Menu

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

Blessing Multipair EA 451 replies

Can anyone convert the Blessing 3 v3.9.6.09 into MT5? 9 replies

Blessing EA by ananthhh 13 replies

'Blessing' EA Modification Wanted 3 replies

Forward testing results of my ea modified from Blessing 3.9.6 18 replies

Attachments: Most Famous Blessing 3.9.6 EA and Setfiles
Exit Attachments

Most Famous Blessing 3.9.6 EA and Setfiles

Quoting emmanuel7788
{quote} what is the LAF Lot Adjustment Factor used in both cases?
LAF 0.5
Quoting cescof
{quote} Ok but why??? look at my screenshot ....same backtest one with balance set to 3000 and same with balnce set to 1.000 {image} {image} Non sense.... how to believe in backtest if we have this huge differences? I've the set posted that seems ok for 2014-2019, but if change start balance gives so different result it's difficult think to go live... What do you think about it? What's wrong? Thanks Regards
Quoting cescof
{quote} LAF 0.5

with the default LAF 0.5

you need to check what the starting Base Lot for both test cases.

with initial balance 1,000, the base lot would most likely be the smallest lot 0.01 tradeable.

however, with the balance 3,000 the base lot with LAF 0.5 would be much larger hence the drawdown would hit 50% by the 7th or 8th level.

run those two test cases again and change the LAF to 0.1 and analyse both results.

you can run tests by gradually increasing the LAF to 0.2 and 0.3 with the same account balance.

I don't use the default LAF 0.5 in my live trading. In fact I prefer to do my Money Management manually and input the first Base Lot manually.
Honesty is a very expensive gift. You wont find it in cheap people.WBuffett
  • #1,403
  • Quote
  • Edited 5:15pm Nov 25, 2019 5:00pm | Edited 5:15pm
Quoting richard96816
Arrgggg. I suppose if you trade only one-hour (H1) or four-hour (H4) timeframes you can get away with using some other data than YOUR BROKER'S DATA. Otherwise, it is my considered opinion that you are making a mistake. I use three different brokers. Their data are all very different from each other. Sometimes a .set file optimized for one will work with the others, but usually not. I'm not a pro or expert. Just a guy with a couple years of experience hacking MT4. And reading everything I can find online about it. I've seen impressive sites explaining...
Dear Richard,

The only reason one would need an extensive period of data samples is simply for statistical reasons.
Surely you learnt statistics and you know that a good sample, and one which can be properly analyzed, should include at least 300 samples, otherwise you include a great margin of error inside your parameters and you overfit.

Basically the market 5 years ago and the market as it is today are not the same, and trends change, but to get a good amount of trades and conclude anything from them, you must have a good source of data, and no, your broker's data is not it, sorry.
I do not know what your sources are, but it is a fact that good data sources are very costly and there is a reason for that.
Ask any bank authority or a hedge fund manager if he uses his liquidity providers data, or is he outsourcing his data for data manipulation and they will all say the same.
The world does not revolve around MT4 and the brokers supporting it, it is the MT4 programmers and traders who think they know better than statisticians and the rest of the free market, and that is why most of them fail.

And please, "It's not hard to get a couple months of good data from your broker. And it's really all you need." ???
Really?
It is VERY VERY HARD!
I have never seen broker's data without holes or broker's data which is more than maybe 85% accurate. Look at your MT4 report, see if you have mismatched bard, and how many, if you have none, share your broker with us as he is a saint.
In real time you will get accuracy of 100%, but history center? It is a joke....
Please guys, anyone who's reading this, try looking at your data with a magnifier glass before you trust it.

Edit: Not history center, and not broker's data from the chart. And Please do verify by looking at the report after a backtest. Even if you choose to use your broker's data, please at least verify it is valid before you trust it.
Thunder OM Flex Secondary All Time Return: -18.2%
I've been thinking about adding an option to set lot size for each level manually. Maybe up to five or ten levels.

Mostly to see what would show up in backtesting.

