Page 51 of 60 FirstFirst ... 414950515253 ... LastLast
Results 501 to 510 of 600
  192 192 Attachment(s)    

Thread: Pattern EA

  1. #501
    Junior Member PyeR2's Avatar
    Join Date
    Apr 2009
    Bondi Beach Australia

    Default Come To Papa

    Looking good for a quick 8% in just over an hour. There's probably more in it but I like to bank and reload. I'll be looking to short again on a retrace if we get one.
    Attached Images Attached Images Pattern EA-eu-gu-jpg 

  2. #502

    Default forex tsd calender

    dear funyoo. can you remodification gartley ea 105 to able using forextsd calendar. because when i use ffcal, my mt4 always crash.
    Attached Files Attached Files

  3. #503

    Default Alternative ZigZag criteria

    Gartley pattern is based on zig zag filtering. If the criteria for ZigZag is erroneous, then you measure wrong patterns. And that's the impression I have. And you can tell, watching this system live, sometimes the patterns it creates are out of market context. Why?
    ZigZag filtering based on % change or fixed #points was created by bone headed statisticians. It is too rigid. Market on the other hand is constantly changing, so set of rules for ZigZag should be adaptive to trading logic as well. Live Market is being moved by trading impulse brakeouts and stop brakes over previous ranges. So that should be an adequate criteria then.
    I don't know if someone thought of this in the past, but that what came to me from dealing with live markets:

    1. Use ATR instead of fixed % or number points. That should be also experimented a bit, perhaps AveragebarRange based solely on Hig-Low would work better.
    2. Use 3 consecutive bar break over previous highest high/lowest low, where higher/lower breaking bar is qualified buy minimum ATR. In other words if bars start breaking in the opposite direction, a minimum of 3 qualified bars should take for ZigZag to form a reversal. And only the bar that moved over minimum ATR from previous qualified bar should be qualified for counting as the next.
    So in this way we always account for:
    1. realistic current range,
    2.for pattern within real trading context
    3. time.
    So in this way some crazy spikes wouldn't affect the pattern measure and breaking formation will be naturally context correlated.
    I tried to constract ZigZag by hand with this method and saw superior results.
    If someone mql literate can try to code this logic. Should be relatively simple and potentially rewarding.
    Last edited by Metatrader7; 07-26-2009 at 01:40.

  4. #504

    Default New High/Low Close BarCounts

    New Higher High/Higher Low count is a classic technique, essp among candle traders. The minimum of 3 sets reversal and trend direction. The reason no one talks about it is perhaps it simply: 1. works 2. ancient, so no one can really put their trade mark and brag in flashy trading magazines like DeMark does. While DeMark is big on selling bragging rights, he is too chicken to pull the trigger himself.
    So his stuff does work, not always though and definitively requires a computer as it is hard to watch.
    Where basic new high/low candle count picks overbought/oversolt zones sharper then any stochastic, more consistent and natural for the human mind as well. Something like that should be definetely coded for EA creation.
    So someone mql savvy can work on that as well. Previous post should clarify the starting point of the count. I strongly belive that new high/low bar counts, ZigZags, patterns - these are the realities that create markets as well as they forcast them. Also higher/lower closes is important to count as well as market likes to balance both for trend completion.
    And the numbers are fibo series 3,5,8,13. I usually focus on 5 and 8, then watch higher time frames. 3 and 5 goes against the trend, 8,13 along. 5 is also strong in sideways.
    Last edited by Metatrader7; 07-23-2009 at 12:40.

  5. #505

    Default ZigZag

    I checked default ZigZag indi in MT4 and found it to be pretty good if tweaked to correlate to meaningful trading pattern. So at least Gartley EA has to have ZigZag variable tweaks.
    Creating a new Atr/pattern based Zig Zag would be ideal. It will be actually almost 100% objective and most relevant within each time frame.
    Bar counts is another story.
    Last edited by Metatrader7; 07-23-2009 at 20:54.

  6. #506
    Junior Member
    Join Date
    May 2009

    Default 5 digits broker

    Thanks for the explanation,

    Here is a first EA.

    Rules :

    Enter long when a long gartley signal appears,
    Enter short when a short gartley signal appears.

    I have edited the indicator, it took me a while.

    I have still to study momo counts to add this filter.

