Bob and Shelley Again

[SpiderX]

SpiderX, I thought this logic is already implemented in the code, if not...I think it is a great idea. Thanks!

Hi,

Just happened that there is a trade on the GBPUSD for me on the 1H that is dragging its feet.
Also looking at some of the history that Shelly posted, it seem like a good idea to pick trades that are able to make higher highs or lower lows within 1-2 candles of entry and cut losses early on those that still remain indecisive after 1-2 candles.

It would nice if Shelley or Steve or anyone else with a lot of experience on BASA can confirm on this before it is implemented.

Cheers

Hi Spider,

could you exactly explain, what difference your idea has with the already implemented two candele closure suggested by Dave (spotdespot)?

Cheers

Tommaso

Hi Tommaso,

Looks like i missed that
Thanks for pointing that out.
Will check it out.


Cheers

[rosst]

Excellent rosst...can you please share those two indicators you are using to get those reports...Thanks!

I got these indis here on the forum...

[rosst]

Excellent rosst...can you please share those two indicators you are using to get those reports...Thanks!

I got these indis here on the forum...

[fxozgirl]

Steve...BASA took a CADJPY trade yesterday after the big candle which I believe the ATR filter should have prevented? Trade went to full SL

ATR was 25, then the next candle's closed at +55 pips, so using 2x ATR should have prevented the trade? In my analysis this is how I viewed each candle, so hopefully this wasn't lost in my original presentation to you

I am using version 2p

cadjpyh4_atr filter.png

Don't know Shelley. The atr on your chart feedback is for the stop loss. There is no display for the candle length atr. The code that decides if the previous candle was too big for the candle length atr is within bool IsTradeAllowed() and seems fairly straightforward:

Code: Select all

  //Previous candle length. Cancel the trade if candle moved too far
  if (ClAtrPeriod > 0)
   {
      if (MathAbs(iClose(NULL, TradingTimeFrame, 1) - iOpen(NULL, TradingTimeFrame, 1)) > ClAtrVal) return(false);
   }//if (ClAtrPeriod > 0)

but I have not included its value in the chart feedback. Shall I do so?

Note that TradingTimeFrame defaults to 0 which means the current chart and is not an input the user can edit without knowing some basic coding.

Note also that I have no idea when or why the candle length atr came to be included in all this or what it was supposed to achieve. Looked at from the point of view of someone who has no idea what it was supposed to do, it is nonsense. No doubt it made sense at the time, but it looks like nonsense now. The 'Cl' in ClAtrVal suggests to me something to do with the length of the candles over a set period of time, but I do not see how the ATR indi returns this.

Is this something that needs recoding?

Playing a bit of catch up so sorry for the delay in responding...too much on at the moment and still have 2 more days at work yet

I suspect that the criteria I provided was probably lacking...so the ATR filter which should prevent BASA taking a trade after a big spike may well need to be recoded

Hopefully I have provided a clearer explanation below and also please see chart attached:

In order for BASA to take a trade the length (in pips) of the body (ie. open/close) of the candle at Shift 1 must be less than 2 x (or user defined) the ATR at Shift 2

Also, it may be helpful to have this ATR reading recorded on the chart feedback separately from the SL ATR feedback

See the example on my chart below

cadjpy_atrfailedtrade.png

[fxozgirl]

Hi,

Uploading here an edited file with the above mentioned changes for 1H timeframe.
Have edited this relatively fast without much testing, so there may be some bugs.
(but highly unlikely given the simple nature of the edit)

Cheers

SpiderX...would you mind editing the name of this file, by indicating that it is for H1 timeframe, so people don't get confused and download it thinking it is the original BASA strategy please?

Done.

