The long run

Dopple

Well-Known Member
#1
I don't have software but have been at this game non-professional for 40 yrs. Could I get a brief primer on the long run. Is it the fact that after x hours of play I will have overcome enough variances to have a good general idea of my EV? What are we marching toward in this illustrious pursuit of the long run. Was I in the long run 20 yrs ago so I don't have to sweat it anymore. Please tell me I am now in the Long Run club, I know Ive paid my dues.
 

gronbog

Well-Known Member
#2
Look up the statistic called N0 (N-zero). It is the number of rounds after which your EV equals one standard deviation. Using what we know about standard deviation for the Normal distribution we can then estimate the probability that you are still in the red after playing this many rounds.

After 1 x N0 rounds, you have approximately a 16% chance of still being in the red
After 4 x N0 rounds, you have approximately a 2.3% chance of still being in the red
After 9 x N0 rounds, you have approximately a 0.13% chance of still being in the red

N0 will vary depending on your style of play, your spread, the rules and conditions and the system you use, but will typically fall somewhere between 10,000 and 20,000 rounds for a card counter playing blackjack. If you've been playing for 40 years, then you have probably exceeded N0 and possibly also exceeded 4 x No and 9 x N0 as well.
 

Dopple

Well-Known Member
#3
Thank you gronbog. I will look into this statistic. I am just getting back into the game after losing last years budget I have had a better variance so far this year and want to relish a slow and smooth rise to riches.
 

Dopple

Well-Known Member
#6
My remark was tongue in cheek. I realize you make little and the swings are great. You are not saying most of the group are losers I hope.
 

Dopple

Well-Known Member
#8
Thanks Hell n Back. I think I am winning also but I need to start keeping records again. It was interesting what Snyder wrote. I find winning streaks can be stressful also, but in a nice way.
 

Dopple

Well-Known Member
#10
So if anyone cares to entertain this query: Lets take a good player w/ perfect BS and the Illustrious 18 plays on a 6D table wonging out at -2. Let him play a full year 8 hrs a day half Kelly. About how many hours could he find himself on a decline. This would the greatest number of hours, or use hands if you prefer, that he would be lower than a previous br level? Could you sim this with some software. Just curious. I suppose you could say just sim it yourself if you are interested.
 

KewlJ

Well-Known Member
#12
Dopple said:
Let him play a full year 8 hrs a day half Kelly.
Ok, wow. 8 hours a day for a full year. That sounds like how players played at the end of the last decade. I don't know many that play that way in 2018. The casinos here (Vegas) won't allow you to play that way....so maybe some other parts of the world. :rolleyes:

So moving on..."the long run". One of my favorite discussions, second to count discussions. :cool:

And sure N0 is important here...very important. But what does it really mean in real life play? How does it translate?

Early on in my career, I think in one of the books about the MIT team, I read that the longrun was roughly 50,000 hands. Ok, certainly a generalization, and not taking into account all the individual things like spread, rules ect. But I kind of liked this simplified generalization and sort of went with that for a few years. Ahh..until I had a 50,000 round losing streak. THEN the 50,000 round number didn't seem to work anymore for me. ;)

I play 80-100 thousand rounds of blackjack a year. 50,000 represents about half year of play. And I have had four (4) different losing periods lasting 40,000 -50,000 rounds. So I have upped my number to 100,000. That's about (the high end) of years worth of play for me. I have never had a losing year or 12 month period, despite those (4) six month losing periods. So that is what this guy is now using as my real life number. Well, until the day comes that no longer works as well. o_O
 
Last edited:

psyduck

Well-Known Member
#13
KewlJ said:
So moving on..."the long run". One of my favorite discussions, second to count discussions. :cool:
Hey KJ, I really like to know which count you think is the best! You have been making a living on BJ play and therefore I assume you do tons of side counts. :p
 
#14
Assuming a higher end mediocre six deck game your N0 is probably about 24,500 hands, therefore, 3 standard deviations is going to be 73,500 hands played. If you're getting 100 hands per hour that's going to take you 735 hours to achieve. The chances of losing after 3 standard deviations is 98.7% (I think). The chance you're going to be losing for the entire year is virtually impossible since there's 2920 hours in a year if you're playing 8 hours a day. You would have to perform well over -6 standard deviations below EV which has a far less than 1% chance of occurring, probably even smaller than 1/10th of a percent.
With that being said, I don't know what the longest losing streak you're going to be able to expect is since I don't know how to quantify that but if I were to take a guess with the given parameters I am going to guess you can expect to perform 2 standard deviations below EV at some point so let's just call it 490 hours.
Edit: I just remembered you're Wonging out at -2 so take in mind that I haven't accounted for that.
 
Last edited:

DSchles

Well-Known Member
#18
JohnCrover said:
Assuming a higher end mediocre six deck game your N0 is probably about 24,500 hands, therefore, 3 standard deviations is going to be 73,500 hands played. If you're getting 100 hands per hour that's going to take you 735 hours to achieve. The chances of losing after 3 standard deviations is 98.7% (I think). The chance you're going to be losing for the entire year is virtually impossible since there's 2920 hours in a year if you're playing 8 hours a day. You would have to perform well over -6 standard deviations below EV which has a far less than 1% chance of occurring, probably even smaller than 1/10th of a percent.
With that being said, I don't know what the longest losing streak you're going to be able to expect is since I don't know how to quantify that but if I were to take a guess with the given parameters I am going to guess you can expect to perform 2 standard deviations below EV at some point so let's just call it 490 hours.
Edit: I just remembered you're Wonging out at -2 so take in mind that I haven't accounted for that.
No, your N0 calculations are wrong. N0 is a square function, not linear. So, two N0s require four times the number of hands of 1 N0. And three N0s require nine times as many hands. And the chance of winning (not losing!!) at the three s.d. level is 99.87%, not 98.7%.

Don
 
#20
DSchles said:
No, your N0 calculations are wrong. N0 is a square function, not linear. So, two N0s require four times the number of hands of 1 N0. And three N0s require nine times as many hands. And the chance of winning (not losing!!) at the three s.d. level is 99.87%, not 98.7%.

Don
Thanks Don, always very helpful.
 
Top