Avalanche - page 521

 
Ftor_007:

The drawdown has become much lower with two order corridors - external and internal, respectively with geometric and arithmetic progression of lot growth. During a flat, there is no longer a large set amount of lots, the probability of MC is also reduced. The USDJPY pair gives the best results. As soon as I will be able to insert indicator into expert function, I will try to increase profitability. Deposit in the tester was set at $70, lot 0.01.


and you can give it to )))))))))))
 
Cheis0590:
and you can give it to me)))))))))))

In the left column , download two avalanches - and the knives fly!

Post your results here - we'll discuss...

 

Hello, I really like avalanche_7, in the tester it works somehow... Well, it works on all pairs except the Japanese yen... My first try was to put it on the real account but it didn't work at all and i couldn't find any error in the log ...

Is it possible to fix it a bit so that it could trade on the real, or is it not realistic?????

Files:
 

This is another problem, I downloaded Avalanche V_6.3_ EA in the tester works on all pairs, but it closes on trailing stop, and everything would be fine, but there is one big bug, trailing stops and starts moving from stoploss boundary.

And before it even reaches the open order it often closes when the price moves in the opposite direction and starts to make lots of losing trades...

Can this be fixed, so that it would start its movement only at breakeven, when the price has already passed its open order and a profit has been made?

Files:
 
Roman Shiredchenko:

In the left column , download two avalanches - and the knives will fly!

Post your results here - we'll discuss...

The EA variants are not working!!!
 
And in the Av02 advisor, there's also a bug, it doesn't work at all, KMartin is multiplying lots by martigail, so it always works with just one lot, and doesn't multiply anything!!!!
Files:
Av02.mq4  17 kb
 
medved-0775:
There are not working versions of EAs out there!!!
medved-0775:

Hello, I really like avalanche_7, in the tester it works somehow... Well, it works on all pairs except the Japanese yen... I had a hard time believing it would work, but I found it wrong and the trading robot would look like a real EA.

Can you fix it a little, so it can trade on real account or it's not realistic?????

I'm not sure how to use it, but I'm sure you'll get the best out of it.

Use these - get the current ones from here:

https://www.mql5.com/ru/forum/139975/page381

and from here

https://www.mql5.com/ru/code/22561

Martingale by RSI indicator - expert for MetaTrader 4


and from here:

https://www.mql5.com/ru/code/20612

RSI Indicator Mediating Expert Advisor for MetaTrader 4

Vladimir Khlystov


I use them myself, looking at my trading conditions....


This one has a double side effect :-) see generally:

https://www.mql5.com/ru/code/25350




учитесь зарабатывать селяне [Эпизод 2] !
учитесь зарабатывать селяне [Эпизод 2] !
  • 2019.08.27
  • www.mql5.com
Прошу селянскую партию писать сюда, а то тормозит очень ваша старая веточка...
 
goga #:

JonKatana 12.03.2010 21:39

By the way, EA writers, instead of almost continuous meaningless messages, we'd better write an auxiliary EA with input parameters "Distance between levels", "Desired profit in pips".
The algorithm is to create infinite horizontal corridor, for example from the price upwards +20 points and downwards -20 points. Then from its borders, profit levels are drawn up and down (in this example, 40+profit, e.g. 40+10). Then the Expert Advisor consecutively moves along the timeline and starts counting successive touches of the initial levels. It stops counting when the price has touched a profit level of the same direction. If the price goes immediately in one direction, the Expert Advisor stops counting not after 50 points, but after the profit level, i.e. 10 points - it is a case without any reversals at all. After that the number of reversals is written in the log-file (preferably with indication of date), the levels are moved to the time of touching the profit level and the cycle continues.
At the end of the run it is desirable to display the statistics. For example, profit (without turning over) - 600 times, with one turn - 900 times, with two - 700, with three - 200, with four - 30, with five - 3, with six - 1, with seven - 0, etc.
This will allow you to choose the most advantageous corridor and establish how many turns are possible at that distance. Add "just in case" to zero (ie do not exist in this corridor), the number of turns 3-5 pieces - and you can safely earn.

JonKatana 12.03.2010 21:49

arnautov >>:

I'm not so sure that you're right, you're right.

I'm just following your recommendations on the poundbucks right on the first run has received 11 reversals.

