MJ: TC Deck Estimation: 1/2 or 1/4?

[MJ]

For all you Hi-Lo users, how accurate does the deck estimation have to be when converting to the true count? Should one estimate to nearest 1/2 deck or 1/4 deck? For single deck it would be to the nearest 1/4 deck I guess. But how about for 2,6, and 8 decks? Would estimating to the nearest 1/2 deck do the trick? Also about how much more of a gain is there in estimating to the nearest 1/4 deck for a shoe game?

Obviously the more exact the estimation is the more accurate the TC will be. It just seems like the division can be a real pain in the neck when dividing by numbers such as 3.25, 4.75, 6.25 etc.

-MJ

[bfbagain]

they're easier for most people.

cheers
bfb

[Sun Runner]

For shoe games, whole deck is fine. Possibly when into the fourth deck played, half deck TCs might be worthwhile.

If dividing, round a lot. Round the decks down to the nearest whole deck then the RC down to the nearest even numerator; i.e. if 4.75 decks left, round down to 5. If the RC is 12, round down to 10.
10/5 = 2 TC.

> Obviously the more exact the estimation is
> the more accurate the TC will be.

Yes, technically it's 12/4.75 or 2.53 TC but the only indices I know are in whole numbers, right?

DD' once said it's better to be 95% right and playing fast than 100% right and playing slow.

He's a smart man.

[Cardkountr]

> For all you Hi-Lo users, how accurate does
> the deck estimation have to be when
> converting to the true count? Should one
> estimate to nearest 1/2 deck or 1/4 deck?
> For single deck it would be to the nearest
> 1/4 deck I guess. But how about for 2,6, and
> 8 decks? Would estimating to the nearest 1/2
> deck do the trick? Also about how much more
> of a gain is there in estimating to the
> nearest 1/4 deck for a shoe game?

> Obviously the more exact the estimation is
> the more accurate the TC will be. It just
> seems like the division can be a real pain
> in the neck when dividing by numbers such as
> 3.25, 4.75, 6.25 etc.

> -MJ
Full deck estimation is all that is needed using hi/lo in a 6 or 8 deck game. That is what the system was designed for and anything else would be counter productive. Don't get hung up with exact estimations, just round down to nearest whole deck....1.5 played is still a 1 etc.

As BFB mentioned below, if you have trouble dividing to obtain the true count, a faster and easier way is to multiply.

For example in a 6 deck shoe with 1 deck played, you know you need a running 5 to equal a true 1, running 10 for a true 2, running 15 for a true 3 etc. With 4 decks remaining you'd need a running 4 for a true 1, running 8 for a true 2, running 12 for a true 3 etc. So a running 14 with 4 decks remaining is still a true 3. This method is much easier than trying to divide and can be done with a simple glance and since you're truncating anyways the fractions caused by dividing aren't needed anyways.

Hope this helps,

Card.

[Brick]

I dont think rounding the running count is a good idea and can leave a counter more apt to make mistakes. It's really not neccesary to mess with the RC. It's better to round the TC. If I have an RC of 12 with 4.75 decks left,I know the TC is not at 3, so I simply use a 2TC and bet accordingly. If the RC goes up to 15 on the next round ,then I bet to a 3 TC. My running count always remains constant.

@

> For shoe games, whole deck is fine. Possibly
> when into the fourth deck played, half deck
> TCs might be worthwhile.

> If dividing, round a lot. Round the decks
> down to the nearest whole deck then the RC
> down to the nearest even numerator; i.e. if
> 4.75 decks left, round down to 5. If the RC
> is 12, round down to 10.
> 10/5 = 2 TC.

> Yes, technically it's 12/4.75 or 2.53 TC but
> the only indices I know are in whole
> numbers, right?

> DD' once said it's better to be 95% right
> and playing fast than 100% right and playing
> slow.

> He's a smart man.

[MJ]

Thanks for the reply. I think with practice I can learn to estimate the discards within a half deck for shoe games.

How much does all this estimation cost the player as far as EV is concerned? Also does the data for BJRM and CVCX assume the player can compute the TC precisely? I would hope the SIMS use the same rounding methodology as humans do rather then calculate a theoretical EV and SCORE based upon a perfect TC conversion.

-MJ

> For shoe games, whole deck is fine. Possibly
> when into the fourth deck played, half deck
> TCs might be worthwhile.

> If dividing, round a lot. Round the decks
> down to the nearest whole deck then the RC
> down to the nearest even numerator; i.e. if
> 4.75 decks left, round down to 5. If the RC
> is 12, round down to 10.
> 10/5 = 2 TC.

> Yes, technically it's 12/4.75 or 2.53 TC but
> the only indices I know are in whole
> numbers, right?

> DD' once said it's better to be 95% right
> and playing fast than 100% right and playing
> slow.

> He's a smart man.

[Brick]

If you round the way I do,the cost of ev is small. I believe BJRM uses rounding with the pre-sims. I suppose you can program CV to use different types of rounding.

@

> Thanks for the reply. I think with practice
> I can learn to estimate the discards within
> a half deck for shoe games.

> How much does all this estimation cost the
> player as far as EV is concerned? Also does
> the data for BJRM and CVCX assume the player
> can compute the TC precisely? I would hope
> the SIMS use the same rounding methodology
> as humans do rather then calculate a
> theoretical EV and SCORE based upon a
> perfect TC conversion.

