MJ: Discard Estimation: Question for Don(and others)

[MJ]

I recently switched from KO to Hi-Lo so I am relatively new to the idea of estimating the # of cards in the discard rack. I will be playing East Coast games.

Basically I want to be able to estimate the cards that remain in the shoe to the nearest quarter deck. The way I see it there are two approaches I can take to practice.

The first method involves estimation that is on the conservative side. I would simply number the cards in the discard rack(in quarters) starting at the bottom of the rack and work my way to the top:

Cards 1-13: 6D remaining

14-26: 5.75D remaining

27-39: 5.50D remaining

40-52: 5.25D remaining

53-65: 5.00D remaining etc.

OR (This approach is the more aggressive)

Cards 1-13: 5.75D remaining

14-26: 5.50D remaining

27-39: 5.25D remaining

40-52: 5.00D remaining

53-65: 4.75D remaining etc.

Well I hope you get the idea. At least this way I can pull cards from the discard tray and then check my estimation instantaneously(I will know about how many cards are left in the shoe). It would seem as though each approach has advantages and disadvantages. So what do you think? Which of the two should I use? Perhaps there is an even better approach I am missing? Thanks for any responses!

-MJ

[Don Schlesinger]

Strongest possible advice: For hi-lo, forget all about quarter-deck estimation. Stick strictly to half-deck. You absolutely don't need or want to do anything more.

Two ways to best estimate decks remaining: 1) Buy enough decks to make half-deck increments from, say, 1/2, all the way to 5. Place them across the table from you and learn what each slug looks like. Mix them up, and practice eyeballing the heights, until you know them cold.

2) Use the lazy approach, which is to ascertain that the discard tray is just slightly more than six decks high (careful! Sometimes they use 8, even for a 6-deck game), and to simply use the fraction of the tray that is taken up by the discards to do your estimates. Example: The tray looks one-third full. That's two decks; so four remain.

Don

[nycfudu]

I use Hi-lo with about 25 index numbers and I've been using full decks to convert the TC unless it's near the end of the shoe for 6 deck games. Should I switch to 1/2 deck conversion? Is there a significant difference in win rate?

> Strongest possible advice: For hi-lo, forget
> all about quarter-deck estimation. Stick
> strictly to half-deck. You absolutely don't
> need or want to do anything more.

> Two ways to best estimate decks remaining:
> 1) Buy enough decks to make half-deck
> increments from, say, 1/2, all the way to 5.
> Place them across the table from you and
> learn what each slug looks like. Mix them
> up, and practice eyeballing the heights,
> until you know them cold.

> 2) Use the lazy approach, which is to
> ascertain that the discard tray is just
> slightly more than six decks high (careful!
> Sometimes they use 8, even for a 6-deck
> game), and to simply use the fraction of the
> tray that is taken up by the discards to do
> your estimates. Example: The tray looks
> one-third full. That's two decks; so four
> remain.

> Don

[Don Schlesinger]

> I use Hi-lo with about 25 index numbers and
> I've been using full decks to convert the TC
> unless it's near the end of the shoe for 6
> deck games. Should I switch to 1/2 deck
> conversion? Is there a significant
> difference in win rate?

It's not "1/2 deck conversion." It's full-deck conversion, with 1/2-deck precision or accuracy. When 2 1/2 decks remain, you divide by 2 1/2 and not by 5.

In any event, you will win slightly more with 1/2-deck precision, but I wouldn't lose any sleep over it.

Don

[MJ]

> Strongest possible advice: For hi-lo, forget
> all about quarter-deck estimation. Stick
> strictly to half-deck. You absolutely don't
> need or want to do anything more.

Thanks for responding Don. There is another method of discard estimation practice which might be simpler. How about numbering the bottom of the cards from 1 to 312 and then place them in the discard rack in ascending order? This way you can pull a pile of cards off the discard rack, make an estimate as to how many cards are left in the rack, and then flip over the top card to get an exact number in an instant! What could be simpler? I got this idea from Kevin Blackwood's new book. What do you think? It also saves you the trouble of buying and setting up piles and piles of decks in various increments.

I am still unsure of how to go about rounding to the nearest half deck. Is this type of rounding schedule correct?

1-26 Cards in Rack .5 decks
27-52 Cards in Rack 1.0 decks
53-78 Cards in Rack 1.5 decks
79-104 Cards in Rack 2.0 decks
105-130 Cards in Rack 2.5 decks
etc...

It would seem as though this method is flawed. 5 cards in the rack certainly is not even close to half a deck. On the other hand if 20 cards were in the rack rounding off to half a deck would suffice. Similarly 29 cards in the rack is way off from 1 deck. Can you give me some type of a rounding schedule which you think will get the job done or is the one above more or less the idea?

-MJ

[Don Schlesinger]

> Thanks for responding Don. There is another
> method of discard estimation practice which
> might be simpler. How about numbering the
> bottom of the cards from 1 to 312 and then
> place them in the discard rack in ascending
> order? This way you can pull a pile of cards
> off the discard rack, make an estimate as to
> how many cards are left in the rack, and
> then flip over the top card to get an exact
> number in an instant! What could be simpler?
> I got this idea from Kevin Blackwood's new
> book. What do you think? It also saves you
> the trouble of buying and setting up piles
> and piles of decks in various increments.

I like it! Seems fine to me and very practical.

> I am still unsure of how to go about
> rounding to the nearest half deck. Is this
> type of rounding schedule correct?

No.

> 1-26 Cards in Rack .5 decks
> 27-52 Cards in Rack 1.0 decks
> 53-78 Cards in Rack 1.5 decks
> 79-104 Cards in Rack 2.0 decks
> 105-130 Cards in Rack 2.5 decks
> etc...

