Originally Posted by
DSchles
Pp. 375-6 of BJA3 show one study comparing the gains from R-A indices vs. e.v.-maximizing ones. The difference is about 2-2.5%. It's possible that any outperformance of the new count compared to Hi-Opt II is due solely to the mismatch between using R-A indices for the former and e.v.-maximizing ones for the latter.
In any event, just as for yet another count that we have just finished studying with Gronbog at the helm, comparison to Hi-Opt II is extremely close. So, the question then becomes not if it's possible to create a count that rivals Hi-Opt II (it clearly is) but rather, given the relative simplicity of using Hi-Opt II, whether it's worth trying to learn either of these other two counts, both of which clearly require more effort to learn.
Don
First I want to thank Gronbog for doing an EXCELLENT job in doing the sim of my KO system. Gronbog is a true professional and a gentleman.
I explained in a previous post the RA indices are used were judgmental based on CC and AACpTCp. Players know not to double or split exactly at the index because the EV is the same and there is only increased risk. How much gain in SCORE was obtained by using my judgmental RA indices you say is 2 to 2.5%. I have no way of quantifying so I have to take your word on that.
But I would like to mention a few points. I have five criteria I used to choose a count system.
You are concentrating on what I call "Power" only and are ingoring the other important points in choosing a count system which I mentioned before but which I will outline again below.
Also Gronbog only did no LS. As I explained earlier, when LS is taken into account, HO2 w ASC will be left in the dust. First the difference in BC using LS EoR of KO + (1/2)*(5m7c) and the HO2 - 2*(Adef) increases form 0.67% to 1.06% when LS EoR are used. Secondly, every single LS CC of my KO system beats the HO2 w ASC. So with LS you can use HO2 with ASC for toilet paper. You will see what I said is true when Gronbog does the LS simulations just as my predictions were true of my KO with no LS and my prediction came true when I said what would happen when HL + (1/3)*(5m6c) was used for betting with HL and AA78mTc and 5m6c system. All of my predictions came true based on comparing BC and on comparing CC of individual plays against the respective HO2 CC for those same plays.
Finally my 2nd book was called KO with 45m79c and my third book was called KO with 45m79c and AA89mTc. I replaced 45m79c with 5m7c for simplicity and ease of use because keeping both 45m79c and AA89mTc are difficulit to do and the player could get exhauseted and cause errors and not be able to play as long. But if 45m79c were used with AA89mTc the for LS game the BC would be 99.6% (with 5m7c BC is 99.0%) and the PE would also increase over using 5m7c. But if you want to increase further I would keep 5m7c and add a third side count such as Am6c, Am8c or 6m2c. Now adding a 3rd side count does make the system more difficult so I will not address a third side count here other than to say that it an option to further increase power.
Power was only one of the five points I had in chosing a system. Power is theoretical assuming you can keep the count accurately and estimate dp and dr and true counts accurately. If you cannot keep them accurately then you will never achieve the theoretical power of the system. As I will explain below I think that the level 2 HO2 with ASC is extremely difficult to keep and also there are huge inaccuracies in calculating true counts and Adef in using HO2 with ASC as both are subject to errors in estimating decks played. Remember XmYc is independent of dp and is EXACT whereas Adef is an ESTIMATE because decks played is an APPROXIMATION. And I showed earlier, HO2 true count calculations when big bets are out are much more inaccurate the my KO system true count calculations.
The problem is people are resistant to change. They are used to the HO2 with ASC which they have been using for years and do not want to change. The look at only POWER and ignore the other four points I have mentioned in choosing a system. And then even when I show my KO system outperforms HO2 w ASC they still try to find a way to knock and degrade my system saying it is not worth 99 cents.
So now I would like to once again touch on the five points I made in choosing a count system and show how my KO system compares with the HO2 system in each of these five points of which POWER is only one of the five points.
(1) Simplicity of use
To me adding and multiplying two small integers (k1*5m7c and k2*AA89mTc) and adding to a third small integer, KO is easy. Also many times either k1 = 0 or k2 = 0 or both k1 and k2 are zero (degenerating to the KO count). There are only a few instances were both k1 and k2 are non-zero. Also I had simplified formulas when crc is outside of the table of critical running counts. For example, stand on hard 15 v 7, 8, 9 and hard 16 v 7, 8, 9 if tc(KO) = 4 (KO = crc(4) = 4*n, n = # of decks) and tc(5m7c) >= 4, 3, 2, 2, 1.5, 1 respectively. So with simplified formulas and the fact that there are not many instances that both k1 and k2 are non-zero the system is simple to use.
Also the KO is a level one count, 5m7c is a level one count and counts only two cards and AA89mTc, although technically a level 2 counts, is really easy to keep since only the Ace is counted as +2 and there are only the Ace, 8's, 9's and Tens being counted. You wait until the cards are on the table and then do a bunch of cancellations with an Ace cancelling two Tens and an 8 or 9 cancelling a Ten. That is not very difficult to do at all.
