The market is a controlled dynamic system. - page 374

 
Олег avtomat:

Surely breakeven is entirely achievable. But it is only possible with a deterministic approach.

If oneadopts the statistical-probabilistic approach, then break-even is fundamentally unattainable by virtue of that approach itself.

Why then spend so much time and nerves on the stat-ver-under ???

 
aleger:

Why the hell would you waste so much time and nerve on a stat-ver-under ???

To whom is this cry addressed?

Let me be clear: I don't use "stat-ver-under".
 
Олег avtomat:

Surely breakeven is entirely achievable. But it is only possible with a deterministic approach.

If oneadopts the statistical-probabilistic approach, then break-even is fundamentally unattainable by virtue of that approach itself.

Not at all. Any auto system always has 3 parameters:

1. The probability of detecting and capturing the target,

2. Probability of missing the target,

3. Probability of false alarms.

The names may vary, but their essence does not change.

It's a classic of the genre.)

 
Yuriy Asaulenko:

Not at all. Any car system always has 3 parameters:

1. The probability of detecting and capturing the target,

2. Probability of missing the target,

3. Probability of false alarms.

The names may vary, but their essence does not change.

It's a classic of the genre.)

No, it's not. Not any of them.

But that's not even the point.

For a system synthesised by statistical criteria, the listed parameters are internal, the synthesis is based on them.

In the deterministic approach, the listed parameters are external, applicable ex post facto to the evaluation of an already working system, while the synthesis of the system is done without the involvement of TViMS.

This is the fundamental difference in the application of these parameters in the different design approaches.

 
Олег avtomat:

To whom is this exclamation addressed?

To clarify: I don't use "stat-ver-under".

I presented this to the main "accidental rambling" controversialist (in another thread).

I found the current statement more useful than those worthless arguments.

I, too, believe that break-even is entirely achievable.
 
Олег avtomat:

No, it isn't. Not any of them.

But that's not even the point.

For a system synthesised according to statistical criteria, the listed parameters are internal, the synthesis is based on them.

In the deterministic approach, the listed parameters are external, applicable ex post facto to the evaluation of an already working system, while the synthesis of the system is done without the involvement of TViMS.

This is the dramatic difference in the application of these parameters in different design approaches.

I have not written anything about whether these parameters are internal or external. I was only talking about their presence.

I, even methodologically, cannot imagine a real system with parameters 2 and 3 equal to zero. I.e. a system without losing trades. Imho, it is unrealistic.

 
I suspect the Automat is labelling without stop-loss, as am I:))) The illusion of infallibility of the chosen model. I can cite dozens of my smart trades, when it seems like I've found it. And one (!) completely kills everything. You think: Well, I'm an asshole, why didn't I use a stop-loss ....? Then an inner voice says: "No, wait, someone is hunting for them... You haven't even considered this and that..." And so on to infinity. Fighting with yourself.
 
Yuriy Asaulenko:

I did not write anything about whether these parameters are internal or external. I was only talking about their presence.

I, even methodologically, cannot imagine a real system with parameters 2 and 3 equal to zero. I.e. a system without losing trades. Imho, it is unrealistic.

I don't use "probabilities" when synthesizing a control system. They do not exist. They do not exist and therefore cannot be zeros. They do not exist.

Although these probabilities can be counted on historical system performance data. Post factum. This information can be used to introduce a correction. But it does not change the original setting.

 
aleger:

This I presented to the main "accidental rambling" controversy (in another thread).

I found the current statement more useful than those worthless arguments.

I too believe that break-even is achievable.

I think the SB thread has made a useful contribution to shaping the view of the issue.

 
Alexander_K:
I suspect that the Automat is doing without stop-losses, just like me :))) The illusion of the infallibility of the chosen model. I can cite dozens of my smart trades, when it seems as if I've found it. And one (!) completely kills everything. You think: Well, I'm an asshole, why didn't I use a stop-loss ....? Then an inner voice says: "No, wait, someone is hunting for them... You haven't even considered this and that..." And so on to infinity. Fighting with yourself.

Well, never mind, we'll get through it. ;))