Alot of faith is put into the mathematics, I've heard sayings such as 'you shouldn't sweat short term results', easier said than done, at what point does short term become long term.
In terms of hands played, what constitutes the 'long run'?
Alot of faith is put into the mathematics, I've heard sayings such as 'you shouldn't sweat short term results', easier said than done, at what point does short term become long term.
In terms of hands played, what constitutes the 'long run'?
> Alot of faith is put into the mathematics,
> I've heard sayings such as 'you shouldn't
> sweat short term results', easier said than
> done, at what point does short term become
> long term.
> In terms of hands played, what constitutes
> the 'long run'?
You are not going to get a definitive answer to that question.
I have seen Norm respond to these types of questions with answers such as: 500,000 hands are insignificant in a statistical analysis.
If you play 2 hours a week, every week for three years (approx 10,000 hands), in 15 years you should have played 1/10 of 500,000 hands. In my opinion you are over the short run and playing for the long haul. This is only 50,000 hands however, and many players will disagree with that statement.
Variance is a shady character. A good basic strategy player, playing those 50,000 hands, and betting within his bankroll, has about a 50/50 chance of coming out on the plus side after all those hands. A good Advantage player could easily loose 2% of his play in that same time period.
The mathematics prove that the longer you play, the closer you come to the metric medium of your ability. The short run is longer than people think of they play all of their adult life.
> In terms of hands played, what constitutes
> the 'long run'?
Search "July 17, 2003" for a post from Don Schlessinger for a beginning point to your question. Following is an excerpt from that post:
"Actually, to be more precise, it's the number of hands you need to play before expectation catches up to (equals) one standard deviation. Some people think that, to define "long run," one ought to achieve a two-standard-deviation result, instead of just one. Unfortunately, this doesn't double the number of hands needed to play, it quadruples it, as SD is a square-root function."
As Don say's it's based on the game your playing and can vary wildly between games.
But I like definitive answers, which in the world of BJ, are sometimes hard to come by, so I'll throw out a number.
For 6D, HiLo using (18+4) indices, decent rules and pen ... 140,000 hands.
How's that for an answer!?
At 75 hands per hour .. roughly 2,000 hours.
If I'm anywhere close, it's a lot shorter than I first imagined.
> Search "July 17, 2003" for a post
> from Don Schlessinger for a beginning point
> to your question. Following is an excerpt
> from that post:
> "Actually, to be more precise, it's the
> number of hands you need to play before
> expectation catches up to (equals) one
> standard deviation. Some people think that,
> to define "long run," one ought to
> achieve a two-standard-deviation result,
> instead of just one. Unfortunately, this
> doesn't double the number of hands needed to
> play, it quadruples it, as SD is a
> square-root function."
That number (n0) can be calculated for thousands of situations using the first calculator at the below link:
Well, when looking at the difference between two strategies you need to run huge numbers of hands. But, see Sun Runner's response for a better Long Run response.
> You are not going to get a definitive answer
> to that question.
> I have seen Norm respond to these types of
> questions with answers such as: 500,000
> hands are insignificant in a statistical
> analysis.
> If you play 2 hours a week, every week for
> three years (approx 10,000 hands), in 15
> years you should have played 1/10 of 500,000
> hands. In my opinion you are over the short
> run and playing for the long haul. This is
> only 50,000 hands however, and many players
> will disagree with that statement.
> Variance is a shady character. A good basic
> strategy player, playing those 50,000 hands,
> and betting within his bankroll, has about a
> 50/50 chance of coming out on the plus side
> after all those hands. A good Advantage
> player could easily loose 2% of his play in
> that same time period.
> The mathematics prove that the longer you
> play, the closer you come to the metric
> medium of your ability. The short run is
> longer than people think of they play all of
> their adult life.
As a person with only time to get in about 250 hours a year at the tables, the term long run is just a vage general term to me. It resembles the term infinity closely.
You just need to understand the deviations possible, accept them and know that if you play long enough you will have a winning game. Also understand that you can go through long periods of loosing but not let it shake your confidence.
Short term you will the shoe you can not loose on and that highly positive one that beats you to a pulp. My last trip I had a string of 16 straight loosing hands. Whats the odds there?
My point is that the Long Term is not a line you cross where bells go off and you are handed a check for getting there. It is a real statistical place but not a real place. It is a math thing in order to judge your method of playing against the house rules over a long enough period of time to be accurate. These sims can play with it. Find that if you spread more you are much more positive in the long term. Input rule changes and give you the long term effect, and things such as this.
That number (n0) can be calculated for
> thousands of situations using the first
> calculator at the below link:
Ohlala! The long run is a lot longer than I always thought! In my case, playing 8 decks fair pen, crowded tables (70 hands/hour), 5 hours/week, 45 weeks/year equal something like 23 years! No wonder, I cannot see the light at the end of the tunnel!
Thanks for the great info.
> That number (n0) can be calculated for
> thousands of situations using the first
> calculator at the below link:
Thanks for ther responses guys, found the post Sun Runner referred to, cheers.
After 400hrs,30000 hands playing an 8 deck game and spending 95% of that time around the -SD1 mark, including a recent $3000 high count hammering I had, my confidence is pretty much shot and the long run does appear to be a long,long way away.
Just a few thoughts on this:
As Victoria pointed out: the Long Term is not a line you cross where bells go off and you are handed a check for getting there. It is a real statistical place but not a real place.
This is very true. Let us suppose, for purpose of discussion, that you have been playing the game for 20 years. A serious part-timer, you have averaged 500 hours/year. Always avoiding crowded tables and slow dealers, you have averaged 100 hands/hour.
This means that you have played roughly one million hands of blackjack. I think most people would agree that you have reached "the long run."
So what doe this mean? Not a whole lot, as it turns out. If you have averaged $100 per hand bet over the entire 20 years, through careful game selection played to an average 1% advantage, put every cent of your winnings into the bankroll, and of course, never made any mistakes, congratulations: You're a millionaire!
Or at least you were before Uncle Sam got through with you. :-)
All it means is that at this point in your playing career, your actual earnings should be pretty much in line with with your expectation. That's all.
Does this mean that you will no longer experience wild winning and losing swings? Of course not. It simply means that those swings, when averaged into your lifetime results, will not significantly alter the results.
For this you waited 20 years?
:-)
Life is too short to play 8 decks.
> ... accept them and know that if you
> play long enough you will have a winning
> game.
You might play a winning game -but somebody out there, who by all rights should have roared past the long run with that $1MM Parker referred to .. won't.
They will have lost.
That's why they call it gambling.
Getting in the long run simply means aquiring a sufficient sample size such that short term fluctuations become increasingly less significant. The closer the measurement you want the bigger the sample size that is required.
For example, how many hands do you need to have a 60% certainly that you will be plus or minus 700% of your theoretical? You obviously need far fewer than would be required to have a 90% certainty that you will be plus or minus 10% of theoretical.
So long run is an imprecise term that refers to having a sufficient sample size that your sampling error is insignificant for the quantity you want to measure.
Posting Permissions
-*-
Bookmarks