Robot Lab Project | Page 15 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Financial Markets Observatory Lab 1,095 replies

Indicators Lab 236 replies

Rel's Lab 103 replies

Nikola's Trading Lab 247 replies

Spud's Lab Accident 22 replies

Attachments: Robot Lab Project
Exit Attachments

Robot Lab Project

  • #281
  • Jan 7, 2015 11:47am Jan 7, 2015 11:47am
Quoting MMGood47
I have been trading the Ichimoku Kijun Sen cross and just wondered if it could be adapted into an EA. Long Entry: Price crosses and closes above Kijun Sen above the cloud and Chikou Span is above price Exit: Price crosses Kijun Sen SL: just below Kijun Sen at open TS: follows Kijun Sen Short entry opposite the above {image}
Kijun_Rider_01

First version attached as described.

Interesting
Attached File(s)
File Type: ex4 Kijun_Rider_01.ex4   12 KB | 512 downloads
File Type: mq4 Kijun_Rider_01.mq4   5 KB | 665 downloads
No guts, no glory
  • #282
  • Jan 7, 2015 11:56am Jan 7, 2015 11:56am
Quoting abokwaik
HA_Rider_04 Robot Attached New Inputs : F1_MA_Period = 200; // Filter 1: MA Period. 0 to disable F1_MA_Type = 0; // MA Type:0:Simple,1:Exponential,2:Smoothed,3:Linear_Weighted F2_Higher_TF_Minutes = 240; // Filter 2: Higher Time Frame in Minutes. 0 to disable Fixes : Trailing Stop Time Frame Issues Even with the Filters (Moving Average and Higher Time Frame), I am still not impressed by this Robot {file} {file}
Thanks abokwaik...will test and report.

Regards,

Ty
My job: I speculate and manage risk.
  • #283
  • Jan 7, 2015 11:57am Jan 7, 2015 11:57am
Quoting abokwaik
{quote} Kijun_Rider_01 First version attached as described. Interesting {file} {file}
Whuuuuuttttt!!!!

That was FAST man!

Are you sure you haven't coded MT4 before!?

Regards,

Ty
My job: I speculate and manage risk.
  • #284
  • Jan 7, 2015 12:25pm Jan 7, 2015 12:25pm
Quoting abokwaik
{quote} Kijun_Rider_01 First version attached as described. Interesting {file} {file}
Ok...Kijun_Rider set up...

I notice that it ALSO opens positions AS SOON AS the AutoTrade button is activated...Is this supposed to be the case?

Keep them coming Sir.

Regards,

Ty
My job: I speculate and manage risk.
  • #285
  • Edited 1:48pm Jan 7, 2015 1:28pm | Edited 1:48pm
Quoting TyLamai
{quote} Ok...Kijun_Rider set up... I notice that it ALSO opens positions AS SOON AS the AutoTrade button is activated...Is this supposed to be the case? Keep them coming Sir. Regards, Ty
Ty, don't you ever back test?
You go directly Forward. That's HOT.
Any way one candle wont make a difference, so why wait.
No guts, no glory
  • #286
  • Jan 7, 2015 1:35pm Jan 7, 2015 1:35pm
You might be interested in this free and open source automated system generator.
  • #287
  • Jan 7, 2015 1:47pm Jan 7, 2015 1:47pm
Quoting TyLamai
{quote} Whuuuuuttttt!!!! That was FAST man! Are you sure you haven't coded MT4 before!? Regards, Ty
I thought, Can I finish this one before those guys finish installing my kids new bedroom ? So I DID.

I am having fun here, don't I ?

Really guys, with respect to all, no Genuine ideas were posted so far.

I very much encourage you to think out of the box, maybe away from indicators.

I am sure many systems were developed around existing indicators. What we need is something New, Genuine, Out-of-the-Box, Never tried before. Or at lease we take an existing strategy (a successful one of course) and improve it further, get around draw downs, ... etc.

Think of simple ideas, for example ... OK this is too much for a Quoted response. Sorry, Ty. I'll complete this in another post.
No guts, no glory
  • #288
  • Jan 7, 2015 1:53pm Jan 7, 2015 1:53pm
Quoting algoTraderJo
You might be interested in this free and open source automated system generator.
Hi, Jo (if I may), Glad to have you here

I checked your Machine Learning thread, Great job your doing there.

