Indicators development for T101's not very simple method

Time for some sleep now after a hard day;s work and also some testing today when time permitted,,,

Hope to test it whenever time permits and contribute my little part towards this EA. Hats off to the coders Don, Kos, Mr Pips, Charvo and others

Be back tom

My reason for changing from i to 0 was to enforce to fifo rule.
The first order places must be the first order closed.
However, changing this back to i should still work OK because the fifo rule will be met unless another trade is open for the same pair.

If you are using another EA this could definitely cause a problem.

I checked the code for MACD close and it does appear to be correct. It uses the last 2 closed candles for the signal and does not consider the current open candle.

Don,
I do not want to get too carried away with changes yet but I am concerned that the code in some of the functions is mixed with closing and opening trades. It would be much cleaner to have functions that close trades when trades are open and not then open new trades. This would require splitting some of the functions. Get the strategy variables as one function, then in the start function check for open if no trades are open, check for close if trades are open.

Robert

The issue with the instant reopening of the basket after your trailing stop is hit is going to be a tricky one. This is compounded further on the m15 timeframe.
It's a tradeoff between the tight trailing stop grabbing locked in pips as much as it can, and deciding when to stop trying to reenter the current basket signal.
A couple of solutions, not sure which would be best
1. only trade a basket x many times per fresh signal
2. loosen up the profitlock so it doesn't throw you out so fast like it does now
3. ?

Playing with 1k on m15, it doesnt take much for it to hit $10 profit, lock in $5 and retrace enough to hit that $5 and close. I think it's too tight, some inputs to adjust this would be a nice idea I think.
JMO

Robert,
Another slight problem I just noticed. I just experienced a nice strong move that actually locked in $20 profit and closed.
But it appears that at least the screen display did not reset all the way (not sure yet whether the profitlock didnt reset either or if its just the display).
So a new basket is open now, but the profit lock now says 'must reach $20 to lock in $10, when it's supposed to say must reach $10 to lock in $5.
Can you take a look?
Thanks

my metatrader crashed

i will show you here the crash report picture

the crash happened when it wanted to close the trade of the tradebasket.

using version 06-09-10 as shown in the pictures.

tazman,

It looks like both you and trade4fun must be using an older EA version, since Robert said he put in Reset function. Maybe I'm wrong about tazman, but trade4fun definitely is using older EA.

If you're "playing" with $1,000, then Lots (given the current numbers in user inputs which got changed not by me) would be:
(1000 x (1/14)*0.01))/(4x10) //on standard account
= $0.7142 / $40
= .01 Lots (one micro lot per pair)

(1000 x (1/14)*0.01))/(4x1) //on mini account
= $0.7142 / $4
= .1 Lots (one micro lot per pair)

same thing, different account type (standard vs mini)

So you're trading a micro lot each pair, about $0.10/pip, so $5 is about 50 pips, and $10 is about 100 pips, so it locks in 50 when it reaches 100. It does seem like a tight stoploss, but there are times I've been glad it was that tight, and other times I wished it was looser.

Regards,

Don

first of all want to express my gratidude for the coders , strategy thinkers and all the other people why have put so much time and effort in this. i didnt thought i mentioned that before. but guys you rock

also i might have an idea to fix the problem we have now with the constant buy sells after sale is closed.

i was thinking of 2 options.

1. so far in my experience, the first trade is always a winner.

so my idea is this. first idea is the save idea.

it takes 4 signals to trigger a buy or sell. after first trade is done,why not ride it. just dont open any sales or buys even it shows 4 signals. wait till 1 signal clears. then EA becomes active again. this way we wont make any money after it. however we can be sure when next clear signal hits we are back on the first trade.

2. after first trade finnished. open a second trade as it goes now. but make the basketprofit and stoploss smaller. do this in percentage.

is the second trade also a winner. then change the stoploss and basketprofilt smaller again smaller for third trade.

because as we know how longer the signals stay, how weaker in the end it will become. there will always be a hitback sometime. so i thought this is maybe a good idea for an exit.

anyway thanks again guys and hopefully more people will come up with some great ideas that help.

Robert,

In CheckProfit(), I would use myTotalOrders = GetOrdersTotal(); instead of myTotalOrders = OrdersTotal(); // GetOrdersTotal() already has magic number, so we would only look through orders with magic number, not all orders when in CheckProfit(). I see it's no big deal, since you are using magicnumber for adding up profit, when looking through all orders.

