Adding AA78mTc side count to High Low

[Three]

Attached is my accuracy of true counts for HL vs KO at various true count points.

Perhaps you don't understand true counting. KO is a running count system. But to true count it, as you have been saying, you need to use deck estimates just like every other TC system or assume the number of sevens removed is equal to the number of sevens you would expect to be removed for the current number of decks remaining. Your chart shows KO error as though you aren't true counting it one way or the other, but some hybrid which can't be the case.

Perhaps the best way to illustrate my point is to look at UBTC of zero (equal to the imbalance per deck in an unbalanced system) for KO and RC of 0 for Hilo. Hilo RC of 0 has a TC of exactly 0, no math involved, and is always 100% accurate. A RC that would have exactly UBTC of -4 for KO (-4 times the number of decks unseen) is only accurate if the number of sevens removed is equal to the number of quarter decks seen.

So basically you have Hilo which allows the skill of deck estimation to determine accuracy versus KO which has how close the number of sevens removed are to what is expected determine accuracy. I think most would prefer their skill determine accuracy rather than random chance determine accuracy. Your chart makes no attempt to factor in where KO's inaccuracy comes from concerning decks remaining. You assume KO is 100% accurate at the UBTC pivot point. This is wrong. It is only correct if the number of sevens removed equals the number of sevens expected to be removed at that point in the shoe. So basically that chart is fallacious and worthless for making any point about accuracy.

Basically all your research is based on false assumptions. What you have is a place to start investigation based on a lot of assumptions that are known not to be accurate but are used as a starting point for investigation. You act like you are at the point where investigation is complete and conclusions can be based on all the assumptions that are known to be somewhat accurate approximations to start with. These assumptions are the starting point for investigation not the end of investigation. Since BJ is non-linear you need to sim to make any conclusion with any acceptable degree of accuracy. Tables, charts, and graphs based on approximations can't be relied on and are certainly not accurate.

[bjanalyst]

Some think KO + k1*(5m7c) + k2*(AA89mTc) is complex because you are multiply and adding two integers (k1*(5m7c) + k2*(AA89mTc) and adding to a third interger (KO). But for most situations either k1 = 0 or k2 = 0 or both k1 = 0 and k2 = 0 (which degenerates to just the KO count) and there are only a few situations where both k1 and k2 are non-zero.

Attached is a list of those situations where both k1 and k2 are non-zero. The most important are the six sitautiosin for no LS and the two LS situations so there are a total of only 8 important situations where both k1 and k2 are non-zero. So for the majority of stuations, you are just using one or no side counts with KO.
KO with non-zero k1 and k2.jpg

[bjanalyst]

You posted the "Accuracy of Hi-lo vs KO at various true counts." What about "Accuracy of Hi-OPT II vs KO at various true counts." That is what I am looking for. So you didn't post the information. If you did where is it??? Posting the betting efficiency chart for Hi-OPT II is not enough.

The problem with a direct comparion of true count with KO and HO2 is that the SD of the two are not the same. The SD of HL and KO are very similar so a comparison of accuracy with true counts can be made. If you hit all of the values of HO2 with a positive constant you still have an equivalent count since the CC is unchanged and the indices the HO2 indices are changed by that constant. So what I would have to do it hit HO2 with a positive constant to make the SD of both KO and HO2 the same and then I could compare true counts with that modified HO2. I will do that later when I get a chance.

[bjanalyst]

Perhaps you don't understand true counting. KO is a running count system..

Attached is the six-deck and eight-deck Table of Critical Running counts from my first book is KO with Table of Critical Running Counts which shows in detail how to calculate KO true counts. If you want more details you can spend $3.99 and buy the PDF online version of this book from www.xlibris.com
https://www.xlibris.com/Bookstore/Bo...arch=bjanalyst
Table of crticial runnig counts 6 decks.jpg
Table of crticial runnig counts 8 decks.jpg

[Three]

The problem with a direct comparion of true count with KO and HO2 is that the SD of the two are not the same.

SD is a measure of accuracy not CC.

So what I would have to do it hit HO2 with a positive constant to make the SD of both KO and HO2 the same and then I could compare true counts with that modified HO2. I will do that later when I get a chance.

So you have a measure of accuracy for Hiopt2/ASC and KO, namely SD. SD is THE measure of accuracy. But you think the way to compare accuracy is to multiply everything by a constant to make the SD the same and then compare CC which is a meaningless comparison to begin with? Try learning something in your thread instead of making the same mistakes you have the entire time as you go forward.

[bjanalyst]

SD is a measure of accuracy not CC.

So you have a measure of accuracy for Hiopt2/ASC and KO, namely SD. SD is THE measure of accuracy. But you think the way to compare accuracy is to multiply everything by a constant to make the SD the same and then compare CC which is a meaningless comparison to begin with? Try learning something in your thread instead of making the same mistakes you have the entire time as you go forward.

