Probability, how do you turn it into a pattern ...? - page 57

 
dentraf >>:


Я так понимаю вы как используете ускорение время исполнения ордеров так их и замедление. Вот очень интересна логика замедления

There is no point in slowing down an already protracted process. Indicating (-) positions correlate with similar ones which are in (+). And is the criterion for calculating the balance discrepancy one way or the other. I should add that if we periodically mark differences as a % of positions with the maximum deviation, as a % of positions with a moderate deviation, as a % of positions with a small deviation (of course both + and -) then the calculations will be more accurate, which is acceptable when working with a small number of instruments. But I've solved this problem in a simpler way, the averaged value of balance will look safer with a larger number of instruments.

I set the acceleration only for that group of positions, which decisively affects the balance in my favour.

 
Neveteran писал(а) >>

There is no point in slowing down an already protracted process. Indicating (-) positions correlate with similar ones which are in (+). And is the criterion for calculating the balance discrepancy one way or the other. I should add that if we periodically mark differences as a % of positions with the maximum deviation, as a % of positions with a moderate deviation, as a % of positions with a small deviation (of course both + and -) then the calculations will be more accurate, which is acceptable when working with a small number of instruments. But I've solved this problem in a simpler way, the averaged value of balance will look safer with a larger number of instruments.

I set the acceleration only for that group of positions which decisively influences the balance of forces in my favor.


Is the tactic of closing positions in the second cycle just about extracting + from all positions, or based on the probability calculated?
 
Neveteran >>:

...он обрывается по правилу достижения суммарного балансового преимущества.

dentraf, your question

 
moskitman писал(а) >>

dentraf, your question


so the resulting plus after two cycles will be determined by the sum of positions that were locked after the first cycle?
 
dentraf >>:


тоесть таким образом результирующий плюс по истечению двух циклов будет определяться суммой позиций которые были локированы после первого цикла?

probably just broke (-) to breakeven and run away - to build the next first one, and most likely the results of the second one give the main profit - there is a lot of lottery there

 
moskitman писал(а) >>

probably just broke even and went on to build the next first one, and most likely the results of the second one will give the main profit


What's the point of breaking (-) to breakeven and then building the first cycle again to get the calculated data, if we already have them from two cycles
 
moskitman scratching his head
 
ha! the second cycle is skewed in terms of pairings, what BEGINNING EQUAL CONDITIONS are we talking about?
 
Neveteran >>:

...Ускорение я задаю только для той группы позиций, которая решающим образом влияет на расстановку сил в мою пользу.

how do you determine the decisive pairs? if they are the ones that are deeper "buried" in the minuses, then according to the thin thread rule, they can give even more negative results by being accelerated...
(scratching his head again)
maybe it only makes sense to accelerate those for which there is a higher probability of going to the plus given the minus they already have on them?

 
moskitman >>:

как Вы определяете решающие пары? если это те, которые глубже "занырнули" в минуса, то согласно правилу тонкой нити, они могу дать еще больше отрицательного результата, будучи ускоренными...
(снова чешет репу)
может имеет смысл ускорять только те, для которых выше вероятность выйти в плюс с учетом уже имеющегося на них минуса?


The decisive pairs, those pairs whose deviation has a value ((-) position correlate with similar ones which are in (+). And are the criterion for calculating the balance divergence in one direction or another. I should add, if we periodically mark differences as a % of positions with the maximum deviation, as a % of positions with a moderate deviation, as a % of positions with a small deviation (of course both + and -)) those that are already hanging near (0) they will remain near (0) or tilt, creating an imbalance, but insignificantly enough . You have a very vague idea of the mechanics of crossing OSI, it is the passing through (0) (the desire to stabilize) that is the basis of the logic. If the balance went to (-) and after the first cycle I corrected (calculated) the axis angle, then the second cycle, is an unambiguous pass over the axis in the opposite direction.

The main factor, is the cycle time, it is absolutely indicative.