Sensation! A profitable strategy for playing beagle has been found! - page 2

 

Anyway, if you change it in the script like this:

// if(MathMod(MathRand(),2)==1)RND+=0.0001;
if( MathRand() > 16383) RND+=0.0001;
else RND-=0.0001;


then everything falls into place. There is no SENSATION and there can be no SENSATION.



This, by the way, is a very common mistake. Instead of checking the logic of using the PRNG, they start criticising the PRNG itself.

 
I will use the PI number as the GCI http://ja0hxv.calico.jp/pai/estart.html. The period is up to 100 billion.
 

Do you want to measure the periods? The generator I use has a period of ^19937 - I don't even have a name for such a huge number.


But in this case the problem was not in the period of the generator, but in the wrong condition of the script. The error is at the intersection of mathematics and programming.

 
C-4 >> :

So, the conditions of the problem:


The rate of an instrument is set as follows.
At the initial moment, the price is 1.0000
At every new tick a coin is thrown, if heads = p*1.0001, if tails = p/1.0001
Spread is equal to 2 pips. The trader has a sum of $100,000. The maximum leverage is 1:100
Prove that there is no profitable strategy for the trader in this situation.


The problem conditions are taken from Elita in his tasks for trader on http://club.investo.ru/viewtopic.php?f=29&t=126082

The problem is poorly formulated, so there are a lot of interpretations of the "right" answer.


1. we need to define exactly what a "profitable strategy" is.

2. we have to define exactly what kind of coin we have.

3. what is "2 pips" based on the dimension of numbers in the problem.

4. etc.

 

That's a lot of stink out of nowhere. I'll put up a test to determine the limit of the MT generator tonight, and I'm sure it will be well over 200,000k.

// if(MathMod(MathRand(),2)==1)RND+=0.0001;
if( MathRand() > 16367) RND+=0.0001;
else RND-=0.0001;

What you have shown in the figure looks like an accumulation of positive values at all. There's no smell of SB here.

I'm just tired of hearing the same thing from SB enthusiasts. Here is our dialogue verbatim:

- the system will be constantly in a narrow corridor of values.

- Bazil, you are completely missing the point of random wandering.

- The value will always tend towards 1.

- the deviation from the starting point is proportional to the square of the time
let's say the deviation when you throw 10 000 times was 12 points
10 million times 1000 times more, the root will be 32
12*32 yourself calculate

- Bazil, don't embarrass yourself and listen carefully to RedRat
there is no corridor here - increase the number of throws by 2-8 times and all your corridors will be broken

- Ok, I will do as you say increase the number of throws by a factor of 100!

After a while...
- Zoomed in and the picture got even more interesting.

- That's because your generator sucks! I'm gonna give you the program and show you a picture of what it's supposed to be like.

After a while...

- That's weird! I got the same results! It turns out my generator sucks too. But if you take 100,000 ticks, the picture looks like the SB! (That's the logic!)

- You said you needed more ticks!

- You need a "good" generator and a "special coin". Checking for odds and evens doesn't work. (As if there is some difference between one probability of 50% and another "correct" probability of 50%).

 

2. нужно точно определить, какая у нас монетка.

Now that's just unbelievable! I take it this is a hidden implication that the coin must have at least an altered centre of gravity. Judging from the drawing you showed, the coin is clearly trying to land with an eagle.
 
C-4 >> :
Now that's just unparalleled! I take it it's a hidden implication that the coin must be at least with an altered centre of gravity. From the drawing you showed, the coin is clearly trying to land with an eagle.

From what you're saying, you have no idea what you're talking about. This coin tends to land heads just as often as tails. Nevertheless, the graph is exactly like this. Guess why.


C-4 >> :


It's not a stinker, just pointing out your amateur mistakes.


Why you are bringing up this fictional dialogue is more than unclear, it has nothing to do with me.


Spell it out - you have an error in your algorithm. You, out of the goodness of your heart, drew universal conclusions from erroneous data. That's all. They are trying to correct you, but judging by your reaction the result will be little more than hopeless.

 

К чему Вы этот вымышленный диалог приводите,

I wasn't referring to you at all. I brought it up in another forum. It's just that there are a lot of SB adepts there. And our approximate dialogue is from there.

Nevertheless, the timeline is exactly like this. Guess why.

Variant 1. RND+=0.0001; it falls much more often than RND-=0.0001; we see the effect of cumulative profit.

Variant #2. The ticks are taken much less, and this is only part of the picture.

OK, I'll take simple true random numbers 0 and 1 in the amount of 200,000 pieces from random.org and see what happens. Just don't say that the numbers should have been taken differently, and that the reliability of the source leaves a lot to be desired.

 
C-4 >> :

I wasn't referring to you at all. I brought it up in another forum. It's just that there are a lot of SB adepts there. And our approximate dialogue is from there.

I don't care who told you what. I'm only responsible for what I say. >> that's it.


C-4 >> :

Option № 1. RND+=0.0001; falls much more often than RND-=0.0001; we see the effect of accumulating profits.


This is easy to check, just count the number of eagles or tails. In your case, but with my corrections in the code, the probability of eagles is 0.500233. This is more than close to 50/50. If you do it your way, the probability is 0.499989. That's also very close to 50/50, but. But, the trouble is, you were trying to use the cycle generator as a stream generator. And you can't do that.


C-4 >> :

Variant #2. The ticks are taken much less, and this is just a part of the picture.

Please, this is what the WHOLE series of your data looks like (used your syd).




And by the way, you have an error in your tick count there. It's not 10000, it's 9999.



C-4 >> :

Ok, I'll take simple true random numbers 0 and 1 in the amount of 200,000 pieces from random.org and see what happens. Just don't say that I should have taken other numbers and the reliability of the sample is poor.

This will not help. Again, your error is not in the quality of the data, but in the logic of the algorithm. Recall that you were trying to check even/integer, which is not the same as 0 and 1.

 

for a "very long" series of games - the easiest way to win is to minimize the odds from 6-9 and using dabbles and sixlines (the scheme is about 1-1-1-1-2-2-2-3-4-5-6-8-10-10-12-14-16-18-20....)