SteveHopwood's scooby-doo Shag'emAll trading robot | Page 6 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

What Would Be The Ideal Trading Robot For Our Markets Today? 2 replies

SteveHopwood EAs 4 replies

Robot trading 120 replies

What’s Differences Between Robot and Manual Trading Money Management Systems 0 replies

FX Engines Robot?? 3 replies

Attachments: SteveHopwood's scooby-doo Shag'emAll trading robot
Exit Attachments

SteveHopwood's scooby-doo Shag'emAll trading robot

  • #101
  • Dec 17, 2009 1:39pm Dec 17, 2009 1:39pm
Steve,

Running this in Demo according to the rules. When I read the code though found this in MakeTradingDecision:

ma1 = iMA(symbol, 60, 2, 0, MODE_SMA, PRICE_LOW, 1);
ma2 = iMA(symbol, 60, 2, 0, MODE_SMA, PRICE_LOW, 2);
if (ma1 > ChannelLow || ma2 > ChannelLow) return;

which doesn't match with running it on a 4 hour chart with the template, which of course uses a 4 hour SMA.

Have I missed something? Or is it a bug?

Thanks
  • #102
  • Dec 17, 2009 2:41pm Dec 17, 2009 2:41pm
Seems I missed something. This is you trying not to trigger every time the price crosses out of the channel.

It doesn't match your original rules so it threw me.
  • #103
  • Dec 17, 2009 2:41pm Dec 17, 2009 2:41pm
Quoting SittingDuck
Steve,

Running this in Demo according to the rules. When I read the code though found this in MakeTradingDecision:

ma1 = iMA(symbol, 60, 2, 0, MODE_SMA, PRICE_LOW, 1);
ma2 = iMA(symbol, 60, 2, 0, MODE_SMA, PRICE_LOW, 2);
if (ma1 > ChannelLow || ma2 > ChannelLow) return;

which doesn't match with running it on a 4 hour chart with the template, which of course uses a 4 hour SMA.

Have I missed something? Or is it a bug?

Thanks
This is for the filter Steve added that ensures the last 2 hours are still tradeable- I haven't tried version out yet- but recognize the code... Just waiting to get home and see how the original version did over the last 12 hours...
  • #104
  • Dec 17, 2009 3:52pm Dec 17, 2009 3:52pm
Quoting SittingDuck
Steve,

Running this in Demo according to the rules. When I read the code though found this in MakeTradingDecision:

ma1 = iMA(symbol, 60, 2, 0, MODE_SMA, PRICE_LOW, 1);
ma2 = iMA(symbol, 60, 2, 0, MODE_SMA, PRICE_LOW, 2);
if (ma1 > ChannelLow || ma2 > ChannelLow) return;

which doesn't match with running it on a 4 hour chart with the template, which of course uses a 4 hour SMA.

Have I missed something? Or is it a bug?

Thanks
This is an attempt to make sure the channel is moving in the correct direction before entering the trade, but it looks as though I have the conditionals the wrong way around - I do this all the time.

In the snippet you quote, the robot has spotted that the market is below the channel low, and is therefore looking to go long. The code you quote here is supposed to look at the 60 min moving average, and allow the trade if this is moving upwards. As you can see from the code, the trade is actually cancelled if the ma is moving upwards. For sure, this explains the number of losing trades the robot has been taking.

Sorry guys. I got my < and > mixed up. Again. As usual.

The fix is in post 1. Download this and see if the bot actually makes some profit.

SittingDuck, sitting ducks usually get shot. On this occasion,

Nice piece of proofreading. Cheers.

  • #105
  • Edited 6:14pm Dec 17, 2009 5:04pm | Edited 6:14pm
Hi Steve,

I'm running the Shag'emAll bot on an FXM demo (4 digit crook) and the GBPJPY pair is giving it fits! All the other JPY pairs are 2 digit except for GBPJPY. It for some reason is 3 digit and the bot reads the channel size as a 4 digit number. I can see your "if" call for 3 digits and it looks correct. Take a look if you get time and I just will not use GBPJPY. Thanks for all your great work........whip


PS: I forgot to tell you that I'm running your updated version from the above post of the bot....It is the only EA I have on this broker
  • #106
  • Dec 17, 2009 7:18pm Dec 17, 2009 7:18pm
Quoting SteveHopwood

SittingDuck, sitting ducks usually get shot. On this occasion,

Nice piece of proofreading. Cheers.

It was at about 6am and I knew something wasn't right, but couldn't quite place it. Certainly explains why it seemed to always wait for the market to turn in the oposite direction before trading. I should be awake enough to fix and test tonight.

I also have a NO BLACK BOX rule. Which means if I can't read and understand the code, I don't use it.

I don't want to be a Sitting Duck, hence the reminder when I log in
  • #107
  • Dec 17, 2009 7:26pm Dec 17, 2009 7:26pm
Quoting SittingDuck
I also have a NO BLACK BOX rule. Which means if I can't read and understand the code, I don't use it.

I don't want to be a Sitting Duck, hence the reminder when I log in
Sounds like A Plan to me.

Maybe I should rename myself BugWaitingToHappen.
  • #108
  • Dec 17, 2009 7:29pm Dec 17, 2009 7:29pm
Quoting hiredwhip
Hi Steve,

I'm running the Shag'emAll bot on an FXM demo (4 digit crook) and the GBPJPY pair is giving it fits!
Couldn't be a tad more specific, could you? You know what you are talking about; I don't.

