Entophy(ish) Automated | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Automated Daily/Weekly Pivots here 117 replies

Automated or semi automated trading? 3 replies

do automated trading systems work or... 21 replies

Automated trading with MetaTrader4 3 replies

Tradestation soon to be offering automated Forex trading 7 replies

Attachments: Entophy(ish) Automated
Exit Attachments

Entophy(ish) Automated

  • #41
  • Sep 23, 2015 9:07am Sep 23, 2015 9:07am
No bash intended but you're kidding, right? 2 weeks winning means JACK.

I've designed EA's that are extremely profitable for YEARS of back testing but then they miserably fail. Oh, and "optimization" is curve-fitting's brother. Trust me, I've been into this for way too long. Show us a back test that is successful for a period of 10 years with 99%+ modeling-quality data and then you are a winner/hero ;-) M1 systems MUST be tested with extremely high-quality data otherwise the tests are no good.

Again, I'm not bashing anyone.

Quoting Magix
{quote} lol You are comparing the quality of backtests on what has admittedly not been optimized. I'm not even sure which EA you have your "Quality Backtests", run on. In any case. The triggering and trend definition on MyENT is a little different than what I currently have in test on a Correlation EA and has been testing live for the last couple weeks. I found a little something interesting when I started playing with the ideas put by Fxenthropy and decided to partially incorporate the ideas into my core bot and released the unfiltered for...
  • #42
  • Sep 23, 2015 10:09am Sep 23, 2015 10:09am
Quoting rkp
No bash intended but you're kidding, right? 2 weeks winning means JACK. I've designed EA's that are extremely profitable for YEARS of back testing but then they miserably fail. Oh, and "optimization" is curve-fitting's brother. Trust me, I've been into this for way too long. Show us a back test that is successful for a period of 10 years with 99%+ modeling-quality data and then you are a winner/hero ;-) M1 systems MUST be tested with extremely high-quality data otherwise the tests are no good. Again, I'm not bashing anyone. {quote}
Kewl story Bro.

I will take any 2 weeks of forward over all of the historical backtest bullshit you can post.

And, that's not personal either. Use the FF search, you will see how many times I've posted the exact words, "Backtests are Bullshit".

Designing a system and writing the code are 2 very different things. Where you look at your data from a backtest, I am more interested in watching the logic trade and probabilities of success. In this, I have created and coded several systems that have done and continue to do very well, all of which I have provided for free/donations to charity.

As much as I do appreciate a comment dropped, if you feel the need to talk about quality of backtests anymore, find another thread.

If you want to talk actual design and possible improvements, please continue.

Consider this a final warning.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #43
  • Sep 23, 2015 10:42am Sep 23, 2015 10:42am
Still going strong
Attached File(s)
File Type: rar Desktop.rar   13 KB | 385 downloads

desktop.rar is history file demo account
  • #44
  • Sep 23, 2015 1:43pm Sep 23, 2015 1:43pm
Quoting Magix
{quote} Kewl story Bro. I will take any 2 weeks of forward over all of the historical backtest bullshit you can post. And, that's not personal either. Use the FF search, you will see how many times I've posted the exact words, "Backtests are Bullshit". Designing a system and writing the code are 2 very different things. Where you look at your data from a backtest, I am more interested in watching the logic trade and probabilities of success. In this, I have created and coded several systems that have done and continue to do very well, all of which...
2 weeks of forward testing are still too short a time IMO. I won't bother you any more. Good luck!

I'm out.
  • #45
  • Sep 23, 2015 1:52pm Sep 23, 2015 1:52pm
Quoting rkp
{quote} 2 weeks of forward testing are still too short a time IMO. I won't bother you any more. Good luck! I'm out.

I just wrote the fucking thing 2 weeks ago.

How much live data can I give without the use of a crystal ball or bullshit backtests?

Yeah...You're out.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #46
  • Sep 23, 2015 2:07pm Sep 23, 2015 2:07pm
timeframe?

Quoting pitcher
Still going strong {file} desktop.rar is history file demo account
  • #47
  • Sep 23, 2015 2:29pm Sep 23, 2015 2:29pm
1m
  • #48
  • Sep 25, 2015 12:27am Sep 25, 2015 12:27am
dont know what im doing wrong loaded it up on a Global Prime demo a la bazze pairs / timeframe and its just steadily haeomrraghing money
  • #49
  • Sep 25, 2015 1:09pm Sep 25, 2015 1:09pm
Quote
its just steadily haeomrraghing money

this is a trend based EA , Best to use only m.tue.wed.th. 30 min after London Open to London close.. GL --

thanks for the great EA magix - u da MAN
Learn THEN Unlearn
  • #50
  • Sep 25, 2015 4:52pm Sep 25, 2015 4:52pm
Quoting HighFivve
{quote} this is a trend based EA , Best to use only m.tue.wed.th. 30 min after London Open to London close.. GL -- thanks for the great EA magix - u da MAN
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #51
  • Sep 25, 2015 4:55pm Sep 25, 2015 4:55pm
860 downloads between the 2 EAs posted on page one and we have 3 and a half traders working with this?

I had forgotten why I get frustrated with the FF crowd.

Now I remember.

The overall week wasn't terrible. A little ruff waters for some of the pairs but after the change from 34-70 on the beltline, the week seemed a little smoother.

There is one observation that I have had and perhaps another modification to the EA...private version anyways...

Thanks to the few testers this week and have a great weekend.

Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #52
  • Sep 26, 2015 2:50am Sep 26, 2015 2:50am
Hello Magix, i have been testing your ea for a week now, had some big drawdawns but recoverd good.
I think it wil be a good idea to put tradingtime in the ea, an perhaps something as a basket close for safety.
Wil do some backtesting on the idea, and after another good week ill try it on a small live account.
keep up the good work, tnks.
.
Attached File(s)
File Type: rar test 21-29.rar   18 KB | 330 downloads
  • #53
  • Sep 26, 2015 3:36am Sep 26, 2015 3:36am
Hello Magix Welcome back . Signed .
  • #54
  • Sep 26, 2015 4:56pm Sep 26, 2015 4:56pm
Quoting Matrix8756
Hello Magix Welcome back . Signed .
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #55
  • Sep 26, 2015 5:03pm Sep 26, 2015 5:03pm
Quoting pitcher
Hello Magix, i have been testing your ea for a week now, had some big drawdawns but recoverd good. I think it wil be a good idea to put tradingtime in the ea, an perhaps something as a basket close for safety. Wil do some backtesting on the idea, and after another good week ill try it on a small live account. keep up the good work, tnks. .{file}
I've re-written the signal based on a little different idea of trend analysis, mostly based on my correlation trader and it's piss poor performance this week.

Working with the idea of the time filter seems like a decent enough idea, but what happens when the clock runs out and there are positions that either could have been closed by logic, or cost averaged to close in profit?

The idea of unprotected funds by either the stop loss or the trading logic, may not be the best idea, but I will continue to ponder it.

I think I will fire up 2 different versions next week.

Version 1 - just a different trend filtering idea, but still working with the belt and trigger.
Version 2 - work with a dynamic stop loss filtering based on the daily trend idea. When the trend changes, take the first possible exit and wait for triggering to the new trend.

With NFP upon us, I am willing to bet that Version 1 will do a little better, but possibly over the long run, Version 2 will be the way to go.

I will be putting these to a little visual backtesting to see if the logic actually pans and see what can be optimized over the immediate.

Will still run the same pairs, but believe the AUD crosses will be the ones to take the lions share.

Will get em cooked up and add a couple more TE links to the thread.

Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #56
  • Sep 27, 2015 11:00am Sep 27, 2015 11:00am
Quoting Magix
Version 1 - just a different trend filtering idea, but still working with the belt and trigger.
Version 2 - work with a dynamic stop loss filtering based on the daily trend idea. When the trend changes, take the first possible exit and wait for triggering to the new trend.
2 New versions written and ready for the new Market Week!
Trade explorers added to the thread.

Version 1 - MyENT2
Version 2 - MyENT3

Off of visually backtesting, the triggering seems to be in keeping with the logic as put forth, but will be able to verify and make corrections as the week moves on. I have dropped the Correlation filter, but may in the next versions, use the same trend recognition the Correlation filter. For now, let's see what the monkey fighters can do on their own.

E Haw!
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #57
  • Sep 27, 2015 1:18pm Sep 27, 2015 1:18pm
An interesting byproduct has been created by the Newer Trend Filtering. Dynamic no trade ranges.

Seemingly, tight markets and over extended markets seem to share in the same dynamic filtering. Tight Markets will trade a lot more restrictively and massively overextended ranges will stop trading.

The image shows the tight ranges that were traded just after daily roll over, really before the TK session. The second trade shows what would seemingly be a Press Spike. After that, and appreciate the reason for the MACD histogram in that it is an easy visual representation of the 14EMA and 70. If the 14 is above 70, and the trend has been determined to be short, it will look for short triggers.

Given the trend analysis and the range of market, the blue box shows where multiple trades could have been taken, but were sat out.

Perfection?

Heck No!

But would this help eliminate selling absolute bottoms and buying absolute tops?

I believe it would.

Very Kewl and only something that I really picked up on by watching the actual trades run in a visual backtest with all indies loaded.
Attached Image (click to enlarge)
Click to Enlarge Name: bipro.png Size: 55 KB
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #58
  • Sep 27, 2015 1:39pm Sep 27, 2015 1:39pm
Quoting Magix
An interesting byproduct has been created by the Newer Trend Filtering. Dynamic no trade ranges. Seemingly, tight markets and over extended markets seem to share in the same dynamic filtering. Tight Markets will trade a lot more restrictively and massively overextended ranges will stop trading. The image shows the tight ranges that were traded just after daily roll over, really before the TK session. The second trade shows what would seemingly be a Press Spike. After that, and appreciate the reason for the MACD histogram in that it is an easy visual...
Looks really good I would love to test your new versions this week are you planning on sharing them?
  • #59
  • Edited 5:20pm Sep 27, 2015 4:34pm | Edited 5:20pm
Hi magix Q; for myENT are you now recommending we change MA1 to 70 from 34 -- also what setting would you recommend if using EA on a 5 min chart --thanx in advance ..
Learn THEN Unlearn
  • #60
  • Sep 27, 2015 5:49pm Sep 27, 2015 5:49pm
Quoting HighFivve
Hi magix Q; for myENT are you now recommending we change MA1 to 70 from 34 -- also what setting would you recommend if using EA on a 5 min chart --thanx in advance ..
70 will force a delay. If you are running uninterrupted, having this delay will definitely keep your buttcheeks from puckering in an overnight reversal.

If you are playing the trend prior to NY open and until after London Close, manually starting and stopping the EA, 34 will be fine.

The settings on my last week TE were 70 and the Stoploss=1000.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
0 traders viewing now