Slingshot 30M & 1H - The 2nd Coming...... | Page 30 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Slingshot-30 Min. PA System 394 replies

ECN && STP && Scalping && Hedging - Broker Review - Oct 2009 24 replies

question about "slingshot" indicators 1 reply

Advice/Criticism on Indicator Settings for 30M and 2H 0 replies

SlingShot Method A 12 replies

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

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

  • #581
  • Feb 17, 2009 5:12pm Feb 17, 2009 5:12pm
it's been a good day, the EA seemed to work better, overall balance is positive for today.
Attached File(s)
File Type: xls DetailedStatement 090217.xls   33 KB | 229 downloads
  • #582
  • Feb 17, 2009 5:32pm Feb 17, 2009 5:32pm
Quoting BlackWolf
Damon,

If Pavatar can get this EA perfected and it starts turning big consistent profits, we'll ALL buy NEW super computers!

Glad to hear you made some new coding discoveries, Gary. The Magic Number function is definitely something you need to trade and manage multiple EA's and orders. One of the reasons why I only run it on one pair right now. This was not a good day for v2.3 on 30M GBPUSD. Results attached. Looking forward to putting v2.4 and beyond through it's paces!
I think Gary will get it. He is working very hard at it. I have found that this system is not good just trading one pair as it is not good just trading one filter. Trading the whole things appears to give us positive days most of the time as far as i have been testing.

Thanks for your input
"Keep your eyes on the helpers" - Mr. Rogers
  • #583
  • Feb 17, 2009 5:37pm Feb 17, 2009 5:37pm
Quoting hsa013
it's been a good day, the EA seemed to work better, overall balance is positive for today.
Love that drawdown...1.44%...might even be a little too low. worth uping the stakes a bit if it continues to be this low. Thanks for your help.
"Keep your eyes on the helpers" - Mr. Rogers
  • #584
  • Feb 17, 2009 6:30pm Feb 17, 2009 6:30pm
Quoting llambert3
On a side note, since 2.3 has some problems still, I am running 2.1 as a test. ..... the 1hr timeframe, with the stoch,zzi filter set to true. Not inidividually, but together and using 7 -20 gmt.
This produced +$18.54 today.
  • #585
  • Feb 17, 2009 6:45pm Feb 17, 2009 6:45pm
Hi All -
Version 2.4 of the ea is attached. I'm really tickled with it! It seems to be a big improvement, at least in the backtesting and forward testing I was able to do today. The more folks testing the better. It opened and closed multiple trades candles in good fashion.

Since we haven't come up with a consensus on the SL issue yet, the code reflects post #1 (+1 or -1 High/Low of the entry bar on the close of the entry bar). I will change it when we settle the issue. For those who can open/change code, I've left the options of adding the spread to the high or low (depending on your preference) in the program, I've commented them out - change to whatever flips your switch! If we've made the progress we've been looking for with 2.4, then we can do any fine tuning needed, really get some good forward tests and get back to several suggestions that are pending (wanting to get the basic system coded correctly before moving on). Gosh I hope this is as good as it feels - if not, we'll stick with it until it works correctly - you've been great with the feedback and it really helps.

Gary
Attached File(s)
File Type: mq4 Slingshot_2.4.mq4   34 KB | 283 downloads
  • #586
  • Feb 17, 2009 6:48pm Feb 17, 2009 6:48pm
Quoting Pavatar
Hi All -
Version 2.4 of the ea is attached. I'm really tickled with it! It seems to be a big improvement, at least in the backtesting and forward testing I was able to do today. The more folks testing the better. It opened and closed multiple trades candles in good fashion.

Since we haven't come up with a consensus on the SL issue yet, the code reflects post #1 (+1 or -1 High/Low of the entry bar on the close of the entry bar). I will change it when we settle the issue. For those who can open/change code, I've left the options of adding the...
Ok buster, this better be for real
  • #587
  • Feb 17, 2009 6:57pm Feb 17, 2009 6:57pm


Quoting llambert3
Ok buster, this better be for real
  • #588
  • Feb 17, 2009 7:00pm Feb 17, 2009 7:00pm
Quoting Pavatar
Hi All -
Version 2.4 of the ea is attached. I'm really tickled with it! It seems to be a big improvement, at least in the backtesting and forward testing I was able to do today. The more folks testing the better. It opened and closed multiple trades candles in good fashion.