If you have a simple robot, try it on the history and then discuss the probabilities.

I understand you know how to program. Well, shove the critics at their throats with the test from 1999 with a maximum of 4 U-turns.

If you've done a "run-through", it means you already have a "robot". Why would I write it then? Or have you "raced" something else?

I only pay attention to constructive criticism - other posts are ignored.

It makes sense to test on periods close to the present - month, quarter, year. The nature of the price behaviour in 1999 will not tell anything now. The market is changing too fast - that seems obvious to me to everyone.

JonKatana 12.03.2010 22:01

arnautov >>:

And I thought you knew how to program.....

Sorry, I did not mean to insult you by suspecting you know how to write experts. It is clear - your destiny is to soar in high matter.

You are mistaken. That's not the point at all. I know how the DC works "from the inside", not from the client side. There are several legal ways to disrupt or terminate a maximum error-proof EA. Even if the terminal is on a dedicated server and not on a normal computer. And if one pending order does not trigger after several price reversals, it will rapidly wipe out the entire deposit.

Using Avalanche, one can trade only manually, with constant monitoring. The Expert Advisor is needed only for calculation of initial volume and distance between levels.

JonKatana 12.03.2010 22:08

arnautov >>:

If the market changes very quickly, you cannot use the history to determine the distance in the future. Everything changes.

But if you can determine it, the test must be effective since 1999.

I've written periods for testing - one month, one quarter, one year. Please read carefully. For the 1999 test we need to calculate the distance between levels for 1998. How - I wrote above using Average True Range indicator.
But you are trying to run Avalanche (if you are testing it - no one has seen your EA, so so far all your claims are not proven by anything) by 1999, using the corridor calculated by 2009-2010. "Back to the future"? And if in 2011 the average daily EURUSD move will be 900 pips and therefore the spread between the levels will be 300 pips - will you also claim that "it should work in 1999"? That is not correct.

JonKatana 12.03.2010 22:15

Roger >>:

And here, if I may elaborate a bit more (and I think colleagues will support me), it's a very hot topic.

It's irrelevant to the Avalanche discussion - already there's 80% of the thread's rubbish. In addition, since you and "colleagues" find the topic "burning", it means that you have already encountered it in practice and understand that I am right. The specific methods are not that important - there are probably more of them than I know, and new ones are being invented all the time. The important thing is that you can only trade on Avalanche manually.

JonKatana 13.03.2010 12:59

arnautov >>:

I don't have uninterrupted history of minutes. That's why testing with 20 pips makes no sense, because it will be inaccurate.

My Expert Advisor triples lots, while your version doubles them.

Since there is no continuous history of one-minute periods, what are you testing? One-hour candlestick reversals? And who needs them?
If we triple the volume, it is unclear how we obtain 10 or even 11 reversals. At threefolds, passing by the price of HALF of the distance between the levels is enough to make you breakeven. Moreover - to exit the "Avalanche" faster, you can immediately quadruple the opposite volume - then breakeven starts from 14 pips (for an initial distance of 40 pips). Want even faster? Increase volume fivefold - break-even starts from 10 points (with the same initial 40). And so on. And you must have very negative luck to allow the price to fail to pass 10 points, turn around and go in the opposite direction. If desired, you may set such a volume that a breakeven point is opened after ONE point has been passed. What do you mean by 10 reversals?

And I am sure your Expert Advisor does not take into account the simplest and most natural option - when the price, having touched the first level, goes further in the same direction, making profit IMMEDIATELY. In this case you don't have to wait for the distance between the levels, because you are not at a disadvantage. And you can close the order at least 5 points after the level - they are profitable.
"Avalanche" is created for extreme cases - when the price immediately goes in the opposite direction, having caught the order - only then you switch on my algorithm. And in the vast majority of cases, the price will pass some distance after the level, bringing you a profit without a single reversal.

You will not believe it! This is the year 2021. Nothing like this has ever been written in 10 years, either on a paid or free basis.

 
Vladimir Gulakov #:

You won't believe it! It's the year 2021. Nothing like this has ever been written in 10 years, either on a paid or free basis.

and I was expecting something explosive from neural networks... And how long will currencies and consequently their rates live - not much left.

 
We're all going to die (