To me keeping the HO2 with ASC is extremely difficult to do which is why many blackjack teams will not use the HO2 w ASC. The HO2 counts the 4's. 5's as +2 and the Tens as -2 and the 2, 3, 6, 7 as +1 - that is a lot of cards as +2, -2 and +1 and counting all of these at once is very difficult as far a I am concerned. Then there is the Ace side count where you have to keep track of an ever increasing value of Aces played and then you need to estimate decks played to calculate Adef = Ap - 4*dp which is an ESTIMATION and it is also very difficult to do. Both HO2 and Adef are difficult to keep and are subject to errors as you need an accurate estimate of decks played to get their true counts. So I will explain this more in my other four points in choosing a system.
(2) Power
This is what simulations show and what BC and CC show. My simulation show that my KO system outperforms HO2 with ASC for the back counted game which is all you should be playing anyhow. So my KO is theoretically more powerful for the no LS game and as I said above, when LS is included he difference between my KO system and the HO2 w ASC will be even greater.
(3) Accuracy
I have touched on this many times before. For true counts > 3, the KO true counts are at least twice as accurate as HO2 true counts and that is where you want accuracy in true count calculations, when large bets are made. The KO system is not very sensitive at all to errors in estimating decks played for true counts > 3 which is a big plus. Player may thing they are estimating decks played to the nearest half deck but they probably are estimating only to the nearest full deck. Deck estimation is difficult. But with KO estimation to the nearest dr is more than adequate when tc > 3 and as I said above the side counts are XmYc side counts and unlike Adef which depends on an accurate estimate of dp, XmYc are totally independent of dp and are EXACT.
So in addition tot the HO2 with ASC being much more difficult to keep than that KO with 5m7c and AA89mTc, it is also less powerful and more subject to errors in true count calculations when it is in actual use in the casino and so the theoretical gain of the HO2 w ASC is never fully achieved.
(4) Camouflage
I also mentioned camouflage many time before. Casinos mainly look for bet spreads and splitting Tens as a give away to counters. But casinos also look at insurance plays and would notice strange plays like hitting hard 12 v 6 when you have a large bet out and the HL count (which the casinos are probably using) is high. I had mentioned both of these before but I will repeat again here. I will mention two situations below which I mentioned before but there are more camouflage plays such as standing on hard 15 v 7, 8, 9 which may draw the casino's attention when a large bet is out.
Consider Insurance if KO + AA89mTc >= crc(4) = 4*n where n = number of decks is 100% accurate insurance decision and is totally independent of decks played since the index is at the KO pivot of a true count of 4 and to the insurance decision is EXACT. If tc(KO) = 6 and you have a maximum bet out and AA89mTc < (-2)*dr then you would NOT insure. The casino would see you with a large bet out and not taking insurance and they would consider that you do not know what you are doing. If you have a good hand (hard 11, 20 of blackjack) and if crc(3) < KO + AA89mtC < crc(4) then you should under insure to reduce risk. Note that the HO2 + Adef has an insurance efficiency of only 92.8% and would never predict this kind of play at not taking insurance when the count is high.
Consider hitting hard 12 v 6 at tc(KO) = 4. Standing on hard 12 v 6 is, of course, a basic strategy play, and then the tc(HL) is positive it is, in the absence of any additional information, even more important that they payer stand on hard 12 v 6. So if the casino sees you hitting hard 12 v 6 when tc(HL) = 4 for example and you have a big bet out, they will think that you do not even know basic strategy and that you do not know how to play. Using my OK system you would hit hard 12 v 6 when KO + 1.5*(AA89mTc) < crc(-1) or using the simplified formula hit hard 12 v 6 if AA89mTc < (2/3)*(1 - t)*dr where t = tc(KO). So if tc(KO) = 2 hit hard 12 v 6 if AA89mTc < (-2)*dr and if tc(KO) = 4 then hit hard 12 v 6 if AA89mTc < (-3.3)*dr. So you have a big bet out and you hit hard 12 v 6 which is a violation of basic strategy. The casino is watching you when yo have a big bet out and if they see this play it may help you play longer as they will label you as an idiot and playing a hunch. Note that the CC of standing on hard 12 v 6 when KO + 1.5*(AA89mTc) >= crc(-1) is 99.8% whereas the CC of standing on hard 12 v 6 when HO2 + 2*(Adef) >= (-1.6)*dr is only 91.5% so this camouflage play would never come up with the HO2 system.
(5) Side Bets
The AA89mTc and 5m7c may come in handy for certain side bets. As I explained many times before, for the Lucky Ladies bet use LLc = KO + AA89mTc. If Blazing 7's side bet is offered, 5m7c can be used for that bet. The HO2 w ASC offers no additional help for either or these side bets.
So just wait until Gronbog comes back from vacation and you will see that when LS is added, my KO kicks the HO2 w ASC ass.
Bookmarks