PolyFitScalper - page 59
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
In my settings (which I am using) this one https://www.mql5.com/en/forum/179781/page10 - this EA is openning the trades at GMT night only, and most profitable pair is EURCHF. So, basicly I do not need news filter as no any significant news for USDCHF and EURUSD at night.
To use this EA during the day time - yes - we need newsfilter. But we need to filter high and medium news events.
By the way - DayTrading EA (this one) is counter trade too, and this EA is using just few standard indicators, and it is not martingale, and having exit (same as stop loss) too. And it is profitable for 1 year trading (not backtesting; just forward testing) for GBPUSD trading at GMT day time incl news and so on. Just 20, 40 or 60 pips in a week.
You can see:
So ... may be - we can change some logic for this Polyfit?
To trade some ... 5 or 10 trades (up to 20 trades) in a week with very low risk?
Just an idea.
By the way - I am not a coder but I think that this Polyfit was coded in some very complicated way by Igorad (something as "old fashion" way). So, it may be much more easier to create new EA than to edit this Polyfit.
ND,
thanks, I am going to look at DayTrading EA, hadnt seen that yet.
What I have noticed is that during off-hours (when Polyfit is set to trade by default for example), these scalpers can trade well counter-trend.
BUT, during GMT day (prime time), it looks like with-trend may be better - but I havent yet seen a good with-trend scalper for daytime. I was looking at tickscalper, is that good?
This polyfit with cfb-log filter is almost perfect during non-prime time, it only ever needs to martingale sometimes 2 or 3 progressions at the most, usually none at all. And that is on all pairs, though I agree some are super-easy to use, like EUR/CHF.
Odysseus
Tickscalper?
Basicly - Tickscalper should be more risky than Polyfit.
But with this settings and with EURGBP pair only - it is very profitable (1 year forward testing): min lot size was 0.01 and max 0.32; and the profit for 1 year was about 13,000 with very low drawdown. Profit factor = 20
Trading (forward testing) for 1 year. I had initial deposit as 10,000 and it is 23,000 for now. But initial deposit can be much lower than 10,000 as drawdown is very low and max lot size was 0.32 only.
It is here.
Other pairs are profitable too (up to 100,000 dollars profit for 1 year) but much more risky.
So, EURGBP is the best for Tickscalper.
As to Polyfit so - if we will use EURCHF pair only (may be - EURGBP too), and if this EA will be recoded (simplified) to some more modern way with newsfilter so it may be very good (as risk to use it will be minimal).
Polyfit should be less risky to use by definition but it was coded in such a complicated way so simplified TickScalper is winning in some cases with profit.
By the way - I am not a coder but I think that this Polyfit was coded in some very complicated way by Igorad (something as "old fashion" way). So, it may be much more easier to create new EA than to edit this Polyfit.
I agree Newdigital, Polyfit was coded very complicated and seems more i do to this version the worse it gets
, and think i will start on a new version, because i'm out of options on fixing this one.
Odysseus was noticing you are using large starting lot size, IMHO very dangerous for this Ea news filter or not, but as long as its demo guess it doesn't matter and none of my business
ND, just went thru your TickScalper statement in detail.
In the last year, there were 4 occassions of the martingale getting up to 1.28 lots.
8/13/2010 (began 8/10 at .01) - FOMC DAY!!!!
9/8/2010 (began 9/6 at .01) - US BANK HOLIDAY
2/15/2011 (began 2/9 at .01) - GBP TRADE BALANCE and BERNANKE TESTIFY
4/18/2011 (began 4/13 at .01) - GBP unemployment big surprise, several USD high-impact news
SO, wouldnt it be reasonable to think that modifying the TickScalper to avoid high-impact news, the result would be better? And I would think we could use all the pairs you had originally tested, werent there 4 or so decent ones? Mr. Tools obviously agreed, as he attempted to implement newsfilter on it himself. Look at it this way, if you avoid high-impact news and therefore dont NEED to martingale up to 7 progressions. You would never get beyond 3 or 4 at the most, and therefore could use a much higher base lot with less risk and more profit. Right?
So we are back to this:
Polyfit (whether coded complex or not) is a VERY good counter-trend scalper for SOME pairs at least and with a WORKING newsfilter and the cfb-log filter I added, it is darn near perfection in motion for NON-PRIME TIME HOURS. So can we fix or re-do it with this in mind? I am not a coder, I was barely able to add the cfb-log min value filter LOL...we need Mr tools or Mladen or Igorad....
And my personal view is that for PRIME-TIME (euro open thru london close), we could use a really good with-trend scalper, I personally love tickscalper because it is always trading and on EUR/GBP it ROCKS (maybe some of the other pairs too depending on broker and ONLY with a working newsfilter IMO). I am curious too if perhaps we couldnt add SOME kind of a cfb based filter to tickscalper to enhance its performance.....
