Nanningbob 10.2 Autotrader's new home

[jiaxingx]

I backtest on pair "EU" from 2,Jan to 30,Aug on Alpari UK using OOTB,It seem to not be good !

[alex_forex]

I did some backtests too and they were bad.
I trade it on demo on about 20 pairs *forward test* with more or less success, see my "myfxbook".

Alex

[alex_forex]

Hi all,

Still using v1.0, will update over the week-end.

Missed about 200pips yesterday before the big move, my EAs were disabled...

Picked up 137 pips today.

NB10.2 AT is on this demo account since Tuesday (20120828) :



Regards
Alex

Hi Alex, are you closing your trades manually?

I have had this on since the 28th (OOTB), on 16 pairs. So far it only closed one trade (AUDUSD +40pips), and still on EURAUD, EURJPY, EURUSD, GBPAUD, GBPUSD and AUDJPY, right now up a total of about 310 pips. All of them Trend Trades, no Range trades yet. Yesterday I saw it going from overall up 200 pips to negative, then up again today.

I see good potential here, but might have to fine tune it a bit with MPTM.

MPTM and manual close, plus some mistakes from me with Empty4....

The market is waking up, and as you said, we will have to fine tune the exits.

About the yesterday movement, it was a news release.

BTW, it's not too bad for a start

Regards
Alex

[worlock]

I did some backtests too and they were bad.
I trade it on demo on about 20 pairs *forward test* with more or less success, see my "myfxbook".

Alex

The problem with closing manually is that, if the conditions of the trade still apply, the EA opens same position again.

[alex_forex]

It's not a big concern because if conditions still apply, the new trade is expected to give some pips

But you are right, it is not clean...
For the moment, I disable EAs when I want to interfere but a sleep
fonction would be better !

Alex

[cuzgeorge]

I did some backtests too and they were bad.
I trade it on demo on about 20 pairs *forward test* with more or less success, see my "myfxbook".

Alex

The problem with closing manually is that, if the conditions of the trade still apply, the EA opens same position again.

Hi All,
Here are two instances of Empty4, both on ava criminal, both with V1.1, one with standard settings as per al3xx release, except risk=0 and 0.06 lots.

second, is with my set file changes hardcoded as per steves instructions on how to best install an ea onto charts.

My results have been so-so, IF i had closed positions when the EA was making great progress, sure this would be better. Looking at alex-forex's results, it's again clear that we need to understand the manual method of 10.2 to determine the best exit point. You might like to use your own, but lets stick to NB's as it's shows the best promise in my opinion so far.

the second Empty4 is brand new and has no trades yet,(i lie, the ea just took 3 positions, ) but check up again in a while and i'm sure it will have taken some positions. There is one pending The widget updates automatically for those who dont know.

As for manually closing positions, i've always like that. If you are toooo early to close, the ea tells you to ....... off and opens it again if conditions are still present. I dont see the problem with this.?
It's nice to lock in the profit into the bank, and carry on trading, but i can understand if one is trying to find the absolute curve of the graph. This sounds very very dangerous practice.

my newly appointed moderator friend, alfisi, congratulations. Please be carefull of banning yourself. I hear it can be tricky to get your self back in!!!





regards
cuzgeorge

[alex_forex]

Hi Alfisti, Hi All,

Hi All,
Here are two instances of Empty4, both on ava criminal, both with V1.1, one with standard settings as per al3xx release, except risk=0 and 0.06 lots.

second, is with my set file changes hardcoded as per steves instructions on how to best install an ea onto charts.

My results have been so-so, IF i had closed positions when the EA was making great progress, sure this would be better. Looking at alex-forex's results, it's again clear that we need to understand the manual method of 10.2 to determine the best exit point. You might like to use your own, but lets stick to NB's as it's shows the best promise in my opinion so far.

That is the hardest part of the work...

the second Empty4 is brand new and has no trades yet,(i lie, the ea just took 3 positions, ) but check up again in a while and i'm sure it will have taken some positions. There is one pending The widget updates automatically for those who dont know.

As for manually closing positions, i've always like that. If you are toooo early to close, the ea tells you to ....... off and opens it again if conditions are still present. I dont see the problem with this.?
It's nice to lock in the profit into the bank, and carry on trading, but i can understand if one is trying to find the absolute curve of the graph. This sounds very very dangerous practice.

I like that too

my newly appointed moderator friend, alfisi, congratulations. Please be carefull of banning yourself. I hear it can be tricky to get your self back in!!!

Ahah Thanks my friend Spam killing is a hobby
Will be cautious with that silly button but who knows ??

Those results are not so bad too George, green pips

regards
cuzgeorge

We have to study those results over the week-end to be better next week

Hey ho ! Do your homework

Regards
Alex

[al3xx]

Hi, I might be wrong but I think there are some mistakes in the code

L.1571

Code: Select all

   if (type == OP_SELL)
   {
      //Find a weekly pivot cross
      if (Bid < WeeklyPivot && bias == down)
      {
         if (DetectWeeklyPivotCross(OP_SELL) )
         {
            SendLong = true;
            TradeOrigin = H4TrendTrade;//TradeOrigin is used by CalculateTakeProfit()
            magic = D1H4TrendTradeMN;
            comment = D1H4TrendTradeComment;
         }//if (DetectWeeklyPivotCross(OP_SELL)
      }//if (Bid < WeeklyPivot) 

It's written SendLong = true; when I think it should be SendShort = true;

----
L 1486

The DetectWeeklyPivotCross function never returns true. It seems to return false when it should return true/

-----

L 1226 in the CalculateTakeProfit function we check if the D1SlopeTrend != buyhold while in the comment it's written that we want to check if D1SlopeTrend != sellhold

Code: Select all

         
//If D1 slope is not sell and hold, then set a tp based on the next full support level
         if (D1SlopeTrend != buyhold)

Should be

Code: Select all

         
//If D1 slope is not sell and hold, then set a tp based on the next full support level
         if (D1SlopeTrend != sellhold)

Let me now if I'm not clear

Thanks a million pips Yoi The fixes will be available before weeks open.

Alex

[al3xx]

I backtest on pair "EU" from 2,Jan to 30,Aug on Alpari UK using OOTB,It seem to not be good !

Some pips ago we had a question about backtesting this bot and the reliability of TMA in those circumstances. I explained why it is not possible to backtest without some changes. I cannot find it now but it is a few pages back this thread or the original. I am not sure. Any one who remebers or finds it please link.

I should put a backtest caveat explanation on page one. Anyways, if look closely, this bot tries to follow Nanningbob 10.2's original entry rules. You probably will not need to double check Nanningbob with a backtest

Kind Regards,

Alex

[al3xx]

Here is for the week on a $3000 finfx demo. I would have lost a AUDNZD short entry but the recovery entry level 2 (1 is the original trade) rolled back south today to close +5 BreakEven on the original.

I am just starting to practice recovery and I am loving it...

Kind Regards,

Alex