I have a suspicion there may be very good patterns that Blessing can't currently cater to. Maybe something like this ...

Inserted Code
UseManualLotSize = true
LotSizeL1 = 0.01
LotSizeL2 = 0.02
LotSizeL3 = 0.03
LotSizeL4 = 0.01
LotSizeL5 = 0.01
...
Quoting friday13
{quote} Dear Richard, The only reason one would need an extensive period of data samples is simply for statistical reasons. Surely you learnt statistics and you know that a good sample, and one which can be properly analyzed, should include at least 300 samples, ...
You seem to be talking about theoretical probabilities. Where you don't know the data or problem very well.

My real world experience (including much familiarity with Blessing and my brokers) is that: better than 75% of my solutions (optimizations) forward test quite profitably. Most with a back testing trade count of less than 200. Often more like 50. This is with Blessing and the various 28 pairs. This is over hundreds of tests.

I can usually place a good setup on a live account without forward testing and achieve good results immediately. (Though I still don't recommend it.)

In my most recent string of 30 set files I've probably only had one or two that didn't perform as well as expected. Failures are rare. All with broker data (FXChoice, Oanda and Forex.com.) I don't even check data much any more unless back testing isn't producing results.

Ultimately, if a setup forward-tests well then it doesn't matter where it came from! No need for tons of magical data blessed by the gods :-)
I'm sure there are some brokers out there with terrible data. Mine apparently aren't terrible. And the data I use these days mostly came in live, real time, over the last month or two and my MT4 software saved it for me. Real data from the same live accounts I'll be trading on. There is no better data available. Period. Data from anywhere else is simply not equivalent to my broker's live data.

DO NOT USE THIRD-PARTY DATA!

Forex is not the Stock Market. Forex is de-centralized. Today's Forex 'brokers' are more like the bucket shops of old. Each one has their own version of the current price action, filtered, delayed, spread-adjusted and modified to their needs and desires. Read your broker's contract and you will see where they reserve the right to modify the data. And they always do. They can't help it.

Best of luck!
Quoting speedkeks
Hello, after testing the version 3.9.6.12 - i found 2 bugs : 1. the holidays function did not work correctly - the date is showing wrong and the EA did not stop during the holidays. 2. the trend indication ( the last line with bigger letters ) dont show the correct trend. with the version 3.9.6.10 and 3.9.6.11 its the same. perhaps its possible to fix it. So far the EA works great for me - much thx
Thanks for the report. I'll take a look.
Quoting cdsaa
I would like share my experience with Blessing 3.9.6.09 This is an EA developed by JTAToday. It is one of the stable EA around. most difficult thing with this EA is set files. I have developed few set files for EURUSD, GBPUSD, XAUUSD. All of them perform well. Good stable returns for anyone who is looking for such. Caution: Use it according to your own risk appetite. Test it well. Also maintain proper Account Balance. Do not use it on small account balances. You can use ex4 version. mq4 is uploaded, for anyone who wants to modify. Not sure why one...
mine did not execute a single trade in GOLD# even laptop is on for more than 24hrs..will it not execute a trade with $481 account?
Quoting richard96816
{quote} You seem to be talking about theoretical probabilities. Where you don't know the data or problem very well. My real world experience (including much familiarity with Blessing and my brokers) is that: better than 75% of my solutions (optimizations) forward test quite profitably. Most with a back testing trade count of less than 200. Often more like 50. This is with Blessing and the various 28 pairs. This is over hundreds of tests. I can usually place a good setup on a live account without forward testing and achieve good results immediately....
Thanks Richard,

Actually what you talk about is only testing stage and forward testing stage, but you never go live, so how come you recommend so many recommendations to so many people here?
I've been trading live with Blessing for the past 2 years on 3 different brokers and testing it using 10 years of well-verified data.
I received average results in my live trading close to what I've received in my backtests.

Trying to use Broker's data gives faulty data!