The point is that KO is exact at a its true count of 4 and is independent of decks played whereas HO2 true count at other than its pivot of zero is dependent on estimating decks played and the farther from is pivot of a true count of zero the more sensitive the true count calculation is to errors in estimating decks played.

I just had an email from Gronbog and he said he can simulate the exact KO count with its unbalance which is what players would use instead of using KO.bal with its fractional values. So your questions of accuracy of true counts around the KO pivot of a true count of 4 will be taken care of in the simulations. See attached file
KO system simulations.jpg

[Norm]

So, what you are saying is that KO is exact where it is exact. But, it is inexact where it's inexact, which is the majority of time. And, EoRs do not apply to KO.

[stopgambling]

So, what you are saying is that KO is exact where it is exact. But, it is inexact where it's inexact, which is the majority of time

This is hilarious . it is a "KO " !. Bj analyst did try very hard , I applaud his/her effort . I will continue to pay attention to this thread and hopefully learn something.

[seriousplayer]

This is hilarious . it is a "KO " !. Bj analyst did try very hard , I applaud his/her effort . I will continue to pay attention to this thread and hopefully learn something.

Try hard to do what? To look like a detestable person?

[BJGenius007]

Try hard to do what? To look like a detestable person?

Try very hard not to be a person who has to admit he is wrong.

[seriousplayer]

If you hit all of the values of HO2 with a positive constant you still have an equivalent count since the CC is unchanged and the indices the HO2 indices are changed by that constant. So what I would have to do it hit HO2 with a positive constant to make the SD of both KO and HO2 the same and then I could compare true counts with that modified HO2. I will do that later when I get a chance.

Hitting all of the values of HO2 with a positive constant is skewing. Why don't you compare the SD of both KO and Hi-OPT II as is?

[stopgambling]

Try hard to do what? To look like a detestable person?

i am quite fair and not take any side . he/she really want people to appreciate his effort but i don't have the know how to judge . All i see is a lot of work for not much gain in side counting in today's game. IMO , a true AP win with less effort and a bigger and sometimes much bigger advantage . Even i win more than him even he has more knowledge and mathematical foundation. A race car driver doesn't know as much as the engineer of the car but she /she can drive better. this person is stuck in the analytical part and doesn't seem to able to get out of there any time soon.

[bjanalyst]

So, what you are saying is that KO is exact where it is exact. But, it is inexact where it's inexact, which is the majority of time. And, EoRs do not apply to KO.

What I am talking about is accuracy. How much does an error in estimating decks remaining affect the accuracy of the true count. I am most interested in accuracy of the true count near the KO's pivot of a true count of 4. By exact I am talking about the KO true count at its pivot of a true count of 4. At the pivot, when KO = 4*n where n = number of decks then tc(KO) = 4 everywhere in the shoe and is totally independent of decks plays. Also I mentioned that the XmYc side counts are exact also. That is because unlike a side counts of Aces, Adef = Ap - 4*dp, an estimation of decks played is needed to calculate Adef. But dp is an estimate and if the dp is off so is Adef. But XmYc does not depend on decks played, So there is no estimation in XmYc and that is what I mean when I said that XmYc is exact.

Also the EoR do apply to KO because the KO can be balanced. The Excel file I sent to Gronbog that ETFAN reviewed used both LSL and PD (Proportional Deflection) and could calculate indices for any number of decks and any count, balanced or unbalanced. I calculated my indices for the HL with AA78mTc and 5m6c using the LSL and EoR. And every time I added more changes, the SCORE increased, which means that my calculation had to be correct.

And I also used EoR for the calculation of the values of k1 and k2 and the indices in KO + k1*(5m7c) + k2*(AA89mTc). Gronbog is going to do the simulations using the unbalanced KO so that the true counts near the KO's pivot of a true count of 4 will be very accurate and is adding in 5m7c and AA89mTc. So let's see what the simulations some up with. They should verify that my technique of using EoR is correct.

Finally, if EoR are so useless, then why has BJA3 listed the EoR to five significant figures? Five significant figures to not mean fudge factors.

So just be patient and wait for Gronbog's results.

I will repeat my prediction. The KO with 5m7c and AA89mTc will be the HO2 w ASC for the no LS game and will definitely beat HO2 with ASC for the LS game.

The KO with 5m7c and AA89mTc was always my choice of counts that I recommended (actually I switch from 45m79c to 5m7c because 45m79c kept with AA89mTc is too difficult so I gave up a bit of power for simplicity). I used HL with AA78mTc and 5m6c as a test of my system because modifying the HL was easier that the unbalanced KO. But again HL was NEVER my recommended count for the shoe game. For the DD game I said use HL with Am6c and 7m9c but for the shoe game use KO with 5m7c and AA89mTc.