Win/Loss/Tie% ?

#1
I read that the averages on any hand of Bj is win 40%, lose 45%, push 5% or something like that. I assume that is at a count of 0.
Does anyone know how those %'s change based on the count?
Is there a table for this?
Count W L P
-1 ? ? ?
0 40 45 5
+1 ? ? ?
+2
+3


thanks for any help
 

Sonny

Well-Known Member
#2
b said:
I read that the averages on any hand of Bj is win 40%, lose 45%, push 5% or something like that.
I think it is around 43% win, 49% lose and 8% tie.

b said:
Does anyone know how those %'s change based on the count?
It doesn’t change much based on the count. You will still win/lose about the same at just about any count.

-Sonny-
 
#3
Thanks

Sonny
My numbers added up to only 90% - good thing I am an accountant by trade.
I guess I am a little surprised by your answer that the %'s dont change with the count. Intuitively I would have guessed that at a count of +4 (+1.5%) that my win % s/b over 50% at least?
If you are correct then where is the logic that we should big bets on high counts come from if our win % doesnt increase?
 

Sonny

Well-Known Member
#4
b said:
My numbers added up to only 90% - good thing I am an accountant by trade.
Now if only you could find a way to pay 90% of your taxes… :)

b said:
If you are correct then where is the logic that we should big bets on high counts come from if our win % doesnt increase?
The reason you bet big on high counts is not because you win more hands but because the hands that you win are worth more money. At a high count you are more likely to get a blackjack, which pays 3:2 instead of 1:1. Even if you only win 43% of the time you still have a +EV situation. Imagine a game where you got a BJ 43% of the time and the dealer got a BJ 49% of the time:

.43 * 1.5 - .49 = 0.155 = 15.5% advantage!

You will also win more double downs and the dealer will bust his stiffs more often. That is the reason that a high count is more profitable than a negative/neutral count.

-Sonny-
 

schismist

Well-Known Member
#5
I believe at high counts, you actually push much more, lose a little more, and win less. However, when you win you win 150% as much when you get a blackjack and 200% to 800% as much when you double or split.
 

Kasi

Well-Known Member
#6
b said:
I guess I am a little surprised by your answer that the %'s dont change with the count... If you are correct then where is the logic that we should big bets on high counts come from if our win % doesnt increase?
Surely the W/L/T percentages are not constant over any and all counts.?!

I don't know by how much they change at what counts but they must change somewhat I would think. Like u say. And even Sonny seems to say.
 

schismist

Well-Known Member
#7
No, like I said, of course they are not constant. Think of an extreme case: a deck with four tens, and only four tens. The push rate here is 100%!
 

Kasi

Well-Known Member
#8
schismist said:
No, like I said, of course they are not constant. Think of an extreme case: a deck with four tens, and only four tens. The push rate here is 100%!
Couldn't agree with u more :)

Just didn't want to hit anyone that thinks otherwise with a sledgehammer lol.

O yeah - I also agree with what u said about the effect of extremely high counts.
 
Last edited:

Cardcounter

Well-Known Member
#10
What if they adjusted the win percentage?

What if instead of how many hands you won they adjusted it for double downs counting as 2 wins or 2 loses. If you split to 3 hands you could get up to 3 wins in the hand or 3 loses. If you got a blackjack it would count as 1 1/2 wins. I wonder what the win, loss, and push ratio would look like than after the adjustments.
 

mdlbj

Well-Known Member
#11
I think it is around 43% win, 49% lose and 8% tie.
It may be hard to stomach but, it actually does not matter if you win the hand in front of you because as a skilled counter, you are only playing with an advantage and you know over time you will come out ahead. Sick thought but true.
 

eps6724

Well-Known Member
#12
It was my understanding that the 43-49-8 was an AVERAGE based upon the million + hands in simulation. That would take into account all of the times the count skyrockets, plummets, and stays in the neutral. (There are a TON of material that gives the stats on each hand and how the % changes based upon the remaining composition). in other words, playing through an entire shoe.

If I understood all of THAT, then I would expect to have a HIGHER win rate if I played starting on a high count, and following it down. I base this on the expectation that the count FALLS when high cards come out, and therefore is favorable to us. I always assumed that this was the reason why ZG, Auto Monk and others have been screaming about never playing the negatives, and why there are 'walkaway' points for wong-out.

Somehow, I suspect that the 'real' pros are the ones who do the shuffle tracking, etc. so that they play ONLY when the 43-49-8 would be a 49-43-8 win for THEM.

I would be interested in knowing if anyone has kept the stats on this. I think QFIT (Man, I miss his input!) had some things on his sim that bears this out...
 

Sonny

Well-Known Member
#13
eps6724 said:
It was my understanding that the 43-49-8 was an AVERAGE based upon the million + hands in simulation. That would take into account all of the times the count skyrockets, plummets, and stays in the neutral.
That is correct. The numbers do change gradually as the count changes, but the change is slow and not indicative of an advantage. Also, the change does not always help the player. As schismist pointed out, at higher counts the percentage of pushes increases, which doesn’t benefit us very much. I believe at a TC of +7 the win/loss ratio actually shifts in favor of the player, but we have an advantage long before that. The win/loss ratio is not the main cause of our advantage.

eps6724 said:
If I understood all of THAT, then I would expect to have a HIGHER win rate if I played starting on a high count, and following it down.
Yes the win rate in dollars is higher but the win rate in percentages doesn’t have to be. You can still make a profit when winning only 43% of the hands.

-Sonny-
 

eps6724

Well-Known Member
#14
Sonny said:
Yes the win rate in dollars is higher but the win rate in percentages doesn’t have to be. You can still make a profit when winning only 43% of the hands.

-Sonny-
I guess what I was most curious about was if the percentages would change if you played strictly when you DID have an advantage?
 

Sonny

Well-Known Member
#15
eps6724 said:
I guess what I was most curious about was if the percentages would change if you played strictly when you DID have an advantage?
It would change, but the player would still be losing the majority of the hands played.

-Sonny-
 

eps6724

Well-Known Member
#16
Sonny said:
It would change, but the player would still be losing the majority of the hands played.

-Sonny-
I am going to take it that it is because of the 'player must act first' rule? How much then would shuffle tracking actually gain? Aren't you basing the whole thing on the idea that more favorable cards are available to you, or does that still only affect the increased chances of BJ's, dbls and splits?

I guess I'm thinking in terms of ev charts at different counts.
 
Last edited:

eps6724

Well-Known Member
#19
Sonny said:
I just ran across this which might be helpful:

http://www.bjmath.com/bjmath/conseq/streak.htm (Archive copy)

-Sonny-
Yes, actually now that I see it, it does make sense. Now, about a proper bet spread when playing ONLY at + counts. I know that QFIT had it when us peons could use his calculator, unfortuantely it's going to be a few more months until I upload bootcamp to take advantage of it!

Still, thanks. It certainly does make sense. (Griffin's book is on its way, so maybe that'll help a bit more to clear up the math, too!)
 
Top