Bolding and/or resizing text does not change this fact. You are deceiving some inexperienced traders and Blessing followers by giving advice which you do not even adhere to on your own live account and have no experience past Demo accounts and tests.

Guys -
I will end by warning again - disregard what Richard or myself are saying, please check your reports and verify you have valid data. MT4 knows how to verify validity of the data but not to correct it.
If you receive accurate data, well done (let's say 85% accuracy and better). I suggest no less than 90% accuracy.
You need to examine and fix your data if you do not see accurate data, or something like:
Modeling quality: n/a,
Mismatched chart errors: larger than 0
Here is an example of wrong backtesting process because of data accuracy:
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 5 KB
Thunder OM Flex Secondary All Time Return: -18.2%
Quoting friday13
{quote} Thanks Richard, Actually what you talk about is only testing stage and forward testing stage, but you never go live, so how come you recommend so many recommendations to so many people here?
Sorry if I hurt your feelings. I appreciate your contributions. I spent a lot of time messing with third-party data too. That's why I'm warning others against it. The folks that sell third-party data make money. Those that use it, not so much. At least not with EAs like Blessing.

I'm not saying you can't make money with other data. Just not as much and not as reliably.

Quote
I've been trading live with Blessing for the past 2 years on 3 different brokers and testing it using 10 years of well-verified data. I received average results in my live trading close to what I've received in my backtests. Trying to use Broker's data gives faulty data! Bolding and/or resizing text does not change this fact. You are deceiving some inexperienced traders and Blessing followers by giving advice which you do not even adhere to on your own live account and have no experience past Demo accounts and tests.
Again, sorry if I hurt your feelings. What I've been saying is so incredibly easy to prove, it's kind of silly to claim otherwise so forcefully.
Sorry if I hurt your feelings.

If you have multiple accounts then you know what I'm talking about. One setup doesn't work nearly as well on more than one account.

I'm sorry if I'm not very diplomatic. This issue is pretty important.

Blessing is a pretty good tool. But using it well requires lots of time spent optimizing. That's why I added the EnableOncePerBar feature that lets you use MT4's Open prices only mode instead of the extremely slow Every tick mode. Tick data was one of the big things third-party data sellers were pushing hard. Now we just filter them out and ignore them ... with no loss of profits. That's why MetaQuotes provided the Open prices only mode in the first place. It just makes good sense. And it works very well if your EA focuses on open prices.

There are bad brokers out there, so I hear. I've luckily not encountered one yet. Yes, my brokers don't give you all that much historical data. And it may not always be perfect. Ultimately, that's what forward-testing is for. No matter what crazy stuff you throw together for a system and data, if it forward-tests well then you can make money. It's that simple. Forward-testing is everything.

I repeat: Forward-testing is everything.

When I first started optimizing Blessing it took days and days to complete one optimization using 'good data' and plenty of history. I kept trying everything, kept optimizing and improving. Looking for time savings everywhere. With 28 pairs, six different timeframes and all kinds of other variables it seemed like an impossible task to test everything at the speed things were going.

Then I noticed more and more folks suggesting Open prices only and learning more about how Forex actually works. Learning that each broker's data is necessarily different since there is no exchange like NASDAQ or NYSE for Forex. No central authority, at least not that most brokers can access. (Besides, that freedom to adjust their data certainly must help brokers be more profitable!)

So, while there are lots of sites selling expensive 'clean' data, and extolling its virtues, and extolling the virtues of tick data, it's simply not the best stuff to use in real life. (I never use tick data anymore, and my results are better than ever before. Especially in my live trading)

Why wouldn't you use the same data you trade with to test? Data is never perfect. But the data you trade with is really all you've got. Anything else is not real. Not accurate to real life. How can you possibly do better than live data from your broker? Anything else is simply wishful thinking. Or good salesmanship by the seller.

Forward or demo testing is the real proof. No amount of advertising copy can change that. Start with one, two or three months of real live data from the same account you intend to trade live and optimize. I suggest M5 or M15 timeframes for best results with most pairs. That's what hundreds of tests have shown me. Do the same thing with third-party data and see which forward tests most profitably.