> -MJ

[MJ]

Hmmmmm....that sounds like a good idea. Its simple enough. Thanks for the tip.

As far as rounding off the # of decks is concerned, do you always be conservative? Lets say the RC is 14 with 3.25 decks remaining. In this case do you round off the # of decks remaining to 3 or 4?

How about RC = 16, and # decks remaining 2.5? Would you round the 2.5 up or down? Whichever way you round the answer will be off by more then 1 unit of TC. 16/2.5 = 6.4

16/2 = 8
16/3 = 5

I realize this does not happen in most cases but errors can certainly creep in when rounding. Not much can be done about it I suppose unless your brain is a calculator.

-MJ

> Full deck estimation is all that is needed
> using hi/lo in a 6 or 8 deck game. That is
> what the system was designed for and
> anything else would be counter productive.
> Don't get hung up with exact estimations,
> just round down to nearest whole deck....1.5
> played is still a 1 etc.

> As BFB mentioned below, if you have trouble
> dividing to obtain the true count, a faster
> and easier way is to multiply.

> For example in a 6 deck shoe with 1 deck
> played, you know you need a running 5 to
> equal a true 1, running 10 for a true 2,
> running 15 for a true 3 etc. With 4 decks
> remaining you'd need a running 4 for a true
> 1, running 8 for a true 2, running 12 for a
> true 3 etc. So a running 14 with 4 decks
> remaining is still a true 3. This method is
> much easier than trying to divide and can be
> done with a simple glance and since you're
> truncating anyways the fractions caused by
> dividing aren't needed anyways.

> Hope this helps,

> Card.

[bfbagain]

6 decks left - Use .2 [actual 1.6]
5 decks left - Use .2
4.5 decks left - Use .2 [actual .22]
4 decks left - Use .25
3.5 decks left - Use .3 [actual .28]
3 decks left - Use .33
2.5 decks left - Use .4
2 decks left - Use .5
1.5 decks left - Use .7 [actual .75]
1.25 decks left - Use .8
1 deck left - Use 1
3/4 deck left - Use 1.3
1/2 deck left - Use 2
1/4 deck left - Use 4

A while back, as I recall, this was another one of those topics that was discussed at Don's Domain

Hope this simplifies things for you.

cheers
bfb

[Don Schlesinger]

Nice chart. One correction:

> 1.5 decks left - Use .7 [actual .75]

Actual is 2/3, or .67, the reciprocal of 3/2, or 1.5.

Don

[bfbagain]

correct. Thanks.

bfb

[Francis Salmon]

At the beginning of the shoe, full deck estimation for your convenience is certainly OK.It doesn't make much difference whether you divide by 5.5, by 5 or by 4.5, so in this region you always divide by 5 or even easier as bfbagain suggests multiply by 2 and put a decimalpoint.
But the closer you get to the end of she shoe the more important correct deck estimation gets.
In your example at the 2.5 deck level I would strongly object to rounding either way, not only because of precision but because of convenience as well: you can simply multiply by 0.4. So RC+16 is exactly TC+6.4 .Now you know that you may split tens against five or six but you still hit 15 against the 9 or 16 versus the 7.
I don't agree with the widespread simplistic attitude which says that because the indices aren't precise the TC calculation don't need to be either. Sometimes these imprecisions will compensate each other but just as often they will just work in opposite directions and this will be at a cost.
Even in shoe games it's important to deal with half decks. Suppose you have a count of +4 and there is just a half deck remaining. If you carelessly divide by 1 you get TC+4 but in fact you have TC+8 which in many cases requires totally different playing decisions.

Francis Salmon

> Hmmmmm....that sounds like a good idea. Its
> simple enough. Thanks for the tip.

> As far as rounding off the # of decks is
> concerned, do you always be conservative?
> Lets say the RC is 14 with 3.25 decks
> remaining. In this case do you round off the
> # of decks remaining to 3 or 4?

> How about RC = 16, and # decks remaining
> 2.5? Would you round the 2.5 up or down?
> Whichever way you round the answer will be
> off by more then 1 unit of TC. 16/2.5 = 6.4

> 16/2 = 8
> 16/3 = 5

> I realize this does not happen in most cases
> but errors can certainly creep in when
> rounding. Not much can be done about it I
> suppose unless your brain is a calculator.

> -MJ

[Sun Runner]

> Even in shoe games it's important to deal
> with half decks. Suppose you have a count of
> +4 and there is just a half deck remaining.

Respectfully, when was the last time that happened?

MJ's post was ultimately about counting by half deck in 2, 6, or 8 deck shoes and I agree that in the last half of the shoe half deck conversions could be/might be useful.

I also still believe that, except for a full on math whiz such as yourself, calc'ing TCs to fractions of integers (i.e. 6.4) is not realistic.

Neither is learning a multiplication table consisting of fractions (quickly now, whats 1.25 decks times .8 equal? If you didn't get exactly 1.0 in the precious few seconds you have to think about it, you blew it.) Multiplying is great, but again, were I going to learn a new multiplication table it would be like this: between four decks and three decks remaining, I'd round to 4 decks and would already know that 4,8,12,16,20 stands for a 1,2,3,4,5 TC.

I'd like to see, in a shoe game, in the real world, how much is really being given up by not wearing yourself out doing the math.

Just me.