Since we haven't come up with a consensus on the SL issue yet, the code reflects post #1 (+1 or -1 High/Low of the entry bar on the close of the entry bar). I will change it when we settle the issue. For those who can open/change code, I've left the options of adding the...
Hey Gary....i have one more thing to do and that is re-issue new passwords for my trading accounts and then i should be back in business. i will be watching this EA closely to see what's up You've been great. I will definitely get back to you with what I know.
"Keep your eyes on the helpers" - Mr. Rogers
  • #589
  • Feb 17, 2009 7:15pm Feb 17, 2009 7:15pm
Good questions/comments about the ExpertID. I've provided the code below for your inspection and comment. Basically the ea determines if the "orderBelongsToMe" based on the trading pair (OrderSymbol) and expertID (default setting = 999999). I've also set the parameter "TimeSpecific" = true. This adds a unique number to the expertID for the particular TF I'm using. I do this because I trade all 5 pair (see post#1) on both the M30 and H1 TF's. By setting the TimeSpecific = true, I have generated 10 unique magic numbers = one for each of my 10 charts! Hope this helps a bit.

Gary


//+------------------------------------------------------------------+
bool orderBelongsToMe()
{
if (OrderSymbol()==Symbol() && OrderMagicNumber()==simpleMagicGenerator() )
return (true);
else
return (false);
}
//+------------------------------------------------------------------+
int simpleMagicGenerator()
{
int truMagic;

if (TimeSpecific)
truMagic = ExpertID + (Period()*17 + 203);
else
truMagic = ExpertID;
return (truMagic);
}
//+------------------------------------------------------------------+
  • #590
  • Edited 11:07pm Feb 17, 2009 7:50pm | Edited 11:07pm
Pavatar,

Should we change the ExpertID number in each EA we attach to the various charts then? Or...are you saying that this should'nt be necessary due to your coding?

Also, which of the filters are you activating in your tests?
- Bowen (a.k.a. Wolf)
  • #591
  • Feb 17, 2009 8:31pm Feb 17, 2009 8:31pm
Quoting BlackWolf
Pavatar,

Should we change the ExpertID number in each EA we attach to the various charts then? Are or you saying that this should'nt be necessary due to your coding?

Also, which of the filters are you activating in your tests?
Regarding the filters.....it is i believe more important at this time that the EA is functioning correctly. Once this base has been established we can do some more thorough testing with multiple people testing different scenarios. I hope you will be a part of that.

For now though, the technicals are more important. is the EA functioning as it is desingned to? That should be first and foremost at this time.
"Keep your eyes on the helpers" - Mr. Rogers
  • #592
  • Edited 8:51pm Feb 17, 2009 8:39pm | Edited 8:51pm
Hi BlackWolf -

With "TimeSpecific = true" you shouldn't have to change a thing in the ea for each of your charts to have its own unique magic number. I'm assuming your not, for example, using two EURUSD M30 charts with different trading options. In this case you would want to change the ExpertID number in one of those two charts. In other words, as long as each of your charts has a unique combination of pair + ExpertID +TF you don't have to change anything in the settings relative to generating a unique magic number. Does that make sense? Or have I messed you up even more?!

Regarding your other question, I backtested using different pairs, random one month time periods (from the last year) and different combinations of filters. I also was forward testing version 2.4 for the last few hours of the trading window today for all 10 (5 pair x 2 TF's) of my charts.

Gary

Quoting BlackWolf
Pavatar,

Should we change the ExpertID number in each EA we attach to the various charts then? Are or you saying that this should'nt be necessary due to your coding?

