The Bat...a strategy by Team Aphid | Page 25 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

My Bat from team aphid trades 82 replies

The BAT Method - The EA 186 replies

BAT system for 15 min chart 28 replies

BAT indicator 2 replies

The BAT mathematical formula for lines? 0 replies

Attachments: The Bat...a strategy by Team Aphid
Exit Attachments

The Bat...a strategy by Team Aphid

  • #481
  • Feb 20, 2007 7:55am Feb 20, 2007 7:55am
GBP/USD backtest Jan - July 2006 IBFX with %risk
Attached File(s)
File Type: zip BAT-GBPUSD TEST- 2006.zip   126 KB | 702 downloads
  • #482
  • Feb 20, 2007 8:46am Feb 20, 2007 8:46am
Quoting TeamAphid
What you are doing is actually something very similar to change Factor parameter in original BAT indicator.

Please have a look at GBPJPY chart. Change distance parameter to 1.0 and keep ATR Period @ 5 and Factor @ 4.0

Then, in original BAT indicator, keep ATR period @ 5, but change Factor to 4.6. You'll see what we mean.

The result will be less noise, bigger stoplosses.

Thanks for your contributions.
I must say all credits goes to TeamAphid, this method really change how mechanic we can be in order to balance discretionary decisions.

I didn't explore the settings yet, I am sure it is similar, but the concept of adding instead of multiplying is simpler, I guess. Perhaps we have "updated curve fitting" value for distance, instead of ATRPeriod and Factor.

Only backtest could prove it, no over optimization is what I am hoping too.

I take a sample chart for better visualization in our head.

Attached Image
  • #483
  • Feb 20, 2007 9:13am Feb 20, 2007 9:13am
[quote=Anderwill;222459]
Quoting mason
I took a little quick visual backtest and fix a little bug.

On the schedule 2 indicators are resulted: BAT ATR v1 and BAT ATR Distance (last version which the author has corrected from a mistake)
BUT! The mistake of new indicator BAT ATR Distance Is observed
EUR/USD concerns the remote indicator (11 pips) with a thorn and departs back, BUT! The indicator reacts (instead of ignores a thorn) and starts to draw from below a dark blue line. The arrow of white color specifies the new corrected indicator.
The previous indicator (first version BAT ATR Distance) was more correct. It ignored a contact of thorns and reacted only to the price of the closed position bar's. I do not understand for what you have altered it. The new indicator badly works, return please old version BAT ATR Distance
I am sorry, I still believe what I did for last version is right, I believe the first version is double-add, and I have replaced it, if you can take a cleaner chart with no-resize resolution, may be I can figure it out what exactly you want.

Thanks for mentioning, we need every possible efforts to make this indicator better.
  • #484
  • Edited 4:06pm Feb 20, 2007 2:12pm | Edited 4:06pm
Hello,

I tried to improve the system by reducing it to 1 lot only, without using fibonacci retracements to add to the initial position. This way, you only look for big moves in the direction of your entry.

Then I searched for those moves (mostly 200 pips or more maximum favorable excursion [MFE] since entry without touching Bat ATR in the meantime) and observed the maximum adverse excursion (MAE) since entry. I did this manually for both 2005 and 2006. Next I calculated how large an initial stop was needed to keep you in 3/4 of those trades or at least in 1/2 of those trades (quartile 3 and 2).

If you use a smaller intitial stop, whipsaws hurt you much less, and you can increase the position size while keeping account risk the same as before.

Maybe someone could do an automatic backtest wether this helps on overall profitability.
Attached File(s)
File Type: zip Bat GBPUSD MAE.zip   2 KB | 569 downloads
  • #485
  • Feb 20, 2007 5:12pm Feb 20, 2007 5:12pm
Quoting Bounty_Two
Hello,

I tried to improve the system by reducing it to 1 lot only, without using fibonacci retracements to add to the initial position. This way, you only look for big moves in the direction of your entry.

Then I searched for those moves (mostly 200 pips or more maximum favorable excursion [MFE] since entry without touching Bat ATR in the meantime) and observed the maximum adverse excursion (MAE) since entry. I did this manually for both 2005 and 2006. Next I calculated how large an initial stop was needed to keep you in 3/4 of those trades or at least in 1/2 of those trades (quartile 3 and 2).

If you use a smaller intitial stop, whipsaws hurt you much less, and you can increase the position size while keeping account risk the same as before.

Maybe someone could do an automatic backtest wether this helps on overall profitability.
Bounty,

Thanks for result, questions :
1. what is the pair ?
2. Are those trades win only ?
3. is MAE a BAT line to BAT line pips ?
4. is MFE a BAT line to maximum pips ?
5. is Quartile3 & 4 a minimum SL ? do you use fibo swing point as SL ?
  • #486
  • Feb 20, 2007 5:29pm Feb 20, 2007 5:29pm
You have explained yourself clearly, no problem there. We were considering next entry strategy when saying ATR factor 4 is the best, sorry not being clear here.

