Trading probability - page 13

 
SProgrammer >>:

It is not about TP or SL probabilities at all. ZigZags on market and generated data are considered. A "pattern" is observed on market data.

 
getch писал(а) >>

It is not about TP or SL probabilities at all. ZigZags on market and generated data are considered. On market data a "pattern" is noticed.


I'm glad, but in my opinion, the answer to my question is the same as the answer to yours - because I'm simply asking a question - why should we fuck if it's clear that nothing will happen :)) ... And you're trying to look for a pattern - that is, trying to somehow change the ETC so that the rules don't work for it anymore?

 
SProgrammer >>:


А вы пытаетесь искать закономерность - то есть пытаться как-то изминить эТС, таким образом чтобы правила для неё перестали работать?

The thinking was structured slightly differently. First, there were results on real prices. Then it became interesting to see the presence/absence of a "pattern" on a simple random time series.
The power dependence showed up on all data. "Regularity" - only on market data.
I wanted to check whether the result of such a "pattern" was a simple consequence of probability theory or something else.

 

Ok - look - there is a TS - let's call it a "benchmark TS" with "inputs" and a primary distribution (with equal ), here we have no strategy at all - and let's say there is another TS with already "So using zigzag you make a TS with strategy from ATS (TS without strategy). And I ask - and you check if the rule persists, and if it persists, then there is nothing to fuck. :) And if it does not - then it already makes sense, being clearly assured ( mat. proven) that there is something to look for. I hope it is clear - because I will not repeat it again. :) Sorry.

 
SProgrammer >>:

А я и спрашиваю - а вы проверьте может правило то сохраняется, а раз сохраняется так и трахаться нечего. :)

This is where I don't understand the non-triviality of the recipe. Well testing, of course, i.e. calculating the balance. Which is clearly related to this "rule" (or changing/unchanging it), only it requires fewer operations to calculate.

 
SProgrammer >>:

Ок - смотрите - есть некая ТС - назовем ее эталонной с "входами" отбалды и первичным распеределением ( с равномерным ), тут у нас как бы нет никакой стратегии вообще - и скажем есть еще некекая ТС с уже "неслучайнеыми" входами ( точнее со случайными но _не_ с первичным распределением, _не_ с равномерным) ну вот вы исполоьзуя зигзаг делаете из эТС ( ТС без стратегии ) ТС со стратегией. А я и спрашиваю - а вы проверьте может правило то сохраняется, а раз сохраняется так и трахаться нечего. :) А вот если нет - тогда уже имеет смысл, будучи четко уверненым ( мат. доказанной ) что есть то что наод искать. Надеюсь понятно - потому что я больше повторять не буду. :) Извините.

Apparently, the reasons for my misunderstanding are the 1st of April... No TC was built. The usual ZigZags.
Pseudo prices can be constructed in a myriad of ways, applying various distributions and other modelling methods. You all know it better than I do.
No task to prove or disprove anything. I see a consistent pattern on real price data. Pseudo-real data - except to look at some assumptions on it.

 

OK - I think this pattern needs to be analysed. Probably it is necessary to state IMHO, all a little more expanded, and more formalized ( in terminology ) and to open a new separate topic. If there is an interesting regularity it deserves a separate subject. The only thing I wouldn't do is to create a zigzag - I would generate a price series, and then I would create a zigzag from it.

 
SProgrammer >>:

я бы сгенерировал бы ценовой ряд, а потом бы уже из него построил бы зигзаг.

Why not build ZigZags directly from real data rather than generated data?

 
getch писал(а) >>

Why not build ZigZags immediately from real data rather than generated data?


Didn't say not to build from real data - that's as much as you like.

 
SProgrammer >>:


А не говорил что надо не строить из реальных - это пожалуйста сколько угодно.

That's how it was done then.
Applying the same 'pattern' to trading is a separate topic...