I do see where in StrategyEntryRules() I close SellAll trades that are open if I get a BuyAll signal. This is on purpose and must be done since we want a buy signal to override an older sell signal, correct? And so the old trades must be shut down and we reverse direction on them. I did the same with SellAll also. So I have 4 if statements:
IF sellall open when get buyall signal, Close SellAll; // we use Trade to give us tradetype, so we don't close BuyAll when getting a BuyAll signal - if it were to happen.
IF no trades are open and get buyall signal, buyall;
IF buyall open when get sellall signal, Close BuyAll; // we use Trade to give us tradetype, so we don't close SellAll when getting a SellAll signal - if it were to happen.
IF if no trades are open and get sellall signal, sellall.

What needs to happen probably, to make sure no closing and opening of orders on the same tick, is what I've said at least a couple times already (although I can understand getting lost in number of posts, and you've done great), is to make 2 if/else statements instead of 4 if statements:
IF sellall open when get buyall signal, Close SellAll; // we use Trade to give us tradetype, so we don't close BuyAll when getting a BuyAll signal - if it were to happen.
ELSE IF no trades are open and get buyall signal, buyall;
IF buyall open when get sellall signal, Close BuyAll; // we use Trade to give us tradetype, so we don't close SellAll when getting a SellAll signal - if it were to happen.
ELSE IF if no trades are open and get sellall signal, sellall.

So on this tick, it may close or open trades, but not both. At least that's what I think, and could easily be wrong and it not fix the issue. Only trying it can tell us for sure.

As far as LSMA in color, I do believe it would be good. I am also testing others out I think would be good. For now, let's try our best to not change the strategy if it's working for Kos. BUT Kos needs to make sure he's testing the latest EA (now at 06-09-10a). Everybody else can test, and they may add their opinion, but I am determined not to get side-tracked before we get this EA done. After that, I am pretty sure all heck's gonna break loose and people will want to add this or that (we do, other people have, but we haven't even gotten the bugs for sure worked out and I see people need to settle down and test the most recent EA and give feedback on if it's doing what it should do, per Kos, or not. If not, we need to fix bugs. If it is working like Kos said, I still have one thing I'd like to clear up, and that's timeframe trend (Weekly or Daily) being different for different brokers, and wondering if we need to implement Broker GMT Offset and set our week and day open time to his broke, because waiting and missing so much of the trend when I'm using MBT I think might be a problem, where maybe Kos is getting in much earlier because of broker he uses? Don't know, just a thought. Anyways, after we get it there and Kos (if he will) tests on multiple brokers with different start times for week and day, and see that it's acting right on these multiple brokers, THEN AND ONLY THEN do I think we should think about changing anything else. And when we think about changing anything else, you know what that will do. You see here how we have so many EAs that some aren't using the right one most likely - and this is only one strategy. I don't know how we'll handle other EAs past this one, but I say let's get this right, and after that make sure the names of further EAs are different from this one.

Having said all that, I'm wondering if there's anything else that you see that needs changed besides what has already been fixed or won't be fixed by what I said above.

Thank you again Robert for your patience. It's already getting too many requests that I would have to tell people to chill out. Either we code this strategy, or blow it to heck with all kinds of changes from different people. We have to decide and stick to what we decide. I have decided what I want to do, and have voiced it. If I didn't think what Kos is doing would be profitable, I would have went with what banana was doing, which looked close to this anyways if my memory serves me correctly. There should be time in future to try multiple EAs FROM ONE GOOD FIXED AWESOME EA.

Regards,

Don

Yeah, I definitely have robert's latest, and it's resetting fine other than the issue I just described. I am sure Robert will see it pretty quick, looks like maybe the profitlock slipped outside the loop again or something.
It's definitely not only display related, as my new basket did just cross over the 100 pip mark, and didn't lock in 50 pips like it should.
So to sum up: profitlock isn't resetting completely now in this latest version

I'd be curious to get Kos' input on how he handles the situation manually. Does he lock in 50 pips after hitting 100 pips, and if so, does he instantly reenter the basket when it retraces to take him out at 50 pips? If not, what does he use to decide whether to attempt the basket again or not?

tazman,

you still using 1 micro lot per pair? If using more, then it will not be $10 to lock $5, but say you're trading 2 micro lots per pair, then it will be $20 to lock $10.

Regards,

Don

P.S. Tired now. Gonna get off here and rest.

P.S.S. Also, don't know how many people see this happen, but it's becoming a weekly thing I'm seeing: Right about the time it's time to trade because we have LONG Strong or SHORT Strong agreeing with other indicators, the trend has run it's course and it's ready to turn right about the time we get in the trade. This just proves what I've said before about the 0 line (Weekly open for basket) being support/resistance. And why I made it to where this trend agreement (conditon 4) can be turned off if someone doesn't want to use it. I'd rather get in the trades earlier myself, but if it wasn't Kos' strategy whose could we have coded. Before him was banana. Other than that, there hasn't been anybody giving a strategy how to trade T101 (other than Julius himself, who was misunderstood on some things I believe).

