Calculate the probability of reversal - page 11

 
This is an interesting topic. So far I've come up with a solution: just calculate how many times a step up was followed by a step down and separately how many times a step down was followed by a step down, then find the average percentage of probability of continuation. As usual I reduced all the maths to "just count". The simplest solution is "outright", just the way I like it.
 
Maxim Romanov:
It's an interesting topic. So far I've come to a solution: just calculate how many times after the step up was a step up and separately how many times after the step down was a step down, then find the average percentage of probability of continuation. As usual I reduced all the maths to "just count". The simplest solution is "outright", just the way I like it.

Yes, the purely scientific approach of a postgraduate student...

But you are not interested in the "average hospital temperature"...?

A trader needs PRACTICAL solutions, which bring profit... And if you approach it from this point of view, it's all "monkey's work"...

 
Serqey Nikitin:

Yes, the purely scientific approach of a postgraduate student...

But you're not interested in "average hospital temperature"...?

A trader needs PRACTICAL solutions that bring profit... And if you approach it from this point of view, it's all "monkey's labour"...

So the monkey did a pretty good job.

You don't know anything about what I do, what it's for, how to apply it, where it comes from, and where it will be used. I wasn't explaining where this distribution came from, why it's needed, how to get it. It was about an abstract construct. Why this construct is needed is not clear to you either. Why are you writing nonsense?

Mathematics is a language for describing processes.
 
Maxim Romanov:

So the monkey did a pretty good job.

It was about an abstract construct.

Sorry, I thought the topic was really for Trader...

 
Serqey Nikitin:

Sorry, I thought the topic was really needed Trader...

I do, you just wouldn't understand)

 
Vladimir:

In order for the relationship to become linear, the coordinates must be transformed non-linearly.

Two different non-linear parabolas have a linear relationship without any coordinate transformation.

 
Maxim Romanov:

I do, you just wouldn't understand)

You're right! I'm not even going to try to get into the peri-scientific stuff... Keep playing...
 

I think it makes sense to clarify the problem again.

If we proceed from the formulation of the question about probabilities of transitions at each step from one cell to another, and that, as a result of simulation of such wandering we obtain a frequency distribution close to the specified one, then the answer variant has already been given by me.

It could be a wandering handful of balls, each of them with probability 1/2 - stays in its hopper (note that this hopper consists of two cells), and with probability 1/4 go to the next.

But for the last (limiting) bunker the probability changes - the ball 3/4 stays in the bunker (because no further to go - the wall) and

1/4 returns to the bunker in the direction of the start of the wander.

The initial histogram gives us an idea of the likely outcomes of such a wander and, assuming exactly 10 steps are taken, my model is very plausible. If the steps are more or less, there will be no match.

So, if the real problem is not reduced to such a model, then another model should be built - otherwise there will be "games of numbers" again...

)

Рассчитать вероятность разворота
Рассчитать вероятность разворота
  • 2020.02.24
  • www.mql5.com
Кто соображает в математике, помогите пожалуйста решить задачку, не могу догадаться как это сделать...
 
Maxim Romanov:
It's an interesting topic. So far I've come to solution: just count how many times after step up was a step up and separately how many times after step down was a step down, then find average percentage of probability of continuation. As usual I reduced all the maths to "just count". The simplest solution is "outright", just the way I like it.

If you take 9 steps, 10 is the transition to a different parameter you will have an offset, and if you take 3, 6, 9, 12, etc., then try to get a better value.

 
Mikhail Dovbakh:

I think it makes sense to clarify the problem again.

If we proceed from the formulation of the question about probabilities of transitions at each step from one cell to another, and that, as a result of simulation of such wandering we obtain a frequency distribution close to the specified one, then the answer variant has already been given by me.

It could be a wandering handful of balls, each of them with probability 1/2 - stays in its hopper (note that this hopper consists of two cells), and with probability 1/4 go to the next.

But for the last (limiting) bunker the probability changes - the ball 3/4 stays in the bunker (because no further to go - the wall) and

1/4 returns to the bunker in the direction of the start of the wander.

The initial histogram gives us an idea of the likely outcomes of such a wander and, assuming exactly 10 steps are taken, my model is very plausible. If the steps are more or less, there will be no match.

Thus, if the real problem is not reduced to such a model, then another model should be built - otherwise there will be "games of numbers" again...

)

Thanks, interesting thought, need to think about it, whether I have this effect on the edges or not.