Lost Sock Evolution - Automated

Hi All,

So, my single pair (GU) 1.3 demo has ended, EA was caught in a trend and eventually killed the account.
During this time it did take some small profits but not enough to negate the early losing trades.

Version N, on a 5 pairs, 1K demo, is currently at 30% DD from a single pair (EG) and continues to buy, so far, the demo account has enough margin so I'm keeping it automatic.

On my small live account the GU move with V1.3 killed the account so I've re-deposit and will launch hte EA again when I have accumulated more funds.


It's nice following this EA, seeing how and when it tkaes trades, still not able to understand it completely but I can, at least, understand some of the trades' logic.

Take care, my friend.

Be well in your journeys.

Rev NN must be a result of genius coding. Magix is absolutely brilliant with this. Thank you.

I see its been quiet in here.
Results below



Everything going as it should I think. Not bad for 3 weeks of test.

I will share a little obvious trick. I checked on most of the TEs and noticed that low ADR pairs close pretty faster than high ADR pairs. So in the new account, redistributed the lots.

50-70 ADR: 0.05
70-100 ADR: 0.03
>100 ADR: 0.02

I however allocated 0.05 to some high ADR pairs e.g. GJ and GC because I noticed the turnaround time for trades were also quick most times.



Play around with the settings, the bot is cool.

Cheers.

hi, magix, why you delete your eas files in your first post?

Hello! I am late to the test party.

Any TE still required here?

Got a couple of modifications to finish...will have something out early on next week!



Happy Easter!!

\m/

Alrighty Then...

I know...finish the code and get it released...but here's the thing...Long day, half in the bag and really, I got enough issues with this coding shit when I'm totally fucking sober...so, in the meantime and in-between time...let's check some of the TEs out and analyze what this all means and how we get shit working better in the up and coming...

Gunna start off with Evolution 1.1.


This is the version that I set to Summ Start Lot...ie, it adds incrementally to each position...1, 2, 3, 4, 5, 6.....

Before the total dogshit action and spread variance of good Friday, this motherfucker was actually holding zero positions and all was in profit.
Which is kewl...

$2k Start, M1 chart, 7 pairs in trade - 20 days of action (minus the weekends) and was only set to increment after the first few days.

This motherfucker has seen some Draw Downs - To be Sure - but I also started it when EURAUD had already taken off and had a few positions that it needed to cost average out.

What does that mean?

Well...the cost averaging is A LOT more aggressive on both evolution and revolution than the original 'start trading logic' is, so it had to bury itself a little deep on some of these before it made good on all of them. In all honesty, and due to the logic and what I've seen with others using the Martingale Setting, I figured it'd be BUST long before now...

Instead...

50% gains in what is really a couple of short weeks worth of trading on a $2k start with never seeing more than 50% in Draw.

That's actually pretty fucking impressive if you ask me...

BUT WAIT!!

There is a change that will be coming that will help on the original trend definition on the primary to help keep a little better in the good graces of market.

For all intents and purposes, it will soon look to start trading more by way of breaking out of range and also be a little more sensitive on rebounds. Given it's current success and what I have for my personal Marti...this should put it in the running as the absolute BEST that Evolution can be!!

In theory of course...we know nothing until my coding proves positive and the ideas actually work out...and of course, there are always exceptional market situations.

Risk is always on!

Here's the thing...as a part of this, it will be looking at Daily Open and Close...the logic should be traded to a TRUE DAY and by that, I mean it should trade NY Close as daily open and work in a 24 hour cycle. This is where shit gets deep...if you have no idea when NY roll over is, or if your broker has their EOD at NY Roll, you will need to figure that shit out. Part of the reason this hasn't been implemented up to now is because it causes a lot of questions that will be repeated over and over that my attention span and patience have little tolerance for...

Yeah...mostly...I am an asshole.

But I am trying to be better for you...just going to have some hair pulling moments ahead of us.

Here's whatcha need to do...

Look at the FF trade day, home page, bottom right hand corner and simply confirm that your brokers end of day aligns with it...otherwise...you will need to adjust.

OR

If you believe that London is the better start to the day...figure that shit out, same way, and adjust!

I've been trading NY roll for years now and have it applied to my personal logic and traders, seems successful enough so I don't bother fucking with it...that's not to say you need to do the same.

AFTER THIS...

Study the difference between Fixed Start and Sum Start lot sizes and make a differentiation on which one you would prefer to use.
Summ Start, a lot more aggressive, a lot more risk, a lot more profit.
Fixed Start, slow and consistent and easy to close out the week at end of week...

Once this is all put together and compiled, I will be adding it to post one and will update the thread.

Revolution NN...

Gawd, I love this motherfucker!!

After having made a change - to my personal versions - not thread updated...it's managed to keep the accumulated positions for redistribution lower and still maintain a fairly good clip on the profiting side of things.

At just over 1 month in...

Sitting with $1000 in profit and very low DD, trading 21 pairs.

No...not huge, but consistent and after adding in a Better Filter for the Homeostatic direction, it keeps itself a lot better in check.

Now...I could just release this motherfucker and be VERY positive about its future...but why leave it just at that if we can shake shit up a bit?

How So??

Well...it's like this...

The original version was written for the M1 charts...which minimizes the range of profiting. Keeping low DD and consistent profit is good...but what if we can stretch out the charts and increase the average closing from 5 pips to 6, 8, 10???

How??

Glad you asked...

If we increase the TF, we can stretch the range.

