Incidentally, it's probably buried in the documentation somewhere; but if you right-click on any of the charts in CVCX/CVData and click copy, you can paste the underlying data into Excel or CVSpread.
Incidentally, it's probably buried in the documentation somewhere; but if you right-click on any of the charts in CVCX/CVData and click copy, you can paste the underlying data into Excel or CVSpread.
"I don't think outside the box; I think of what I can do with the box." - Henri Matisse
There was a time I side counted by block 6,7,8 for shoe games but I found that side counting the ace in untraditional ways was much more powerful and a much better use of brain power. Like I said yesterday I wasn't advocating side counting shoe games, although I like Norm think it has merit, I was pointing out that defining the gain by how often it changes the decision is wrong. By side counting gain after the index is exceeded is much higher for the plays that would be affected. That same decision that wouldn't change now has more EV. This is because you don't know the deck composition but only the gain for the bin you are in and with a better correlated decision the gain is more. There is gain in every deviation you make and some you don't make. Whether or not the decision changes has nothing to do with the gain. Imagine using the ace-5 count for decisions. Think of the gain for side counting the T's. The T's don't need to be out of whack for the gain to be huge by comparison. The decision need not change. The gain comes from basing your decision on better information. That is what you get when you side count. Some decisions will be based on better information and gain EV much quicker after the index is exceeded. Whether the side counted card is out of whack has nothing to do with the difference in EV.Originally Posted by mofungoo
I am not sure how many get this. Now the gains may not be worth it for other reasons but not because the rank isn't out of whack enough. It is because extreme TC's are experienced less in shoe games. Extreme situation frequency determines the gain from playing efficiency. At more frequent extreme counts the gain from index plays in general is higher. The same is true for side counting. It is about the general frequency of extreme counts not the frequency the side counted cards is out of whack. Every matchup that is affected by the side count will have a higher EV but the effect is much greater if there is a higher frequency of exceeding the index by a lot. Which is why pitch games benefit more from side counting.
No argument there. It just bugs me when people make statements that indicate they don't understand how side counts increase EV. They don't change EV only when you have a deficit or surplus of the side counted card. They change the EV difference between the possible decisions for every decision that uses the side count information. The concentration of side counted cards has nothing to do with whether or not there is EV change. It is about increased correlation of the count used to make the decision to the matchups playing EoR's which changes the slopes of a line in the decision graph.
If one is to comment on side counting the comment shouldn't make it seem you don't understand this or make others believe something that isn't true. Too many trite expressions stated consistently indicate people don't understand the topic they are commenting on.
Tthree, it's some of you that don't get it. It's impossible. Some of you are at a level, liike Scientists with Ph'ds in nuclear physics, sitting here in this forum where a lot of high school juniors with their first Physics text are walking in, thrilled that they understand the first couple of chapters in an "Intro to Physics" book, and wanting to be scientists too. You are going to hear silly chatter that you just cannot believe.
then there are others ranging from undergraduate college students in Physics, some wth a bachelor degrees and some into their Master's degree trying to teach the high school juniors.
Don S., is like you but does not attempt to get in as many threads or attempt to get into the teaching.
Most all, the high school juniors, the teachers (I am probably like a first year undergrad helping the high school junior) who are not at your level, will never get there. Maybe 5% will be here 5 years later.
Just a side point. Obviously this is true because most playing decisions, by far, require no thought at all. You aren't going to double on 12 or higher. You aren't going to stand on 11 or lower. You aren't going to hit on hard 17 or higher. You aren't going to soft double against 789TA. You only surrender a handful of hands. The point is that you will also follow BS most of the time even with a TC of +10 or -10. But, that doesn't mean that indices have no use; any more than it means side counts have no use.
So, we should only look at hands where there exists a reasonable chance that a playing decision could be aided by information on seen cards. Now, suppose you are counting for betting, but using a counter's basic strategy. In this case, a large percentage of such hands do not follow BJ BS.
All I'm saying is that side counts can add additional hands to the group of hands where some extra effort can change improve the EV of the play. I personally don't think it's worth the effort. But, that's just my opinion.
"I don't think outside the box; I think of what I can do with the box." - Henri Matisse
Well, duh, of course the side counts offer an increased correlation on certain decisions. Why else would one do this sort of thing?
I do fully understand what you are writing about, but the goal isn't to make pretty lines on a graph. You seem to fail to understand that the graph doesn't correlate well to what makes money without also considering how often a profitable matchup occurs. That's why you say you have to run 100 billion hand sims. These events that cause sufficiently whacked decks occur rarely and it takes that many trials before an accurate difference can be discerned. Events that happen only rarely do not make a person rich.