  • #109
  • Dec 17, 2009 7:43pm Dec 17, 2009 7:43pm
Hi Steve,

The GBPJPY line on the front page reads as follows:
200 sma is 4 digits
channel high is 4 digits
low is 4 digits
depth is 4 digits
direction is 4 digits

All should be 2 digits, the depth should be 50 not 2450. I hope I'm making myself understood. And that pair never trades! FXM shows that pair as a 3 digit (123.345) pair. All other JPY pairs are 2 (123.45). Thank's for questions.

whip
  • #110
  • Dec 17, 2009 9:24pm Dec 17, 2009 9:24pm
Hi Steve,

After doing a little digging, I think this FXM demo and GBPJPY pair situation is not an EA malfunctioning problem but a broker problem and I think I will switch brokers. The GBPJPY pair is setup as a 5 digit crook would have it and the rest of FXM is setup as a 4 digit crook. I do not think that will work Thanks for your time and patience and keep those great ea's coming...........whip
  • #111
  • Dec 18, 2009 3:04am Dec 18, 2009 3:04am
Stops are too tight. Changing to = sma200 helps. Need to manage them a bit better though to stop drawdown getting out of hand.

Maybe:
Break even @ 0.5 * Channel Depth
  • #112
  • Dec 18, 2009 3:41am Dec 18, 2009 3:41am
Quoting SittingDuck
Stops are too tight. Changing to = sma200 helps. Need to manage them a bit better though to stop drawdown getting out of hand.

Maybe:
Break even @ 0.5 * Channel Depth
I am going to leave things as they are for now. The robot has sent some trades that are looking good, so I want to see how it performs after the bug-fixing from last night.

  • #113
  • Dec 18, 2009 4:09am Dec 18, 2009 4:09am
I've fixed my Money Management problems, now it needs to be tweaked.

As to the EA, i've changed the Channel Depth to 800, just to weed out some of the more undesirable trades.
  • #114
  • Dec 18, 2009 4:17am Dec 18, 2009 4:17am
I'm just getting up to speed here...Loaded up the Ea and got my first trade right away! Pretty cool except that this pair is not in my pairs to be traded list.
Attached Image (click to enlarge)
Click to Enlarge Name: shag em.gif Size: 61 KB
  • #115
  • Dec 18, 2009 4:20am Dec 18, 2009 4:20am
Steve,

The following calculation for obtaining base lotsize seems to be much better regarding DD.

Base.Lotsize = ((AccountBalance() / 50000) * 1.5)

Also,

TP = (AccountBalance() / 100)

Scoobs
  • #116
  • Dec 18, 2009 12:33pm Dec 18, 2009 12:33pm
Quoting SteveHopwood
This is an attempt to make sure the channel is moving in the correct direction before entering the trade, but it looks as though I have the conditionals the wrong way around - I do this all the time.

In the snippet you quote, the robot has spotted that the market is below the channel low, and is therefore looking to go long. The code you quote here is supposed to look at the 60 min moving average, and allow the trade if this is moving upwards. As you can see from the code, the trade is actually cancelled if the ma is moving upwards....
OOO that explains why when I swapped the buy and sell positions earlier in the week I got soo many winners...
  • #117
  • Dec 18, 2009 5:32pm Dec 18, 2009 5:32pm
Quoting Jefftrader
OOO that explains why when I swapped the buy and sell positions earlier in the week I got soo many winners...
Hehe. This reminds me of Steve's Patent Trading Systems Signals Service, that I posited a while ago in the BOJ thread.

It goes like this:

 

  1. I do all the analysis. MACD us rising. The 3 sma is above the 9 sma which is above the 200 sma. The gaps in between each is growing. There have been at least 5 higher high's followed by a retrace for 3 candles. The national bank of the second in the pair has announced that its currency is too strong and that it is about to dump trillions in order to weaken it to help exports.
  2. I announce, "Buy immediately".
  3. You sell and make a killing.

If one of my coding conditionals is >, then it is usually best to assume that < will be the correct one. Hey ho.

Once I made the conditional change, the robot performed much better.

I am going to make two changes tomorrow, in time for trading next week:

 

  1. Replace market orders with pending trades set to fill at x pips away from the current price. The idea is to make sure the market is already moving in favour of the trade when it is filled.
  2. Move the tp inside the channel by x%, just to make it that bit easier to hit.

Also on the radar is a BE function, if this proves to be necessary.

  • #118
  • Dec 18, 2009 8:24pm Dec 18, 2009 8:24pm
Quoting scooby-doo
Steve,

The following calculation for obtaining base lotsize seems to be much better regarding DD.

Base.Lotsize = ((AccountBalance() / 50000) * 1.5)

Also,

TP = (AccountBalance() / 100)

Scoobs
Will you be incorporating this formulae in KG too Steve? Mypenny a/c (thank God) has been having humungous DD and to tell you the truth, so is profit. It goes way beyond what it should and I have to close those baskets manually.
  • #119
  • Dec 19, 2009 11:31am Dec 19, 2009 11:31am
Latest version of the bot in post 1. Read the details in Updates.

Let's see if this helps the profitability.

  • #120
  • Dec 21, 2009 3:08am Dec 21, 2009 3:08am
Quoting SteveHopwood
Latest version of the bot in post 1. Read the details in Updates.

Let's see if this helps the profitability.

Dear SteveHopwood
I'm new of this forum, sorry for my English,
I'm not a programmer, but I need someone to make an EA expert about of my strategies that I think this strategies wins forever for many pairs , I put here my pdf file it will say what I whant mean.
I'd like your opinion
Thanks a lot
Attached File(s)
File Type: pdf EA Bluestrader.pdf   17 KB | 357 downloads
0 traders viewing now