I still disagree that HI-LO is the best overall level one count system. It is not the best! Anybody wants to prove me wrong they can.Originally Posted by mofungoo
With any data I have seen, Hilo is at the top of level 1 counts. This does not mean it is at the top in all circumstances. It means it is at the top under more circumstances than any other level 1 count. It beats the unbalanced level 1 systems under nearly all conditions. What happens when these systems are true counted I don't know. Neither, apparently, do you.
Well, you would be thinking wrong. Let's see what Don said about the $Won/100:
"Next - and we hope this will not be too problematic - we assumed, for the sake of easy comparison, that all games would be played at the rate of 100 hands played or seen per hour. The problem here is that, again, to be as realistic as possible, we put two players at the single deck games, three at double deck, and four at shoes. Well, it simply isn't true that the two-player games and the four player games will be played at the same speeds. Borrowing from the charts supplied by Stanford Wong, in Professional Blackjack (see, specifically, pp 234 and 237) we have attempted to suggest certain "factor adjustments" to the SCOREs provided in the charts. Next to each level of penetration is a number, such as 1.15 or 0.90, that represents our best estimate as to how the SCOREs might best be adjusted, by multiplying".
So that explains the lack of wong in numbers in the 2D charts. The 2D games weren't wonged. It also explains the ".95 factor" next to "62/104 H17 DAS" at the top of that chart.
Last edited by mofungoo; 02-17-2013 at 07:09 PM.
For mofungoo
You said "what happens when these systems are true counted I don't know. Neither, apparently, do you." I do know because some sims have been done in Modern Blackjack comparing some True Counted unbalanced count with HI-LO but somehow I believe that it is better to simulate unbalanced count fully with True Counts. In page 227 of Modern Blackjack Volume I second edition. It clearly states that the True count version of REKO does perform 7.5% better than HI-LO using the same number of index. Now tell is HI-LO still the best overall level one count system.
http://www.qfit.com/Book/z8406Chart.gif
http://www.qfit.com/Book/z8405Chart.gif
http://www.qfit.com/Book/z8392Chart.gif
Last edited by seriousplayer; 02-17-2013 at 07:38 PM.
The only chart you provided that compared Hilo to true counted REKO was for a 6 deck game, and even so the graph lines are nearly on top of each other. That hardly means it is the best overall level 1 count. The title of this thread is "Which counting method should I USE (2DECKS)" For those games Hilo seems to be the leader of the level 1 counts, except with a 1-4 spread against HiOpt I which has a SCORE 5 cents higher. If it's victory in an argument you seek, send me your address and I will mail you the nickel
I've only played DD once, and I wonged the living hell out of it for hours. Had a Uston lookalike at my table who seemed to believe in clumping, so he'd ask me to sit out until lots of small cards had come out and then happily fall back to one spot so I could eat his 'bad cards'. One of the more fun counting games I've played, except that I'm fairly sure by the end of the second hour that the elderly Vietnamese woman dealing to that table had me pegged.
That sort of disturbance is par for the course given the amount I've played, though. Except for my most recent play I always left thinking a dealer had made me and would dime me out. This has not been tested as I have yet to play the same store twice.
"Wait a minute. How do you beat someone to death with their own skull? That doesn't seem physically possible." "That's what Jimmy kept screaming: 'This doesn't seem physically possible!'"
I'm not saying Norm or Don is wrong, I am saying it appears you did something wrong in the sim you ran. Don's sim that he ran for that game was for play-all. Once again, here is what the info on the header means:
"Next - and we hope this will not be too problematic - we assumed, for the sake of easy comparison, that all games would be played at the rate of 100 hands played or seen per hour. The problem here is that, again, to be as realistic as possible, we put two players at the single deck games, three at double deck, and four at shoes. Well, it simply isn't true that the two-player games and the four player games will be played at the same speeds. Borrowing from the charts supplied by Stanford Wong, in Professional Blackjack (see, specifically, pp 234 and 237) we have attempted to suggest certain "factor adjustments" to the SCOREs provided in the charts. Next to each level of penetration is a number, such as 1.15 or 0.90, that represents our best estimate as to how the SCOREs might best be adjusted, by multiplying".
The SCOREs for wonging all say so, and a wong-in count is given. That is clearly not the case with the 62/104 H17 DAS SCORE chart. Check the settings on your simulator, the ramps may not be correct or something like that.
You are fortunate that I don't have CVDATA 5.0 since CVDATA 4.0 won't let me simulate unbalanced True Counts and nobody wants to assist me in doing this simulation. When I get CVDATA 5.0 I will continue on the dispute with you. Since I am confidence that HI-LO is not the best level one card counting system neither is the other are the level one counts.The only chart you provided that compared Hilo to true counted REKO was for a 6 deck game, and even so the graph lines are nearly on top of each other. That hardly means it is the best overall level 1 count. The title of this thread is "Which counting method should I USE (2DECKS)" For those games Hilo seems to be the leader of the level 1 counts, except with a 1-4 spread against HiOpt I which has a SCORE 5 cents higher. If it's victory in an argument you seek, send me your address and I will mail you the nickel
Last edited by seriousplayer; 02-18-2013 at 04:16 PM.
The original SCORE article has "$Won/100, 0.95 factor" heading the SCORE chart for 62/104 H17 DAS with no "observed" mentioned. There was probably a typo when the charts were put into BJA3. In any event I left a message for Don to see if he could clarify that point, although I think we know the answer already - I just ran a sim on Statistical Blackjack Analyzer and got a SCORE of 31.25 for Hilo with a 1-6 spread, 3 players. The SCORE article gives that game a SCORE of 30.36.
With CVCX Online I got a win/100 for that game of 33.08 using Hi-Lo Full Indexes. The player number is unknown. Perhaps the different simulators calculate differently or there are rounding errors. I still don't know where that 17.59 number you got came from, that's waaaay out of line.
The number of players affects the $/hour because the additional players slow the game down. The difference when results are given in "Win/100" is not as great because you still win approximately the same amount over the same number of hands, it just takes longer time-wise. That's why Don used the multiplication coefficient, to get a SCORE/hour number.
Last edited by mofungoo; 02-18-2013 at 04:46 PM.
Posting Permissions
Reply With Quote
Bookmarks