Thanks,
No guts, no glory
  • #289
  • Jan 7, 2015 2:01pm Jan 7, 2015 2:01pm
Quoting abokwaik
{quote} I thought, Can I finish this one before those guys finish installing my kids new bedroom ? So I DID. I am having fun here, don't I ? Really guys, with respect to all, no Genuine ideas were posted so far. I very much encourage you to think out of the box, maybe away from indicators. I am sure many systems were developed around existing indicators. What we need is something New, Genuine, Out-of-the-Box, Never tried before. Or at lease we take an existing strategy (a successful one of course) and improve it further, get around...
abokwaik..im prepared to work with you but surely dont wanna play it live on this stage.

that is all I can say on this issue.
  • #290
  • Jan 7, 2015 3:34pm Jan 7, 2015 3:34pm
Quoting abokwaik
{quote} *** when you say Chikou Span is above price , you don't mean current price, Right ? I think it should be above the price at the bar where Chikou Span is. Right ?
That's correct .
  • #291
  • Jan 7, 2015 10:43pm Jan 7, 2015 10:43pm
Quoting 21vs7
{quote} abokwaik..im prepared to work with you but surely dont wanna play it live on this stage. that is all I can say on this issue.
Not sure what you mean by this....
We miss 100% of the shots we don't take
  • #292
  • Jan 7, 2015 10:48pm Jan 7, 2015 10:48pm
Quoting MMGood47
{quote} That's correct .
Did you try the Kijun_Rider_01 Robot ?
I hope you didn't miss that post (281).
No guts, no glory
  • #293
  • Jan 7, 2015 11:11pm Jan 7, 2015 11:11pm
I very much encourage everyone to think out of the box, maybe away from standard indicators.
I am sure existing indicators were exhausted by many systems .

What we need is something New, Genuine, Out-of-the-Box, Never tried before.

Think of simple ideas, for example ...

  1. Time Based

    1. Trade in the Direction of London 1st Hour at NY open
    2. Trade in the Direction of the Daily Candle at Mid-Day
    3. Trade in the Direction of the Daily Candle after London Close
    4. Trade in the Direction of the Weekly candle on Wednesday
    5. ....



  2. GAPS

    1. Trade Intra-day gaps
    2. Trade Weekly Gaps



  3. Percentage Based

    1. Trade instrument percentage change (lower and higher time frames)
    2. ...



  4. Bargain Hunt
  5. Range Break Outs
  6. .....

No guts, no glory
  • #294
  • Jan 7, 2015 11:21pm Jan 7, 2015 11:21pm
"Cut your losses short and let your winners run" was a very simple idea, no-indicators at all, just position management, which filters out bad trades :

Choose 3 consecutive time frames, (i.e. : H1,H4,D1 OR D1,W1,MN1)
Lets call them T1,T2,T3 where T1 is the smallest time frame.

  1. Open Buy and Sell on open of T1 with same lot size.
  2. At the close of T1, close the looser and keep the winner
  3. Repeat Step 1, and Step 2
  4. At the close of T2, close the looser(s) and keep the winners
  5. At the close of T3, close the looser(s) and keep the winners
  6. Protect you Big Winners with distant trailing stop

Back test results were impressive in my opinion

I don't know why it didn't capture your attention

No guts, no glory
  • #295
  • Jan 7, 2015 11:35pm Jan 7, 2015 11:35pm
Quoting abokwaik
I very much encourage everyone to think out of the box,

Hi Abokwaik

I like the idea of new ideas being tested

But was wondering what the overall goal is here.Is it to try and be profitable as fast as possible or to discover new ideas...

If you want to be profitable then there is nothing wrong with building old ideas and making them work IMHO but if we are doing research then I understand trying many different things.

I think there are many basic ideas that can be profitable but just like in all trading we need to accept Draw Down periods

All the best

Rooi
We miss 100% of the shots we don't take
  • #296
  • Edited Jan 8, 2015 12:22am Jan 7, 2015 11:37pm | Edited Jan 8, 2015 12:22am
Bargain Hunt Example ...

This was my comment on "EUR/USD Forecast May 5-9"/2014, written on May 4th.

Attached Image (click to enlarge)
Click to Enlarge Name: eur_usd_comment.PNG Size: 221 KB


And this is what happened to EUR/USD on May 8th.

Attached Image (click to enlarge)
Click to Enlarge Name: EURUSDDaily.png Size: 59 KB


That down move (continuing until now), is more than 2200 Pips in 8 months, (275 pips a month) !!

