Pippo: a Murrey Maths retrace auto-trading EA by SteveHopwood | Page 19 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Pippo: A Murrey Maths Retrace EA 188 replies

Maths (ratio) problem 16 replies

Do i need to be a maths genius? 12 replies

How important is a knowledge of maths? 96 replies

Stupid maths 19 replies

Attachments: Pippo: a Murrey Maths retrace auto-trading EA by SteveHopwood
Exit Attachments

Pippo: a Murrey Maths retrace auto-trading EA by SteveHopwood

  • #361
  • Edited 5:58pm Jul 9, 2011 5:47pm | Edited 5:58pm
Quoting Viv108
I've backtested the other versions and they run fine. Anyway, lets see...
It is not the trading decision code, but the stop loss insertion that is wrong. There were errors in Paul's logic that I have corrected - at least I think I have, so check on demo next week. It is really hard to divine how Pippo makes trading decisions, so this is no reflection on Paul's coding ability.

Update in post 1.

Remember my comment about ltf trading with this ea guys; I think anybody trying this is baying at the wrong moon.

  • #362
  • Jul 9, 2011 5:50pm Jul 9, 2011 5:50pm
Steve,

I see your generous work is going to make it rain.

Regards,
Jeff

Quoting SteveHopwood
It is not the trading decision code, but the stop loss insertion that is wrong. There were errors in Paul's logic that I have corrected - at least I think I have, so check on demo next week. It is really hard to divine how Pippo makes trading decisions, so this is no reflection on Paul's coding ability.

Remember my comment about ltf trading with this ea guys; I think anybody trying this is baying at the wrong moon.

Attached Image
  • #363
  • Jul 9, 2011 5:51pm Jul 9, 2011 5:51pm
Quoting azjeff
Clarification to one part of my below post. Where I typed Steve indicated he will make changes in time means that he makes changes when he sees something worthwhile for making changes to. He has done this in other threads that he has started. Steve is easily in the top 1% of givers in this forum.

Regards,
Jeff
Hi Jeff....been awhile....Steve has a great thread here....sandovich and droland both have some great ideas....Steve has that magic "user input" that can help everyone.....this should be really interesting....

Have a great weekend

whip
  • #364
  • Jul 9, 2011 5:52pm Jul 9, 2011 5:52pm
Quoting azjeff
Clarification to one part of my below post. Where I typed Steve indicated he will make changes in time means that he makes changes when he sees something worthwhile for making changes to.
Hehe. Not always he doesn't. Sometimes he makes them because someone wants them and he is bored, has a few minutes to spare and wants something to do.

In this case, I suspect Sandovich and the relevant moon are not even in the same universe, but will be delighted to be proven wrong.

  • #365
  • Jul 9, 2011 5:56pm Jul 9, 2011 5:56pm
Quoting azjeff
Steve,

I see your generous work is going to make it rain.

Regards,
Jeff
Well, I have been banging away at this blasted Forex business for what feels like a lifetime. It is either make it work for me or shoot myself.

Given that I am probably a useless shooter, it is a good thing I have finally started to make it work. I wonder if I would have continued to hammer away at it when I started five years ago, had I known how long it would take.

Hey ho.

  • #366
  • Jul 9, 2011 5:56pm Jul 9, 2011 5:56pm
Quoting azjeff
Hello Sandovich,

I for one would like you to continue to post your results and findings in this thread. After all you are using the pippo ea that Steve wrote. Steve indicated in this wonderful thread that he will be making changes to it in time.

Regards,
Jeff
Yes , I also think we can make a super EA out of Pippo if we share ideas. If we split we might miss a good idea that can be used on both time frames.

Steve is definitely in the top 1% of givers that is another reason why I would like to see Pippo EA succeed.
  • #367
  • Jul 9, 2011 5:59pm Jul 9, 2011 5:59pm
I inadvertently uploaded the wrong version in my update notification a few posts ago - now edited and removed.

The relevant update is in post 1.

  • #368
  • Edited Jul 10, 2011 1:42am Jul 9, 2011 8:14pm | Edited Jul 10, 2011 1:42am
Quoting Sandovich
Yes , I also think we can make a super EA out of Pippo if we share ideas. If we split we might miss a good idea that can be used on both time frames.

