> If resolution is 1/4 deck, how do you figure I am
> dividing by
> "whole" decks?
Honestly, it's not your fault, but I just don't have the patience to answer this again, after having discussed and answered this point at least 100 times.
When you have one deck remaining, do you divide by 1 or by 4?
Don
> I stated that the MIT Team uses rounding for the TC. I
> would venture to say that the optimal bet ramps you
> posted are for FLOORING the TC. Obviously, with
> flooring you have a greater advantage per each unit of
> TC as compared to rounding. That is why your ramp
> mandates placing a max bet at a lower TC.
The difference between flooring and rounding is one-half of a true point. Placing the max bet for the interval +6.5 to +7.4999), instead of +7 to +7.9999, is almost just as bad, no?
> Now, do you have any optimal bet ramps for when you
> ROUND the TC?
Sure. Shift everything by one-half TC.
> If so, please go ahead and post it.
Use CVCX.
> Otherwise, the comparison you are trying to make is
> invalid.
It's perfectly valid; it just isn't perfectly accurate, because you're rounding. All the principles remain the same. ALL of them.
> Also, was the statistic you sighted of a TC
> frequency of +7 occurring 1 in 119 hands for a floored
> or rounded TC?
Floored.
> If it assumes flooring, do you know how
> often a TC of +7 would occur when using rounding?
> Thanks Don!
Yes, I'd guess a whopping 0.5% more! :-)
Don
Sorry about that Don. I didn't realize this question was so frequently asked and answered. It could have been worse, I could have asked about playing multiple hands. I know how much you like to answer that one! :-)
On another note, you might want to consider adding an archive where you answer FAQ. This way it saves you the trouble of repeatedly answering the same questions over and over and over......
But to answer your question, if there is 1 deck left to be played I divide by 1.
MJ
> Honestly, it's not your fault, but I just don't have
> the patience to answer this again, after having
> discussed and answered this point at least 100 times.
> When you have one deck remaining, do you divide by 1
> or by 4?
> Don
> But to answer your question, if there is 1 deck left
> to be played I divide by 1.
Then, you're dividing by whole decks, whether you divide by 1, 1.25, 1.50, or 1.75. You're dividing by whole decks, and you may choose to use whole-deck precision, half-deck precision, or quarter-deck precision. Understood?
As for the rounding/flooring question, see BJA3, pp 82-83. This is a very old question, as well. There is virtually no difference whatsoever between the two approaches. One produces more frequent counts with smaller edges, the other less frequent counts with larger edges. Bet optimally, the SCOREs are virtually identical.
Finally, I reiterate: Using quarter-deck precision, rather than half-deck, in a 6-deck shoe game, will take a lot of effort of your part, for virtually no extra gain whatsoever.
And, you're perfectly free to ignore my advice and do it MIT's way, but don't delude yourself into thinking that it matters; it doesn't.
Don
> Finally, I reiterate: Using quarter-deck precision,
> rather than half-deck, in a 6-deck shoe game, will
> take a lot of effort of your part, for virtually no
> extra gain whatsoever.
> And, you're perfectly free to ignore my advice and do
> it MIT's way, but don't delude yourself into thinking
> that it matters; it doesn't.
> Don
Most professional use a far more efficient method than divididing by remaining decks. If say 1.25 decks remain they will multiply by 8 and put in the decimal point.Having like most people learnt the multiplication tables in their first few school years they can recall the results instantly.
Now let's take as an example an RC of +5. With 1.25 deck remaining, we know immediately this is exactly TC +4 (8*5=40 => 4.0).Your students will first have to round conservatively to the nearest half deck which is 1.5. Now they have to perform the akward division 5/1.5.While the slow ones will miss their bet the smarter ones will figure that its rather 3 than 4 before the dealer has dealt for the next round.
So with more trouble they arrived at the wrong result which will not only make them miss some important index plays but also a considerable amount of action.
It's not a question of opinion whether this matters.It does.
Francis Salmon
You can see why Don and I do not respond to Francis.
I have always used the following approximation simply for ease of use.
Remaining decks / Multiplier
6.0 x 0.15
6.5 x 0.15
5.5 x 0.15
5.0 x 0.20
4.5 x 0.20
4.0 x 0.25
3.5 x 0.30
3.0 x 0.30
2.5 x 0.40
2.0 x 0.50
1.5 x 0.70
1.0 x 1.00
May not be all that accurate to the 4th decimal point but gets the job done.
> You can see why Don and I do not respond to Francis.
but would you agree that the TC is more sensitive the further you get into the shoe? If there is 1.25 decks in the discard rack, it would NOT make the least bit difference if you use 1/2 or 1/4 deck resolution. But if you look at my realistic example with 4.25 decks in the rack, it would seem like 1/4 deck resolution might make a slight difference in SCORE(at least intuitively). See my example above.
I certainly cannot argue with that chart you posted a while back comparing the SCORES of full, half, quarter, and exact card resolution. Again, all the curves were practically on top of one another. However, I am surprised that when less then one deck is left to be played, the curves were still nearly on top of each other!!!!!!!!!!
Seeing that I have never run a simulation in my life, I guess I'll have to trust your chart/simulation. As I learned with BS plays, not everything is intuitive, hence we use simulation.
MJ
I'm not saying we don't respond to him because he's wrong; but because he's extremely nasty. I spend the time putting together a free system and he uses it as an excuse to call me and Don delusional and falsely accuses me of treating you with contempt.
That aside, index precision just isn't that important in shoes. I think the REKO system is a dramatic example. Forget tenths or exact cards in precision, I can get good results with indexes that are way off the ideal. Most people don't estimate the correct number of remaining cards anyhow since most people play with many players and don't bother to subtract the cards on the table.
> I'm not saying we don't respond to him because he's
> wrong; but because he's extremely nasty. I spend the
> time putting together a free system and he uses it as
> an excuse to call me and Don delusional and falsely
> accuses me of treating you with contempt.
This has nothing to do with your "system" but with the way you and Don reacted to MJ's critical remarks.
If you want to know what being nasty and contemptful really means you just have to read Don's response to MJ's quote from an MIT member.Your one-liners earned the qualification platitudes.
It should be obvious to everybody that I was just using the same language as Don uses when somebody disagrees with him.
Francis Salmon
> I have always used the following approximation simply
> for ease of use.
> Remaining decks / Multiplier
> 6.0 x 0.15
> 6.5 x 0.15
> 5.5 x 0.15
> 5.0 x 0.20
> 4.5 x 0.20
> 4.0 x 0.25
> 3.5 x 0.30
> 3.0 x 0.30
> 2.5 x 0.40
> 2.0 x 0.50
> 1.5 x 0.70
> 1.0 x 1.00
> May not be all that accurate to the 4th decimal point
> but gets the job done.
You might want to add:
1.75 x 0.6
1.25 x 0.8
Francis Salmon
I guess the MIT teams ran their simulations and calculated the CE numbers using EXACT card resolution and then required their BPs to use 1/4 deck estimation to try and get a very accurate TC.
These techniques do suffer from the same delusions as Mr. Salmon?s one-decimal ?precise? indices. A non-rewarded mental effort.
Sincerely,
Zf
> For increased remaining card estimates to be of full
> value, you need to generate the indexes using the same
> accuracy.
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
Posting Permissions
-*-
Bookmarks