Win percentage on the trainer?

[WhiteJill]

I have spent a lot of time using the blackjack trainer on this site. I keep the coach on and use BS. However, recently after hundreds of hands, my win percentage is about 37%. Is that normal? With perfect BS, what should the win percentage be after, say, 1000 hands? How many hands, roughly, are required for probability to stabilize at baseline?

Thanks

 

[EasyRhino]

One of the propellerheads can tell you better on the math part, but when I first started on the BS trainer I was on quite a winning streak for many hundreds of hands. It really seemed like an easy game.

And then on the next extended session of the BS trainer, I had a very extensive losing streak, of greater magnitude than any wins.

Which is often how it goes in real life too.

 

[blackjack avenger]

Not A Propellerhead

A thousand hands is about 10 hours of play. That is nothing. Check your numbers once you reach 100 hours of play. However, that is not enough either. Things may start to even out some at 500 hours of play. That is about 50,000 hands. If I told you 100,000 hands (1000 hrs.) is a more realistic number you may doubt me. This is assuming you are flat betting and playing all. The more you play the closer you get to expectation but it is a very wild rollercoaster.

 

[WhiteJill]

So I ask again, what is the percentage I should be looking for when things "even out", as you say?

Why does it matter if one flat bets or not? What you bet does not influence the win percentage.

 

[Kasi]

So I ask again, what is the percentage I should be looking for when things "even out", as you say?

Depends on the game and rules of course but a standard multideck game probably like about 43%W, 8%Tie, 49%L.

It more or less evens out money-wise because when you win maybe you win 1,2 units but when you lose maybe you lose 1.05 units kind of thing.

If you're flat-betting, knowing how many bets you are up or down and how many dealer upcards you played against (hands played if it keeps track of that) makes it easy to figure out how "unlucky" you are so maybe you get used to what might seem like incredibly bad luck but really maybe it's not, etc. It's the units up or down that maybe is more important than win% to keep track of.

 

[WhiteJill]

Thanks Kasi, I have experienced 43% wins most of the time, but the last 300 hands have stabilized at 37%, a huge variation considering the house advantage should be only 2% or so.

Another thing, I frequently get losing streaks of 6,7,8, or more. But never get winning streaks of the same length to match. A four hand winning streak is about the best I get, while 7 and 8 hand losing streaks happen as often. For example, one series of 15 hands, I won only one hand, and lost 14. Yet, never does the reverse ever happen. With a 2 % advantage, why is this the repeating pattern? I note down the results of every hand.

 

[blackjack avenger]

Thanks Kasi, I have experienced 43% wins most of the time, but the last 300 hands have stabilized at 37%, a huge variation considering the house advantage should be only 2% or so.

Another thing, I frequently get losing streaks of 6,7,8, or more. But never get winning streaks of the same length to match. A four hand winning streak is about the best I get, while 7 and 8 hand losing streaks happen as often. For example, one series of 15 hands, I won only one hand, and lost 14. Yet, never does the reverse ever happen. With a 2 % advantage, why is this the repeating pattern? I note down the results of every hand.

If you are playing a negative expectation game; -2% is very high, you will lose more hands then you win, so your streaks will consist of more losing hands then winning ones. If you only win 43% vs 49% loss you would expect longer losing streaks. Your number of hands played is to low to consider much of anything. As you play on you will find even longer streaks and bigger swings in your bankroll.

Your longer streaks will be negative, your biggest bankroll swings will also be negative.

 

[WhiteJill]

Right, with a 2% disadvantage, I expect to lose more than I win. But by rights, that should work out to about 48% win, not including pushes. Anything significantly different should be an aberration that should even out, or should vary up and down and average on the baseline. But, more to the point, my losing streaks have always been longer than my winning streaks regardless of the win percentage.

 

[godeem23]

Right, with a 2% disadvantage, I expect to lose more than I win. But by rights, that should work out to about 48% win, not including pushes. Anything significantly different should be an aberration that should even out, or should vary up and down and average on the baseline. But, more to the point, my losing streaks have always been longer than my winning streaks regardless of the win percentage.

A 2% disadvantage means you lose 2 cents for every dollar bet. Or that you'd win 49% (not 48%) of hands, if you flat bet and all bets payed even money, BUT all bets don't pay even money, your EV is maximized by making extra money on doubles and the bonus for blackjack. This all amounts to winning LESS THAN 49% of hands (counting pushes as half win, half loss). I can't tell you how much less, I don't have any programs to run.

 

[Kasi]

Thanks Kasi, I have experienced 43% wins most of the time, but the last 300 hands have stabilized at 37%, a huge variation considering the house advantage should be only 2% or so.

Another thing, I frequently get losing streaks of 6,7,8, or more. But never get winning streaks of the same length to match. A four hand winning streak is about the best I get, while 7 and 8 hand losing streaks happen as often. For example, one series of 15 hands, I won only one hand, and lost 14. Yet, never does the reverse ever happen. With a 2 % advantage, why is this the repeating pattern? I note down the results of every hand.

What the heck game are you playing that has a 2% house advantage?

Sometimes when people say they lost 14 in a row what they really mean is it was a non-win streak. In other words maybe there are some ties in there.

Sometimes, if, say, you split to 4 hands and lose them all they think they have lost 4 hands in a row as a streak and also maybe count it as 4 losses in their overall loss percentage. Really it's a loss of only one hand of say 4 units.

