Vindication of my insecurities

[boneuphtoner]

http://i219.photobucket.com/albums/cc145/boneuphtoner/ddstacks.jpg

I think this photo says it all. See attached.

For years, I went back and forth between unbalanced and balanced counts. I've tried nearly all of them - level II and Level I. I always ended up preferring the feel of balanced counts overall, with FELT that I switched to last year feeling the most natural for me. Compared to playing with an unbalanced count, I feel that it is so powerful to instantly know whether you have the advantage or not with a balanced count. Indexes like 16 v. 10 and 12 v. 4 are also extremely easy and the answer is always obvious. I never found the math of dividing by whole decks and flooring the result to be difficult either. Neither was dividing by fractions like I do with pitch games...piece of cake right? However, what I have always felt insecure about were my deck estimation skills. After thinking about it at length and after reading Modern Blackjack, scanning the effects of estimation error, I put my mind to rest....the effects of 10%deck estimation error on every calculation makes very little difference...at least it put my mind at ease for the moment. Like any decent counter in training who uses a balanced count, I ordered a bunch of used casino playing cards off of ebay. I followed the advice of several reputable authors, and made a bunch of different stacks to try to fine tune my visual acuity. The problem is, I began to perceive differences in the heights of stacks of cards from the different casinos. So, to test this, I counted 104 cards from each of the three different casino cards that I received. The results are in the attached photo. Now, I don't think I have that great visual acuity, but even I can see these differences! The one from Ceasar's in particular looks to my eye to be nearly a half deck taller than the others!

To me, this is confirmation that my insecurity about deck estimation was warranted all along. At this point, I'm seriously thinking of switching back to an unbalanced count. What do you guys think? In my mind, for those of you who estimate decks to the nearest half or quarter MUST be aware of these differences to have reasonable accuracy. I use half deck resolution for pitch games, and that is looking really shaky to me right now. Let alone quarter deck resolution.

 

[metronome]

Photography skills

I(we) get the point, which I find at once interesting and also disturbing.
But, your focal plane is WAY past the cards. The monitors however, are rendered quite sharp. The decks... not so much.
In all, a very interesting quandry... I see a mass theft of cards from as many casinos as possible.

 

[metronome]

deck estimation

Boneuphtoner,
I've never seen anything as to card manufacturing variances on this, or any other site. I do know the meaning of "collusion". Every gaming corporation knows their competition intimantetly. I would not be surprised at all to see different "recipes" for playing cards from the same supplier to various "stores" in the same market. Happens in the concrete industry every day... but can't be proven.

 

[Automatic Monkey]

....
To me, this is confirmation that my insecurity about deck estimation was warranted all along. At this point, I'm seriously thinking of switching back to an unbalanced count. What do you guys think? In my mind, for those of you who estimate decks to the nearest half or quarter MUST be aware of these differences to have reasonable accuracy. I use half deck resolution for pitch games, and that is looking really shaky to me right now. Let alone quarter deck resolution.

It makes little difference in pitch games, especially SD.

Here's a suggestion: pick up a nice strong unbalanced count (BRH-1 or Unbalanced Zen) and learn to use it both in true counted mode for shoes, and in running count mode for pitch games or odd situations like covered shoes or discards.

 

[metronome]

check slots

It is like GameKing and IGT.
They (almost) exclusively control the slot machine market share in the U.S.

 

[boneuphtoner]

It makes little difference in pitch games, especially SD.

Huh? I thought deck estimation mattered MORE in pitch games than in shoes! I did sims on my rounded FELT using your Ben Franklin style approach using quarter deck, half deck, and whole deck resolution. Quarter deck was the best, but it wasn't much better than half deck. Both were better than whole deck, especially at single deck. Still, I thought deck estimation resolution was more important for pitch than shoes.

I(we) get the point, which I find at once interesting and also disturbing.

Yes, I find this extremely disturbing! I ordered 64 decks of used cards off of ebay, and they sent me cards from Ceasar's, Planet Hollywood, and Palace Station. I never imagined that you would see this kind of variation. The PH and PS cards are closer in thickness, more in line with variation I would expect, but the Ceasar's appears to be a major outlier! This is potentially a strategy changer for me. Unless one of you experts out there can explain how you can deal with the inaccuracies these differences can introduce into your true counting, I'm ditching FELT for KO or UBZ-II.

