Help.

#1
[note: x-posted in advanced forum for exposure]

I'm a member of a 3-member team of card counters. After training for the better part of a year, we recently began making actual casino runs. The results have been perplexing and I was hoping someone from the counting community might be able to shed some light on our predicament.



Originally we started with a $6,000 bankroll. So far we have made five casino trips and we've lost money four of the five trips. Each time we've lost, we lost around $500-$800 dollars. On the trip that we won, we won a little over $1,000. I know that losses are an expected part of counting, but losing 4 out of 5 times seems excessive. Each trip we've spent roughly 24 hours in the casino counting. During each trip, we've fluctuated wildly between winning and losing money. At one moment we'll be up $500 or so, then we'll be down $700, make the $700 back and get up a couple hundred, and then, more often than not, lose it again. The way I've heard counting described by other counters, the profits seem slow but sure. This is what perplexes me the most: How could we have logged so many hours in the casino and still be losing money unless we're doing something wrong? How could we spend so many hours in the casino and not be completely bankrupt unless we're doing something right?



Here's what we've been doing: At the beginning of the night we each take $2000 of the bankroll. Inside the casino we each find separate tables to count at. If a shoe is very hot, we may signal another member in, but for the most part we operate alone and total are wins/losses at the end of the night. Our betting increments for playing a single hand have been $20 for a true count of +1, $40 for +2, $60 for +3 and $80 for any count +4 or greater. If we're playing two hands we use $15 increments. We usually backcount until a shoe is positive and then join in, but sometimes we'll just sit down and place the minimum $10 bet until the shoe becomes positive. Whenever the count drops below a true count of negative two we leave the table. When calculating the true count, we subtract 1 to account for the house's edge.



Here are the rules of the casino we count at:

Blackjack pays 1.5 to 1

Resplit up to four hands

No resplitting aces

Double down on any two cards

Dealer peeks for aces and tens

No surrender

Minimum bet is $10



The casino places the cut card 4 - 4 ½ decks deep.



The only thing I can think of that may be the problem is that usually there are 4 to 6 other players at the table with us. Could this account for the fluctuations we've seen? Can anyone think of anything else that might account for our losses?



We've tested and retested our counting under extreme conditions so we are certain that the problem is not lack of skill. Any help will be greatly appreciated.
 

Kasi

Well-Known Member
#3
soft18 said:
[note: x-posted in advanced forum for exposure]We've tested and retested our counting under extreme conditions so we are certain that the problem is not lack of skill. Any help will be greatly appreciated.
I know nothing but here are a few thoughts anyway.

Rarely is card-counting skill the culprit if you ask me. It's always betting how much and when.

Sometimes you play separately spreading 1-8? Sometimes you call in a BP? If and when you do does he know exactly what to bet? Sometimes you spread to multiple hands with a different increment?

Might mean nothing, might be something to consider, as each method will have a different EV, ROR etc. You know what's expected from each?

When you say 24 hours, is that per person or total hours for team? Session ROR's is something to consider too. Obviously if you had 100 guys on your team, you wouldn't want each of them to wander off with $60, tell them to play for 4 hours, and report back with winnings. Not many will.

Get an estimate of how many hands actually played each time from each player since it all ultimately comes down to per-hand crap anyway. More players, fewer hands played kind of thing.

Pick one way you guys bet, maybe the one you bet most often, - you at least know what to expect from that? ROR, etc?

Ever sim anything or use something so you have something to measure against?

How detailed are your logs? It may be possible to figure things out retroactively, to some degree anyway.

Going forward, crack that whip - play one way to one plan so it's easy to measure. Keep it simple. Or at least make sure any deviations to different plans are noted.

Maybe shorten sessions, even maybe giving them fewer units to play with because they are not allowed to play for more than an hour before reporting back, requiring more frequent assesments and more liklihood of identifying what maybe going wrong sooner, if anything.

Whatever. Never been on a team in the first place lol. Just stuff to think about maybe lol. Never said it was easy.
 
#4
Not a good game

and you are in real life, not the movie 21.

I am not a proponent of 6 or more deck games unless the rules are great and the pen is also great.

As to your team play there are much better ways, not impressed with your tactics at all.

Re-think.

CP
 

EasyRhino

Well-Known Member
#5
soft18 said:
The way I've heard counting described by other counters, the profits seem slow but sure.

Slow and unsure would be much more apt.

24 hour sessions? I really hope you don't mean all players are playing for 24 hours, or if you are, it's at a place like AC where no one gives a damn.

The bet ramp seems okay. And wonging would seem to be necessary.