If you use, say ATR factor 6 to give the trade more room and exit trade #1, what will be your entry strategy for next trade as BAT always keeps you in position? If you use ATR factor 6 again for next trade entry (say it is not a counter trend entry), it may be late as entry signal will already be triggered by ATR factor 4.

Please see post # 263 for GBPJPY spreadsheet.

Sorry for the confusion.



Quoting charlies
I might not have explained myself clearly. I agree that ATR factor 4 is best overall. However, I was proposing to change it only after a threshold is reached (say 200 pips gain). This way, you keep the initial risk the same, but give the market more room when trending (see chart below).

In a similar way, you would only avoid the contra trend trade after a large (say 200 pip) gain. As a result, the proposed change would have taken the short signal on 3 Jan. I only looked at this on GBPUSD, so don't know if the same rule / threshold would improve the results on GBPJPY. If you have a spreadsheet on GBPJPY, I could check this fairly quickly.
  • #487
  • Edited 6:23am Feb 21, 2007 12:43am | Edited 6:23am
The strange schedule. (Look attach).
Yesterday and today I look as on GBP/USDGBP in the beginning does a thorn upwards punching a level "1" and coming back. Then at night again attempt to punch "1", the first bar punches level Enter 1 lot, the second bar probably rests against resistance, the third bar returns again back. Indicator BAT ATR during unsuccessful attempt of breakdown of a level "1" (it Enter 1 lot) draws a dark blue line as a trend, at that that the red line is not punched. Look the attached chart. I have paid attention that at attempt to pass a level "1", the price has rested against key level Fibo 1,236. On a scale I have designated it conditionally Level for Start. This level, also has shown 1,236 me that the price could not go above up to 1,618.
Here such supervision.

ПИЗДЕЦ КОРОЧЕ! недотянул график до контрольной 1,618
Attached Image
  • #488
  • Feb 21, 2007 12:55am Feb 21, 2007 12:55am
Quoting mason
Bounty,

Thanks for result, questions :
1. what is the pair ?
2. Are those trades win only ?
3. is MAE a BAT line to BAT line pips ?
4. is MFE a BAT line to maximum pips ?
5. is Quartile3 & 4 a minimum SL ? do you use fibo swing point as SL ?
Hello Mason,
1. The pair is GBP/USD
2. Yes, losing trades have been skipped in the spreadsheet. My intention was to optimize the initial stop loss width by determining how small it can be set without stopping you out of more then 1/4 or 1/2 of the relevant winning trades. Since there's no harm to be stopped out of a losing trade, I disregarded losing trades in my calculation.
3. MAE is a statistical number for each trade. It tells you how much the trade went against you after your entry.
4. MFE is a statistical number for each trade. It tells you how much the trade went in your favor after your entry.
5. I didn't use any fibo swing points at all. Quartile 3 for a group of numbers tells you at which number 3/4 of the group are smaller than the resulting number. For me, all favorable trades' MAEs were the group of numbers, and Quartile 3 of that group tells me how small I can set the stop without being stopped out more than 1 out of 4 times for those favorable trades.
  • #489
  • Feb 21, 2007 4:41am Feb 21, 2007 4:41am
Quoting Bounty_Two
Hello Mason,
1. The pair is GBP/USD
2. Yes, losing trades have been skipped in the spreadsheet. My intention was to optimize the initial stop loss width by determining how small it can be set without stopping you out of more then 1/4 or 1/2 of the relevant winning trades. Since there's no harm to be stopped out of a losing trade, I disregarded losing trades in my calculation.
3. MAE is a statistical number for each trade. It tells you how much the trade went against you after your entry.
4. MFE is a statistical number for each trade. It tells you how much the trade went in your favor after your entry.
5. I didn't use any fibo swing points at all. Quartile 3 for a group of numbers tells you at which number 3/4 of the group are smaller than the resulting number. For me, all favorable trades' MAEs were the group of numbers, and Quartile 3 of that group tells me how small I can set the stop without being stopped out more than 1 out of 4 times for those favorable trades.
Thanks for explanation, any conclusion on how we should put SL and TP, such as if swing is less than 57 pips from BAT line, we stick to 57 pips.

Or our RRR should be around 1 : 10, because average of MAE : MFE is 32.75 and 324.
  • #490
  • Feb 21, 2007 5:07am Feb 21, 2007 5:07am
Quoting mason
Thanks for explanation, any conclusion on how we should put SL and TP, such as if swing is less than 57 pips from BAT line, we stick to 57 pips.

Or our RRR should be around 1 : 10, because average of MAE : MFE is 32.75 and 324.
Hello Mason,

I ought to add that the MFE numbers might be misleading, they only tell you in hindsight how much profit a trade might have generated. Caculating the average number for MAE values can be dangerous since it doesn't necessarily tell you how often the average will be hit: Average numbers can be skewed by outliers, so I better like to work with quartiles (there's an excel function for that, as you can see from my spreadsheet).

You might want to calculate the quartile 1 from all MFEs of the big winning trades. The resulting number tells you which target would have been hit by 3 out of 4 big moves.