Such bargains happen over and over again ...
No guts, no glory
  • #297
  • Jan 7, 2015 11:46pm Jan 7, 2015 11:46pm
Quoting Rooi
{quote} Hi Abokwaik I like the idea of new ideas being tested But was wondering what the overall goal is here.Is it to try and be profitable as fast as possible or to discover new ideas... If you want to be profitable then there is nothing wrong with building old ideas and making them work IMHO but if we are doing research then I understand trying many different things. I think there are many basic ideas that can be profitable but just like in all trading we need to accept Draw Down periods All the best Rooi
Of course the idea is to Make Profitable Robots. But I don't like to end up by making a robot for every indicator in MT4.

I don't mind building on old ideas, in fact most of the robots done so far were based on old ideas.

We should treat trading as a REAL Business, "Competition" is so high ... So why not be creative ,
No guts, no glory
  • #298
  • Edited 12:38am Jan 8, 2015 12:28am | Edited 12:38am
Quoting abokwaik
{quote} Of course the idea is to Make Profitable Robots. But I don't like to end up by making a robot for every indicator in MT4. I don't mind building on old ideas, in fact most of the robots done so far were based on old ideas.
I agree.

Hope you dont mind me saying this but.

My view on systems are like this....

Find a pattern or a scenario or a market type that repeats itself and then build entry and exit conditions to take advantage of it as best as possible. That's simple and we all get that BUT now we must identify what market conditions are bad for the those entry requirements and minimize the effect.

The MA_Cross gives us nice results the way you have coded it to stack winning potions and to try and cut losers short and back tests give fairly good results even though there is plenty to be improved upon.

1min
Attached Image (click to enlarge)
Click to Enlarge Name: graph 1min.png Size: 16 KB


60min same period
Attached Image (click to enlarge)
Click to Enlarge Name: graph.png Size: 21 KB



Now if we think about why the system works....

A trend starts and then entries happen on the re-trace of the trend.Nice and simple and basic.I know it sucks in ranging market.We will get to that.

Look here.Same type of idea.Enter on the retrace of the trend
http://www.forexfactory.com/showthre...15#post7980115
http://www.forexfactory.com/showthread.php?t=518257

I could find plenty of systems like that work when the market breaks out and the we just jump on the retrace.

All we have to do is create "the best" filters we can to help with the trend , limiting entries in ranging markets , stopping the EA buying at worse levels when trend exhausts itself , and building trading it as part of a basket with the idea that we can diversify out of being dependent on just one pair and two currencies.

There is plenty we can add into the EA to filter out stuff.

Thats what I am thinking currently.

The most important is that market either trend or range and the best way to protect yourself is to enter on retraces.because re-traces work ok in ranging markets as well.

Let me know what you think.

Rooi
We miss 100% of the shots we don't take
  • #299
  • Jan 8, 2015 1:00am Jan 8, 2015 1:00am
Quoting Rooi
{quote} I agree. Hope you dont mind me saying this but. My view on systems are like this.... Find a pattern or a scenario or a market type that repeats itself and then build entry and exit conditions to take advantage of it as best as possible. That's simple and we all get that BUT now we must identify what market conditions are bad for the those entry requirements and minimize the effect. The MA_Cross gives us nice results the way you have coded it to stack winning potions and to try and cut losers short and back tests give fairly good results...
I agree with your view on systems, and we'll continue this,
Some Robots will shine and some will fade

Pinocchio for example, a Pin Bar strategy promoted every where. Failed on test, so we drop it and move on
Ochuko, an "Accurate Trading Strategy", same thing, we drop it and move on

We might test hundreds of robots, and end up with very few winners.
No guts, no glory
  • #300
  • Jan 8, 2015 1:17am Jan 8, 2015 1:17am
Robots can be divided into 2 categories, based on their exit.

  1. SL/TP
  2. Close and Reverse on opposite signal

A Simple way of evaluating 1st category Robots would be by testing it with 1:1 R/R ratio.
For 2nd category robots, we test without any TP or SL or Trailing .... .

Then we look at testing results, over different periods, tfs, pairs

  1. If its winning rate is over 50%, we continue enhancing (Money and Position Management, Filtering ... etc)
  2. If its winning rate is less than 50% we drop.

No guts, no glory
2 traders viewing now