Odysseus
PS Mr tools yea I would never use that lot size on live acct, it was just to "push boundaries" on demo while testing....I DO have my mod of polyfit running on 2 live accounts tho - check it out heres one of them: up 50% in a week or so....(starting 5/23 it was only polyfit on that account)
Hi odysseus11,
TickScalper is good for EURGBP. As to all the pairs using so - may be ... but scalping is not TSD/MTF EAs where we should trade as many pairs as possible. Scalping is something to trade 1 or 2 pairs only simultaniously.
So, to summarize:
- yes, it may be good to modify TickScalper for news filter. In this case - we will have more profitable pairs to use. And to improve by some indicators for example.
- as to Polyfit scalper so, as I said - the code for this Polyfit is very complicated and very knotty (coded by Igorad) so that is why it takes so long time to improve this EA. If it will be done by some new indicators and newsfilter too so it may be fine.
Besides, it may be good to see some trend following scalper (something similar logic with Kaufman AMA break out EA.
In this case - we can use both: counter trend Polyfit and trend following Polyfit on the same account.
So, if it may be done so it will be very good. If not so - it may be easier to create new scalping EAs instead of improving those scalpers.
OK, understood.
Well, the issue I had with polyfit 1.05 appearing to take trades that werent outside the polyfit bands was simply my own error, I had adjusted the length on the bands indicator (not the EA), so again it was simply my own error.
HOWEVER, the newsfilter even on v 1.59 1.05 is still NOT functional.
So again we need that fixed somehow.
And I would like clarity on 2 issues from ANYONE who can help:
1) "time condition" - this simply uses the EA start/end times and is based on the broker server clock, yes? Then why did I see a "no" during hour 00 as posted above in the pic on a version running that was set to start and end times of 24/5 operation?
2) what is the meaning of and how does the following "polyfit bands" line work ?
extern int PS_Mode = 1; // PolyTrend Mode: 0-off,1-trend(confirmation),2-signal
it is set to 1 by default, which is "trend"...what does that mean? does it require trend confirmation for entry? trend in which direction? and what values is it reading from? I am just trying to understand why it enters when it does..
Odysseus
1) Yes, it is broker's server time.
I am not using this EA in 24/5 way so I have no idea. But from what I see - this timefilter works well.
2) I do not remember explanation ...
may be - it is the same as FitMode in Polyfit indicator?
may be ...
As far as I see from the code of the EA - it is some virtual parameter of Polyfit indicator (PolyFitSignal) which exists only inside the code of this EA. If I am right so it may be good to see this indicator coded separatedly.
Improved version of PolyFitScalper_v1.59_noHedge EA with FXCM broker: new settings
I am trading EURUSD, GBPUSD, USDJPY, USDCHF, EURCHF, GBPJPY and GBPCHF.
It is EURCHF (in case we are trading this pair only):
GBPCHF:
GBPJPY:
For all the pairs (statement is attached):
============
- For PolyFitScalper EA all the versions, to download settings and so on: https://www.mql5.com/en/forum/179816
- improved version of PolyFitScalper_v1.59_noHedge EA with new settings is on this post.
- PolyFitScalper_v1.59_noHedge_newsfilter-1 EA is on this post: added TSD news filter to the improved Polyfit 1.59, can be turned on/off, can adjust minutes before news to stop Ea from trading, and can choose from high, medium, or low impact news to filter, along with currency pairs news we want to filter.
- PolyFitScalper_v1.59_noHedge_newsfilter-cfb norm-log EA and PolyFitScalper_v1.59_noHedge_newsfilter-cfb norm-sigmoidal EA are on this post.
- what is sigma/risk and which values to use (safe, medium and risky settings): https://www.mql5.com/en/forum/179781
- the setting/input for settings for Alpari UK, LiteForex and MB Trading (and any other broker): https://www.mql5.com/en/forum/179781
- 1.58 and 1.59 versions with the settings: https://www.mql5.com/en/forum/179781/page4
- 1.60a version of the EA is on this post.
- PolyFitScalper_v1.61_noHedge EA is on this post: this version using Daily HiLo Jurik Smooth only taking trades in direction of breakout if in agreement with Daily HiLo Jurik Smooth.
- v1.7.3 version with cfb indicator is on this post with this settings.
- old statements for 1.57 version, EURCHF with sigma = 2.8: this post.
- old statements for 1.59 EA, moderate settings: this post.
- old statements for 1.59 EA, agressive settings: this post.
- old statements for 1.60a EA, moderate settings: this post.
- visual statement is forward testing/trading results. More about it: Signal Performance Factor and visual statements: this post with the explanation and on those 2 posts as detailed comments: this post and this one.