Yes, its still trading .01 per pair. But something has changed after closing about 3 baskets in profit, something made it change to lock 10 after 20 now, but just not sure why/what.
Maybe that's the intention after having x number of successful baskets? lol
I do know that I like it, I am at about $12 right now, so at least the trailing profitlock has loosened up for me, just not sure why it happened lol

Using GetOrdersTotal here would cause a problem if another trade is open by a different EA or manually. It would cause some of the trades open by this EA to be missed in the profit calculation.

I might have missed a letter somewhere in the versions but the one attached here has hopefully solved some problems.

I have added the trading sessions code, this code has been working correctly for years now.

I also adding using lsma with 25 by default for determining trend.
I also split the functions a little but the logic is still the same.

Basically get the buy and sell conditions using GetBuySellConditions function.

These conditions are then checked for exit if trades are open or for entry if no trades yet.

StrategyExitRules function has the code moved from StrategyEntryRules to check for close of trades.

StrategyEntryRules now only checks rules to open trades.

Don,
If you like check where I now use TradeDirection and the function TradeDirectionFromTrend. This will make it easier to test other ideas for trend if needed.

Robert

Don,

My unsolicited opinion...

I don't understand why all these new features are being added to the EA when the original EA still has bugs to be fixed.

I think you need to stop the insanity now and get back to original goal of a 100% functional EA based on Kosomolate's trading strategy.

And don't share any more EA source code until you have a properly functioning product.

yeah, definitely. i second this.

i found quite a few readers showing up suddenly within last 48 hours. without too much mediation, they provide some other creative ideas----let's NOT to.

let's focus on clean the logic of the EA of Don and MrPip, complete its functions first.

Feedback your observation of the EA's performance/issues, instead of suggesting new entry/exit rules.

if Don had support from MT4 guru like you, ZZ, the code doesn't have to spread like this.

just sometimes we tend to get desperate when picking bugs. i found even a perfectly Alpari EA will suck ass when running on IBFX. these kinds of problems cannot be easily solved by the author alone.

luckily MrPip offered his MT4 expertise to help, i'd dare to appear being rude to request that all programming efforts should focus on Kos idea, and to be complete functional ONLY. Right now, let's put away distractive idea.

Well, there is only 24 hours in a day...I have my own EA development, testing, and monitoring to spend time on. If there were 28 hours in a day, we all could get so much more done!!!!


I think this was the point of my comments. It does not appear you all need help adding features...rather, you need help getting the features of the original EA to work correctly. And that is where initial efforts should be directed.

Just offering advice based upon 16 years of programming experience. I'm sure you guys will get it all worked out one way or another.

Back to the sidelines....

Don,

i Used a newer version, you can see that in the pictures i show when my EA crashed.

and i am saying nothing about changing stategies. i am only suggestting a way how it could be possible fixed in the EA. nothing more nothing less.

there is obviously a problem and i only like to help out. so please dont take my suggestions the wrong way.

thank you.

Functionally, it shouldn't have changed even though Robert added those options.

Having said that, I would ask Robert to refrain from adding anything else at this point, and let's stay on task of fixing existing bugs.

I can still read my code, so I'm cool with how things have went. Having said that, I am seeing that were right are now not working (good reason to not add anything new yet). For instance, Buy condition 3 should be true if Buy Conditons 1 and 2 are true and BI Red is more than Bi Blue line. Yet Buy Condition 3 is false right now with no explanation from me as to how.

Having said that, I will still leave c version (last one uploaded by Robert) on the chart to see if Kos' strategy EA makes trades like it should. That's all I can do at this point.

The only thing I see blatant is a change from 40 stoploss per pair to 4, and RiskPercent has gone from 10 to 1. I ask that these be changed back to what I had them. If someone wants to change only their copy, that's fine. But I uploaded the EA with the settings I wanted, and wish them to stay that way. If Kos' strategy is all I think it is, there is no reason to risk only 1% per basket trade, unless you're already a millionaire and have an a problem with making more money.

Regards,

Don

trade4fun,

The reason I said you were using an old EA version was because I know what the newer EA version (which is now old) was showing on screen at the time, and yours showed an older name, so it was just obvious.

No harm intended, just don't like to see things spinning out of control when we all know what should be going on, and nothing more at this point. I just hope people keep up with the updates if they do decide to test.

Regards,

Don