should one round to figure the TC

Hi,

I am fairly new to counting an play an 8 deck shoe with a 1.5 (sometimes one deck with the right dealer) cut, DAS, up to 3 splits allowed and double on any 2 cards.

I normally Wong in at a TC of +2 using the Hi-Lo lite from Arnold Snyder's Blackbelt in BJ book.

Today I Wonged in with a RC of +19 and 9 half decks remaining to be played giving me a TC just over +2, with 6 players at the table including myself. After winning that hand, the RC was +13 with 8 half decks remaining.

This brings me to my question. I had figured my TC to be just over 1.4 with the rc of +13 and 8 half decks remaining to be played. I now know after breaking out the calculator I had a TC of +1.625. Of course after the 2nd hand was dealt, the dealer had an A showing. I knew to take insurance at a TC of +2, but decided against it as the TC was exactly 2. Of course, the dealer had a BJ.

So should one round up when figuring the TC? If so, do you round up only if the tc is above the .5 mark? or the .75 mark?

Any advice would be greatly appreciated.

Steve

general consensus seems to be "flooring". That is round down to the next lower integer. truncation is ugly in that TCs from just above -1 to just below +1 all truncate to zero, which makes that count twice as wide as any other.

Main issue is to calculate the true count the same way your playing indices and betting ramp were computed. If you use CVCX/CVData to generate indices or betting ramp for a specific game, then use the same round/truncate/floor policy when you play.

Not sure how math is taught now, but I was taught to round up or down from .5 It's easier and more natural for me than "flooring" or "truncating".

I say pick the method that allows for the least amount of brain drain while playing. The difference is practically negligible.

What do you mean by:

"not player optional for best individual game?"

Did you substitute 'optional' for 'optimal'?
What did you mean by "individual game"?

What you need to know is:

Bet-sizing is a VERY important aspect of card counting.
Rounding your counts will lead to sub-optimal bet ramping.
It means that sometimes you are betting too much and sometimes too little.

The former equates to unjustified increases in risk.
The latter equates to failure to maximize profits proportional to the advantage.

This means that you are increasing the flux of the game.
Increasing luck and decreasing skill.
That is not what you need to do.

My question was poorly worded but I was asking why you said to avoid rounding as compared to the other methods. Do you favor flooring or truncating? Isn't there a certain degree of error no matter how you arrive at a whole number?

Or are you saying to bet your units as a fraction equal to the true count and never arrive at a whole number. If so, I can understand going for optimal accuracy in bet sizing and playing decisions as long as one is capable and comfortable doing so.

Frenchman

Flooring, truncating, and rounding are ALL sub-optimal.

FLOORING errors will cost you the least, without much increasing of your Risk of Ruin.

Today I Wonged in with a RC of +19 and 9 half decks remaining to be played giving me a TC just over +2,"

Your math is a bit wrong. A +19 with 9 half decks would be the same as +19 with 4 1/2 full decks left or a TC of a little more than +4, not +2.

"After winning that hand, the RC was +13 with 8 half decks remaining."

Here we have a TC of +3.25.

Hi,

Thanks for your reply.

I actually used the wrong wording here. I am playing using the Hi-Lo lite in Arnold Snyder's Blackbelt in Blackjack book.

I should have used True Edge. In this hi lo lite, Arnold teaches one to divide the running count by 2 times the remaining decks. The resulting answer is the true edge and the hi-lo lite indices are actually half of those of the normal Hi Lo ones.

Steve

Except for insurance, I use rounding in 6 deck games. I know my running to true count conversion contains a certain amount of error so I use the easiest method for myself. I'm not going to take an estimated 1.8 t.c. and revert back to a one. The error would just be compounded.

Vmiman is fairly new and plays 8 decks. The margin for error in this case is such that my advice was to play in as comfortable a way as possible no matter what method to use when arriving at a whole number. I don't believe the differance is significant enough to warrent playing under harder circumstances.

One or two deck is different. Accuracy is more important in smaller spreading games IMO.

Imagine the following:

Two players.
One departs from Basic Strategy whenever the index is approached.
The other NEVER departs from B.S.

The second player will outperform the first.

Surprised ?
I was !

Surprised that you keep refering to results that are computer driven with players at the 100% accuracy level.

Everyone can not play like a machine, and at 8decks, not many people can duplicate the rainmans results. Do you not give any leeway on optimal playing as it pertains to individual talent or playing ability and conditions?

Instead of nit-picking on trying to play or bet every hand perfectly, shouldn't the focus be on an overall winning strategy based on each person's playing world?

I do not find it hard to play, count, use the I18+surrender indices and others that I have learned "just because", and play without errors. How do I know? One way is to let CVBJ nag me every now and then. Play say 600 rounds in an hour, and see if you can get thru error-free. No playing mistakes, no betting mistakes, etc.

I consider playing as perfectly as possible to be one of the tenets of using a counting system. Anybody will make a mistake here and there, but they should be in the .1%-1.0% range at worst, IMHO. The edge is already slim enough, and chipping away at it by using indices incompatible with the way you round/floor/truncate is just another way to fall behind.

All my $.02 of course... Since I first saw the discussion, I have found flooring to be easy to deal with and some practice can make it automatic. It probably is less important which method you use than to make sure the method you use matches the method used to generate the playing indices.

All of this is in the fractional-percentage-point range. But then so is the edge obtained by card counting.