Last edited by mofungoo; 07-02-2016 at 07:54 AM.
Actually the 100 billion round sims are for betting mostly because of my particular use of information. I know you get my point. I am trying to get others to understand how things work. You are right about the benefits in shoe games for playing decisions to be of modest value and worth deciding if the way you will use the extra info is worth the effort. I am not talking about that. It is a decision that each has to make for themselves. Most choose not to side count for playing decisions in shoe games and I no longer side count for playing decisions in shoe games. I do use my side count information for insurance but that is it except for betting decisions which is why I side count (ace neutral count).
The point is each decision is based on the average of all the situations that are equated into the same bin or number on the decision number line for all conventional counts. By getting the situations that are equated in the same bin (usually TC) and averaged to be much more similar (well correlated) in terms of whatever decision you are making you change the difference between the average for the strategy deviation for that bin and the other strategy option being considered. This difference between a playing deviation average for the bin and the next best option of standing or hitting etc is the EV for that decision at that time. IF the bin population changes the difference in EV changes even if the decision doesn't.
To understand what I am talking about the example of comparing Hiopt2/ASC to Hilo. They might make the same decision an awfully lot of the time but the EV gain from any individual decision made the same will be different. This is why Hiopt2/ASC can bet more with the same RoR. Most playing decisions and betting decisions are much better correlated to the decision. The result is most decision bell curves around the average for the bin for each individual deck composition that populates the bin doesn't range as far or as frequently away from the average.
Compare the betting decision bell curves for Hilo TC +1 (Post number #17) and Hiopt2/ASC equivalent TC (post #30) from an excellent thread made by one of our top posters who contributes on rare occasions (note the conditions in the combinatorial analysis are not what you will see often). The bell curves look identical until you look at the EV scale at the bottom. Then you see how poorly correlated Hilo is to making the decision compared to Hiopt2/ASC. The BC difference looks very small but the increase in betting accuracy is jaw dropping. The increase in correlation increases the EV from 0.2% fro Hilo to 0.4% for Hiopt2/ASC. That is not the big difference though. The big difference is how far the the individual deck compositions range from the average.
At 0.5:
The range of advantage for Hilo count is -0.5% to +1.1%.
The range of advantage for Hiopt2/ASC is +0.15% to +0.75%.
At 1.0:
The range of advantage for Hilo count is -0.20% to +1.10%.
The range of advantage for Hiopt2/ASC is +0.25% to +0.65%.
https://www.blackjacktheforum.com/sh...-at-tc-1/page3
Basically by increasing correlation of the decision you are getting a much more certain answer to your decision query rather than getting one that looks more like an educated guess. This results in a higher EV (based on the average EV of all equated deck compositions) and certainty (based on the range around the average) for each decision made. When you make a decision you are at one point somewhere in the bell curve. You have no ability to determine where beyond using all the info you have gathered. By changing the population of the bin that is averaged (equated by reducing them to the same number on a number line usually called the TC for most systems) to one that is better correlated to the decision you both increase EV and certainty as the above example demonstrates.
This says nothing about whether side counting is worth it. It just shows the nuts and bolts of how decisions are made. The increase in EV doesn't come from how many times you change your decision. It comes from the difference in average EV for each decision bin for the two methods. Whether a decision is changed or not has nothing to do with it because the bin populations have changed so the same decision has different EV gains because your EV is based on the average of all decisions that populate the bin you are in not the current deck composition because you don't know the current deck composition. The gain in EV for 1 tick higher than the decision on the decision number line increases with better correlation and the range of EV possibilities around that average becomes more tighter indicating the EV gain is more certain each time you make the decision.
The tighter you can make your decision bell curves (variance) for any decision and the higher the EV for the decision the more aggressively you can bet with the same RoR. So gain from better decisions isn't just in the EV of the decision but the higher your optimal bet which is proportional to EV/Var. Increase EV and you increase optimal bet for any advantage bin. Decrease variance and you increase optimal bet for any advantage bin. The accumulation of tweaks that do these things can have a significant impact on optimal bet ramp where individual tweaks may not. You are tasked with whether any individual tweak is worth the effort at the cost of other tweaks. Most feel side counting in a shoe game is not a good use of brain power except for ace side count.
Posting Permissions
-*-
Reply With Quote
Bookmarks