 

[boneuphtoner]

Here's a suggestion: pick up a nice strong unbalanced count (BRH-1 or Unbalanced Zen) and learn to use it both in true counted mode for shoes, and in running count mode for pitch games or odd situations like covered shoes or discards.

I wouldn't disagree with this until I actually picked up a copy of the Second Edition (not the free online version) of Modern Blackjack. It was there that I learned that true counting any unbalanced strategy is much much harder than a balanced count. I didn't realize you had to use the RC-4(decks played)/decks remaining. No wonder my older CVCX sims using RC/decks remaining gave such poor results, and I only recently realized what I was doing wrong.

So, if I switch to an unbalanced strategy, I have no desire to true count it, other than using perhaps a Color of Blackjack style fudge maneuver.

Isn't BRH1 essentially equivalent in performance to UBZII? I've read about that count, and it only differs from FELT with counting the 5 as +3....shouldn't be that much harder, although if it isn't any better than UBZ why move to level III?

 

[LIB]

For years, I went back and forth between unbalanced and balanced counts. I've tried nearly all of them - level II and Level I. I always ended up preferring the feel of balanced counts overall, with FELT that I switched to last year feeling the most natural for me. Compared to playing with an unbalanced count, I feel that it is so powerful to instantly know whether you have the advantage or not with a balanced count. Indexes like 16 v. 10 and 12 v. 4 are also extremely easy and the answer is always obvious. I never found the math of dividing by whole decks and flooring the result to be difficult either. Neither was dividing by fractions like I do with pitch games...piece of cake right? However, what I have always felt insecure about were my deck estimation skills. After thinking about it at length and after reading Modern Blackjack, scanning the effects of estimation error, I put my mind to rest....the effects of 10%deck estimation error on every calculation makes very little difference...at least it put my mind at ease for the moment. Like any decent counter in training who uses a balanced count, I ordered a bunch of used casino playing cards off of ebay. I followed the advice of several reputable authors, and made a bunch of different stacks to try to fine tune my visual acuity. The problem is, I began to perceive differences in the heights of stacks of cards from the different casinos. So, to test this, I counted 104 cards from each of the three different casino cards that I received. The results are in the attached photo. Now, I don't think I have that great visual acuity, but even I can see these differences! The one from Ceasar's in particular looks to my eye to be nearly a half deck taller than the others!

To me, this is confirmation that my insecurity about deck estimation was warranted all along. At this point, I'm seriously thinking of switching back to an unbalanced count. What do you guys think? In my mind, for those of you who estimate decks to the nearest half or quarter MUST be aware of these differences to have reasonable accuracy. I use half deck resolution for pitch games, and that is looking really shaky to me right now. Let alone quarter deck resolution.

Every time you hesitate you lose your turn. Aren't the years that you've spent switching back and forth with different systems enough? I also think that choosing a system is a matter of what you are capable of and what the circumstances require rather than what you like and what you feel.

It seems to me that you're making this more complicated than it is.
Situation: You're unsure about your visual estimate.
Solution: Work on it. If it doesn't work, then you know for sure. You know what to do. It's the unknown that's difficult to deal with.

I'm no decent counter, but the 4th bold line doesn't seem like something that requires such level. Again, it's a simple situation.
Situation: Discrepancies within different brands of cards actually used in real casino.
Solution: Measure the cards and go for the cards with thicknesses whose discrepancies are statistically insignificant.

It's not a confirmation of the origin of your insecurity. It's a justification for your peace of mind.

 

[Automatic Monkey]

I wouldn't disagree with this until I actually picked up a copy of the Second Edition (not the free online version) of Modern Blackjack. It was there that I learned that true counting any unbalanced strategy is much much harder than a balanced count. I didn't realize you had to use the RC-4(decks played)/decks remaining. No wonder my older CVCX sims using RC/decks remaining gave such poor results, and I only recently realized what I was doing wrong.

So, if I switch to an unbalanced strategy, I have no desire to true count it, other than using perhaps a Color of Blackjack style fudge maneuver.

Isn't BRH1 essentially equivalent in performance to UBZII? I've read about that count, and it only differs from FELT with counting the 5 as +3....shouldn't be that much harder, although if it isn't any better than UBZ why move to level III?

