Adding AA78mTc side count to High Low

[Three]

Previous posts stated that specialty software is needed for these simulations so your software will not work.

He never specified if he had software that could do it. He may have such software. You just can't buy software that can do it as far as I know. People often write their own software or modify purchased software.

Finally I did not ask you specifically do the sims. My request was general to ANYONE interested who wanted to run simulations. If you do not want to do sims that is fine. I am not twisting anyone's arm.

I hope the software that runs my system will have this capability some day as I use multiple combined playing counts. If it doesn't it isn't any big deal. I am not trying to sell anything or convince anyone of anything. I just try to inspire people to think outside the box and use my private approach to make money in casinos. I can just base my play on the strongest version the software can do and know I will outperform it. Since I am not selling anything that is good enough for me. If I was selling something I would do whatever it took to get the sims run. It isn't like it can't be simmed, it is just that custom software is required. I am not sure how you can say you have no access to sim software and your clearly wrong assumptions that you say you have proven and expect anyone to have confidence in what you are saying.

I know your approach is an improvement. It will get you a lot back for the insurance bet. But my sense is going to full indices rather than the I18 will get you more than the other indices you site, besides insurance, will get you. They just affect your smallest bets that you should probably have already wonged out to avoid having to make. By just affecting small bets in general it doesn't have much affect in terms of risk aversion or increasing certainty of BR growth. The approach has the potential to affect larger bets with different counts chosen. You targeted the strongest index play, insurance, and did a great job there, but what came along with it isn't worth much. The side count allows you to adjust the weight of the T and A which would set the neutral card weights. But the possibilities don't mesh well were an important impact can be made except for insurance, which you targeted.

You worked the problem backwards. Instead of starting with an arbitrary count and working towards a strong combined count for insurance, start with a great insurance count and work backwards to have pieces that add together with some multiplier and leave you with the strongest main count while factoring in what plays the side count choice will allow to come with insurance (I recommend the balanced version of a great insurance count, there are a few to choose from. A level 5 insurance count limits the possibilities). You ultimately want the strongest betting count possible, a near perfect insurance that brings important plays as different multiples of one count added to the other, and the ability to make several playing counts with a diversity of uses. The betting count can be a combined count as well. You don't have to use the multipliers just on the side count, you can also use them on the main count or both. The 9 is usually not counted as a high card because it messes up insurance and the plays on your 12 totals, which are among the easiest index values to reach. But the 9 can be counted as a high card in one count and a low card in the other count so it can be weighted either way or as a neutral card in one of the combined counts. This way of attacking the problem would reap the highest reward from your approach and allow you to see many more possibilities for choosing two counts to be used.

[Three]

My weighted average CC was for S17, DAS, LS game not H17. Also I only included around 25 playing strategy situations for someone back counting, not play all. Yours was play all. And Hi Opt 2 weighted average CC was somewhat larger than UAPC so HO2 should be slightly stronger as it was. Finally the results were still close as were the weighted average CC.

So what you are saying is you made an observation that worked out by coincidence. Then you made the errant assumption it would have a general application.

Using AA78mTc with HL increases HL insurance and hard 12 v 2, 3, 4, 5, 6 hit/stand decisions by 20% to 30%. This will make a huge difference that should show up in simulations.

Insurance, 12v2, and to a lesser degree 12v3 will apply to larger bets often and be worth a good amount, but the others will not be worth much and rarely used by a backcounter unless the deck is really out of whack. Out of whack deck compositions is where our opportunities are found, but the only way to know is to sim.

[bjanalyst]

So what you are saying is you made an observation that worked out by coincidence. Then you made the errant assumption it would have a general application. Insurance, 12v2, and to a lesser degree 12v3 will apply to larger bets often and be worth a good amount, but the others will not be worth much and rarely used by a backcounter unless the deck is really out of whack. Out of whack deck compositions is where our opportunities are found, but the only way to know is to sim.

I would like to explain hard 12 v 2, 3, 4, 5, 6 decisions with HL and AA78mTc a little more thoroughly here to clear up misunderstandings such as yours where
I will show that your statement that “Insurance,12v2, and to a lesser degree 12v3 will apply to larger bets often and be wortha good amount

but the others will not be worth much” is not true. Plus hard 12 v 2, 3, 4, 5 and 6 are part of theIllustrious 18. So all hard 12 v 2, 3, 4, 5, 6 decision are important and the HL CC is increased 20% to 30% using AA78mTc for these situations.

I use KO with AA89mTc but if sims were done it is easier to use all balanced counts and so I am using HL with AA78mTc which is similar to KO with AA89mTc and will prove my point. So for simplicity I will analyze HL with AA78mTc here.