Steve is definitely in the top 1% of givers that is another reason why I would like to see Pippo EA succeed.
My only thinking is not necessarily a separate thread although that is possible but maybe keep two versions running in parallel rather than in series (keep on changing the only version we see which is Pippo.mq4 for everything is not logical). At least for now. The original and the Sandovich version keep running in parallel on Page 1.

Running an H4+ version versus an M15 version is, IMHO running two different trading styles. H4 is really long term and M15 is almost a scalper. I love scalpers (make no mistake here) and I personally make a lot of real money using scalpers so I am happy to see this line of development continue, but I also love the basic concept of the original Pippo as well.

My worry is we keep modifying something that works with major concept changes until it doesn't work anymore. Why?? Let me clarify my thinkings!!

I belong to Funyoo's forum and I contribute a lot of time and input there but I have seen many of his EA's get modified by every idea that comes along until it doesn't work anymore (I do not say every idea is bad but follow my logic). Each new innovation that does not work, then gets forgotten and people move onto the next newer, better and greater EA when the one they already have is actually really really good if they just step back four or five revision paces in the logic.

Anyway, I am not just putting out ideas and will be delighted to be personally proven wrong. I am here to help make a great EA that can make us all some decent good money, not just to be proven right or wrong or anything else. Right or wrong is insignificant here.

I do not claim to be in Steve's league or anyone else for that matter but my observations of what has happened in so many forums and in so many threads, seems abundantly clear to me.

Let me give you an example. I have been using an older version of Funyoo's EA called RSI MA Scalper. It was modified by one of Funyoo's forum members called Lordordly. It is one of the main scalpers I use on my LIVE SCALPER ACCOUNT. Here is a one year demo forward test using this EA starting with a $10k demo account (not back testing!!!) and which demo account is still live (by the way, this is my account, no one elses):

https://www.myfxbook.com/portfolio/r...orly-102/72189

Guess what. The RSI MA Scalper thread is almost dead on Funyoo's website now because the EA has been subsequently changed and changed until it doesn't work anymore then everyone went to the next best thing. I have been PM'd questions on certain logic in this EA which I then posted as questions on the thread and even Funyoo hasn't answered. He may eventually but in the meantime, I am still making superb money with this EA mod on my Scalper LIVE ACCOUNT. For me this is unbelievable. BUT PEOPLE ARE STRANGE!!

Anyway, what I love about Steve is he is always there and may give me hell but I love to poke and prod but in the end my goal to ensure this potentially great EA continues.

By the way, those who have never met Azjeff who has recently posted here, he is everywhere, very knowledgeable and contributes incredibly to so many forums. He is well worth listening to.

Dave R.
  • #369
  • Jul 9, 2011 8:39pm Jul 9, 2011 8:39pm
Quoting droland
I belong to Funyoo's forum and contribute a lot of time and input there but I have seen many of his EA's get modified by every idea that comes along until it doesn't work anymore. It then gets forgotten and people move onto the newer and hopefully better EA when the one they have is really really good if we step back four or five paces.

Anyway, I am only putting out ideas and will be delighted to be proven wrong. I am here to help make a great EA that can make us all some good money, not to be proven right or wrong or anything else.

Dave...
Agree 100%. I died a little bit inside when a lot of people abandoned TB but I'm still getting good results with it.

Thanks Steve and Droland for all your valuable input
  • #370
  • Jul 9, 2011 9:16pm Jul 9, 2011 9:16pm
Quoting SteveHopwood
He cannot. I am the only one who can update post 1. I cannot do this with Paul's Sandovich version until use of a stop loss is a user option - it enforced in Paul's code. Either Paul or I will add the option to use the stop or not in time for Monday, then I will replace the post 1 version.

Paul, fantastic contribution. Thanks very much.

I should have made the Sandovich variant a user option in the first place! I'll try and add this later.

EDIT
Hi

I see that Steve's done this already with the UseIndividualTradeStopLoss option.

Thanks Steve!

Note to self: Must read all the posts before I respond

Paul
  • #371
  • Jul 10, 2011 1:25am Jul 10, 2011 1:25am
Hello Dave,

Very nice information that you share with everyone. Your participation is something that everyone should give thanks to.
People can learn from the results and information that share.

Regards,
Jeff

Quoting droland
My only thinking is not necessarily a separate thread although that is possible but maybe keep two versions running in parallel rather than in series (keep on changing the only version we see which is Pippo.mq4). At least for now. The original and the Sandovich version.