    Do you use also a confirmation with the size of the rectangle compared to the distance A-D (minimum 1/3 of this distance, something like that) ?[/QUOTE]

    Hi funyoo

    I want to know if exist or you can make a version of v101 or v103 that works with 5 digits broker, because i am having problems with the backtest and optimitation.
    I sincerely appreciate your help


  7. #507

    Default This is too raw to trade live

    ...was inspired by ViennaTrader to run forward test on V101.
    Overall it looses money, but it does it periodically. So one week it may appear profitable, another week start loosing big, much bigger then it makes money.
    Just watching it, noticed when even it picks pattern correctly it opens trades far outside target red box, so what's the point then?
    Even with all shortcomings fixed - Gartley should be taken in a trend context and it doesn't form that often and is to be confirmed by other factors. So to trade solely on Gartley isn't the brightest idea.
    P.S. Sorry if I spoiled the enthusiasm of back testing programmers and MT4 interface aspiring tweakers. I should give some credit for impressive exercise in coding. But from a trader perspective - I would never run it live, even as assistant indicator. Totally useless perversion of swing/Gartley theory and it seems nobody really cares for accuracy. It seems everybody fall in love with butterfly pictures on the screen.
    Try photoshop or some animation - more thrills.
    It may pass for a demo for Funyoo programming services
    I'll try it in reverse, could be another reverse psychosis candidate.

    one of important qualities of a successful trader rarely mentioned- being brutally honest with themselves and others.
    It isn't that popular I know, so is the ability to pull $$$ from the market.
    Last edited by Metatrader7; 07-24-2009 at 19:37.

  8. #508

    Default Gartley is a repaint indicator

    By design Gartley can develop 2 and perhaps more variations of fibonacci ratios within each setup, the question is which one will hold and even one of the ratios does work, there is a huge reversal cluster. So if you take multiple ratios and put these huge variation boxes at each cluster they will paint a vague area of few hundred points. So this is nothing else but another fancy repaint indicator that looks good only in retrospect. It is one of those that prove fibonaci govern markets in retrospect, but useless for accurate forecasting and live trading.
    I enter my trades with 25 pip stops and most of them hold. I can never do that with Gartley and that's why don't use it.
    There is a popular saying "...but I don't care, bottom line if it makes $ or not".
    Seems logical, but insufficient in longer run, because even it makes money for a day, week or month on forward test it can still blow up at some point.
    If system uses abnormal stops or no stops that's the first sign of a blow up candidate. Keep looking.
    No amount of "hoping" or "finger crossing" will help if the numbers are wrong, it's a time bomb by design and will blow up.
    The only way to tell is by checking risk reward ratios. And if they are real bad, then system can be potentially a reverse trade candidate if number of loosing/winning trades allows. These are first 2 things to check evaluating trading system . Actually sole purpose of evaluating indicator effectiveness, running back test and forward tests is to determine these 2 ratios and improving them.
    Running forward test just on hopes that a winning streak excitement continues indefinitely somehow is a dead end approach and very unprofessional.
    Last edited by Metatrader7; 07-26-2009 at 01:30.

  9. #509

    Default If anyone is real serious about pattern trading

    If anyone is real serious about pattern trading, should start isolating patterns on new higher high/lower low/close criteria and building zig zags and and new high/low index indicator.

    Without that all effort is a waste.
    Sorry to burst your bubble, but that's what the numbers show.

  10. #510
    Junior Member
    Join Date
    Jul 2009

    Default 2 days test

    hi guys,

    here is 2 days forward test on 8 pairs so far 34% , - Share Your MT4 Statements Free



Similar Threads

  1. Chart Pattern - The Hanging Man
    By painofhell in forum Trading discussion
    Replies: 0
    Last Post: 10-03-2014, 22:28
  2. Chart Pattern - The Hammer
    By painofhell in forum Trading discussion
    Replies: 0
    Last Post: 09-26-2014, 09:51
  3. 1-2-3 Pattern indicator
    By funyoo in forum Indicators
    Replies: 29
    Last Post: 02-02-2012, 14:43

Tags for this Thread

100, adx, alarm, alpari, automatic, average, backtest, backtesting, bb-macd, breakeven, broker, brokers, candle, change, check it out, closing, coding, color, crossover, demo, divergence, eas, ecn, eur, eur/usd, eurjpy, eurusd, expert, expert advisors, experts, fibonacci, forecast, forex, forex trading, free, gbpjpy, gbpusd, german, gold, guide, high, home, how to, index, indicator, indicators, learn, live, live account, low, macd, magic, magic number, main, managed, manual, manual trading, martingale, metatrader, mirror, mq4, mql, mt4, news, offer, osma, parabolic sar, pattern, pivot, probability meter, profit, profitable, programmer, real, request, resistance, scalp, scalper, signal, signals, simple, spread, step by step, stop, stoploss, strategy, study, support, system, systems, test, thank you, time, tool, trader, trading, trading system, trailing, trend, usd, video

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts