Offering the game !!!!! - page 8

New comment
 
FION писал (а):
 I agree - it's not a ToR, but an algorithm designed for a more or less trained trader.
Wow... You said it yourself ... For a trader, not a programmer... It's not an algorithm at all. Just a verbal description of an idea. Far from an algorithm, even farther from the ToR. Apparently, now the programmer's skill is largely determined by his or her ability to write programs based on a verbal description. And the qualification of customers is falling.

FION wrote (a):
Actually I also tried to build complex systems with precise formalization, but experience shows that making the system more complex by refining the formalization does not improve the result, i.e. there is a certain quality limit to the initial idea. Hence a simple conclusion - we must discard unviable ideas in time.
I agree... But you need a criterion for selecting ideas. This criterion is formed with experience. But experience is a dangerous thing. It makes you throw the baby out with the water.


FION wrote (a):
 But the principle, the more ideas, the better - remains relevant.
There is another principle. Less is more :)) You can drown in a large number of ideas and remain an eternal student.
 
KimIV писал (а):

FION писал (а):
 Программисту для написания кода необходим формализованный алгоритм процесса торговли советника.

Согласен с Вами. Что называется ППКС (подпишусь под каждым словом). Но, увы... То, что Вы описали ниже, не является формализованным алгоритмом. Алгоритмом я считаю такое описание процесса, который сможет выполнить человек, не владеющий предметом. Например, из Вашего описания нужно знать, что такое каналы и как они строятся. Что означает подход цены к границе канала? Разные люди вкладывают в это разный смысл. Для одного подход - это вход в 10-ти пунктовую зону, для другого пересечение, для третьего ещё что-нибудь. К чему мы в итоге приходим? К тому, что программист сам занимается формализацией словесного описания процесса. Результат - неправильная реализация того, что хотел заказчик. Причина - общение на разных языках, вкладывание в одинаковые слова разного смысла. Как этого избежать? Нужно и заказчику и программисту участвовать в процессе формализации, алгоритмизации. Еще лучше обоим разработать свои алгоритмы и сравнить их, найти несоответствия и устранить. За весь мой 15-тилетний программистский опыт у меня был только один заказчик, который пытался разговаривать со мной на понятном мне языке. В ТЗ было всё необходимое: учтены все условия, все формулы, умолчательные значения и прочее. Мне практически ничего было формализовать. Всё уже было сделано. Видимо, это идеал. А реалии таковы, что большинство деталей из заказчика приходится вытягивать, то есть типичный заказчик не заботится о том, что может понадобится программисту для реализации его идеи. Какие цифры, какие графики, таблицы, формулы или просто пояснительные иллюстрации.


In most cases, the customer is not even aware of all these details and subtleties. I found one option - program all possible (more or less realistic) options, make options and write instructions - it is much faster and less nerve-wracking;-)
 
KimIV писал (а):
FION wrote (a):
 I agree - it's not a TOR, but an algorithm designed for a more or less trained trader.
Wow... You said it yourself... For a trader, not a programmer... It's not an algorithm at all. Just a verbal description of an idea. Far from an algorithm, even farther from the ToR. Apparently, now the programmer's skill is largely determined by his or her ability to write programs based on a verbal description. And the qualification of customers is falling.

FION wrote (a):
 Of course I also tried to build complex systems with precise formalization, but experience shows that making the system more complex by refining the formalization does not improve the result, i.e. there is a certain quality limit to the initial idea. Hence a simple conclusion - we must screen out unviable ideas in time.
I agree... But you need a criterion for selecting ideas. This criterion is formed with experience. But experience is a dangerous thing. It makes you throw the baby out with the water.


FION wrote (a):
 But the principle, the more ideas, the better - remains relevant.
There is another principle. Less is more :)) You can drown in a large number of ideas and remain an eternal student.
I used to be engaged in development of precision electronic devices for vibration measurement - consequently, I can imagine the process of searching for solutions. To understand better or worse, if it is not immediately clear (from experience) it is necessary to try, strangely enough, at the championship very simple algorithms turned out to be the most successful, although I approximately expected it.
 
FION писал (а):
.... strangely enough, very simple algorithms were the most successful at the championships, although this is about what I would have expected.

This is not strange at all.
The goal of the Championship is not to identify the most resilient Expert Advisors, but to attract a wide set of Expert Advisors to participate in the competition.
That's why the prize fund is so big. This, in turn, attracted not only seasoned EA writers, but also freshly minted ones. And what do these newcomers want from the Championship? They only want to win prizes, not to prove the viability of their experts. Hence the simplicity of their Experts and their excessive risk-taking.
As for the Expert Advisors with a long experience, they are naive. They are fighting for an idea. They want a "beautiful" Expert....
 
Michel_S писал (а):
FION wrote (a):
.... strangely enough, very simple algorithms were the most successful in the championship, although that's about what I expected.

That's not strange at all.
The goal of the Championship is not to reveal the most resilient Expert Advisors. The goal of the Championship is to attract a wide audience of EA developers to take part in it.
That's why the prize fund is so big. This, in turn, attracted not only seasoned EA writers, but also freshly minted ones. And what do these newcomers want from the Championship? They only want to win prizes, not to prove the viability of their experts. Hence the simplicity of their Experts and their excessive risk-taking.
As for the Expert Advisors with a long experience, they are naive. They fight for the idea. They want a "beautiful" Expert....
It is often the case that simplicity is where beauty lies.
 
Drinking and peeing.
 
Rosh:
Drinking and peeing.
If you drink but can't pee, yes!
 
Rosh:
Drinking and peeing.

eating and pooping :)
living...
 
FION писал (а):
 If you drink and can't pee, yes!
What's that got to do with it? Roche showed simplicity in an original way, not pathology
 
Sergey, I warned you not to get too serious here :).
By the way, everyone has their own definition of 'nice/not nice'. Therefore, an EA written using only MACD with an exorbitantly long period, plus errors in code, plus persistent failure to get out of the top 10.... it's undeniably beautiful! Not wise at all though. ...

OK, I'm leaving now, sorry for the flooding, but everyone's getting a bit carried away here.
12345678910
New comment