Is long run testing senseble?

[schoenpoetser]

The best long run test is the profit of the casino.Millions of spins  and the payback is about  97,3%.All mathematicians did know this for centuries.
In my opinion long run results of every system is predictable.You maynot deny the law of the big numbers.It is therefor that testing very large ( >1000) samples is senseless.The knowledge of the deviations in small samples are useful  for developing systems and strategies.There is a different for the several chances. A small sample is not the same for EC,Dozens etc.

The diagrams of the analysed chances are more useful than long run test.

[col1879]

Agree.

Individual A wins 1000 EC bets in a row
Individual Z loses 1000 EC bets in a row

What do all the individuals in between get? How do they bet? What are they betting on? Are they flat betting or using progression? Are they betting intelligently or not?

[Kelly]

A long run test consists of 1000s of short term sessions. Its easy to construct a betting scheme with a progression where you win 9 out of 10 sessions and when  you hand test this system you might easyli end up thinking you got the holy grail and in the end go risk some real money.  In reality, for some odd reason, you are likely to hit one of the bad streaks (which is embedded in the long run test), in one of the early sessions in your brick and mortar casino and if you are not carefull, will cost you your entire enheritance or monthly payroll.

So the long run test is there to show you that what you are playing has a negative expectation and WILL end up negative sooner or later.  Its then your choice to go and play anyway and hope for  a bit of luck and that the good streaks is on your side today.  If you analyse a long term test you will find some negative streaks you wouldnt believe existed, even if you use real spins as data base.

So the long run tests is basicly just there to show you that you can moon walk back and forth but only be in profit for so long, sooner or later the devil will catch you and your search for fresh money to play for will take a dramatic turn for the worse.  I have seen them come and go (to the cash machine) and come back.

[darrynf]

short runs+short runs+short runs+ short runs = long term

isnt this true, saying a system will only win short term is saying it wont win at all.

if a system wins lots of short terms then it should win long term, though how may short runs make long term.

[schoenpoetser]

Ken my conclusion is you agree my statement.In a other topic I wrote ,I had tested strategies in 1M samples.This seems not logic, but I make a difference in strategies and systems.
A strategy consist of many small special samples in a along row spins.The sum of all small samples is not a random row.
This is the clue a strategy can be a winner.

[darrynf]

and yet there isnt one

[schoenpoetser]

Darrynf what do you mean with " ONO  ".

[schoenpoetser]

-