> That aside, index precision just isn't that important
> in shoes.
and examine bet variation rather then index generation. Bet variation is where 80%-90% of a counter's edge comes from. WHY wouldn't there be a gain in SCORE using 1/4 deck resolution instead of 1/2 deck resolution in the second half of the shoe? Sometimes this can make a difference in the bet of more then an entire unit!!
I tried posting your CV chart but don't quite know how.
MJ
I re-read the quote from MIT and it appears they're using TC precision,not Index precision. I'm not sure why you're using index precision,could you explain?
thanks,
Brick
> I re-read the quote from MIT and it appears they're
> using TC precision,not Index precision. I'm not sure
> why you're using index precision,could you explain?
There's an awful lot of confusion in this thread -- a lot more than need be. There are only a couple of principles involved, and there is nothing new here at all.
1. When you describe how you are going to calculate the true count, it is important to generate the indices that you will use when making departures in a manner consistent with the methodology used to reckon the true count.
2. Estimating decks remaining in a 6-deck shoe to the nearest quarter-deck is overkill. Estimating to the nearest half-deck will produce virtually identical results -- as any sim will clearly show.
3. Waiting until a rounded, truncated, or floored (I don't care which!) TC of +7 to make a top bet (in play-all mode) is vastly sub-optimal and will lose more money (by a huge factor) than could possibly be gained by reckoning quarter-decks instead of half-decks.
To claim that you're employing a 1-x spread, when the top bet of x isn't made until +7 is plain foolishness. You do'nt have the spread you think you do, because you virtually never make your alleged top bet. What's the point?
Don
> To claim that you're employing a 1-x spread, when the
> top bet of x isn't made until +7 is plain foolishness.
> You do'nt have the spread you think you do, because
> you virtually never make your alleged top bet. What's
> the point?
Assuming your TC frequency of .84% for a TC of +7(flooring)is correct, then you said a +7 TC would occur .5% more for rounding. This yields a TC frequency of 1.34% (rounding). So, rounding makes the difference of placing a top bet from once every 119 hands(flooring)to once every 75 hands. That is not as infrequent as you make it sound. If 60 hands are dealt per hour, that translates into a top bet placed every 1 hour and 15 minutes.
MJ
> This has nothing to do with your "system"
> but with the way you and Don reacted to MJ's critical
> remarks.
I made no such remarks. The fact is that you like insulting people and stirring up trouble. Over the last year you have called me dozens of childish names simply because you disagree with me (and nearly everyone else in the field.) This is of no help to anyone.
An interesting stat might be the number of times improving an already high precision makes a difference. Can't be many times. And when it is different, it's only different by one. It's not like you are betting min instead of max or vice-versa. And it isn't all that accurate anyhow since we don't bet exactly optimal to the penny. There is a gain; it just isn't much when you add in all the other factors.
> and examine bet variation rather then index
> generation. Bet variation is where 80%-90% of a
> counter's edge comes from. WHY wouldn't there be a
> gain in SCORE using 1/4 deck resolution instead of 1/2
> deck resolution in the second half of the shoe?
> Sometimes this can make a difference in the bet of
> more then an entire unit!!
> I tried posting your CV chart but don't quite know
> how.
> MJ
Indexes are generated according to a specified method of TC calculation. If you use a different method, you don't get the full benefit. Albeit the difference is slight.
> Why would index numbers need to change when the only
> difference is we're using fractions of a whole deck to
> "fine tune" the TC? I understand if we're
> dividing whole decks by or 4,but this is not the case.
> thanks,
> Brick
> Assuming your TC frequency of .84% for a TC of
> +7(flooring)is correct,
It's taken from the BJA3 chapter 10 charts, so I assume it's correct. You may assume differently, if you so chose. :-)
> then you said a +7 TC would
> occur .5% more for rounding.
Actually, that's a guess. It's probably less. I looked at the +6 frequency, and it was 0.81%, so I now think I was too generous with 0.5%, as the interval we need is in the second half of that +6 entry (from 6.5 to 6.999), and that will occur LESS than half of the total interval frequency (frequency not being linear). So, 0.3% is probably more like it. Total, therefore, would be more like 1.14%, or once every 88 hands.
> This yields a TC
> frequency of 1.34% (rounding). So, rounding makes the
> difference of placing a top bet from once every 119
> hands(flooring)to once every 75 hands.
See above.
> That is not as
> infrequent as you make it sound. If 60 hands are dealt
> per hour, that translates into a top bet placed every
> 1 hour and 15 minutes.
I wish I could understand your point, but I don't. There is a formula for calculating optimal bet ramp, both for the size of the bets, with regard to the bankroll, and the TCs at which those bets are placed. I keep telling you that, in play-all mode, top bets should be placed at +4 or +5, and we're arguing about +6 and +7. What, exactly, is your point?
You, MIT, and the Queen of England are free to place your top bets at any true count that makes you happy. Why are we discussing this? If you'd like to know the optimal point at which to place such a bet, with Hi-Lo, then I've told you that. If you'd like to tell me where MIT makes their bet, then you've told me that, too.
May we move on?
Don
No doubt there is something to gain by using 1/4 deck resolution for the end of a shoe.And this is even more true for Don/Norm-disciples.I just noticed that with an RC of +4 the TC+3 doesn't exist in their philosophy no matter how many cards are left(and the TC+4 is impossible with an RC+5)!So at one point there is a jump in TC from +2 to +4.This is not serious and some fine tuning cannot harm.
RC+4 with 1.25 decks to go means TC+3.2 but with 1/2 deck resolution and truncating it's only TC+2.
For me,a TC of +3 at the end of the shoe is clearly worth a max bet and even with more conventional bet schemes it calls for a bet size that differs from the TC+2 bet by at least two units.
At this point of the shoe (1.25 decks remaining)any positive count between 4 and 7 (and this is not rare!) will yield additional ev for a player using 1/4 deck resolution through higher bets.
I have explained in another post why 1/4 deck resolution at the end of the shoe doesn't require more effort. So the extra money is for free.Just take it!
Francis Salmon
I am aware if the index numbers were generated using flooring and the TC's are rounded the precision will be off,but what if the index numbers were generated using rounding also? Does it still matter if we divide by fractions of a whole deck?
thanks,
Brick
It's possible the MIT used a sort of risk-averse betting to allow for margin of error,not quite sure though.
Posting Permissions
-*-
Bookmarks