Are your results in units when you record them or is a BJ win the same as any other and just maybe gets a "W"? Are you maybe basically just wondering how "bad" your apparently unrealistic "bad luck" has been?

Anyway, what really matters overall is not the number or length of winning streaks but how many flat-bets you are up or down (even if you don't always flat-bet you can always assume you bet $1 hand if you have all the results) versus how many dealer upcards you played since that's how you ultimately figure things. I mean who knows maybe the one hand you won was a 4-split with doubles for 8 units and the 14 you lost were 1 unit so over 15 hands you're only down 6 units kind of thing.

Now if you're talking about a real-life game, internet or otherwise, and suspect something, stuff like that happening too often could mean something.

 

[blackjack avenger]

Streak Chances

Given:
43% won
49% loss
not counting ties

your chances of losing 7 hands in a row is .67821%
.49 to the 7th power

your chances of winning 6 hands in a row is .63213%
.43 to the 6th power

If after a thousand hands this is not quite matching your results I am not surprised because a thousand hands is not very many. If you play more hands you will have even longer streaks.

So you will lose 7 hands in a row about as often as you will win 6 hands in a row.

I hope you are not confusing this with expectation.
Expectation is :
bet * advantage * number of hands played.

example:
$100 * -2% * 1000 hands = -2,000

 

[sagefr0g]

So I ask again, what is the percentage I should be looking for when things "even out", as you say?

.........

http://www.blackjackinfo.com/bb/showpost.php?p=70925&postcount=24

 

[WhiteJill]

With basic strategy (showing a typical 42% win ratio, including pushes), the numbers of decided hands are 107 vs 95 for the house. At a simulator win rate of, say, 42% including pushes, the house wins about 53% of the decided hands, the player wins about 47%. So the house advantage is 3% above 50/50 odds.

Now, I understand, with the ability to double bets, or split hands, the house doesn't win 3% more money, because you can increase your bet when BS says it is to your favor. So in terms of dollars, the advantage is reduced to, from what I read, about .5%.

Kasi, When I lose on a split, I still count it as one loss. So when I say X loses in a row, that is not including those split hands. My losing streaks are always longer than my winning streaks. Significantly.

Has anybody here run the simulator for a thousand hands or more and recorded their win/loss percentage?

 

[godeem23]

With basic strategy (showing a typical 42% win ratio, including pushes), the numbers of decided hands are 107 vs 95 for the house. At a simulator win rate of, say, 42% including pushes, the house wins about 53% of the decided hands, the player wins about 47%. So the house advantage is 3% above 50/50 odds.

Now, I understand, with the ability to double bets, or split hands, the house doesn't win 3% more money, because you can increase your bet when BS says it is to your favor. So in terms of dollars, the advantage is reduced to, from what I read, about .5%.

Kasi, When I lose on a split, I still count it as one loss. So when I say X loses in a row, that is not including those split hands. My losing streaks are always longer than my winning streaks. Significantly.

Has anybody here run the simulator for a thousand hands or more and recorded their win/loss percentage?

BS will never tell you that it's in your favor to increase your bet. I think you're still missing something important. I tried to explain it before, I guess I didn't do a good job. Hopefull someone else can explain it in a way that makes more sense to you.

 

[Kasi]

the house wins about 53% of the decided hands, the player wins about 47%. So the house advantage is 3% above 50/50 odds.

Now, I understand, with the ability to double bets, or split hands, the house doesn't win 3% more money, because you can increase your bet when BS says it is to your favor.

If they win 53% and you lose 47% when all bets are even money it's a 6% HA not 3% lol. (After you wager $100 you are $6 down,)

Anyway, nobody could run a sim without knowing the specific rules. And whatever the HA is, it's always applies to only your initial bet whether you double or not.

Why the fascination with win percents anyway lol. It doesn't tell very much.

 

[blackjack avenger]

Progressions = Bankroll Suicide

I hope you are not flirting with the idea of any type of progression. No progression works. Every major blackjack expert will say the same thing.

Progressions don't work.

Basic strategy tells you what you can expect to lose if you play long enough. It includes when you double, split, double after split, blackjacks and any other rules that apply. The generic -.5 disadvantage for using basic strategy takes all of the plays into consideration. The rules used will determine your actual disadvantage. There are different strategies for different rules.

A simple example of standard deviation (sd):
sd is equal to the square root of the number of hands played times (1.1). The 1.1 takes into consideration your splits and doubles.

(sqrt. 1000) * 1.1 = 34.78
Let's say you are flat betting $100.
So 35 times your average bet of $100 = $3500
Now you add and subtrace your sd to your expectation to get your range of results.
So $100 * 1000 * -.5= -$500 is your expectation.

1sd range is = -4000 to -3000
Now you can expand your results further:

Your results over 1000 hands
68% of time -$4000 to $3000
95% of time -$7500 to $6500
99.7% of time -$11,000 to $10,000

So in the short run a basic strategy player can win but they are not likely to. This is how those casinos pay those electric bills and all those employees.

10,000 hands?
expected value = -$5000
sd = $11,000

range

Your results over 10,000 hands
-$16,000 to $6,000 68% of time
-$27,000 to $17,000 95%
-$38,000 to $28,000 99.7%

Remember when I said it would take many many hands for things to even out? This is what I meant.