By being a little less sensitive on the time frame and closing logic, we can possibly move this fucker to WHOLE NEW LEVELS!!

Up to date, this could have run 21 pairs on a lot lower account value...meaning, we can either increase the lot size or decrease the account start and really make this fucker work!!

Personally...I LOVE what bishopdotun wrote a couple of posts back!!

Dude is a fucking GENIUS!!

ICYMI ->HERE

Where we have been treating all as equals, he has figured out the HACK to pairs based on their ADR - Average Daily Range!!

Brilliant!!

If we know we are working in shorter ranges, we are minimizing DD and can increase the lotsize.
If we know Pairs typically close out faster, we can adjust range, profit and account size...

I've been working through some of this manually - as applied to AUDUSD, but never really considered different settings for the EAs...

THIS is why I love working with a community.

I am a total fucking moron, so when the rest of you are chiming in, no matter how trivial your posts may seem to you, they CAN make HUGE strides in our community of progress!!

LOVE IT!!

Working on Rev2...again...will update as is posted!!

IN THE WORKS...

Now, I know I said I wasn't going to release a version of GridMeOut...but see that there may be a great market potential for a lot of manual traders with this idea.

I am working at a public release of a CLOSE ALL GMO EA that can hopefully work as well for others as does for me!

There really is a lot to this, but the potential is fucking AMAZING!!

Imagine working with ideas from Davit's Pivot thread, setting your first trade either manually or as a pending order, and having the EA add as is needed and close out at your predetermined profit??

I will have a very GENERIC version in play and will be opening a new thread for this and hopefully will be leaning on a Brother Shitposter who fed us the idea for NN -> Redeflect ->TE-> HERE

Small setback on this week, but check out his manual EQ!

AWESOME!!

This is what we are working on with our ideas and if we can further garner ideas from him, as well as other manual traders, imagine what we can achieve?

Q1 is OVER and we are just getting started...

CAN YOU FEEL IT???

It's going to be a fucking AMAZING YEAR!!!

Anyways...

Apologies for delays in replies...you may have seen me chiming in on other threads and ignoring this thread and emails...need to do this every once in a while to just get some fresh perspective...clean out some cobwebs!!

Long story short...

Will be getting back into shit after the family weekend and we will all be hitting this market like a FUCKING ATM!

\m/

Awesome Magix, fully understand where you coming from.

Enjoy this special time with your family, and I look forward to next week, and to the future exciting developments ahead.

Happy Easter to you bud.

Along these same lines I was thinking the the "Minimal Profit Close" could be tied to ADR - like a fraction or % of the 5 day average ADR. It would be an input. This would make the capture for every pair dynamic and more relevant I think. For that matter I guess ADR would also have relevance to the "Step Order". I'm sure that using ADR in the right ways would tune things better than a 1 size fits all approach.

Things gonna get exciting

Just added an old tool I used to play with on Post 1.

i-Sessions indicator.

This will help you plot out the individual sessions, may be useful with the new EAs coming.

It does not automatically configure broker to actual Market Time or align sessions, those are manual inputs.

SPINS:
For NY...if your true time for NY Close is 24:00, you will have to enter 23:59...
IF you only want a single session to show, change the color of other sessions to None.

An interesting market view to be sure.

NY Session for EURUSD:

I have looked at these as ideas as well...

My major issue in past was, what happens if the daily trend reverses and you are caught on the wrong side?

But!!

Given that we are now playing bi-directional with group ranged profiting while still allowing other positions to remain open, may give us another lift on profit range.

Good stuff, brother!!

Let's get these in play and I will work on adding as we grow!

codes actually doing ok. hard work. one simple q, why u dont back test 1st but forward?

I do run Backtests in the "Strategy Tester"...but for only the purpose of verifying triggering...ie->testing the strategy.

Reports by the Strat Tester do not show an accurate depiction of account growth and draw, in fact, even on the logic portion, they can be a little misleading.

You may come to realize that for all of the flawless EAs out there that turn $10 into 12 million in 2 months, typically will turn 10k into Ash in less than 2 weeks...there are no backtest champions who made millions off of the Due Diligence of working with the strat tester and I fear that anybody that puts any faith into these results that goes live without having any conceptual idea of the Draws against the account in a live environment will, in most cases, not like the results.

Long story short:

Backtests are Bullshit!

short in long

back test is not bs especially when it comes to alvorithms. 100% mechanical system will do same in d future what it made in d past.
u r just losing time. bt each imstrument then cobsolidate reports into one see what will apprx happen in d future.

u r hoping an algo will work in d future that wont work in d past. or forward testing an algo where u dont like bt results. this is such thing like doing exact same thing w expectations of different outcomes.
difference can only happen because of market randomness but not because of algo itself. then if ur expectations is based on randomness ur code can be as short as 1 line includes rnd ()

sorry but u r actually/partially foolin ur self. no insult. just my thought.

bt n optimization r different things.

short in long
if u expect it to work in d future then u should expect it to work on d past too

good luck

Edit after Ignored by magix:
blocking me wont change d fact that what I have said above is right

yes happier now

I think we will both be happier this way.

back at the home front---- I had to start a new account for the testing -- not sure what happened that I could not get the original account up and running again.

My new explorer is up with Rev 1.3 with two pairs this time -- E/U and U/CHF 2K starting and 0.1 lot everything else is default.

Got some catching up to do

A little slow on finishing touches...daughter-in-law has been in the hospital dealing with some brain issues.

Been holding down the fort and watching the baby...

Will catch up on market stuffs as time allows.

Enjoy the week!