WTF is going on

rollem411

Well-Known Member
#1
Alright, to summarize, here is what I am attempting. On this site I've been using the Blackjack strategy trainer to practice my counting and see if it is actually working with bet spreading.

I've been using the 8D, S17, NS, DOA options and obviously took off the strategy helper so I can use indices. I also stop playing each hand at around 70%-80% penetration because in AC you will rarely see 7/8.

I know the long run is where you see results and in shoe games it comes around 10,000-15,000 hands (read it in another post), but I have played about 2,000 hands so far and it does not seem like I'm getting any results worth my efforts. I build a BR over a few hundred hands or so and then when the count hits my favor it drops significantly. It almost seems like betting in the higher counts is hurting me. I am going to keep playing to see if it is just bad luck at this point, but I would expect to see a little something except for a loss of BR--which is what I am at right now.

I'm beginning to think I'm missing something. I'm really starting to get p*ssed off and I am not even playing with real money.
Is the BJ trainer on this site completely random?
 

k_c

Well-Known Member
#2
rollem411 said:
Alright, to summarize, here is what I am attempting. On this site I've been using the Blackjack strategy trainer to practice my counting and see if it is actually working with bet spreading.

I've been using the 8D, S17, NS, DOA options and obviously took off the strategy helper so I can use indices. I also stop playing each hand at around 70%-80% penetration because in AC you will rarely see 7/8.

I know the long run is where you see results and in shoe games it comes around 10,000-15,000 hands (read it in another post), but I have played about 2,000 hands so far and it does not seem like I'm getting any results worth my efforts. I build a BR over a few hundred hands or so and then when the count hits my favor it drops significantly. It almost seems like betting in the higher counts is hurting me. I am going to keep playing to see if it is just bad luck at this point, but I would expect to see a little something except for a loss of BR--which is what I am at right now.

I'm beginning to think I'm missing something. I'm really starting to get p*ssed off and I am not even playing with real money.
Is the BJ trainer on this site completely random?
I can't speak for the trainer on this site, but I can tell you of my experience with my own program. It's the one I've mentioned before that is not yet available. It has the advantage of knowing the overall expected value of each deal using basic total dependent strategy before any hand has been dealt and can determine the right total dependent play after the hand and each card is dealt. I've had really bad streaks even knowing this. I've had some really good ones as well though.

Because I have on older, not too fast computer I have used it mainly for single deck. I mentioned that it can display 2 counts at a time. Sometimes a count shows negative EV when the actual EV is appreciably positive and sometimes the reverse is true. I think this happens more for single deck. I have used it for double deck on my computer some and it seems to happen less. I have used it for 6 decks on a friend's faster computer, but I was mainly interested on how fast it ran and it was pretty fast.

I even used it to test Licentia's system, just ignoring the math. (You can bypass computing overall EV before each hand. It will still compute the EVs for each hand and dealt card and that doesn't take much time.) I actually was up about 50 units initially, but kept playing and began losing. After about 3 hours I was down well over 100 units and quit. I'm sure if you consulted the math in the program, you would see that there is no correlation to actual advantage and when his system says to raise his bet.

Anyway, and I'm sure this has been mentioned on this site before, you can lose and lose some more even though you have the mathematical edge. I've had some winning streaks in actual casino play as well as some losing ones. The losing ones (and one in particualr) tend to make you think, "something may be rotten in Denmark," and put me on a path to try to see EXACTLY what is happening mathematically.

k_c
 

jack.jackson

Well-Known Member
#3
k_c said:
I even used it to test Licentia's system, just ignoring the math. (You can bypass computing overall EV before each hand. It will still compute the EVs for each hand and dealt card and that doesn't take much time.) I actually was up about 50 units initially, but kept playing and began losing. After about 3 hours I was down well over 100 units and quit. I'm sure if you consulted the math in the program, you would see that there is no correlation to actual advantage and when his system says to raise his bet.
k_c,
If you ever feel like going out on a limb, could you give my system a try. Maybe an hour or two. Thats if you can make heads or tails of it, of course. It can be used with any counting system. And as I mentioned, only if you feel like it.(2DPreferred)

See:Any successful progression players
 

jack.jackson

Well-Known Member
#4
rollem411 said:
Alright, to summarize, here is what I am attempting. On this site I've been using the Blackjack strategy trainer to practice my counting and see if it is actually working with bet spreading.

I've been using the 8D, S17, NS, DOA options and obviously took off the strategy helper so I can use indices. I also stop playing each hand at around 70%-80% penetration because in AC you will rarely see 7/8.

