They don't. They make up crap just like the civilians do to make them "sound like" an expert. We've all hear the ploppies say, "I don't hit my twelves because I ALWAYS get a ten!"johndoe said:
Commercial simulators can't sim this. You need a custom simulator written for this way of using information. Basically you collect data into more groups. You would be looking at the success rate of doubles and splits in those subgroups and looking for a subgroup that does poorly. Now just dividing into smaller groups does nothing unless you the smaller group is a family of related deck compositions. If your groupings do nothing more than gather a random sampling of unrelated deck compositions you will find no additional information because the subgroup is representative of the parent group. But by thinking about how to make sub-groups that are populated by highly related deck compositions what happens to that sub-group might not be reflective of what happens in the parent group. When you do that you have harvested additional information that may have a use. You at least have more specific information that applies to the group and may be very different from the parent group average. There is a lot of ways to use this better assessment of related deck compositions. Some affect things in opposite ways than others. Like raising your bets because that sub-bin has higher stats than the parent bin average. I would recommend examining the effectiveness of individual doubles and splits and giving up the weak EV close to the index on the weak ones. Raising bets increase variance, especially for doubles and splits. You can give back some of that EV gain to be more patient on the doubles and splits that are weak. That hopefully makes variance less by giving up more of it as lost negative variance rather than lost positive variance. It is the same concept as risk averse index plays only you have EV gain from the increased bets to give back and still have increased EV instead of just giving up EV for reduced variance where variance is more likely to be negative. Doubles and splits are a big cause of variance. When EV is gained from correctly betting larger it will increase variance across the board from doubles and splits. But when you have little gain from doubling and splitting and a lot of negative variance as indicated by not much change in EV from just hitting or standing you can still have a net increase in EV by foregoing those risky weak doubles or splits and just doubling and splitting stronger doubles and splits.dogman_1234 said:
You seem to have an infinite capacity, somewhat like a person with Alzheimer's, to repeat the same thing over and over again and think that you are saying it for the first time. It's one of the strangest traits I've ever seen in an otherwise intelligent person. If I had to win a bet, I could take the time to find the above post, in one form or another, a HUNDRED times over the years. Do you think there's any possibility that you'll stop? Or do you think you can BEAT it out of us? That if you persist to do the same thing over and over again, you'll achieve a different result, and that suddenly, all those who read you will see the light? You understand that trying to do this is Einstein's definition of insanity.Dummy said:
I know that. See this is why I repeat things. People say I said something other than what I said. Obviously they didn't get what i was trying to say. How many times has that happened in these threads. The answer is as many times as I have repeated an answer in a slightly different way in a new post. I said variance is blind to whether the swing was positive or negative. I also said that variance can't tell how swings tend to clump. A 100 trials that have 50 straight wins and 50 straight losses with all being the same data points in a different order have the same variance as if they alternated wins and losses. The variance would be exactly the same but swings would go from extreme to minimal. Your comments indicate you agree with both but said you disagree with what I said. I was pretty careful to refer to short term as swings and global stats as variance. I kept referring to lessening downswings. Variance is irrespective of sign. When EV drops slightly from betting less where doubles and splits perform worse, that decreases negative variance for the portion of play that you make this adjustment but increases positive variance slightly due to the average EV being slightly less making all the differences between other positive variance results more. The net change in variance is minimal but the short term change is not. Variance might go up slightly or down slightly. Remember I explained variance can't tell what the direction a swing is. Variance only makes measurements from the average which now has changed. Less of a difference for results below EV and more of a difference for results above EV. All the addends in the variance equation alter slightly. Some of addend the squares are for bigger differences and some smaller differences. The biggest affect on variance is the big bet doubles and splits. I was describing would to make a positive impact in the way of downswings that didn't affect variance in a significant way. I think you took what I said too literally. But I believe the common statement I made was variance is mostly unaffected or something like that. When you have a move that is 5 significant digits down from the most significant digit most people call that unchanged even though it has in fact changed an insignificant amount.DSchles said:
This is exactly what I have been pointing out. people keep demanding stats to show it makes more money but I keep saying it makes less money. Most of what people like to call super-duper systems don't win that much more money and may not win any more money. You want two things from your system. Winnings and being able to get away with it. I keep saying they aren't going to make you a significant amount of extra money, but it will be quite useful for helping you get away with playing in casino. That is the most important battle to win unless you are a slash and burn player. Then that doesn't matter at all to you. Bjo is right that you shouldn't care. That is why I think it is so funny that people are caring and making all these comments elevating what I said to making a lot more money. I said you have to give up EV to get what I consider to be most valuable, smoother results and more longevity. If you put little or no value on that then what are you arguing about. Smoother results means less of a BR requirement but again that difference here is fairly insignificant, like 4% less. At least according to the sims.bjo32 said:
No. Dummy said it can reduce downswings, which can reduce the frequency and severity of losing sessions and increase the frequency of winning sessions which steadies BR fluctuations. There is no significant change in variance. Based on what I see there are some deck compositions whose contributes to variance by poor performance on doubles and splits and others in the same TC bin that contribute to variance by strong performance on doubles and splits. Variance can't distinguish between variance from more positive versus more negative variance contributions. The difference between the average EV and results get squared so they are both positive when they get added to variance. They both contribute about equally to variance if they are bet and played the same. Hand frequency for split and doubled hands is what separates them in their contribution to variance. The key to grouping these related deck compositions that have worse performance together is picking their common element missed by the main count and separating them from other deck compositions and treating them differently. Griffin described moving them to another bin and treating them as the average of all those unrelated deck compositions are treated. That distinction is what allows improvement. The initial problem is this family of deck compositions aren't reflective of the bin they are in. Moving them to another bin isn't going to change that. They will still be in a bin populated by much more frequent and very different deck compositions and treated as the average of all those deck compositions suggested in the sim that generated the data. But the other key is when you know doubles and splits perform poorly for this density of neutral cards you don't bet them optimally or make doubles and splits optimally. That latter is what caused the problem to begin with. The density of neutral cards made some doubles and splits negative EV plays and others having too little added EV to make the double worth the added risk. Altering playing strategy for the sub-bins that display this while at the same time betting less trade EV for making these downswings less severe from both bet size and playing strategy. Both decrease EV a little. If you ever had a big bet situation where the dealer kicked your ass on all your doubles and splits there is a chance that losing all those big bets was due to playing through one of these poor performance deck compositions rather than just bad luck. If so betting less and doubling less by this criterion would have reduced the damage significantly.dogman_1234 said:
I am not sure you wrote what you meant. But acumen is needed for the traditional approach to using the extra information gathered. memorization is required for doing what I described. There are no computations at the table other than keeping the count. You just make the play or bet that is appropriate for the bin you are in. You have memorized that before you make a bet. Now if you mean acumen for keeping the count. I think column counts are beyond me but believe I can do them with enough practice if I wanted to. And Moss and Tarzan think my two count with significant count overlap method is beyond them. But I am confident they could do it with enough practice even though they aren't. I can keep two side counts and the main count but adding another side count, as the column counts do, seems harder by an order of magnitude to me.dogman_1234 said:
Slightly ironic considering the first thing you ever said to me was "I'm 10 times the counter you are."ZenKinG said:
Au contrair, mon ami...DSchles said:
See ZGI page 4-6dogman_1234 said:
seedogman_1234 said:
