Not the Grail, just a regular one - Bablokos!!! - page 74

 
LeoV:
Doesn't the uncooked stuff get more of a kick out of it? )))
Eating natswai there with freshly mined chicken ingredient. It doesn't smoulder.
 
Vlads:
Lastrer:

What can you say about this formula

about the depth (aka depot size in c.u.) of the random walk

D = ln(z) / ln(q/p), where
z - acceptable probability of loosing (e.g. 1 - 0.956)
q is the price of loss (e.g. 1 c.u.)
p - the winning price (e.g. 2 c.u.)

I have seen this formula many times. Math. I have not encountered any justification. There is a great suspicion that the expression is imperial, and for a strictly defined MM.
 
prikolnyjkent:

In a real situation, the number of heads and tails MAY be absolutely ANY (!!!). Even if we do not take large samples, but look at simple eight-bit binary sequences, even out of 256 available combinations, as many as 70 (!) (if I am not mistaken) have numbers of zeros and ones that ARE MATCHED. And if some difference in quantities is allowed, the PROCENT of combinations satisfying our requirements will become even more overwhelming.

So it turns out that in a large sample (for example - a million shots) the number of heads and tails is EXACTLY the same EXACTLY ONLY WHY AN EXACT NUMBER OF MILLION BINARY COMBINATIONS HAVE EXACTLY THE SAME NUMBER OF ZEROS AND UNITS. And the probability of falling out with such ratio is incredibly higher than with a combination with a huge difference between heads and tails, simply because there are MORE of them in number (although with each particular million-bit combination the probability of falling out is the same).

And it certainly doesn't affect the probability of the outcome FOR ONE particular roll in this giant series. It was, is and will be 50/50 (!!!)...

I wholeheartedly support that. The only thing that surprises me is the fact that there are gentlemen who believe that if a coin doesn't matter and is always 50/50 then it doesn't obey any laws.

We simply do not need to play martin until the end of the capital (which NeCollah has repeatedly said, by the way) one must sometimes feel free to take a loss, and then win back. I.e. we set a goal, take it and play from zero. And if we don't take it, then we try to win back the loss, but not in one move (as suggested by standard Martini), but in several. In this way we can stretch the series almost indefinitely with limited capital, which Martin cannot offer.

 
Lastrer:

We should not play martin until the end of the capital (which NeCollah has repeatedly said, by the way) and sometimes you should not hesitate to take a loss, and then win it back. I.e. we set a goal, take it and play from zero. And if we don't take it, then we try to win back the loss, but not in one move (as suggested by standard Martini), but in several. Thus we can stretch the series almost indefinitely with limited capital, which martin cannot offer.

In a virtual game with zero expectation you can of course stretch your demise indefinitely. But only a real game, there is always a commission (or some other way to shift probability in favor of dealer), which will inevitably eat up your deposit. So martin or no martin, the outcome is the same.

 

Here we go again with the old song. I wrote about unequal probabilities at a=3 and b=4. You may be mistaken, so correct me. In the meantime we may think that the MO != 0.

The issue of commissions, requotes, slippages, non-market quotes, spikes, and sneaky dealers' corrections of spreads, as well as opening/closing deals without disclosing information to the client and so on, so on, so on. As they say, flies to flies and cutlets to cutlets.

Zy I keep thinking mo = 0, and is that good or bad? Take three bits - a probability of 1 / 8. And why, in fact, if there are 7 losses and 1 win (on a sufficient sample, of course), this win must be the last in the series all the time. It can ALWAYS be somewhere in the middle. And then there is a new game and there is no guarantee that it will ALWAYS be with the same starting lot as the previous one.

 
Lastrer: ... Let's take three bits - probability 1/8. Why, in fact, if there are 7 losses and 1 win (on a sufficient sample, of course), this win must be the last in the series all the time. It can ALWAYS be somewhere in the middle. And then there is a new game and there is no guarantee that it will ALWAYS be with the same starting lot as the previous one.

I've never thought about the spacing of series repeats. Thanks for the interesting thought...
 
Does anyone know where Alexander is?
 
Lastrer:

There's that old song playing again.

That's for sure... Well, good riddance. If you want to argue with mathematics, the flag in your hands. Although I personally find it strange to hear such things on a programmer forum, because anyone, even slightly proficient in coding, can easily check these crazy fantasies and see that they are inconsistent. But you don't even need to code anything, you can just build tables in Excel.

 
DmitriyN:
Does anyone know where Alexander is?

He's on some other forum, talking about how he ripped off las vegas.
 
Yeah. I wrote everything strictly according to the mathematics (theorist). As for Excel, it's not as simple as it seems. In fact, such things are easier to write in something else as multi-storey formulas in it, believe me, not ace.