Which count is better depends on the game. Overall you'll be playing a long time before the difference between the two becomes significant. Both are related to RPC/Ben Franklin/FELT; one achieves the unbalance by increasing the weight of the 5 and the other by decreasing the ace. If you're playing mostly pitch UBZ will be slightly better.

What I mean is: in a pitch game a running count mode system will perform about as well as a true counted one, unlike a 8D shoe where there's so much space between the beginning and end of the shoe you really want to be true counting. True counting also makes Wonging in and out much easier, but you're not going to be doing much of that in pitch games. So if you're playing just straight BJ pick a good counting system and learn different ways to use it, and that will be more effective than learning a bunch of different systems.

 

[blackjack avenger]

A Banana For the Cute Little Monkey

So if you're playing just straight BJ pick a good counting system and learn different ways to use it, and that will be more effective than learning a bunch of different systems.

Fantastic:joker::whip:

 

[Sucker]

Unless one of you experts out there can explain how you can deal with the inaccuracies these differences can introduce into your true counting, I'm ditching FELT for KO or UBZ-II.

If you had a friend standing behind you and counting the EXACT number of cards that were dealt/undealt, at the end of one year you would do BARELY better than if you were to just sloppily make quick estimates of the cards. The difference, as far as money won, would be almost unnoticeable.

Card counting is NOT rocket science. You're trying to build a spaceship that will take you to another star system & back. You need to stop wasting your time trying to find this "magic bullet". Pick ONE system, ANY system; and stick with it. Just try to do your best with that system, and the money WILL come.

 

[blackjack avenger]

A Mountain Over a Little Card Hill

1. You can be off on indices and bet ramps with little to no effect. So being off in deck estimation; regardless of reason, is not that important. Do you use any betting camo? How is this any different? your bets are not perfect.
Do you really think if you bet $90 to $110 instead of $100 it's going to make a lot of difference? Do you really think if you stand on X indice at a range of 1.1 through .9 instead of 1 it makes much difference?

2. When you do deck estimation you are doing it off the cards available. For 2 out of 8 decks you are looking at 25% of the cards that are available, not 1.25 inches equals 2 decks for every pack of cards in every different casino.

3. Quit switching counts, the switching can lead to errors and is a waste of time.

4. I have no idea how much you are paying for the cards, but for this thickness of cards question you have already hurt your bank enough.

5. Unbalanced counts use a deck estimation technique also, they take an average. This is a far worse estimation then an imperfect human TC calculation. All one has to do is look at specific situations for unbalanced counts to see the indices and bets are sometimes quite off.

The level and precision of tags used is probably far more important then if balanced or not.

:joker::whip:

 

[blackjack avenger]

I Agree

If you had a friend standing behind you and counting the EXACT number of cards that were dealt/undealt, at the end of one year you would do BARELY better than if you were to just sloppily make quick estimates of the cards. The difference, as far as money won, would be almost unnoticeable.

Card counting is NOT rocket science. You're trying to build a spaceship that will take you to another star system & back. You need to stop wasting your time trying to find this "magic bullet". Pick ONE system, ANY system; and stick with it. Just try to do your best with that system, and the money WILL come.

Well said
:joker::whip:

 

[Sharky]

agreed, it is not that difficult when you understand that decks vary in thickness, just like everyone's, well, you get it.

all the stores I play put all the cards into the discard tray b4 shuffling- even those with ASM to square them up. This is when you see how close to the top of the tray the whole enchilada is...then you estimate from there.

not too tough.

Also, if hand shuffling pay attention to how many stacks they break down the shuffles and you can further estimate what a single deck looks like.

 

[boneuphtoner]

Do you really think if you bet $90 to $110 instead of $100 it's going to make a lot of difference? Do you really think if you stand on X indice at a range of 1.1 through .9 instead of 1 it makes much difference?

In a situation like that, absolutely not. In that situation, I wouldn't care. However, that isn't what concerns me. Take this example.....Say you have a six deck shoe, RC is +12 and there is somewhere between 2-3 decks left. And the dealer is showing an Ace up and asks for Insurance. What do you do? With FELT in a six decker, you are supposed to take insurance at a TC of +6....but if you were looking at thicker cards one might think that 2 decks are remaining when in fact there were closer to three....you would then have bought insurance at a real TC of +4 which would be a mistake! This has to impact performance in a meaningful way, right?

