Discussion of article "Developing a self-adapting algorithm (Part II): Improving efficiency" - page 4

[Maxim Romanov]

Sergey Voytsekhovsky:

Good evening.

I suspect that in line 256 sign 87 there is a typo, instead of "-" it looks very well "=".

Please confirm, I'll correct it myself, or maybe some hidden meaning I do not understand, error not burn, burns as Warning.

I don't understand the codes very much, I can't check it, but there may be a typo. I'll try to change it myself and see how it goes.

[Sergey Voytsekhovsky]

Maxim Romanov:
I don't understand the codes very well, so I can't check, but it could be a typo. I'll try to change it myself and see how it goes.

Good afternoon, the meaning of the line

if(mas_par[i].Pause>0 && mas_par[i].Series_Close_Time>0) mas_par[i].Pause = iBarShift(mas_symbols[i],mas_inp[i].TF,mas_par[i].Series_Close_Time);

if I understand correctly, is that:

WHEN. Pause is greater than zero (How many bars to wait for the opening in the same direction).

And... Series_Close_Time is greater than zero (Series closing time).

THEN... Pause is equal to the index of the candlestick on which the last series closed.

And since there is always a last closed series (except for the first run), so Pause will always be equal to the index of the candlestick on which the last series closed. So setting Rauza is meaningless. Nonsense.

But the "-" sign is even more meaningless, because then the line counted something and did not use it anywhere, again nonsense.

So here I'm wondering if maybe I'm misunderstanding something ?

I would be grateful if you could describe the changes after the correction. Ato I've already changed so much that it is easy to rollback, and to put the original downloaded version is still not in the theme. Thank you for your labour.

[Maxim Romanov]

Sergey Voytsekhovsky:

Good afternoon, the meaning of the line

if I understand it correctly, is that

WHEN Pause is greater than zero (How many bars to wait for opening in the same direction)

And Series_Close_Time is greater than zero (Series Close Time)

Then Pause is equated to the index of the candlestick on which the last series closed.

And since there is always a last closed series (except for the first run), so Pause will always be equal to the index of the candle on which the last series closed. So setting Rauza is meaningless. Nonsense.

But the "-" sign is even more meaningless, because then the line counted something and did not use it anywhere, again nonsense.

So I'm wondering if I've misunderstood something ?

I would be grateful if you could describe the changes after the correction. Ato I've already changed so much that it is easy to rollback, and to put the original downloaded version is still not in the theme. Thanks for your hard work.

The point of the pause is that after closing the previous series, the new one cannot start immediately, it has to wait for n candles before it starts a new series. And how many candles to wait is set in the settings. I will only try it on Monday, I'm having a weekend off). Share later what you have changed, how it works, it's interesting.

[Sergey Voytsekhovsky]

Maxim Romanov:
The point of the pause is that after the previous series closes, the new one cannot start immediately, it must wait for n candles before it starts a new series. And the number of candles to wait for is set in the settings. I will only try it on Monday, I'm having a weekend off). Share later what you have changed, how it works, it's interesting.

In your case (as it was) - the pause was always as set in the settings. The idea was to use a self-adjusting pause equal to the number of candles that passed after the candle on which the last series closed. It didn't work firstly because there was a typo, and secondly, if there was no typo, the algorithm would make the pause equal to the number of candles that passed after the last series closed, i.e. there would be no pause, deure it would be, but de facto it is always already passed - if the counting starts from the moment of closing the last series.

In general, I came to the conclusion that the line is unnecessary, no matter how you turn it, it will not work properly. I commented it out, I don't use it yet.

[Maxim Kuznetsov]

black/white is not 50/50, black is usually systematically more common. Not by a lot, but more often. It's close to a paradox.

if you plot +1-1, you will usually see a weakly rising trend.

[Maxim Romanov]

Maxim Kuznetsov:

black/white is not 50/50, black is usually systematically more common. Not by a lot, but more often. It's a joke, close to a paradox.

if you plot +1-1, you will usually see a weakly rising trend.

Yes, the price will move away in a distance roughly proportional to the root of the number of steps. Here just generated a million data -1+1, loaded into mt5 as m1 and switched to h4

But!!! fortunately the market is not -1+1)) there are differences.

[Maxim Kuznetsov]

Maxim Romanov:

yes, the price will go away by a distance roughly proportional to the root of the number of steps. Just generated a million -1+1 data, loaded it into mt5 as m1 and switched to h4

But!!! fortunately the market is not -1+1)) there are differences.

You don't need to put artificial quotes in.. I think that you have long passed this stage :-))

You can draw a chart on the real received quotes - black candle=+1, white candle=-1.

And you can deduce a lot of interesting things. In particular, the inequality of black vs white.

They are more or less equal only on small TFs (M1, for example), this is due to measurement error. a la "thermal" noise. Also on MN1 but for other reasons (no proper history plus exponential growth of volumes).

In the publication it is possible to make a fix about "for XX instrument and YY TF it is normal (i.e. flat) when 55 blacks on 45 whites". For gold the ratio is crazy

[Maxim Romanov]

Maxim Kuznetsov:

You don't have to put artificial cotyres in there. I think you're way past that :-)

plot the real quotes - black candle=+1, white candle=-1.

And you can deduce a lot of interesting things. In particular, the inequality of black vs white.

They are more or less equal only on small TFs (M1, for example), this is due to measurement error. a la "thermal" noise. Also on MN1, but for other reasons (no proper history plus exponential growth of volumes).

In the publication it is possible to make a fix about "for XX instrument and YY TF it is normal (i.e. flat) when 55 blacks on 45 whites". For gold the ratio is crazy

Got it, yeah. You can try, but I've almost given up on candlesticks. There's too many factors that aren't accounted for. Of course, you can take into account everything that affects the size, shape and direction of the candle, but then it will be a completely different level.

[Maxim Kuznetsov]

Maxim Romanov:
Got it, yeah. I can try, but I've almost given up on candles. There are too many unaccounted factors in them. Of course, you can take into account everything that affects the size, shape and direction of the candle, but then it will be a completely different level.

and I am not talking about figures of 2-3-4-5 candles :-) they have too small a window no matter how you turn it.

the article is correct that the ratio is important, but it is not quite justified that 50/50 is the norm. This conclusion is drawn from "general postulated" considerations, which it is not

---

diverting from the topic, to the equally popular topic of "trends/waves". In an impulse move, quotes are up and b/w up, in a pullback they contradict. A pullback movement is formed by a smaller number of counts.

The "pullback" is highlighted - the quote goes down significantly, but there are more counts/candles up. The whole downward movement was formed by a counting number of candles, and only during the Euro-session.

It is difficult to make out such a pattern in real time, but it is worth taking into account when marking the history

[Andrey Khatimlianskii]

Maxim Kuznetsov:

but it is not well established that 50/50 is the norm. This conclusion is drawn from "general postulated" considerations, which it is not

Analysed in the first article - https://www.mql5.com/en/articles/8616