TC Conversions

[Memphis10Tigers]

I've been doing quite a bit of practice with true count conversions lately, and have been researching some of the different types of rounding for TC conversions. I saw the following on QFIT's site:

Truncate – For positive numbers, round down and for negative numbers round up. This is the method used in the 1994 and later editions of Professional Blackjack. 1.5 is rounded down to 1. -1.5 is rounded up to -1.

Floor – Numbers are always rounded down. This is the method used in the early versions of Professional Blackjack. 1.5 is rounded down to 1. -1.5 is rounded down to -2. This is the most popular method used now.

Round – After the True Count division, the result is rounded to the nearest integer. If the number is exactly between two integers, it is rounded up. 1.5 is rounded up to 2. -1.5 is rounded up to -1. This method is also common.

Statistical Round – After the True Count division, the result is rounded to the nearest integer. If the number is exactly between two integers, it is rounded to the nearest even number. 1.5 is rounded up to 2. -1.5 is rounded down to -2.

My questions would be which is optimal and which do you use? I would assume the floor would be the best as it is the most conservative for betting correlation purposes?

Also, what is the default in QFIT software?...I haven't gone through the settings to figure it out yet...

Lastly, how do I add an Avatar to my profile??

 

[Wookets]

For the avatar:

1) Click "User CP" located in the top left of the menu bar

2) On the lefthand tab, click "Edit Avatar"

3) Select the "Use Custom Avatar"

4) Either link to the URL of your image or upload it from your PC

 

[tthree]

Flooring is most optimal. It keeps you from hurting your advantage gained by the index. Rounding a large TC up and hitting the index routinely will make your expectation for that index negative. That is because of the much larger frequency of the erroneous TC compared to the frequency of the TC at or above the TC index. This effect is nominal for negative indices because the error is made a tiny fraction of the number of times the index is used and at minimum bet. The important thing is not to overestimate a large TC. It is a very costly mistake when you use an index you otherwise would not have.

 

[tthree]

Flooring your discard tray deck estimate is better. There is no need to round a TC estimation. TC = 7 and change, you know to use your 7 or less index play but not your 8 or more.

For betting purposes I imagine you could do either without to much problem. But it is always better to be conservative. You could also pick a nice round bet amount in between if one exists.

 

[Memphis10Tigers]

Flooring your discard tray deck estimate is better. There is no need to round a TC estimation. TC = 7 and change, you know to use your 7 or less index play but not your 8 or more.

For betting purposes I imagine you could do either without to much problem. But it is always better to be conservative. You could also pick a nice round bet amount in between if one exists.

So...Assuming 6-deck, RC = 7 and Discard tray = 2.5. You floor Discard tray to 2, leaving 4 decks remaining and arrive at 1.75. Do you also floor this to 1? Or do you round up to 2 since you have already applied an element of conservatism in your discard tray estimation?

This element really throws me off since I suck at doing fractions in my head.

 

[assume_R]

Flooring is most optimal. It keeps you from hurting your advantage gained by the index. Rounding a large TC up and hitting the index routinely will make your expectation for that index negative. That is because of the much larger frequency of the erroneous TC compared to the frequency of the TC at or above the TC index. This effect is nominal for negative indices because the error is made a tiny fraction of the number of times the index is used and at minimum bet. The important thing is not to overestimate a large TC. It is a very costly mistake when you use an index you otherwise would not have.

I'm a bit lost, tthree. When I generate index values (and betting ramps), I specify how I will do a TC conversion. So let's say I do a ceil. Then every TC from +3.01 to +4.0 will be considered an effective TC of 4.0. My index numbers and betting ramps (ev, freq, var) will reflect that, and give me the average correct decision for TC's between 3.01 and 4.0.

 

[tthree]

Whether you floor your estimate of cards played depends on your accuracy of your estimation ability. If you are very accurate stick with your estimate and floor the TC estimation. Flooring the estimate in the tray may be to a half deck if that is your accuracy. The idea is to use what corrects for your limitations. I am always worried about the accuracy of my estimate of the cards played so I start with an adjustment there. Depending upon how much that adjustment was determines my adjustment to the the TC estimate. The important thing is your final TC errs on the conservative side especially for high positive counts. I hope you understand the cost of over betting and using an index prematurely. If you feel strong about your discard tray estimation it is less complicated to simply floor your TC estimate.

 

[tthree]

I'm a bit lost, tthree. When I generate index values (and betting ramps), I specify how I will do a TC conversion. So let's say I do a ceil. Then every TC from +3.01 to +4.0 will be considered an effective TC of 4.0. My index numbers and betting ramps (ev, freq, var) will reflect that, and give me the average correct decision for TC's between 3.01 and 4.0.

MY ramp is designed for flooring TC estimates. Maybe advice about how to do TC conversions is not transferable to other peoples bet ramps and playing indices. The cost of an error is however. Using a large positive index before it is actually profitable on a routine basis can cost all the advantage gained by the index.

 

[assume_R]

MY ramp is designed for flooring TC estimates. Maybe advice about how to do TC conversions is not transferable to other peoples bet ramps and playing indices. The cost of an error is however. Using a large positive index before it is actually profitable on a routine basis can cost all the advantage gained by the index.

Indeed, but that can be true for any estimation method. I guess my point was only that as long as your index and betting ramps were designed for your estimation method, it will be accurate. You can't necessarily transfer index numbers and ramps from ceiling to flooring.

 

[Memphis10Tigers]

So I decided to create a quick chart of flooring to give me a ballpark idea of when my bets should ramp up. I highlighted all points where TC Floored >= 5 as this is where I am planning to make max bet and I will only be using the Illustrious 18 indices. I'm not going to worry too much about negative true counts as I plan to wong out at negative TCs.

Anyway, am I rounding everything correctly?

 

[Memphis10Tigers]

So I decided to create a quick chart of flooring to give me a ballpark idea of when my bets should ramp up. I highlighted all points where TC Floored >= 5 as this is where I am planning to make max bet and I will only be using the Illustrious 18 indices. I'm not going to worry too much about negative true counts as I plan to wong out at negative TCs.

Anyway, am I rounding everything correctly?

On a side note, reviewing the chart really put the importance of penetration into perspective for me. In a 6-deck shoe, you really won't be hitting that "sweet spot" very often unless you are deep into the shoe or have a really high RC.

 

[BJFan]

It should be wise to take a look to this link: (Dead link: http://www.qfit.com/blackjackblog/?p=199)

 

[tthree]

I agree except for high positive indices. These not only have your highest bets out but a higher frequency of hands lie just short of the index were it is still a negative EV than lie at or beyond the index. This effect is further amplified when doubling or splitting at a negative EV. Making sure you modify your play only after the index is exceeded is very important. If you can quantify your error factor make the adjustment so all your error slop has you only modifying your play when including the error factor has you above the index.

 

[blackjack avenger]

Some Consideration

I believe floored is considered the most accurate, mostly with -TC's.

It would be preferred to use the same TC method for indicies and bets.

That being said. There is no reason to get worked up over how accurate you are with TC conversions or which method you choose. There is not much value here especially if you use any betting camo. and bet conservatively. If you have a method learned it's probably best to stick with it.

Over vs Under using your indice?
Probably best to err on being over your indice, RA indicies show us this.

 

[zengrifter]

My questions would be which is optimal and which do you use? I would assume the floor would be the best as it is the most conservative for betting correlation purposes?

In actual play there is no real difference, so use what is easiest/fastest -
- rounded +/- one integer. See ZGI page 4-5. zg