Running an H4 versus an M15 is IMHO running two different trading styles. M15 is almost a scalper. I love scalpers and make a lot of real money using scalpers so I am happy to see this develop but I also love the basic concept of the original Pippo as well.

My worry is we keep...
  • #372
  • Jul 10, 2011 1:27am Jul 10, 2011 1:27am
Yes, both are very nice to share their results and ideas.

Regards,
Jeff

Quoting hiredwhip
Hi Jeff....been awhile....Steve has a great thread here....sandovich and droland both have some great ideas....Steve has that magic "user input" that can help everyone.....this should be really interesting....

Have a great weekend

whip
  • #373
  • Jul 10, 2011 2:45am Jul 10, 2011 2:45am
One more tiny addition in a post 1 update.

Playing out on strategy tester on the M15 tf, I notice hedge trades being closed on a market retrace to within the yellow lines. These closures are draining the account.

I have added a CloseHedgeOnReturnToGrid input. Setting this to false stops Pippo from closing these trades.

  • #374
  • Jul 10, 2011 3:30am Jul 10, 2011 3:30am
I just noticed a lot of trades closing at one or two pips profit.

LUC invited his pals around for morning coffee when Paul added the TakeProfitBasket thingy. I have sent them all home - fix in post 1. Paul, you will see the extra conditional I needed to add.

David, I keep meaning to say that I adopted your change of variable name to differentiate it from the TakeProfitBasket thingy.

  • #375
  • Jul 10, 2011 7:41am Jul 10, 2011 7:41am
Quoting SteveHopwood
I just noticed a lot of trades closing at one or two pips profit.

LUC invited his pals around for morning coffee when Paul added the TakeProfitBasket thingy. I have sent them all home - fix in post 1. Paul, you will see the extra conditional I needed to add.

David, I keep meaning to say that I adopted your change of variable name to differentiate it from the TakeProfitBasket thingy.

Hi Steve

Thanks for spotting and fixing my 'School boy coding error'!

Also, for my Sandovich changes, I modified your original: -

if (LargestLot >= Lot4) return;

to the following in order to match Sandovich's manual system: -

if (LargestLot >= Lot4 || LargestLot >= Lot3 || LargestLot >= Lot2 || LargestLot >= Lot1) return;

Perhaps this should be changed to the following in order to keep the two systems separate?

if (UseIndividualTradeStopLoss) if (LargestLot >= Lot4 || LargestLot >= Lot3 || LargestLot >= Lot2 || LargestLot >= Lot1) return;

if (LargestLot >= Lot4) return;

Thoughts?

Paul
  • #376
  • Jul 10, 2011 3:07pm Jul 10, 2011 3:07pm
Quoting batemap
Hi Steve

Thanks for spotting and fixing my 'School boy coding error'!

Also, for my Sandovich changes, I modified your original: -

if (LargestLot >= Lot4) return;

to the following in order to match Sandovich's manual system: -

if (LargestLot >= Lot4 || LargestLot >= Lot3 || LargestLot >= Lot2 || LargestLot >= Lot1) return;

Perhaps this should be changed to the following in order to keep the two systems separate?

[b]if (UseIndividualTradeStopLoss) if (LargestLot >= Lot4 || LargestLot >= Lot3 || LargestLot >= Lot2 || LargestLot >= Lot1) return;

if...
Hehe. Not a clue. Two large pre-dinner whiskysh washed down with some dinner wine leaves me rolling around on the floor wondering who my mummy is.

You know as well as I do that excess antake of ilcohol and coding do not mix. Would you mind simply posting the fix and I will replace the post 1 version.

If I can find post 1.

Cheerrrsshh
  • #377
  • Jul 10, 2011 4:58pm Jul 10, 2011 4:58pm
Firstly,

Thank you Steve for allowing me to post here. I hope that I can add something of use to you all.

I've been quite taken with this EA, concept of which sounds great as I have been investigating MM lines as an alternative to fibs and pivot points. However, I have been having the same issues in demo as the rest of you....

1) The lines can redraw, often for no apparent reason.

I can understand that a close outside the +/_[2/8] lines will almost surely result in a redraw but I couldn't quite get my head around why a redraw would suddenly occur when price is nowhere near the extremes, or, worse, a profitable series of trades are in the offing ansd suddenly, the price goes from being around band 9 to being around band 4

2) You cannot see the history of MM levels

I've not been able find an implementation of the MM indicator which adequately show how the lines have changed over time which makes any kind of back testing pretty difficult.