I showed earlier that SD(AA78mTc) = SD(AA78mTc) andCC(AA78mTc, HL) = 20%. So for the 5 outof 6 deck dealt HL can range from -30 to +30 so that means AA78mTc can alsorange from -30 to +30. I have been using KO with AA89mTc for over 4 years and Ihave had occasion where AA89mTc > 20 and AA89mTc < -20 so it does happen. SD(AA89mTc) = SD(AA78mTc) = SD(HL).

Since CC(HL, AA78mTC) is only 20% then AA78mTc can have verynegative values and HL have very positive values. That means that you can have high HL countand so a large bet out and if AA78mTc is significantly negative you would behitting hard 12 v 6 for example.
Please look at the following chart

01 HL with AA78mTc chart.jpg

So for hard 12 v 6 it can be seen that HL CC is 65.5% but HL+ 1.5*(AA78mTc) has a CC of 96.8% with an index of -2 (-1.7 rounded down to -2)and AACpTCp = 1.0%. So I will explainbelow with an example how this may come into play to give a hit decision whenHL count is large and a large bet is out – something you would never do if youwere just keeping the HL count and the HL count was large. Also, this is another example of casino camouflageplay where if the pit is tracking your action with the HL count and they seethe HL is large and you hit your hard 12 v 6 they will you are an idiot and mayleave you alone when hitting the hard 12 v 6 may actually be the correct play . Previously I gave other examples of camouflage play if AA78mTc is usedwith the HL which you may want to refer to.
So here is an example. Suppose n = 6 decks and dp = decks played = 4 and so dr = decks remaining= 2 and further suppose HL = 8 and AA78mTc = -10. Remember SD(AA78mTc) = SD(HL) and AA78mTc is highlyuncorrelated with HL so AA78mTc can easily be -10 when HL = 8.
Now with HL = 8 and dr = 2 then tc(HL) = 8/2 = 4 so a largebet should be out and in the absence of AA78mTc you would stand on hard 12 v 6 whichis a basic strategy play also. SO if thepit sees you hit hard 12 v 6 when tc(HL) = 4 and your large bet is out theywill think you are a real fool and may leave you alone since they will thinkyou do not even know basic strategy.

Now I will show that hitting hard 12 v 6 is actually the correct play in the situation and I will show the gain you get in this situationby hitting instead of standing.

Let psrc = playing strategy running court. For hard 12 v 6 psrc = HL +1.5*(AA78mTc). Since HL = 8 and AA78mTc= -10 then psrc = 8 + 1.5*(-10) = -7.
The index is -2 so stand on hard 12 v 6 is psrc >=(-2)*dr = -4 since dr = 2.
But psrc = -7 < (-2)*dr = -4 so player should hit hard 12v 6 in this situatiiton
Now tc(psrc) = psrc / dr = (-7) / 2 = -3.5 and AACpTCp =1.0%.
If pa(t) = player’s advantage at true count “t” then pa = (t– Idx)*(AACpTCp)
Here t = tc(psrc) = -3.5 and Idx = -2 So pa of standing over hitting is (-3.5 – 2)*(1.0%)= -1.5%. So if player stands instead of hitting his hard 12 v 6 in thissituation he actually is 1.5% worse off than if he hit his hard 12 v 6. Or alternatively, if the player hits his hard12 v 6 he is 1.5% better off than if his stood on his hard 12 v 6.

[lij45o6]

I would like to explain hard 12 v 2, 3, 4, 5, 6 decisionswith HL and AA78mTc a little more thoroughly here to clear up misunderstandingsuch as yours where .
I will show that you statement that “Insurance,12v2, and to a lesser degree 12v3 will apply to larger bets often and be wortha good amount” it not true and if situations were done it would prove thatstatement was not true. Plus hard 12 v 2, 3, 4, 5 and 6 are part of theIllustrious 18.

I use KO with AA89mTc but if sims were done it is easier to makeit to use all balanced counts and do it with HL with AA78mTc which is similarto KO with AA89mTc. So for simplicity Iwill analyze HL with AA78mTc here.

I showed earlier that SD(AA78mTc) = SD(AA78mTc) andCC(AA78mTc, HL) = 20%. So for the 5 outof 6 deck dealt HL can range from -30 to +30 so that means AA78mTc can alsorange from -30 to +30. I have been using KO with AA89mTc for over 4 years and Ihave had occasion where AA89mTc > 20 and AA89mTc < -20 so it does happen. SD(AA89mTc) = SD(AA78mTc) = SD(HL).

Since CC(HL, AA78mTC) is only 20% then AA78mTc can have verynegative values and HL have very positive values. That means that you can have high HL countand so a large bet out and if AA78mTc is significantly negative you would behitting hard 12 v 6 for example.
Please look at the following chart

01 HL with AA78mTc chart.jpg

So for hard 12 v 6 it can be seen that HL CC is 65.5% but HL+ 1.5*(AA78mTc) has a CC of 96.8% with an index of -2 (-1.7 rounded down to -2)and AACpTCp = 1.0%. So I will explainbelow with an example how this may come into play to give a hit decision whenHL count is large and a large bet is out – something you would never do if youwere just keeping the HL count and the HL count was large. Also, this is another example of casino camouflageplay where if the pit is tracking your action with the HL count and they seethe HL is large and you hit your hard 12 v 6 they will you are an idiot and mayleave you alone when hitting the hard 12 v 6 may actually be the correct play . Previously I gave other examples of camouflage play if AA78mTc is usedwith the HL which you may want to refer to.
So here is an example. Suppose n = 6 decks and dp = decks played = 4 and so dr = decks remaining= 2 and further suppose HL = 8 and AA78mTc = -10. Remember SD(AA78mTc) = SD(HL) and AA78mTc is highlyuncorrelated with HL so AA78mTc can easily be -10 when HL = 8.
Now with HL = 8 and dr = 2 then tc(HL) = 8/2 = 4 so a largebet should be out and in the absence of AA78mTc you would stand on hard 12 v 6 whichis a basic strategy play also. SO if thepit sees you hit hard 12 v 6 when tc(HL) = 4 and your large bet is out theywill think you are a real fool and may leave you alone since they will thinkyou do not even know basic strategy.

Now I will show that hitting hard 12 v 6 is actually the correct play in the situation and I will show the gain you get in this situationby hitting instead of standing.

Let psrc = playing strategy running court. For hard 12 v 6 psrc = HL +1.5*(AA78mTc). Since HL = 8 and AA78mTc= -10 then psrc = 8 + 1.5*(-10) = -7.
The index is -2 so stand on hard 12 v 6 is psrc >=(-2)*dr = -4 since dr = 2.
But psrc = -7 < (-2)*dr = -4 so player should hit hard 12v 6 in this situatiiton
Now tc(psrc) = psrc / dr = (-7) / 2 = -3.5 and AACpTCp =1.0%.
If pa(t) = player’s advantage at true count “t” then pa = (t– Idx)*(AACpTCp)
Here t = tc(psrc) = -3.5 and Idx = -2 So pa of standing over hitting is (-3.5 – 2)*(1.0%)= -1.5%. So if player stands instead of hitting his hard 12 v 6 in thissituation he actually is 1.5% worse off than if he hit his hard 12 v 6. Or alternatively, if the player hits his hard12 v 6 he is 1.5% better off than if his stood on his hard 12 v 6.

Are you assuming that we don't understand multi-parameter thinking? That is not what we are saying.

Let me get this to you: The issue is that we don't understand. We understand what you are doing. What we are saying is that you explaining how your system is better than straight-up High Low is far-fetched, based on the fact that you don't have any sim data.

You must be trolling us.

Your mission is futile if you think we are going to blindly accept your system when everyone here has possession of winning systems that have been simmed and verified by many people. Consider rethinking your approach to this board.

P.S. Consider writing posts with brevity in mind. You go off way to far...even under Three's standards!

[refinery]

For whatever the flaws, bjanalyst is no troll.

[Tarzan]

At a glance, I can tell you are missing a few pieces of the puzzle for it to be possible for this to be on par with HiOpt2ASC, full indices. I believe it falls short, with a lot of work to do it. I'd have to look at it closer. I'm questioning ease of practical application for what you get out of it. Adding some things to Hi-Lo or KO will improve your performance in the very long run if you put in a lot of hours at the tables, but it's quite a stretch to say it's on par with, or make it on par with HiOpt2ASC, and outperforming HiOpt2ASC requires composition dependent perfect play, which is not what you've got there.

[bjanalyst]

You understand multiparameter thinking but apparently not everyone does as did the person whose post I replied to who said basically (I am paraphrasing) help with hard 12 v 2 was important, maybe hard 12 v 3 but the others were not important. My reply was directed to that statement.

Everyone has made it very clear that they want simulations of HL with AA78mTc. I have not disagreed with that. My purpose in the above post was to clearly a misunderstanding.

If I had a canned HL program that was written with modules and easy to follow that I could easily modify, then I would modify that program to include AA78mTc. But it makes no sense for me to try to create an entire simulation program from scratch. Why reinvent to the wheel plus I could make a lot of mistakes? I would rather modify a proven HL program with easy to follow source code.

So at this point, unless I find such a canned HL program. I am at a dead end. And if no one wants to run simulations then the HL with AA78mTc will be left without simulation proof.

Finally if HL with AA78mTc is to be verified through simulations, it is best that it be independently verified. I should not do my own verification.

[bjanalyst]

At a glance, I can tell you are missing a few pieces of the puzzle for it to be possible for this to be on par with HiOpt2ASC, full indices. I believe it falls short, with a lot of work to do it. I'd have to look at it closer. I'm questioning ease of practical application for what you get out of it also. Adding some things to Hi-Lo or KO will improve your performance in the very long run if you put in a lot of hours at the tables, but it's quite a stretch to say it's on par with, or make it on par with HiOpt2ASC, and outperforming HiOpt2ASC requires composition dependent perfect play, which is not what you've got there.

I should have concentrated on saying adding AA78mTc to the HL helps the HL count. I added in that I believe it would help so much that it would outperform the HO2 with ASC.

So exactly how much AA78mTc improve the HL, without simulations, is not known.

Now I use KO with AA89mTc but concentrated on analyzing HL with AA78mTc for simplicity of using balanced counts in my analysis.

But I play mainly because of the LL bet. I have shown that KO + AA89mTc is a perfect Ten count which is what you need for the LL bet. I am making money mainly because of the LL bet which is what I mainly use AA89mTc with the KO for.

Again, I realize everyone wants simulation results and if I could find a canned HL program with source code that I easily follow then I would put a stab at modifying it for simulating HL with AA78mTc.

[Three]

You have had done tell you that you are 100% wrong. I have been using a more elaborate version of the same technique and researching it for a long time and I tell you how you can do better, because you worked the problem backwards. And why the coefficients and individual stats for plays don't mean anything until bet size is factored in. Then I show you why some or your stuff is not worth much, because they will only affect small bets that you probably won't make because you have already wonged out. Meaning the high frequency that gets the matchup into the I18 for play-all play will have you rarely playing the matchup because you are wonging. Then you say your calculations are for back counting which means you will not be playing when you are at the index for most of the matchups I said were not worth much due to bet size. It is like you can't understand what anyone is telling you. You go back to the same stats that we were already taking into account and acting like 20% to 30% improvement on almost nothing suddenly becomes a lot. You do realize that the 12 total index plays mostly make the I18 due to frequency of use. The ones that are negative indices rely on you playing those negative counts to get the frequency that gets them into the I18. Some other indices make it more on bet size or EV gain than frequency of occurrence, like splitting TT, insurance, and some others. When you say you are back counting that means you miss all the occurrences that would be be BS deviations without your side count and only get to use the index when your side count moves the count enough that you wouldn't have wonged out but you still make the deviation. That is a fairly rare occurrence.

To understand the gain for a counter that something is worth you must factor in frequency of occurrence, bet size and gain increase for the matchup after the index is exceeded. The flat bet gain is meaningless. You are talking about losing almost all the frequency to back-counting or wonging for the negative indices, the bet size will almost always be your min bet, just the gain after the index is exceeded is left to boost up some of what is lost to lower frequency from wonging. If we played-all and flat bet your argument would be valid. But counters don't do both of them at once. The may play-all, or the may back-count and flat bet, but they never do both. That is why sim data is necessary. It allows bets size, frequency of occurrence, and rate of gain after the index is exceeded to be factored in to see the true gain. Run separate sims that forces a bet of 0 except for that match-up for your wonging style for Hilo and Hilo/(your side count), in which case it makes the bet your ramp would have made. The difference in the EVs for each sim would give the gain for each matchup. Run a sim pair for each matchup. You will see what we are talking about. You will probably use the insurance index more than the worst negative index if you are wonging like most shoe players do.

[Gronbog]

So exactly how much AA78mTc improve the HL, without simulations, is not known.

bjanalyst, I can modify my software to simulate your system. You're not a paying member so i can't PM you. Please post an email address where I can reach you to discuss the details.

[refinery]

bjanalyst, I can modify my software to simulate your system. You're not a paying member so i can't PM you. Please post an email address where I can reach you to discuss the details.

Out of pure curiosity, is it because of the work you did with Tarzan that you are able to sim bjanalyst's method? Or did your software already have this capability?

[Gronbog]

The design of the card counting interface in my software allows me to write code to simulate arbitrarily complex counting systems fairly easily.

[seriousplayer]

Out of pure curiosity, is it because of the work you did with Tarzan that you are able to sim bjanalyst's method? Or did your software already have this capability?

Don't forget that Gronbog is a professional programmer, too. So writing a program to do complex simulation shouldn't be a problem for him.