I have no idea how much you are paying for the cards, but for this thickness of cards question you have already hurt your bank enough.

No, the cards were really cheap (less than 50 cents a pack) and didn't come out of the bankroll anyway. Buying off of ebay is the best way I've seen to buy large quantities for cheap!

Unbalanced counts use a deck estimation technique also, they take an average. This is a far worse estimation then an imperfect human TC calculation. All one has to do is look at specific situations for unbalanced counts to see the indices and bets are sometimes quite off.

Yes, I have heard this mentioned before, and in QFIT's second edition of Modern Blackjack, you can see the actual indices for REKO at each depth, and it is amazing how much the optimal index changes over the course of a shoe. But the fact remains that even with perfect true counting, unbalanced strategies of similar level perform remarkably similar to their balanced counterparts. I participated in creating the UBZ-Open Source strategy on the sticky under card counting. Yes, my FELT strategy is stronger than that one but it also uses many more indexes, including a number of negative indexes which don't work well across all numbers of decks for unbalanced strategies (in other words you could come up with the negative indices for unbalanced strategies, but you couldn't use the across all numbers of decks). But when I simmed the strategies with comparable numbers of indexes (by reducing my FELT strategy to the same number of indices for UBZ), the performances were nearly the same using UBZ in running count mode (ok, the balanced strategy was about 1% stronger, close enough for me).

Having said all of this, I still prefer the feel of balanced counts. I can go back and forth between FELT and KO nearly seemlessly, and having used a balanced count for the past two years, whenever I temporarily use KO, I feel like I'm running blind. Yes, there are nice wong out points provided to me in a great chart by one of our kind fellow forum members, but I feel like wong-out information especially seems more natural in a balanced count.

On the OTHER hand, I see no reason why I should work harder at deck estimation and bother with true counting if the estimation errors I introduce weaken the system by 10% or more. Take away 10%, and I'm basically looking at what I would get with a true fudged version of KO.

So, the question is, do you guys think that estimation sloppiness on top of errors due to card thickness would hurt the system by more than 10%?

According to QFIT's book, being consistently 10% off in your estimation probably hurts you less than 5%, eyeballing the chart below suggests it may be closer to 3%. And I'm not THAT sloppy....I may be 10-20% off every once in awhile, but not consistently. So, do you think errors due to card thickness would hurt the system more than 5%-7%?

http://www.qfit.com/book/z93Chart.gif

 

[boneuphtoner]

If you had a friend standing behind you and counting the EXACT number of cards that were dealt/undealt, at the end of one year you would do BARELY better than if you were to just sloppily make quick estimates of the cards. The difference, as far as money won, would be almost unnoticeable.

Card counting is NOT rocket science. You're trying to build a spaceship that will take you to another star system & back. You need to stop wasting your time trying to find this "magic bullet". Pick ONE system, ANY system; and stick with it. Just try to do your best with that system, and the money WILL come.

Thanks for the encouragement. As I noted in my previous message, I think my errors due to general estimation sloppiness hurt my system's performance well under 3%. If folks think that errors due to card thickness will hurt me less than an additional 7% or so, I'm inclined to stick with my very powerful FELT count. I prefer the feel of balanced counts, and I can count level II tags as fast as level I tags.

 

[blackjack avenger]

Answered Your Own Question

I think my errors due to general estimation sloppiness hurt my system's performance well under 3%. If folks think that errors due to card thickness will hurt me less than an additional 7% or so, I'm inclined to stick with my very powerful FELT count. I prefer the feel of balanced counts, and I can count level II tags as fast as level I tags.

Card thickness differences do not hurt you that much

Why not use the count you prefer? Does it make sense to use a count you prefer less?

Run an optimal sim of a balanced count, then mix up the bets some to introduce betting errors, you won't see a big drop in SCORE.

Again you are talking about TC errors when an unbalanced count introduces the highest error rate based on deck averaging.

The real differences in counts appear when you consider tag values and the level of count.

