Truncating and Flooring at 0 counts

[BJtoronto]

Hey Guys,

I had a question about the results of truncating or even flooring certain counts at 0 that seem unintuitive to me.

Suppose we are playing a 6 deck game and exactly 1 deck into the game we are dealt 12 vs a dealer 4 and the running count is 4 meaning we are at a TC 0.75 (4/5). The basic strategy play is of course a stand with the departure being at a 0. Now, whether we are flooring or truncating, this TC 0.75 becomes a TC 0 which means, as per our index, we should deviate from basic strategy and hit.

I suspect my intuition is wrong however because if I was right the floored index for 12 v 4 would be -1, which it is not.

Is this right? Can anyone expand a little further on this.

[bejammin075]

For indexes that are zero, there's an easier way. Just use the running count, no TC conversion is needed. If I recall right for Doubled Halves, the index for 12 vs. 4 is about -1/3. Almost zero, but just on the negative side. If the RC is zero, I stand. If the RC is -1 (or more negative) I hit.

16 vs 10 is similar, although the index is about +1/3. If RC is zero, I hit, if RC is +1 or more, I stand.

And the same approach applies to doubling A2 vs. 5, A4 vs. 4, etc.


Sent from my iPhone using Tapatalk

[Gronbog]

In HiLo, 12 vs 4 (hit <= 0) and 16 vs 10 (stand >= 0) are examples of indices which are so close to zero that we make the play based on the running count alone.

[Three]

Suppose we are playing a 6 deck game and exactly 1 deck into the game we are dealt 12 vs a dealer 4 and the running count is 4 meaning we are at a TC 0.75 (4/5). The basic strategy play is of course a stand with the departure being at a 0. Now, whether we are flooring or truncating, this TC 0.75 becomes a TC 0 which means, as per our index, we should deviate from basic strategy and hit.

There are a lot of ways to generate indices. If simulation was the way indices were generated there is a big RC 0 bias for a fresh shoe since every fresh shoe has a RC of 0. It can be better to have 2 indices for TC 0. One way is index 0 where you deviate if the playing count is not negative. The other is Positive where you deviate when the RC is positive and play BS when the RC is 0 or less.

You asked about truncating. The TC 0 bin spans 2 TC's, -0.999 to +0.999. I don't like this as it is less accurate than flooring and rounding in that range. Obviously better decisions can be made if this were two bins rather than one. My above suggestion breaks it into 3 bins so you can deal with sim generated index TC 0 bias for fresh shoes.

[DSchles]

No one seized on the real reason you're confused, and it isn't in any of the answers you got. You wrote: "Now, whether we are flooring or truncating, this TC 0.75 becomes a TC 0 which means, as per our index, we should deviate from basic strategy and hit."The italicized part is wrong because you seem to feel that if you are above the index number (in this case, 0), you deviate from basic strategy. That is NOT the case. The rule for any hit/stand decision, using an index, is that if you are greater than or equal to the index value, you STAND. The proper BS play has NOTHING to do with the decision. In this case, it happens to be that attaining the index simply confirms what the BS play already is; it doesn't tell you to deviate from that play (stand).

Now, it is different for a play like, say, 16 v. 10. Again, if TC >=0, you stand (that part never changes). But, in this case standing happens to represent a departure from the BS play, which is to hit. The point is, you don't have to know what BS is; rather, you just have to know that, if you're at or above the index, you stand. Period.

Don

[Three]

No one seized on the real reason you're confused, and it isn't in any of the answers you got.

This is a much nicer way of stating your frustration with this failing than you have in the past. I know the other way just tended to piss me off as it created an adversarial atmosphere. LoL

This is a much nicer way to put it. By now I think we all understand the rest without having to constantly say it which could have adverse reactions you don't necessarily want to inspire. Thanks Don.

[BJtoronto]

No one seized on the real reason you're confused, and it isn't in any of the answers you got. You wrote: "Now, whether we are flooring or truncating, this TC 0.75 becomes a TC 0 which means, as per our index, we should deviate from basic strategy and hit."The italicized part is wrong because you seem to feel that if you are above the index number (in this case, 0), you deviate from basic strategy. That is NOT the case. The rule for any hit/stand decision, using an index, is that if you are greater than or equal to the index value, you STAND. The proper BS play has NOTHING to do with the decision. In this case, it happens to be that attaining the index simply confirms what the BS play already is; it doesn't tell you to deviate from that play (stand).

Now, it is different for a play like, say, 16 v. 10. Again, if TC >=0, you stand (that part never changes). But, in this case standing happens to represent a departure from the BS play, which is to hit. The point is, you don't have to know what BS is; rather, you just have to know that, if you're at or above the index, you stand. Period.

Don

Ah, thanks for the reply Don, really hit the nail on the head.

Sent from my SM-G920W8 using Tapatalk

[seriousplayer]

If you are interested there is a formula you can use to firm up these RC decisions that are at TC 0. But a spreadsheet makes it easier. CV Data required.

Interested in the formula that you are talking about. Could you show me, please.

[seriousplayer]

I'd have to go from memory. Check back later today when I can get to my spreadsheet.

Ok, no problem. Thanks anyway.

[Three]

I assume TC SIM has already been run. So you go define/edit strategy. Open new strategy. Click on the No true count box under Counts.
1.) Go to CV Index under Set up.
2.) Go to indexes and select the green box on the TC 0 items. I also select TC 1 and TC - 1.
3.) Start Index generation
4.) Select either multi tracking or Standard in Set Up.
5.) Press start generation. Run at least 200m rounds
6.) Now start your spreadsheet. The first column is Percentage of hands dealt.
7.) The next is Hands win, followed by Hands lost, and hands pushed. Then Average Bet.
8. You also have a column of Total hands played which you can change to your liking.
9.) Now take the percentage of hands dealt times the Total Hands played. This will give an idea of the number of hands played for that particular indice.
10.) Repeat the process in 9 for Hand won, hands lost, and pushed.
11.) Take the difference of hands won minus the number of hand lost.
12.) Take the difference times the average bet. The gives you a total of the money won or lost for that indices over the course of the hands you chose.

13.) Now run a new SIM after reduceing your respective indices by one. Be sure to save the old one.
14.) Now run another SIM upping the original SIM by one.
15.) You can select Two SIMs in order compare the percentages.

See, easy peasy, one two three... 4, 5, 6,... 7, 8 ,9 ... 10, 11, 12... 13, 14, 15 and viola you have it. LoL