But each trader should test this for him or herself.

And ignore the damn 'Modelling quality' on the report. That's nonsense. Mine often says 'n/a', but tests just fine. 'Mismatched charts errors' is important, but mine is almost always 0. We don't want 'modeled' data. We want REAL data. We don't care about ticks anymore, so we don't need to 'model' them! Try it a few times, you'll see what I mean.

I keep looking for better EAs. And I keep coming back to Blessing. I keep making money with Blessing, and giving it back with other EAs. :-(

I'm not a master trader or Forex wizard. Just a retired programmer/analyst with too much time on his hands. :-)

I'm not selling anything. Just trying to use this forum to collect my thoughts, learn new things from others and maybe get some occasional bug reports.

Best of luck!
Quoting po1yap
{quote} mine did not execute a single trade in GOLD# even laptop is on for more than 24hrs..will it not execute a trade with $481 account?
Try adjusting your spread.
Adding to the conversation on testing with different amounts of history, I'd like to reiterate something I covered some time back in this list. I don't recommend trading a single Blessing instance on a whole account. Two, three, four or more instances and different pairs is a very good idea. The PortionPC parameter can help with that.

Crazy stuff can happen in the market. Blessing protects 50% of your account by default (MaxDDPercent). But you never want to experience a loss that large. If you split the risk between multiple instances you reduce your risk to more tolerable levels.

You can reduce the MaxDDPercent setting, but after lots of testing and real-world trading especially, I've decided to leave mine set to the default 50%. Many difficult events within that 50% seem to usually end up in profits. With lower settings things don't perform as well.

As someone pointed out to me, using lots of history for testing is more to do with these 'black swan' (rare) events. I prefer to reap the larger profits in the mean time and reduce risks through diversification.

Again, best not to trust your whole account to just one Blessing instance and one pair ... my opinion. If you're only trading one instance at least consider reducing PortionPC (with appropriate testing of course.)

(If your account is only $1K and that's easily replaceable for you then let 'er rip. But consider reducing your exposure as things grow. Just don't be surprised if you lose a chunk of it next time Trump does something stupid.)

Please refer to the manual if you don't know what these settings do.

Best of luck!
Quoting fareed51
Hi all, Well my testing continues, I've closed all trades multiple times therefore you can see many RED marks. I've finally reduced MAX DD to 5% :-) and Next week I will test with MAX DD 2%. Overall, GOLD & GBP pairs did fantastic regardless % was 20, 10, and now 5. Next week will be fun with 5% and then MAX DD 2% only :-) Yes testing in DEMO. MOST of red trades are because of my manual interference and testing by opening closing trades manually. Guys don't go live, yes I'm live with GBP & Gold only but you make a move on your own, otherwise i recommend...

Dear Fareed51,

Can you please advise how to get the open positions summary indicator that you have in this image? It would be very much appreciated!
Does anyone have opinions, ideas or experiences on Blessing's B3Traditional setting?

It turns on and off the use of limit and stop versus market orders.


Thanks.
Quoting richard96816
Does anyone have opinions, ideas or experiences on Blessing's B3Traditional setting? It turns on and off the use of limit and stop versus market orders. Thanks.

Hi richard

I trade Blessing3 manually i.e. I activate my buy or sell using manual FMC setting. This way I do use Traditional setting.

I recall the Traditional setting used to place stop orders when the market condition is ranging and B3 will either trigger buy stop or sell stop.

I have not tried you version 12 yet.
Honesty is a very expensive gift. You wont find it in cheap people.WBuffett
Quoting emmanuel7788
{quote} Hi richard I trade Blessing3 manually i.e. I activate my buy or sell using manual FMC setting. This way I do use Traditional setting. I recall the Traditional setting used to place stop orders when the market condition is ranging and B3 will either trigger buy stop or sell stop. I have not tried you version 12 yet.
Thanks for the response, Emmanuel.