:joker::whip:

 

[blackjack avenger]

Precision Not Necessary

Say you have a six deck shoe, RC is +12 and there is somewhere between 2-3 decks left. And the dealer is showing an Ace up and asks for Insurance. What do you do? With FELT in a six decker, you are supposed to take insurance at a TC of +6....but if you were looking at thicker cards one might think that 2 decks are remaining when in fact there were closer to three....you would then have bought insurance at a real TC of +4 which would be a mistake! This has to impact performance in a meaningful way, right?

Let's say you place max bets at TC3, so if you are at TC 3,4,5+ then being off on deck estimation does not matter because you are making the same bet, right?

Being off by 1 or 2 does not matter that much in indices right? Go over again the theory of lite indices or grouped indices.

Depends on the composition of your hand and if you are considering EV or RA indices on what actually the correct play is, right? So there is a lot of room for a range of correct indices.

If you arrive at TC by deck or half deck it does not really matter right?

So there is a lot of room for errors, for whatever reason.

Do you sometimes use betting camo? Then you are introducing errors yourself so why get wound up over card thickness errors?

If you introduce betting constraints in any form then you are again introducing error.

If you are comparing a similar in SCORE TC vs unbalanced system then the extra work in TC may not be worth it, especially since a human TC will be off and have some small effect. However, if the TC system is quite a bit stronger then the TC may be worth it.

I use halves and halves is so strong in applying a TC that it's worth it even if I pay a small cost due to human error.

As a general statement probably the level of count is the first consideration in how strong a count is, a secondary consideration is if its unbalanced or not.
Hmmmmm
Which class of counts is stonger?
high level unbalanced
or
lower level balanced
If they are similar in SCORE then which requires more effort?
I think higher level unbalanced would be stronger

It seems overall you are looking for ultimate precision when it does not exist.

:joker::whip:

 

[UK-21]

Card counting is NOT rocket science. You're trying to build a spaceship that will take you to another star system & back. You need to stop wasting your time trying to find this "magic bullet". Pick ONE system, ANY system; and stick with it. Just try to do your best with that system, and the money WILL come.

Hear hear. . . .

The margin of error when estimating the number of decks in the tray will make very little difference to the number of £££/$$$s moving back and forth across the felt, unless your betting ramp is virtually vertical ! ! ! (and even then I'm sure it would average out over time).

 

[Deathclutch]

http://i219.photobucket.com/albums/cc145/boneuphtoner/ddstacks.jpg

I think this photo says it all. See attached.

For years, I went back and forth between unbalanced and balanced counts. I've tried nearly all of them - level II and Level I. I always ended up preferring the feel of balanced counts overall, with FELT that I switched to last year feeling the most natural for me. Compared to playing with an unbalanced count, I feel that it is so powerful to instantly know whether you have the advantage or not with a balanced count. Indexes like 16 v. 10 and 12 v. 4 are also extremely easy and the answer is always obvious. I never found the math of dividing by whole decks and flooring the result to be difficult either. Neither was dividing by fractions like I do with pitch games...piece of cake right? However, what I have always felt insecure about were my deck estimation skills. After thinking about it at length and after reading Modern Blackjack, scanning the effects of estimation error, I put my mind to rest....the effects of 10%deck estimation error on every calculation makes very little difference...at least it put my mind at ease for the moment. Like any decent counter in training who uses a balanced count, I ordered a bunch of used casino playing cards off of ebay. I followed the advice of several reputable authors, and made a bunch of different stacks to try to fine tune my visual acuity. The problem is, I began to perceive differences in the heights of stacks of cards from the different casinos. So, to test this, I counted 104 cards from each of the three different casino cards that I received. The results are in the attached photo. Now, I don't think I have that great visual acuity, but even I can see these differences! The one from Ceasar's in particular looks to my eye to be nearly a half deck taller than the others!

To me, this is confirmation that my insecurity about deck estimation was warranted all along. At this point, I'm seriously thinking of switching back to an unbalanced count. What do you guys think? In my mind, for those of you who estimate decks to the nearest half or quarter MUST be aware of these differences to have reasonable accuracy. I use half deck resolution for pitch games, and that is looking really shaky to me right now. Let alone quarter deck resolution.

Since the difference between being able to estimate to whole decks and half decks is negligible in straight counting you won't get a ton of difference in your results if you're off a bit with the estimation. You can begin to sweat it a bit more once you need your deck estimation for other skills.