A Cautionary Tale

[blackjack avenger]

Are Many Missing Something?

If a card counter starts playing and if he takes a large hit, even if he is still winning, retires? Then he would almost certainly stop playing at some point!:joker::whip: This is what the OP is doing.

Sagefrog is on the right track.

What was it?
24 hands in a row lost?
one hand won
and then 19 hands in a row lost?
Improbable.

To lose half of bank ($25.000 lost) in 4 days? max bet $300? How soon a max bet? He only played 22 hours. Improbable if playing correctly.

The OP I believe mentioned a pseudo progression?

Red flags are everywhere!:joker::whip:

 

[caramel6]

If a card counter starts playing and if he takes a large hit, even if he is still winning, retires? Then he would almost certainly stop playing at some point!:joker::whip: This is what the OP is doing.

Sagefrog is on the right track.

What was it?
24 hands in a row lost?
one hand won
and then 19 hands in a row lost?
Improbable.

To lose half of bank ($25.000 lost) in 4 days? max bet $300? How soon a max bet? He only played 22 hours. Improbable if playing correctly.

The OP I believe mentioned a pseudo progression?

Red flags are everywhere!:joker::whip:

agree,looks like he did not do all right, pity, so much money lost in a such short period of time.

Better win 100 every day, hoping that a day when 12000 will be lost (according to math) will be as distant as possible. May be not in a lifetime...

Curious , where this chap who won 12000, dissapeared? wonder why his success did not bring many comments?

 

[aslan]

Why not to end a session when winning certain amounts of units from the casino?

A fine idea if you don't start off losing some sessions and never regain your footing. Many a player has gone broke chasing the elusive "quit while you're ahead" wisdom.

 

[aslan]

Sagefrog is on the right track.

What was it?
24 hands in a row lost?
one hand won
and then 19 hands in a row lost?
Improbable.

Tarzan had some posts a month or so ago about the reason for the house's ability to win large numbers of hands in a row in certain situations. He discovered that when the middle cards (I think he was referring to the 6, 7, 8, and 9's) are 50% or more depleted, this is exactly what can happen. He designed a count, impossible for the average counter, that tracked low cards, middle cards, high cards and aces. When the middle cards get depleted, he's out of there! I suppose the middle cards could be deleted in a positive count when accompanied by a lot of little cards also falling, which could account for large losses at $300 max bet. Anyway, all agree that it did happen--I'm just trying to put some flesh on the bones as to why it can happen, and how it might be detected in advance.

 

[FLASH1296]

Yes, I have seen Tarzan in action — in his natural habitat — prowling the boardwalk.

Think of it this way — The dealer's bust chances dry up when the middle cards are severely depleted.

 

[aslan]

Yes, I have seen Tarzan in action — in his natural habitat — prowling the boardwalk.

Think of it this way — The dealer's bust chances dry up when the middle cards are severely depleted.

Right. Assuming all the middle cards were gone: Two tens, fine for the dealer. Two or three small cards and a ten, fine. The only bummer for the dealer is one small card and two tens. So yes, 2 out of three ain't bad.

 

[Sucker]

Anyway, all agree that it did happen--

Not everyone agrees that it did happen.

 

[caramel6]

Tarzan had some posts a month or so ago about the reason for the house's ability to win large numbers of hands in a row in certain situations. He discovered that when the middle cards (I think he was referring to the 6, 7, 8, and 9's) are 50% or more depleted, this is exactly what can happen. He designed a count, impossible for the average counter, that tracked low cards, middle cards, high cards and aces. When the middle cards get depleted, he's out of there! I suppose the middle cards could be deleted in a positive count when accompanied by a lot of little cards also falling, which could account for large losses at $300 max bet. Anyway, all agree that it did happen--I'm just trying to put some flesh on the bones as to why it can happen, and how it might be detected in advance.

o,key, average counter cannot count that-however, if 2,3 times on a high counts still not joy-just stop! I mean of course, a real BJ warrior should go ahead with a count, but here is ROR!

 

[caramel6]

A fine idea if you don't start off losing some sessions and never regain your footing. Many a player has gone broke chasing the elusive "quit while you're ahead" wisdom.

Exactly, so that ,s why it should be a stop -loss. Most important is to have much more winning sessions than losing sessions, isn,t it?

 

[k_c]

Tarzan had some posts a month or so ago about the reason for the house's ability to win large numbers of hands in a row in certain situations. He discovered that when the middle cards (I think he was referring to the 6, 7, 8, and 9's) are 50% or more depleted, this is exactly what can happen. He designed a count, impossible for the average counter, that tracked low cards, middle cards, high cards and aces. When the middle cards get depleted, he's out of there! I suppose the middle cards could be deleted in a positive count when accompanied by a lot of little cards also falling, which could account for large losses at $300 max bet. Anyway, all agree that it did happen--I'm just trying to put some flesh on the bones as to why it can happen, and how it might be detected in advance.

Years ago when I was playing I hit a streak that seemed to defy likely probabilities. It seemed like whatever could go wrong did. How could I bust on 12 so often and finally when I didn't it was because I was dealt a 2 or 3.

Previous to this I had done fine using traditional HiLo. I just accepted that plus counts were good and bet accordingly and used some indices.

Anyway this streak seemed to have no end and I decided maybe it's because something's rotten in Denmark. I decided to count cards using 2 groups: (1,2,3,4,5,10) & (6,7,8,9). Although I only had hand calculations I reasoned that if all (6,7,8,9) were removed then player would have an advantage using best strategy and also if all (1,2,3,4,5,10) were removed player would also have an advantage using best strategy, although I had no way to completely confirm this.

Miraculously as soon as I adopted this count the streak ended and it seemed for a while that I couldn't lose. Could it be that no matter which way the count swung player had an advantage? Well the winning eventually ended so I figured the answer was probably no.

This was how I started analyzing blackjack.

It turns out player advantage with all (6,7,8,9) removed is ~ +7%-8% and with all (1,2,3,4,5,10) removed ~ +19%-20$ using best strategy so my premise was right in that at both extremes of the count, assuming equally balanced distribution of ranks, that player has the advantage using best strategy.

I thought this might be an interesting side note to how medium cards might affect streakiness. Their absence or too many of them could cause an unknowing player to make EV reducing misplays.

 

[sagefr0g]

Exactly, so that ,s why it should be a stop -loss. Most important is to have much more winning sessions than losing sessions, isn,t it?

the key is caramel, that you need an edge to make sure that your money will grow long term, with out an edge one would be with out control over the vagaries of luck.

having a stop loss is no guarantee that the next session, or some combination of sessions on down the road wont wipe out one's gains that were made by quitting when ahead in the past. if you have a stop loss, then it could just be a series of stop loss's values that over time add up to be enough to wipe out any gains you may have made on past stop wins. all we can really know that can help us control the situation is the nature of the edge that we face, is it disadvantageous or advantageous, and the degrees thereof which can help us with risk of ruin considerations.

if you make plays that have an edge, then you have an expectation that as those plays are made over time that the culmination of the results of those plays will bring you close to expectation when one adds up the results of a series of sessions. the same is true in a negative sense if you don't have an edge.

the thing i like about stop loss and stop win actions is that you can look at those results and determine where you are at theoretically with respect to expectation and ranges of standard deviation. that way you can have an idea of how lucky, unlucky or if things seem to be working out as expected. that way if things don't seem correct one can think about what might need to be done to either fix or improve the situation.

 

[caramel6]

the key is caramel, that you need an edge to make sure that your money will grow long term, with out an edge one would be with out control over the vagaries of luck.

having a stop loss is no guarantee that the next session, or some combination of sessions on down the road wont wipe out one's gains that were made by quitting when ahead in the past. if you have a stop loss, then it could just be a series of stop loss's values that over time add up to be enough to wipe out any gains you may have made on past stop wins. all we can really know that can help us control the situation is the nature of the edge that we face, is it disadvantageous or advantageous, and the degrees thereof which can help us with risk of ruin considerations.

if you make plays that have an edge, then you have an expectation that as those plays are made over time that the culmination of the results of those plays will bring you close to expectation when one adds up the results of a series of sessions. the same is true in a negative sense if you don't have an edge.

the thing i like about stop loss and stop win actions is that you can look at those results and determine where you are at theoretically with respect to expectation and ranges of standard deviation. that way you can have an idea of how lucky, unlucky or if things seem to be working out as expected. that way if things don't seem correct one can think about what might need to be done to either fix or improve the situation.

Fully agree.I try to count as much as possible (if circimstances allows). But here is a guy-no counting, everyday 100 , 12000 not easy, but possible.O,key, it is just a positive variance, isn,t it?

 

[sagefr0g]

Fully agree.I try to count as much as possible (if circimstances allows). But here is a guy-no counting, everyday 100 , 12000 not easy, but possible.O,key, it is just a positive variance, isn,t it?

exactly, since he isn't counting or using an advantage technique that gives him a positive expected value, he is in fact playing to the house edge.

his concern is not that in the past he won $100 a day or that over time he has realized $12,000.
his concern is that each time he makes a play he can expect a range of results influenced by the house edge he is playing against, while his actions support the house's edge.

 

[Deathclutch]

Fully agree.I try to count as much as possible (if circimstances allows). But here is a guy-no counting, everyday 100 , 12000 not easy, but possible.O,key, it is just a positive variance, isn,t it?

Here is something to take a look at when discussing stop losses.

http://www.advantageplayer.com/blackjack/forums/bj-main/webbbs.cgi?read=18995 (Archive copy)

 

[assume_R]

It turns out player advantage with all (6,7,8,9) removed is ~ +7%-8% and with all (1,2,3,4,5,10) removed ~ +19%-20$ using best strategy so my premise was right in that at both extremes of the count, assuming equally balanced distribution of ranks, that player has the advantage using best strategy.

k_c, are you talking about (eor of 6) + (eor of 7) + ..., times the # of cards removed??

 

[aslan]

Exactly, so that ,s why it should be a stop -loss. Most important is to have much more winning sessions than losing sessions, isn,t it?

Basic strategy is a losing proposition. You cannot ensure more winning sessions than losing sessions by use of stop losses or any other form of money management.

 

[blackjack avenger]

Losing 24 Hands in a Row?

What is the % chance of losing a BJ hand?
What is that to the 24th power?
Very improbable he lost 24 hands in a row!:joker::whip:

Is the 6,7,8 discussion an attempt to explain bad play?
You will have losing streaks! If you overbet you can lose a large % of your bank in a short period of time.

The OP stated he used a progression.
He also apparently overbet. If you resize your bets but bet 2Kelly your bankroll will shrink.

His sometimes underbetting and sometimes overbetting (his pseudo) progression on high counts raised his NO so he increased his variance. He might have been unfortunate and lost most of the hands where he overbet and won most of the hands where he underbet during those high TCs.:joker::whip:

 

[k_c]

k_c, are you talking about (eor of 6) + (eor of 7) + ..., times the # of cards removed??

No, all I did was to tag (6,7,8,9)=+2 and (1,2,3,4,5,10)=-1 relative to what remains in the shoe. Largest theoretical TCs possible would be +104, meaning shoe consisted of all (6,7,8,9), or -52, meaning shoe consisted of all (1,2,3,4,5,10). In either case if ranks are balanced and best strategy is used then player has a good advantage.

At the time I was in a seemingly endless negative streak. I figured at minimum any count would be better than none. Luckily it worked very well for a while.

It may not have been a great idea, but it was interesting and I learned from doing it.

 

[iCountNTrack]

Years ago when I was playing I hit a streak that seemed to defy likely probabilities. It seemed like whatever could go wrong did. How could I bust on 12 so often and finally when I didn't it was because I was dealt a 2 or 3.

Previous to this I had done fine using traditional HiLo. I just accepted that plus counts were good and bet accordingly and used some indices.

Anyway this streak seemed to have no end and I decided maybe it's because something's rotten in Denmark. I decided to count cards using 2 groups: (1,2,3,4,5,10) & (6,7,8,9). Although I only had hand calculations I reasoned that if all (6,7,8,9) were removed then player would have an advantage using best strategy and also if all (1,2,3,4,5,10) were removed player would also have an advantage using best strategy, although I had no way to completely confirm this.

Miraculously as soon as I adopted this count the streak ended and it seemed for a while that I couldn't lose. Could it be that no matter which way the count swung player had an advantage? Well the winning eventually ended so I figured the answer was probably no.

This was how I started analyzing blackjack.

It turns out player advantage with all (6,7,8,9) removed is ~ +7%-8% and with all (1,2,3,4,5,10) removed ~ +19%-20$ using best strategy so my premise was right in that at both extremes of the count, assuming equally balanced distribution of ranks, that player has the advantage using best strategy.

I thought this might be an interesting side note to how medium cards might affect streakiness. Their absence or too many of them could cause an unknowing player to make EV reducing misplays.

What is also interesting is how the effects of removals of the neutral cards changes drastically depending on the deck composition. The following shows the normalized EORs of 6, 7, 8, 9 for two different. I use normalized EORs to eliminate any effects from the change in the size of the deck while doing the comparison. The EORs were calculated using combinatorial analysis using the Igor/CDCA hybrid

EOR_6 (Full deck)=+ 0.4429%
EOR_7 (Full deck)=+ 0.3031%
EOR_8 (Full deck)=+ 0. 0310%
EOR_9 (Full deck)=- 0. 1457%


EOR_6 (Zero 7, 8, 9)=+ 0.7013%
EOR_7 (Zero 6, 8, 9)=+ 0.6600%
EOR_8 (Zero 6, 7, 9)=+ 0. 4928%
EOR_9 (Zero 6, 7, 8)=- 0. 1028%


What is interesting is how the removals of a neutral cards is cooperatively working to drastically enhance the EOR of the remaining neutral cards, so while the 8 is typically treated as a neutral card on average there are certain deck compositions where its removal is very valuable.

 

[k_c]

What is also interesting is how the effects of removals of the neutral cards changes drastically depending on the deck composition. The following shows the normalized EORs of 6, 7, 8, 9 for two different. I use normalized EORs to eliminate any effects from the change in the size of the deck while doing the comparison. The EORs were calculated using combinatorial analysis using the Igor/CDCA hybrid

EOR_6 (Full deck)=+ 0.4429%
EOR_7 (Full deck)=+ 0.3031%
EOR_8 (Full deck)=+ 0. 0310%
EOR_9 (Full deck)=- 0. 1457%


EOR_6 (Zero 7, 8, 9)=+ 0.7013%
EOR_7 (Zero 6, 8, 9)=+ 0.6600%
EOR_8 (Zero 6, 7, 9)=+ 0. 4928%
EOR_9 (Zero 6, 7, 8)=- 0. 1028%


What is interesting is how the removals of a neutral cards is cooperatively working to drastically enhance the EOR of the remaining neutral cards, so while the 8 is typically treated as a neutral card on average there are certain deck compositions where its removal is very valuable.

I guess my first question is, "What are normalized EORs and how do they eliminate effects from changing deck size?" (My knowledge of statistical methods is approxiamtely 0 . )