I know the long run is where you see results and in shoe games it comes around 10,000-15,000 hands (read it in another post), but I have played about 2,000 hands so far and it does not seem like I'm getting any results worth my efforts. I build a BR over a few hundred hands or so and then when the count hits my favor it drops significantly. It almost seems like betting in the higher counts is hurting me. I am going to keep playing to see if it is just bad luck at this point, but I would expect to see a little something except for a loss of BR--which is what I am at right now.

I'm beginning to think I'm missing something. I'm really starting to get p*ssed off and I am not even playing with real money.
Is the BJ trainer on this site completely random?
What system ya playin? What are ya spreads? At what count ya raisin your bets? Are you capable of determining the TC on the trainer?
Sounds like you could be overbetting. Also keep in mind these game are extremely tough to beat in a play-all situation.
 

Kasi

Well-Known Member
#5
k_c said:
It has the advantage of knowing the overall expected value of each deal using basic total dependent strategy before any hand has been dealt and can determine the right total dependent play after the hand and each card is dealt. I've had really bad streaks even knowing this. I've had some really good ones as well though.k_c
I was just wondering if you think that playing at a greater advantage somehow lessens your ROR. I'm kinda thinking all it does is make you win faster.
 

Kasi

Well-Known Member
#6
rollem411 said:
I'm beginning to think I'm missing something.
Probably not lol.

But you may want to take advantage of the practice to determine how you measure just how bad your luck is.

You'll probably find what you're experiencing is quite common.
 

rollem411

Well-Known Member
#7
jack said:
What system ya playin? What are ya spreads? At what count ya raisin your bets? Are you capable of determining the TC on the trainer?
Sounds like you could be overbetting. Also keep in mind these game are extremely tough to beat in a play-all situation.
Well, I was wonging out as I would in TC of -3 and beyond, but it seemed like I wasn't losing my money here...it seemed to fluctuate evenely.

I posted in other threads as to how I should spread my bets, but here it is:

0,+1 = 10
+2 = 20
+3 = 40
+4 = 60
+5 = 75
>=+6 = 100

I know for sure my TC is correct, but I might be overbetting/underbetting at times to try and camo. the counting as I would in the real casino (keep in mind I am trying to keep the atmosphere as real as possible). I was first going slow and making sure the count was correct and making sure I was making the correct index plays. Now I am trying to make it more realistic and go at the pass of dealer in the casino, but I am confident in my ability.

The thing is, at certain times I may overbet/underbet, but by only a half-unit, which in this case would be $5...maybe $10 at most. Would that be enough to screw up the long run efficiency?
 

rollem411

Well-Known Member
#8
k_c said:
I can't speak for the trainer on this site, but I can tell you of my experience with my own program. It's the one I've mentioned before that is not yet available. It has the advantage of knowing the overall expected value of each deal using basic total dependent strategy before any hand has been dealt and can determine the right total dependent play after the hand and each card is dealt. I've had really bad streaks even knowing this. I've had some really good ones as well though.
k_c
K_c, I am using the indices at hand and know the correct move every single time...it just seems as if I am not coming out on top. As you say it could just be that I am having bad luck, but it just sucks that it is happening when I am practicing right before I am about to play the real deal.
 

k_c

Well-Known Member
#9
Kasi said:
I was just wondering if you think that playing at a greater advantage somehow lessens your ROR. I'm kinda thinking all it does is make you win faster.
I think a more +EV lessens ROR. I think this is true at least for a fixed bet at a fixed EV with an even money payout, which is not the case for blackjack where EV is variable and bets aren't fixed due to doubles and splits and you get odds for blackjack.

I did some research and found an approach to compute ROR for the case of a fixed EV, fixed bet, and even money payout. My web site's not much and not too sophisticated, but (Dead link: http://www.bjstrat.net/cgi-bin/ror.exe) _this ROR calculator _ is there. Fool around with it and see what you get.

