Cost of Deck Estimation Errors

[AcesFool]

Does anyone know what would be the loss in EV due to incorrect deck estimations when using high-low. For instance, if I average being off by 1 in an 8 deck shoe how much EV would this cost?

I realize High-Lo outperforms KO, but I'm skeptical about being able to be 100% accurate in estimating remaning decks down to the nearest half. In practical application does High-Lo still outperform KO or is this just theoretical based on perfect true counts.

mainly for 8D, dealt 6 deep. DOA, DAS, LS
I know this isn't a good game, to put it mildly, but its the best i have access to at the moment

 

[Felix Rue-de-Guerre]

I had the same doubts about my ability to estimate remaining cards. What I did was this:

I bought a clear discard tray online for 9 bucks, put a piece of masking tape along the back with with .5 deck graduations on it.

I drilled by grabbing a clump of cards, putting them in the tray. I checked my estimations by just turning the tray a bit so I could seee the tape. After a couple of 40(ish) minute sessions I was consistently accurate within a few cards. Just make sure you use new cards that aren't fluffed up.

-Felix

 

[Adam N. Subtractum]

re: deck est., KO v. Hi-Lo, etc...

"Does anyone know what would be the loss in EV due to incorrect deck estimations when using high-low. For instance, if I average being off by 1 in an 8 deck shoe how much EV would this cost?"

I'm not sure what you are alluding to by saying "off by one". What I can tell you is that with sloppier deck estimation win rate MAY ACTUALLY GO UP. This is due to the slight overbetting that occurs, but is of course outweighed by the increase in variance.

"I realize High-Lo outperforms KO..."

A common misconception.

"...but I'm skeptical about being able to be 100% accurate in estimating remaining decks down to the nearest half."

100% accuracy is of course not necessary. Being able to estimate within +/- 6 cards will more than suffice.

"In practical application does High-Lo still outperform KO or is this just theoretical based on perfect true counts."

As I hinted before, I have plenty of conclusive sim data that shows KO in running count mode outperforming Hi-Lo over a broad range of common conditions. Yes, there are a couple instances where Hi-Lo edges KO out very slightly, but as you presumed, this will be nullified, and then some, by the effect of errors.

One more point, if you did want to progress to TC conversion in the future, there is no reason to avoid KO, as it can be TC'ed just as Hi-Lo can, in fact easier.

There really isn't a need to differentiate between balanced and unbalanced. Either one can be used in TC or RC form, it just so happens (for obvious reasons) that unbalanced performs better in RC mode. As Pete Moss used to say, "There is no such thing as a balanced count."

"...mainly for 8D, dealt 6 deep. DOA, DAS, LS
I know this isn't a good game, to put it mildly, but its the best i have access to at the moment."

In that case, I would highly recommend the Red Sevens count. It is a simple unbalanced RC mode system that provides the optimal pivot point for backcounting ease, as well as accuracy. If wonging is not an option similar results can be achieved by aggressive table departure.

ANS

 

[Adam N. Subtractum]

also, use casino cards

 

[The Mayor]

In shoe games it costs almost nothing to be off by 1/2 deck, or even a full deck. Suppose you have a RC of 14 with 5 decks out. The correct TC is 14/5 = 2 (using the "floor" method). If you evaluate it as 4 decks you get a TC of 14/4 = 3. If you evaluate it as 6 decks out you get a RC of 14/6 = 2. The cost of being consistently off by one index number is negligible over your lifetime.

In pitch games, because strategy deviation is such a big factor, correct conversion of wildly fluctuating TC's is a must, and failure to be accurate can cost a bundle.

--Mayor

 

[The Mayor]

You might also read this essay I posted...

(Dead link: http://www.cardcounter.com/Essays/Mistakes.htm)

 

[The Mayor]

Question

Do you know Pete Moss?

 

[Adam N. Subtractum]

Re: Question

Unfortunately no, I haven't had the pleasure of knowing Pete. From what I understand you were (still are?) a friend of his, Mayor, are you still in touch with him? From what I heard he started getting serious with a "real" job sometime in '99 and was cutting down his BJ play. I've learned alot from his old posts, he truly was one of the few genuine innovators of the game in the last ten years, IMO.

ANS

 

[The Mayor]

Re: Question

Yes, Pete and I are good friends. He has always had a real job, and was a theorist more than anything in BJ. One of the most brilliant individuals I have ever met, we have known each other since 1983. Other that BJ, he also is a superb Sax player, a great poker/backgammon/chess player, a poet, a screen writer, an artist (painter), and of course a skilled Computer Scientist. And when I say that he has these skills, that is an understatement. He is world class in many of them.

--Mayor

 

[Adam N. Subtractum]

Wow...

...and I thought he was just a great BJ theorist, lol. Does he still play? And does he still contribute to any the paid forums, GC perhaps?

ANS

 

[CanKen]

Re: re: deck est., KO v. Hi-Lo, etc...

ANS:
The last line in the above post got my attention. I use KO with exit counts at each deck level equivalent to about TC<=-1. I've been trying to estimate the increase in EV or DI or SCORE compared to "play all" . I assume the values would be greater than "play all" but less than for "wonging". Neither BJA2 nor BJRM2002 gives me an answer. What's your opinion?

 

[Felix Rue-de-Guerre]

Common sense, I guess

Oh yea, and not the ones they sell in the gift shops with a hole drilled through them. (Like in Canada)It leaves a lip that makes them fluff. The liquor store on Freemont street sells them where the edges have been rounded. Mine are from the Bellagio.

I guess I took this to a bit of an extreme. I figured the cards that I bought were at the end of their casino lifecycle... their fluffiest playable state. Factor in any fluffieness produced by the grinding wheel that rounded off the corners?... Very clean job. No noticeable lip. A good squish will simulate a virgin pack.

So, my piece of masking tape has two marks for each half-deck of resolution. One as they sat loosely when I counted them. And, the other as I forcibly squished them into the tray. The range between the two gets slightly larger as the number of decks increases. But not much at all. Maybe two cards as it hits 6 decks (from 1 deck).

-Felix

 

[Adam N. Subtractum]

CK...

To get a good idea of what you are asking, check out T-Hopper's awesome free e-book, "T-H Basic BJ", it has some great system reports including play-all, backcounting, and table departure, as well as various spreads. Check out pages 60-63 specifically, but if you don't have a good understanding of the figures used, read the text, T-H gives an excellent summary of everything.

ANS

 

[T-Hopper]

Some casinos offer unmodified cards

I think I have got some with no holes or shaved corners at the Horseshoe before.