Having looked in detail at how the indicator works, I think I have found there are a few problems with the MM lines indicator. In order of increasing importance...

1) The levels are based on a highest high/lowest low range which is calculated over a period which is set to 64 days by default. The original Murrey method required that candle wicks were ignored so we should be using highest close/lowest close

2) I suspect that the 64 period default for the range is also incorrect in the land of intraday forex. My reasoning is thus. Murrey Math was originally designed as a daily method and the 64 days was used because it allowed for a convenient division of the year. This division does not work as well in 24 hour trading. I've found that for H1, using 118 (Bars in a week for my broker) is more in tune with the rythms of forex.

3) The lines are too sensitive to variations in the value of the high low range. The result of this is that when an old high or low goes out of the picture, the range can change very rapidly and you can see a doubling or halving of the octaves on a day when the price may have barely moved. I regard this as incorrect behaviour.


Anyway, the end result of all this is that I've created an indicator which we can use to view the past MM lines using the original methodology.
This is without disparaging the original authors of the algorithm and the indicator which generate the lines, because it is a brilliant pice of work which I couldn't dream of creating from scratch. In fact, the algorithm itself is unchanged. I have just changed when it is used

Set UseOriginalCalcs = true Calculates the lines the original way.

I've also coded some new behaviour which is switched on when
UseOriginalCalcs = false

The major change is that the lines are only calculated when:
a) Price CLOSES outside the yellow outer lines (This always results in a redraw)
b) The Hi/Lo range falls to less than 2 MM levels apart(This usually results in a redraw, But not always)


UseFractalMultiplier = true
to use the pip value for calculating the octaves. (I've read that for prices below 12.5, the ocatve calculations are a bit of a frig although i haven't seen this makes much difference


UseCloseForRangeCalcs = true to use closes for the range calculation as per the original Murrey method

The end result is a much more stable set of lines which visually encompass the price action and I believe also technicaly conform to the MM method.


Take a look at the attached charts to see the difference between the smoothed lines and the original.
The white lines on the charts are the Hi/Lo range.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: gbpusd_h1_64day.gif Size: 38 KB
Click to Enlarge Name: gbpusd_h1_118day.gif Size: 30 KB
Click to Enlarge Name: gbpusd_h1_64day_smoothed.gif Size: 26 KB
Click to Enlarge Name: gbpusd_h1_118day_smoothed.gif Size: 26 KB
Attached File(s)
File Type: mq4 #MurreyMathLines_ShowHistory.mq4   19 KB | 648 downloads
  • #378
  • Jul 10, 2011 7:30pm Jul 10, 2011 7:30pm
Quoting SteveHopwood
Hehe. Not a clue. Two large pre-dinner whiskysh washed down with some dinner wine leaves me rolling around on the floor wondering who my mummy is.

You know as well as I do that excess antake of ilcohol and coding do not mix. Would you mind simply posting the fix and I will replace the post 1 version.

If I can find post 1.

Cheerrrsshh
Hi Steve

Hope your headache is short lived!

Here's the changes.

Thanks

Paul
Attached File(s)
File Type: mq4 Pippo (4).mq4   58 KB | 503 downloads
  • #379
  • Jul 10, 2011 7:50pm Jul 10, 2011 7:50pm
Quoting batemap
Hi Steve

Hope your headache is short lived!

Here's the changes.

Thanks

Paul
Cheersh.

Hic.

You are a shtarr.

Hic.

Up.

Update added to post 1.

  • #380
  • Jul 10, 2011 11:06pm Jul 10, 2011 11:06pm
Steve, beware the head ache that follows..... lol The post of DietCoke certainly gave me some food for thought as I must have skipped some important information the last time I read about Murray Octave construction.... It also explains why the Sixth always reminded me of Murray.... Steve / DietCoke: a) I am wondering whether there will be any benefit to adopt the Beast's MaxBar calculation mechanism to Pippo EA (I am strictly a MacMan person and never adopted the ChaserL volatility based calculation myself) b) In line with the dotted magenta line in the Beast, I am wondering whether the adaptation of same principle above / below any octave line would help entry. E.g.: split distance from one octave to the next above / below again in octaves, entry conditions same as for the Beast at respective (max / min) dotted magenta line Cheers & keep the goodies coming
0 traders viewing now