It's very very likely this is just normal fluctuation. It's very easy to win or lose a several hundred dollars if your max bet is $80 (you can do that one one hand). But, there's the things you should check for on your game:

1) Mistakes
2) Tipping
3) Playing negative counts (I'd probably try harder to avoid playing in neg counts)
4) Drinking

Also, be careful about calling another player in. Two players at the same table are going to have results linked by covariance. It's effectively like playing 2-4 spots all by yourself. You'd need to drop the bet per hand even lower to have it properly risk adjusted. I don't know precisely how much for 4 hands. Maybe 50% of single hand. Maybe less. It wouldn't necessarily be the best use of time.
 

golfnut101

Well-Known Member
#6
Here's what we've been doing: At the beginning of the night we each take $2000 of the bankroll. Inside the casino we each find separate tables to count at. If a shoe is very hot, we may signal another member in, but for the most part we operate alone and total are wins/losses at the end of the night. Our betting increments for playing a single hand have been $20 for a true count of +1, $40 for +2, $60 for +3 and $80 for any count +4 or greater. If we're playing two hands we use $15 increments. We usually backcount until a shoe is positive and then join in, but sometimes we'll just sit down and place the minimum $10 bet until the shoe becomes positive. Whenever the count drops below a true count of negative two we leave the table. When calculating the true count, we subtract 1 to account for the house's edge.



Here are the rules of the casino we count at:

Blackjack pays 1.5 to 1

Resplit up to four hands

No resplitting aces

Double down on any two cards

Dealer peeks for aces and tens

No surrender

Minimum bet is $10



The casino places the cut card 4 - 4 ½ decks deep.
Over betting your br(80.00 big bet), along with poor(4-4.5) to avg pen=disaster.
 

Sonny

Well-Known Member
#7
Read through the FAQ at the top of this forum. Specifically, read the sections about variance and team betting strategies. If you think that losing 4 out of 5 sessions is excessive then you won’t last long. As others have mentioned, you are also overbetting your bankroll (playing with a very high risk of ruin), playing a bad/mediocre game and not playing optimally. You spent a year developing your card counting skills, which is a very good first step. Now you need to spend the next year learning how to play. Learn how to maximize your EV, minimize your standard deviation, calculate your bets to maintain a reasonable RoR, play efficiently as a team and have realistic expectation of how long it will take to reach the long run using the strategies that you chose. There is a lot to learn and you would need a lot of pure luck to succeed without it.

You can search this website for some of the terms I mentioned above. Read Don Schlesinger’s Blackjack Attack. Also read every free article on these websites:

http://www.blackjackforumonline.com/content/TOClibrary.html
http://www.bj21.com/
http://www.bjmath.com/bjmath/toc.htm (Archive copy)

By that time you should have most of the answers you need to get started. And, of course, keep posting here if you have any questions. :)

-Sonny-
 

moo321

Well-Known Member
#8
I don't think you're over-betting your bankroll IF you were only playing a back-counting game. $6k is 600 $10 units.

However, you cannot play-all with a 1-8 spread with only 600 units; even if you leave in the negative counts, you'll kill yourself with the variance. Switch to back-counting only, and you'll see much smaller variance. This game is only really beatable with back-counting given the penetration and your bankroll.
 

callipygian

Well-Known Member
#11
First of all, I disagree with most of the posters in that I don't think this is a horrible game (assuming you're on a 6 deck shoe, which you actually never specified). 75% penetration is OK, 66% is bad, but that just means you have to risk higher bet spreads (= higher variance and heat) for the same EV. I also think people on this board are more cautious about bankroll sizes because they play more seriously than most people (and thus, demand lower ROR's).

I do agree it's unlikely you're playing with positive expectation, though, based on your bet spreads and patterns.

But your main question is about variance, not EV. (Thanks to Sonny who walked me through these calculations the first time)

5 trips x 3 people x 24 hrs play x 60 hands/hr = 21,600 hands

With your bet spread, your variance is likely 3 units. You have an SD of 3*sqrt(21600) = 440 units. If you have a slim advantage over the house (+0.5%), your EV is +0.005*21600 = +110 units and you have a 68% chance of being between -330 and +550 units, and a 95% chance of being between -770 and +880 units. If you're not playing with an advantage over the house (-0.5%), your EV is -0.005*21600 = -110 units and you have a 68% chance of being between -550 and +330 units, and a 95% chance of being between -880 and +770 units.

You look like you're at -150 units, which is well within one standard deviation of your EV ->whether you are playing with an advantage or not<-.

You'd need to lose 330 units (-$3,300) to honestly question whether you're doing it right, and be down by 770 units (-$7,700) to be reasonably sure you're not. Everything between -$3,300 and +3,300 is anyone's guess.
 
