Slingshot 30M & 1H - The 2nd Coming......

Damon

I'm testing with IBFX. All three individual filters set to True on 30m and 1Hr.

Besides having to manually close a couple trades, it appears to be working fine.

Cfi and hsa,

Thank you both for the advice on the daily report. At first, Cfi, I was puzzled because it only generates in HTML and gives me the entire history of all trades. I just wanted those from today.

Then hsa clarified for me how he was using the "Custom Period" to first filter the dates, THEN run the report as "Save as Detailed Report". It still kicks it out as HTML but then I right-clicked the file and selected "Open With" = Excel. Now it looks perfect! Thanks guys, you taught me something new! Could have used this little trick 4 years ago - LOL!

BlackWolf
yes, the custom report is not bad, eh?

the only thing I'm still absolutely unsatisfied with is: MT4 can't export PIPS. The single most important thing. I just don't get that. Relatively serious platform, and can't export in pips. Aaarggh.
Having to manually do that in excel is somewhat tiring. Besides I still don't know how to automate it, as the yen related pairs have two digits I can't just pull a formula down along multiple rows.

nickj, would you be as kind as to attach your pip calculating excel template, maybe I could be enlightened?
(I know you have it for a real account so just a few rows with two and four digit pairs.) Thanks!

Well you can set up a template, and yes a nested IF statement for yen pairs.. quite cumbersome, but can be done. I've done it.

I dunno if I totally agree with you on the pips being the most important, hsa. I do know what you're saying (that we need pip movement to make profits) but IMHO (and that's all it really is), it's the number of lots x the pips that earns you the money and you don't need a lot of pip movement for that to happen.

My idea of a Utopian trading system, for me personally, is to find a market pattern (setup) that is reliable, consistent and nearly always moves in a predicable direction up or down (even if just a little). Hit it hard with a high number of lots and then get the hell out fast. Sniper trading, in and out quickly, one-shot, one-kill. A lot of traders think you have to stay in a longer trade to make money. I beg to differ. In fact, I would argue that you assume more risk that way then a very short term trader. Just my thoughts and why I like the "Damon Slingshot method".

Thank you cfi - I have made that adjustment. What indi creates the vertical lines for trend changes?

No - I'm with FXDD. What are your spreads? Mine today are: EURJPY =4; EURUSD = 2; GBPJPY = 9; GBPUSD = 4; USDJPY = 3.

Many thanks Bob -

Nice fix to the issue, I've never traded with a 5 digit broker so haven't had to face the issue in my own programming. I'll get your code put in the ea very soon.

Gary

Gary,
Sorry, I can't really follow the code as I have no idea about it.
Only thing I know is as I've put 2.7 mod1 and set SL to 350 (5 digits) the EA still puts the message in the upper left corner "Stoploss is 35".
Does this mean I can't use it on 5 digits? Or does my manual setting override the initial setting in the code?
Shall I edit the code, put 350 instead of 35, 450 instead of 45 and so on?
thanks a lot

EDIT: I'm simply blind, worked too much today. I've noticed the "UseRecommendedSetting"
---------Discard my post.

Set the SSMultiFilter input "ShowMacdVlines" from False to True.

Hi All -

There are several posts I need to reply to - some really great ideas. But first I need to share some thoughts and get feedback.

What has really been bugging me for days (but really got under my skin today = several new versions) about the ea is the random nature of missing trades, missing a proper close, etc. This is a technical issue and not a trading system issue. The lack of a pattern to the errors was a real mystery to me and I just couldn't find the error in the code so in the last version, I replaced the code. In backtesting I got a perfect result......but this had happened to me before. This led to great expectations and disappointing result. So what to do?..............well, the answer is obvious - get away from the computer and go for a long, thoughtful walk on the beach! And that is exactly what I did!! I want to share a cause, effect and a solution that came to me along the way. I'm going to walk (no pun intended) a middle ground in the techno talk I use so programmers and non programmers can think along with me and express an opinion and ideas you have. As I mentioned before, I'm a retired scientist, and as such tend to be a linear thinker...........so here goes.

First, I had been concentrating on the random nature of missed trades, missed closing, etc. So, I thought about the opposite! What did all the failures have in common. Besides the random nature of the errors, the answer was that the last several versions of the ea backtested perfectly. Ok, so what is the difference when we backtest versus when we forward test? One answer is that when we backtest, the tester is only trading one ea but when we forward test we are trading several pair on more than one TF = we have several ea's open at the same time. It dawned on me that perhaps I wasn't seeing the forest (action of multiple ea's operating at the same time) for the trees (testing only looks at one ea operating at a time).

A tiny light came on which could (I repeat could) be the cause of the problem. You programmers out there - any bells starting to ring?! I have never experienced the issue I'm about to talk about (as far as I know) but I remember reading about something a long while back that may apply to us and I'd like your opinion as I continue my research on the issue.