> It would seem as though this method is
> flawed. 5 cards in the rack certainly is not
> even close to half a deck. On the other hand
> if 20 cards were in the rack rounding off to
> half a deck would suffice. Similarly 29
> cards in the rack is way off from 1 deck.
> Can you give me some type of a rounding
> schedule which you think will get the job
> done or is the one above more or less the
> idea?

1-13, 6 decks remain
4-39, 5.5 decks remain
40-65, 5 decks remain, etc.

Another way is:
1-25, 6 decks remain
26-51, 5.5 decks remain
52-77, 5 decks remain, etc.

I think the manner in which you will USE the TC at the table (floor, round, or truncate) dictates how you actually estimate the TC in the first place. The two methodologies should be the same.

Norm may have some comment as to how he does this for his sims.

Don

[Norm Wattenberger]

> Norm may have some comment as to how he does
> this for his sims.

CVData/CVCX allow you to set up a table of ranges so you can specify it any way you wish both for simulation and index generation. For the CVCX canned sims, I try to figure out what each particular author did. Sometimes this isn't clear from the books and I just use what I believe was common at the time.

Serious Blackjack Software

[MJ]

> I like it! Seems fine to me and very
> practical.

I concur! :-)

> 1-13, 6 decks remain
> 4-39, 5.5 decks remain
> 40-65, 5 decks remain, etc.

I like this rounding schedule you propose! It underestimates the first quarter of the deck(good to be conservative), then the next 2 quarters are centered around the middle of the deck, and the final quarter rounds up to the next deck. Works for me. Thanks for getting the ball rolling. Well here is the rest of it(should be correct):-)

66-91, 4.5 decks remain
92-117, 4.0 decks remain
118-143, 3.5 decks remain
144-169, 3.0 decks remain
170-195, 2.5 decks remain
196-221, 2.0 decks remain
222-247, 1.5 decks remain
248-273, 1.0 decks remain
274-299, .5 decks remain
300-312, 0.0 decks remain

Let's just say one wanted to use this type of methodology to round to the nearest quarter deck(maybe they are professional players).
I realize you feel it is unneccessary but can you please tell me if this is correct:

1-7, 6 decks remain
8-20, 5.75 decks remain
21-33, 5.50 decks remain
34-46, 5.25 decks remain
47-59, 5.0 decks remain

The problem arises because you really cannot divide a deck evenly into 1/8ths. If the schedule is incorrect can you write a few lines like last time so I get the idea? Also since the schedule for rounding to the nearest half deck was 12 lines I guess this next schedule will contain 24 lines in total. Thanks.

-MJ

[Don Schlesinger]

> Let's just say one wanted to use this type
> of methodology to round to the nearest
> quarter deck(maybe they are professional
> players).
> I realize you feel it is unneccessary but
> can you please tell me if this is correct:

> 1-7, 6 decks remain
> 8-20, 5.75 decks remain
> 21-33, 5.50 decks remain
> 34-46, 5.25 decks remain
> 47-59, 5.0 decks remain

Lower every number on the chart by one. the idea is to have the number of cards in the tray to be the exact middle of the interval. So, for, say, 5.5 decks remain, that means 26 is in the middle. Your way, 26 is 5 from the left (21) but 7 from the right (33), hence not centered. Make it 20-32, and 26 is right in the middle.

Get the idea?

Don

[Sun Runner]

> How about numbering the
> bottom of the cards from 1 to 312 and then
> place them in the discard rack in ascending
> order?

> I got this idea from Kevin Blackwood's new book.

A great idea that has been around since at least Snyder's ST'ing Cookbook.

> It also saves you the trouble of buying and setting up piles and piles of decks in various increments.

Still, not a bad training tool.

[MJ]

Thanks Don I get it. So the chart will start off like this:

1-6, 6 decks left
7-19, 5.75 decks left
20-32, 5.50 decks left etc

Anyhow, I guess I will stick with a half deck schedule for now and see how that works.

Once a TC has been computed, is it ok to round off to the nearest integer or should I round in increments of .5? For example, if the RC= 6 and there are 2.5 shoes remaining, then the TC = 2.4.
Then I subtract an offset of 1 to determine the # of units to bet. Now do I round 1.4 up to 1.5 or down to 1.0?

It probably won't make a difference in the long run but I figured I would ask just to be on the safe side. Thanks.

-MJ

[Don Schlesinger]

> Once a TC has been computed, is it ok to
> round off to the nearest integer or should I
> round in increments of .5? For example, if
> the RC= 6 and there are 2.5 shoes remaining,
> then the TC = 2.4.
> Then I subtract an offset of 1 to determine
> the # of units to bet. Now do I round 1.4 up
> to 1.5 or down to 1.0?

Flooring (the preferred methodology) means always rounding DOWN to the next-lowest integer. TC of 2.4 becomes 2. TC of -2.4 becomes -3.

Don

[ES]

1. Why is flooring better than rounding?

2. In an earlier post in this thread, you recommend using the same method for deck, i.e. divisor, estimation that one uses for TC calculation. If one uses full deck accuracy and is flooring the TC, does this mean that if he is playing a 6-deck shoe with few players at the table that the divisor for the second round is 5, rather than 6, i.e. that the divisor of "slightly fewer than 6 decks remaining" is floored to 5 and then divided into the RC, which is inturn floored to calculate the TC? BJ Attack reference?

3. Were the High-Low indices in "Professional Blackjack," Humble's published Hi-Opt I and II indices and the Revere's published indices (book and pamphlets) calculated with the intention that divisors and TCs be floored, or, is index calculation independent of determination of divisors and TC calculation?