Also, which of the filters are you activating in your tests?
  • #593
  • Feb 17, 2009 8:42pm Feb 17, 2009 8:42pm
Damon -
I completely agree - lets get the basics down and go from there.

Gary

Quoting damonl
Regarding the filters.....it is i believe more important at this time that the EA is functioning correctly. Once this base has been established we can do some more thorough testing with multiple people testing different scenarios. I hope you will be a part of that.

For now though, the technicals are more important. is the EA functioning as it is desingned to? That should be first and foremost at this time.
  • #594
  • Feb 17, 2009 9:06pm Feb 17, 2009 9:06pm
Ok, I get it, Gary. I'm now testing all 5 pairs but each pair has a 1H and 30M chart/EA test for a total of 10 charts/EA's.

I also agree with you, Damon, about getting the basics down first so I will leave the filters alone (default value as Gary has them in the EA now).
- Bowen (a.k.a. Wolf)
  • #595
  • Feb 18, 2009 6:55am Feb 18, 2009 6:55am
I don't know if it's the market or the v2.4 EA (don't have time to go through it), but I started today morning at $5,050 and now it's $4,500... that's $500 loss in half a day, trading 0.05 lots.
It's good we have demo accounts!
  • #596
  • Feb 18, 2009 7:07am Feb 18, 2009 7:07am
Good morning All -

Version 2.4 is trading like it is on steroids! Technically it seems to be performing correctly (have not done lots of checking yet, just a little with the morning orange juice).

One feature of version 2.4 was to get the ea to max out the trading opportunities per some discussion with Damon many pages back. I interpreted this to include the following: on an entry candle, if the first trade hits the SL, open a new one (signal has not changed, so this seemed reasonable if part of the trading system is to max out the number of trades). One very negative aspect of this approach is shown in the attached image. In this case there was a a buy signal and I was trading the first three filter options independently - this means that each time the ea responded to the buy signal, 3 trades were opened. In this case the first group of 3 trades hit the SL and the ea opened 3 more, etc for a total of 4 signals on the same candle = 12 trades. All these trades hit the SL = lots of losers.

Had the first signal produced a sell, for example, on the same candle, only the intial 3 trades (first signal) would have remained open until the open of the next candle = 3 nice large winners.

The bottom line is the ea traded correctly as written - but, perhaps this definition of maxing out the number of trades is wrong (= my interpretation of Damon's comments). The original idea was that the signal was still long, so the candle would eventually reverse its momentum and the second, etc, trades would be winners. I've looked back over many of these type of candles and only a tiny few resulted in a 'reverse of fortune' so to speak where the overall net result was positive.

My initial reaction is to change the ea so it only allows one loser per candle - once you hit the SL, the ea doesn't trade again until the next candle (assuming it get a buy/sell signal). What'cha think?

Gary
Attached Image
  • #597
  • Feb 18, 2009 7:21am Feb 18, 2009 7:21am
Hello Gary,


I think you are right on with that analysis.



Regards,
Jeff


Quoting Pavatar
Good morning All -

Version 2.4 is trading like it is on steroids! Technically it seems to be performing correctly (have not done lots of checking yet, just a little with the morning orange juice).

One feature of version 2.4 was to get the ea to max out the trading opportunities per some discussion with Damon many pages back. I interpreted this to include the following: on an entry candle, if the first trade hits the SL, open a new one (signal has not changed, so this seemed reasonable if part of the trading system is to max out the number of...
  • #598
  • Feb 18, 2009 8:19am Feb 18, 2009 8:19am
ok - continuing on with the morning observations from above, I changed the code (a version I use for testing) to shut down trading on any bar after the signal triggers the first trade(s). Lots of losers were avoided. However, the image below shows an example of a missed trade that would have been positive. In this image the first trade was a sell, the price went up and hit the SL and the ea closed the trade - no further trades were opened for this bar. However, after the first trade(s) hit the SL, the price turned down. If the ea had opened a second trade(s), the result would have been a winner(s). In this case the winner(s) would have been bigger than the loser(s). The result could have just as easily been a winner(s) but smaller than the loser(s).