#12
callipygian said:
First of all, I disagree with most of the posters in that I don't think this is a horrible game (assuming you're on a 6 deck shoe, which you actually never specified). 75% penetration is OK, 66% is bad, but that just means you have to risk higher bet spreads (= higher variance and heat) for the same EV. I also think people on this board are more cautious about bankroll sizes because they play more seriously than most people (and thus, demand lower ROR's).

I do agree it's unlikely you're playing with positive expectation, though, based on your bet spreads and patterns.

But your main question is about variance, not EV. (Thanks to Sonny who walked me through these calculations the first time)

5 trips x 3 people x 24 hrs play x 60 hands/hr = 21,600 hands

With your bet spread, your variance is likely 3 units. You have an SD of 3*sqrt(21600) = 440 units. If you have a slim advantage over the house (+0.5%), your EV is +0.005*21600 = +110 units and you have a 68% chance of being between -330 and +550 units, and a 95% chance of being between -770 and +880 units. If you're not playing with an advantage over the house (-0.5%), your EV is -0.005*21600 = -110 units and you have a 68% chance of being between -550 and +330 units, and a 95% chance of being between -880 and +770 units.

You look like you're at -150 units, which is well within one standard deviation of your EV ->whether you are playing with an advantage or not<-.

You'd need to lose 330 units (-$3,300) to honestly question whether you're doing it right, and be down by 770 units (-$7,700) to be reasonably sure you're not. Everything between -$3,300 and +3,300 is anyone's guess.

The sad truth.....

its a lot to spend just to wonder if your doing something wrong

and a lot to lose to know your doing it wrong
 

Kasi

Well-Known Member
#13
callipygian said:
Thanks to Sonny who walked me through these calculations the first time
21,600 hands

If you have a slim advantage over the house (+0.5%), your EV is +0.005*21600 = +110 units ...
You look like you're at -150 units,
I'm guessing you're thinking the 110 units are minimum units? Anyway, the ev would be that * average units bet per hand. Since he's spreading, his avg bet will be greater than his min unit. Like , on average, maybe he'll have 2 min units in an avg bet.

If it's the thread I'm thinking of, I asked Sonny about that in what he was showing you but he never answered. So I could be wrong, then as now.
 

callipygian

Well-Known Member
#14
Kasi said:
I'm guessing you're thinking the 110 units are minimum units? Anyway, the ev would be that * average units bet per hand. Since he's spreading, his avg bet will be greater than his min unit.
I can't remember Sonny's post with absolute clarity, but I always express things in minimum units. There's no need to multiply by the average bet because that's already used in a calculation of the EV. The EV is Sum{(Probability of count i) * (EV at count i) * (bet at count i)}.
 

Kasi

Well-Known Member
#15
callipygian said:
There's no need to multiply by the average bet because that's already used in a calculation of the EV. The EV is Sum{(Probability of count i) * (EV at count i) * (bet at count i)}.
Well I assumed that +.005 was an overall advantage for the game.

Anyway, if you think this guy's EV after 21600 hands enjoying an overall advantage for the game of 0.005 is 110 units * a $10 unit, or $1110, I just don't see how it's possible if he's spreading 1-10 or whatever it was.

Like you say the average bet is used in the calc of EV and the avg bet is almost always greater than the min bet.

Anyway I think that's how it's usually done. If u have Don's book, check it out.

But maybe we all have slightly different ways of doing the same thing.
 

Harman

Well-Known Member
#16
Grr, i hate it when people tell us all this stuff, and then never tell us what happens.. What did you change soft18?? Did you change games or bet spreads, or was it just variance that explained your losses, come on, finish your thread man!!
 

eandre

Well-Known Member
#17
soft 18

Why a team? Maybe you are not losing, but your team members are. The team is only as good as the weakest player and if you are not the weakest, divide the money up and go out on your own.
 

Sonny

Well-Known Member
#18
Harman said:
Grr, i hate it when people tell us all this stuff, and then never tell us what happens...come on, finish your thread man!!
His first day was also his last day:

soft18
Member
Last Activity: May 6th, 2008 01:33 PM


I guess he was another casualty of the movie 21.

-Sonny-
 
#20
mdlbj said:
So sad
:cry:
I see what you did there.

Sonny said:
His first day was also his last day:

soft18
Member
Last Activity: May 6th, 2008 01:33 PM

I guess he was another casualty of the movie 21.

-Sonny-
While I cannot say the same for the book, the movie has had no relevance in my life in any way- counting or otherwise. My teammate initially approached me after reading the book (long before the movie was known) and it all began. Still fighting the good fight.

Neither of you probably care, but I have created an update thread if you would like to see it: http://www.blackjackinfo.com/bb/showthread.php?t=10482

Late though it is, it's the least I can do.
 
Top