There's an approach to using this fixed EV type calculation in approximating ROR in blackjack outlined in Theory of Blackjack. I don't understand the statistical math involved, but I can use the method. What would be needed is external data based on a counting system probably from a sim. The data would in the form (as an example):
TC<=0, bet x0 unit(s), occurs y0% of the time, EV=z0%
TC=1, bet x1 unit(s), occurs y1% of the time, EV=z1%
TC=2, bet x2 unit(s), occurs y2% of the time, EV=z2%
.....etc.
y0+y1+y2+....+yn needs to sum to 1

Also, here's a link that may be of interest.(Dead link: http://www.bjmath.com/bin-cgi/bjmath.pl?read=6514)

k_c
 

k_c

Well-Known Member
#10
jack said:
k_c,
If you ever feel like going out on a limb, could you give my system a try. Maybe an hour or two. Thats if you can make heads or tails of it, of course. It can be used with any counting system. And as I mentioned, only if you feel like it.(2DPreferred)

See:Any successful progression players
One thing can be said for Licentia's system: It's simple :grin:.

k_c
 

jack.jackson

Well-Known Member
#11
rollem411 said:
Well, I was wonging out as I would in TC of -3 and beyond, but it seemed like I wasn't losing my money here...it seemed to fluctuate evenely.
I posted in other threads as to how I should spread my bets, but here it is:Try this Rollem +1 = 10
+2 = 15
+3 = 30
+4 = 60
+5 = 120


It looks pretty good, especially since your wongin out. Unfortunately, you cant spread to two hands on the trainer. I would think @ +2 you would have the advantage.
Of course you could also try wongin out more and spreading less. Also, what I like to do is parlay on wins. For example, if you bet 15$@TCof+2 and the TC jumps up to +4, ONLY bet 30$, instead of jumping straight to 60$.

You can try this with my spread or yours.
 

Sonny

Well-Known Member
#12
rollem411 said:
I know the long run is where you see results and in shoe games it comes around 10,000-15,000 hands (read it in another post) but I have played about 2,000 hands so far and it does not seem like I'm getting any results worth my efforts.
Depending on the game it might be closer to 40,000 hands or more. In a run of 2,000 hands (about 20 hours) just about anything can happen. It’s not uncommon for players to be in the red after a hundred hours of play or more. Card counting is all about perseverance.

Let’s assume you have a $20 average bet. After 2,000 hands you would expect to be within $2,063 of your EV about 95% of the time. With a 1% advantage you would expect to be somewhere between -$1,663 and +$2,463 after 2,000 hours. A loss of $2,700 wouldn’t be unheard of either. As long as you are somewhere above that you can consider your results normal.

-Sonny-
 

rollem411

Well-Known Member
#13
Sonny said:
Depending on the game it might be closer to 40,000 hands or more. In a run of 2,000 hands (about 20 hours) just about anything can happen. It’s not uncommon for players to be in the red after a hundred hours of play or more. Card counting is all about perseverance.

Let’s assume you have a $20 average bet. After 2,000 hands you would expect to be within $2,063 of your EV about 95% of the time. With a 1% advantage you would expect to be somewhere between -$1,663 and +$2,463 after 2,000 hours. A loss of $2,700 wouldn’t be unheard of either. As long as you are somewhere above that you can consider your results normal.

-Sonny-
I will keep practicing on the trainer and hopefully as I play more hands, I will begin the see the results. Also, is it possible to run a sim to figure these results instead of playing all of the hands out? I am trying to factor in my error with playing so if I were to screw up the count or bet incorrectly, I can see how it effects me. If a computer were to do it, the results would be exact so I'ma grab a 12 and keep playing.
 

Sonny

Well-Known Member
#14
A good simulator like CVData will let you program in how many mistakes you want to make so that the results will be more accurate. However, in order to know what kind of mistakes you are making and how often you are making them you need to keep practicing...which hopefully gets rid of those mistakes. Do you see where I'm going? :)

-Sonny-
 

rollem411

Well-Known Member
#15
Sonny said:
A good simulator like CVData will let you program in how many mistakes you want to make so that the results will be more accurate. However, in order to know what kind of mistakes you are making and how often you are making them you need to keep practicing...which hopefully gets rid of those mistakes. Do you see where I'm going? :)

-Sonny-
Yes I do. I like the practice because it builds my confidence for when I will actually go back to the casino. I had a really positive count and I managed to bank a few hundred and now I'm back on top. I'm hoping it stays this way though.
 

sagefr0g

Well-Known Member
#16
Sonny said:
A good simulator like CVData will let you program in how many mistakes you want to make so that the results will be more accurate. However, in order to know what kind of mistakes you are making and how often you are making them you need to keep practicing...which hopefully gets rid of those mistakes. Do you see where I'm going? :)

-Sonny-
how much flexability does cvdata have as far as 'programing' in mistakes?
can you set it up where bets vary for a given tc and over a range of tc's for a single simulation?

edit........
oop's forgot.. lol ....... did a little homework and went to the site. sounds really complex.
http://www.qfit.com/blackjack-simulator-update-cvdata4.htm
 
Last edited:
Top