Reuploaded in my original post. Renamed as BASA MTF.
Technically speaking, this now can work on 30 min and 15 min charts.
Maybe someone can try that out.

http://www.stevehopwoodforex.com/phpBB3 ... 151#p83151

Cheers

Thanks SpiderX...I'll set a demo up after Christmas (just don't have time right now)

[fxozgirl]

Steve...BASA took a CADJPY trade yesterday after the big candle which I believe the ATR filter should have prevented? Trade went to full SL

ATR was 25, then the next candle's closed at +55 pips, so using 2x ATR should have prevented the trade? In my analysis this is how I viewed each candle, so hopefully this wasn't lost in my original presentation to you

I am using version 2p

cadjpyh4_atr filter.png

Hi Shelley,

In your mind, you are comparing the candle length with Atr Shift 2.
If there is no mistake, the code compares candle length shift 1 with Atr Shift 0, which should be the Atr value that is developing after the long candle shift 1.


BASA Mtf has been upgraded with the following addition variable:

Code: Select all

extern int      ClAtrShift = 0;//Shift 2 should be correct to compare prev candle to prev2 atr

Code: Select all

   //Atr for previous candle length
   if (ClAtrPeriod > 0) 
   {
      static datetime OldAtrBarTime = 0;
      if (OldAtrBarTime != iTime(NULL, ClAtrTimeFrame, 0))
      {
         OldAtrBarTime = iTime(NULL, ClAtrTimeFrame, 0);
         ClAtrVal = GetAtr(ClAtrTimeFrame, ClAtrPeriod, ClAtrShift) * ClAtrMultiplier;         
      }//if (OldAtrBarTime != iTime(NULL, ClAtrTimeFrame, 0))
   }//if (ClAtrPeriod > 0) 

http://www.stevehopwoodforex.com/phpBB3 ... 151#p83151

Have defaulted this variable to 0 in the new version, but think that it should rightly and permanently be 2.
Probably Steve have a better idea than me on this.

Do note that BASA Mtf should work same as the original in 4H chart, only difference would be in 1H or even lower tf chart.

Cheers

Sorry SpiderX...my brain is not really in gear just yet, but please refer to my explanation at this post
http://www.stevehopwoodforex.com/phpBB3 ... 276#p83276

[fxozgirl]

Gyys, there is a humungous problem in comparing results that is well known and understood by us crabby old buggers but not so well understood by noobs. It is this: different brokers = different feeds = different results. It is even worse than that; as I described a few pages ago, different platforms trading the same ea with the same settings and broker stell = different results.

So I reckon the onus is on us as individuals to find the best version to use with the broker we are trading with (GP) or trading against (most of the other so-called 'brokers'). I am going to set up 3 separate demos after the break:

• one trading BASA as originally coded on the H4.
• one trading the original BASA but on the H1.
• one trading SpiderX's version.

I should have a reasonable idea of which one is going to be most profitable for me after a month or so.

Unless Entropy is fully released and turns out to be a profitable as we are all hoping, in which case I doubt whether I will care.

Definitely agree Steve and as most of us are aware live results to demo results are different as well.

My findings to date on this are:

1. BASA is OK on demo
2. BASA is much BETTER on live
3. I'm finding original BASA 4H strategy is better on Cowboy Go Markets (Live)
4. I'm finding original BASA on 1H chart is better on GP (Live)
5. I too will set up a demo for SpiderX MTF version to see how it compares

6. Hang my hat on Entropy

[fxozgirl]

I started Basa at Cowboy Cowboy IBFX on 10/29 so far it is doing well on H1

Very nice rosst

[fxozgirl]

BASA 1hr. 2013 performance result...2014 would be an interesting year...Happy new year and Merry Christmas to all.

perf.GIF

[SpiderX]

Sorry SpiderX...my brain is not really in gear just yet, but please refer to my explanation at this post
http://www.stevehopwoodforex.com/phpBB3 ... 276#p83276

Hi Shelley,

Updated BASA MTF in my post.
Default CLAtrShift to 2 as confirmed by you.
This version will still be same as the original in 4H or another other time frame as long as useLowerTf is not enabled.

http://www.stevehopwoodforex.com/phpBB3 ... 151#p83151

In addition, it seems that for the StopLoss...AtrVal is calculated with Shift 0.
Can you confirm if it should be Shift 0 or Shift 1 ?

Btw, another thing which i have spotted is that SpotDeSpotCandleCloseTf should be set to 0, otherwise it might not be helpful to have it in 1H or lower time frames.

Cheers