategy; tag values are only an approximation of the actual Effect of Removal of the cards. When you are close to an EV of zero, the count simply is not enough data to determine positive or negative EV. You would need a combinatorial analysis of the remaining cards. The good news is that it really doesn't matter. Any edge would be too tiny to care about.

[MJ]

For the CVCX canned Hi-Lo I-18 Fab4 SIM, was it set up to round the TC to the nearest integer? Were any multipliers used for the TC conversion?

Also, I read on your website(see quote below) that different methods can be used to estimate the discards for the SIM. I understand you used 1/2 deck precision in the SIM, but what specific style did you use to estimate the depth? What range of cards that would constitute 1/2 deck played? Thanks for any clarification.

-MJ

"You can now set up an exact table of divisors by user defined deck segments to more accurately match the sims to your specific style of estimating depth and calculating. You can also use multipliers instead of divisors."