The little light was the fact that in metatrader, only one script or expert can execute at a time. The more ea's (different ea's or multiple uses of the same ea, etc) you have trading the greater the chance you'll run into a 'traffic jam'. We increase our odds of a traffic jam since we are, for example, looking to close or modify trades on the hour or half hour, often for several trades opened at the same time (= trying to close at the same time). If another expert is trying to trade at the same time it will be stopped (and skipped) by the error 146 ("Trade Context Busy"). Since our trades open, are modified and close in a random nature depending on how many charts, trading option, TF's, etc we are using, there are many opportunities to conflict with each other in what would appear to in a random fashion.

The more I walked, and the frustration gave way to clearer ideas/thoughts. The little light/idea became a word and the word was "semaphore". I couldn't remember the exact definition while on the walk so I looked it up when I got back and parapharsing Wikipedia I offer the following: A semaphore (in programming) is a means for supporting mutual exclusion in concurrent programs. In other words, I think we need a "go" and "no go" signal - when one ea is trading, the others line up so to speak and wait until each ea in turn gets a "go" signal to 'do their thing' instead of getting stopped.

Further research lead me to a great discussion of this issue on the net and some code to solve the problem. Basically when an ea wants to execute a trade (open, modify, close) it checks the status of a global semaphore. If it gets a "go" signal it changes the status of the global semaphore to "no go" (= signals other ea's to wait their turn) and when finished with the trade, changes the status back to "go" - as one author put it, "green light" versus "red light".

I didn't see enough forward testing results from v2.7 to know if the changes I made work. As stated earlier, it worked in the tester. However, if the problem is as stated above (= a problem with the forest and not the tree), then the same issues will reappear with v2.7 when it is traded for any length of time.

So...........you guessed it , another version of the ea. I've nowed added a "global stoplight" and functions which will change the light from green to red and back again so the ea's line up waiting their turn to trade. Again, it tested perfectly in the tester but there is only one ea tested at a time in this artificial trading environment. It did however produce the same results for the same dataset as it did without the new semaphore code that plays green light, red light! Only forward testing will let us know if this was the problem or not. I will post version 2.8 shortly.

Sorry for the long, wordy post but i wanted to share my thoughts so it perhaps sparks some comments/ideas on your part to solve this random, perplexing problem. If you have had experience with this problem and think I'm totally off base, I really need to hear from you because I've not dealt with this issue before.

Gary

Hi hsa -

Strange indeed! I changed ea's during the trading window more than.....well, a bunch. Traded all 10 charts I've described before using the first 3 filters traded independently ....and the bottom line was a gain of almost 200 pips! Go figure??

Gary

i know that hayseed on here would add soemthing called "order reliable" management to his EA's to stop the confusion. You would have to add it to the libraries part of mt4. is this close to what you did? it sounds like it but i'm not a programmer at all!

My dear Mr BlackWolf -

Surely you didn't mean little old me?! I'm wounded beyond description!

......Ok, all better - back to work on the next version!

Gary

This blows my mind. I am glad it is working for you. I will have to go and check things out and once again, make sure that all indicators are in the proper folders.

Cheers....Damon

Very good info my friend . One step at a time and we will get there.

All in good humor and fun, Gary. You know you have the highest admiration and respect from me and from us all. I was just yankin' your chain, Bud.

Might I make a suggestion to your earlier e-pip-hany post? I like the whole idea a lot, very well thought out - which doesn't surprise me considering the source. Maybe run it past our good friend Ronald Raygun (Kid Wonder), just to be sure. Two sets of eyes are always better than one and two amazing sets of eyes like you and him would be kill-ahhh!

Looking forward to v2.8

llambert3 -

"Please" was very nice but "Pretty Please" would have been better.......or a paid trip to Hawaii....for two! Ok, I'll do it anyway but just this once....maybe twice.

Gary

ps - real big of me since a very nice person already wrote the code for us.

Hey sayembara!

Very nice, very nice indeed! I'll get your wonderful contribution in a new version I'm working on now.

Gary

Gary,

I know exactly what you mean...sometimes we just get too close to the problem to see the answer (although I prefer my long walks along a nearby trout stream with fishing rod in hand).


You may well be on to something here, although I've never run across it yet myself. Perhaps you can do something with the IsTradeContextBusy() function, maybe in a while loop.

Part of my problem, and the reason I've not been more active in the thread, is that like Ronald Raygun I think of this as more of a reversal stratagy than a trend following one. The EA I'm currently working on only takes trades if they are in the opposite direction of the previous bar.

Best regards,

Bob

Edit: That's not to say that I don't think there is some fine work going on in the thread. Many thanks to Damonl, Pavatar and everyone else involved here.