Seems to me, that examples like this don't warrant keeping a bar open for more trades (after the first trade). Why? The possible results seem to be (1) keep the bar 'live' and risk more losers after the first loser, (2) keep it open and get a reversal of fortune resulting in a winner but the winner is smaller than the loser or multiple losers, and (3) keep it open and get a reversal of forturne resulting in a winner that is larger than the loser or multiple losers. While I haven't done statistical analyses of these options, my gut feeling (and experience) tells me the odds of ending up with a reversal of fortune that results in positive trades over the long haul isn't going to happen. Bottom line .... I think I need to change the code so the bar is closed to further trading after it trades the first time. Comments?

This does not close out the issue that Damon commented on earlier about maxing out trades. The other issue discussed several pages back was about opening new trades when trades triggered on earlier bars are still open. Version 2.4 will open new trades with the start of new bar (assuming it gets a buy or sell signal) when trades triggered from prior bars are still open. I haven't studied the effects of this change in the ea and think its another system/technical issue we need to discuss.

Thanks - Gary
Attached Image
  • #599
  • Feb 18, 2009 9:25am Feb 18, 2009 9:25am
For new traders or anyone just trying to understand what they see on the chart with this ea, here is an example. It dawned on me there hasn't been much in the way of posting screen shots as we've had our discussions, so I'm going to add a few to hopefully help increase our understanding of all this with the help of pictures.

On the little attachement below, you'll notice 3 crudely drawn numbers. I didn't have the slingshot bar indicator on - didn't want to many colors on for the example. I was using the first two filters (maxmacd and stoch) turned on and trading independently. In the example both filters were indicating a long trend; therefore for any buy signal, 2 trades should have been opened at the same time (and they were for all 3 examples below = ea worked properly). The pair traded = EURUSD with the SL = 35 per post #1.

#1 - upon receiving a buy signal, the ea opened two buy trades with the SL placed 35 pips below the entry price. At the end of the entry bar, the two trades were in the negative so they are kept open and the SL is changed to -1 below the low of entry bar. A short time later the SL was hit and both trades were closed out at the exact same time (result = 2 losers at -28 pips each for a net loss of -56 pips).

#2 - again upon receiving a buy signal, the ea opened two buy trades (trend for both filters was still long) with the SL placed 35 pips. In this case both trades were in a profit situation greater than 1 pip at the close of the entry bar.....so, on the opening of the next bar, both trades were closed. In this case, you'll notice the two small arrows indicating the close of the trades are separated slightly (you can see two arrows, not one). This is the result of the trades closing at slightly different times. The first trade closed with a profit of +3. By the time the computer submitted the close order request and it was executed, the price had changed. When the second close request was submitted by the ea (milliseconds later), the second order closed with a profit of +5. Two winners for a net +8 pips.

#3 - same as #2 except for the following. Like the example 2 above, by the time the ea could execute closing the first order, the price had changed in those few milliseconds. Unlike example two, the price went down. So the first trade closed at +3 while the second trade closed at +2.

This situation is more common when your using a demo account - I hope we are all using a demo account. I don't care what the brokers say, servers for demo accounts never work as fast or as consistently as servers used for live accounts. That is not to say, this situation will not happen on live accounts, just that it should happen as often.

Hope this helps a bit - Gary
Attached Image
  • #600
  • Feb 18, 2009 9:26am Feb 18, 2009 9:26am
Gary, I downloaded 2.4 and forgot to set it up, so...

Anyway, I am not quite following you. I would think??? that if on 1hr and it goes below the low and then 1 pip above the close for a buy that it would take all trades whichever filter is there at that time, so the filters change within the hour? Meaning that if the macd fails is it possible that the stoch filter can be in effect later on in that hour when price goe above the close 1 pip again? I guess it can be so... sorry if you don't understand me.. lol

If macd filter trades and then hits the stop (fails), you are not trading the macd filter again in the same candle correct?
0 traders viewing now