I started out with B3Traditional off. It just seemed cleaner not having those orders hanging around. And more in keeping with Blessing's other attempts at not letting the broker know what you're doing ... A few months ago I added that option to my initial optimization set file and started letting Strategy Tester decide. Half the time it would turn it on and other times not.

I have been trading Blessing without specific market hours set. So it trades 24-hours. Which I kind of like. Lots of times when I sit at my computer here in Hawaii the main markets are closed. But I still get to see trades happening. And they're usually profitable. Great for small time-frames.

Stop and limit orders kind of suggest 24-hour trading anyway, since they hang around until hit.

But I've had a few times when limit orders were hit during fairly wide spreads, which reduced profitability. Now I'm thinking that the Stop and Limit orders aren't such a good idea for hands-off trading. Using market orders allows Blessing to open trades only when spreads are within your set limit.

Anyone have any insights here. Maybe Stop and Limit orders should be removed when spreads widen.
  • #1,416
  • Quote
  • Dec 10, 2019 11:56am Dec 10, 2019 11:56am
@richard96816 I wanted to ask you a couple of quick questions relating to your testing of Blessing.

1. Have you encountered pairs that optimise a set of results to a certain return, and then once those results are simply backtested the return differs greatly? I found this on the previous month for EURUSD (optimise gives multiple settings that give $400 return, set the inputs and run a simple backtest and the backtest reports $80 return). Confusing!

2. When backtesting with open prices only, are you ever running live with this set, or do you always revert to turning this switch off before forward testing?
Quoting n229vxhbx
@richard96816 I wanted to ask you a couple of quick questions relating to your testing of Blessing. 1. Have you encountered pairs that optimise a set of results to a certain return, and then once those results are simply backtested the return differs greatly? I found this on the previous month for EURUSD (optimise gives multiple settings that give $400 return, set the inputs and run a simple backtest and the backtest reports $80 return). Confusing! 2. When backtesting with open prices only, are you ever running live with this set, or do you always...
I'm not sure what you're changing between backtests. MT4 is a little quirky at times. If you're getting inconsistent results you might try restarting MT4.

I leave EnableOncePerBar set to true.

I most often test with about two months of history. Tuning up live instances about once a month to keep them as profitable as possible.

I do a first pass on the six primary time frames, m1 m5 m15 m30 h1 h4, and choose the one that looks best according to profit, profit factor, number of trades, average consecutive wins and losses, and max drawdown percent.

Good luck!
Quoting richard96816
{quote} I'm not sure what you're changing between backtests. MT4 is a little quirky at times. If you're getting inconsistent results you might try restarting MT4. I leave EnableOncePerBar set to true. I most often test with about two months of history. Tuning up live instances about once a month to keep them as profitable as possible. I do a first pass on the six primary time frames, m1 m5 m15 m30 h1 h4, and choose the one that looks best according to profit, profit factor, number of trades, average consecutive wins and losses, and max drawdown...
Just a question..how could you understand and decide that your setting are no more updated? I mean you try to optimize just using last 2 month and often happen that same setting could no survive for a year and more cause market is challinging.... How could we find signal that we have to change setting?
Best regards
Thanks
Quoting cescof
{quote} Just a question..how could you understand and decide that your setting are no more updated? I mean you try to optimize just using last 2 month and often happen that same setting could no survive for a year and more cause market is challinging.... How could we find signal that we have to change setting? Best regards Thanks
I watch performance. If daily or weekly performance starts to fall off then I think about re-optimizing.

My shorter timeframe setups often do multiple trades per day. When that starts to slow down it's usually time for a change ...
I have tried version 12 in demo. I opened graph on M15 and attached Blessing version 12. When I have 5 BUY and floating profit was about $50, I changed Time Frame on MT4 from M15 to H1 and then Blessing suddenly open 1 Buy Stop and 1 Buy Limit with lot size 0.01.

Is that normal for Blessing to do so?
0 traders viewing now