However, I only tried to find a reasonable way to minimize the initial stop, while I didn't intend to optimize take profit exits. You have to keep in mind that my spreadsheet omits all the losers, break-even trades and minor winners. I don't know if you could achieve a hit ratio better than 10% for all trades, which you would need to be profitable with 1:10 RRR. That would be an interesting study though. This way, you could use the bat ATR line for entry only, and attach bracket orders (stop loss, take profit) based on your statistical observations for the exit without having to monitor the trade.

For my own forward test I decided to use an initial stop of 40 pips. That should keep me in more than 1/2 of all major moves.
  • #491
  • Feb 21, 2007 5:38am Feb 21, 2007 5:38am
Quoting TeamAphid
We were considering next entry strategy when saying ATR factor 4 is the best.
I have taken a look at GBPJPY. Both suggestions (higher ATR & skip trades after big win) produce significantly worse results for GBPJPY. I think I was guilty of curve fitting for GBPUSD only.....

I think it is going to be difficult to improve the exit for trade 1 - any change will either limit the gains, or impact on the next trade.
  • #492
  • Feb 22, 2007 8:29am Feb 22, 2007 8:29am
I have seen that we need to take partial profit when Pin Bar break against BAT direction.

Let's find the most reliable setting to detect Pin Bar, you can apply this indicator more than one.

Credit to Jim from PF, he is really an eye opener to me.

Attached Image
Attached File(s)
File Type: mq4 PB detection.mq4   3 KB | 670 downloads
  • #493
  • Feb 22, 2007 1:50pm Feb 22, 2007 1:50pm
somebody using price action to better the BAT? brilliant and something ive been patiently watching for. ive been demiong BAT with price analysis and it helps. but then it always does. thanks for the kind words.

jim

Quoting mason
I have seen that we need to take partial profit when Pin Bar break against BAT direction.

Let's find the most reliable setting to detect Pin Bar, you can apply this indicator more than one.

Credit to Jim from PF, he is really an eye opener to me.

Attachment
  • #494
  • Feb 23, 2007 1:46am Feb 23, 2007 1:46am
Hi James,

No problem with using price action as an exit. We just need to create a mathematical rule or something to make sure that we treat every scenario exactly the same. If we don’t do that then we are not treating this as a mechanical trading system. We’ll try to describe the pin bar in math terms. Some of the “magic” might get lost because we will not be able to bring in pivot areas, trend lines and the other tools that the “chart gazers” use but let’s see how it works out.

Thanks


Quoting james16
somebody using price action to better the BAT? brilliant and something ive been patiently watching for. ive been demiong BAT with price analysis and it helps. but then it always does. thanks for the kind words.

jim
  • #495
  • Feb 23, 2007 10:58am Feb 23, 2007 10:58am
yeah i dont want to muddy up the waters. for team aphids sake and the mechanical nature of the system i dont want my stuff becoming a topic of discussion here. your description is why. when you combine my stuff it is no longer totally mechanical at least as far as entries. everyone who is following this mechanical system should continue to do just that and not start altering it.

for those that understand my stuff its certainly possible to refine a few areas but once again when you do that its not the BAT in its true form anymore. as i said all im personally doing is demoing the system and using price action to help refine or determine entries and exits. another words im just playing around and researching. since i know team aphid and have had access to the complete system for some time i know how good it is. good enough to take it thru my normal testing and refining stages which can sometimes take up to a year before i risk real money.

i have enjoyed it a great deal and i value and enjoy my relationship with the team members. some of which i have known now for a couple of years since the very early months of forex factories creation. where does the time go?
Quoting TeamAphid
Hi James,

No problem with using price action as an exit. We just need to create a mathematical rule or something to make sure that we treat every scenario exactly the same. If we don’t do that then we are not treating this as a mechanical trading system. We’ll try to describe the pin bar in math terms. Some of the “magic” might get lost because we will not be able to bring in pivot areas, trend lines and the other tools that the “chart gazers” use but let’s see how it works out.

Thanks
  • #496
  • Feb 23, 2007 4:58pm Feb 23, 2007 4:58pm
Quoting james16
...demoing the system and using price action to help refine or determine entries and exits.
I can't believe the master is here, it is an honour.

It comes to me when I saw many charts, especially cleaner ones, I saw another tool such as resistance, support and flip of res/supp line could become early warning/confirmation. (yes, you know where it comes from)

I also believe EA can help this, with more focus, enhancement and correct setting, we can have more robust strategy in an already very strong method.

Off course we have to monitor always the chart and EA, but we could put an early warning EA before the EA really takes an action.

I don't want to talk about EA in this thread, but since our goal is to be mechanical, it is not a bad idea to have an assistance such as EA, but not to be confused as a reason if the method is profitable or not. EA is just as dumb as rock. (not to mean harsh)
  • #497
  • Feb 23, 2007 7:06pm Feb 23, 2007 7:06pm
your very kind but the only thing im the master of is two kids, two dogs and two cats. LOL.

Quoting mason
I can't believe the master is here, it is an honour.

It comes to me when I saw many charts, especially cleaner ones, I saw another tool such as resistance, support and flip of res/supp line could become early warning/confirmation. (yes, you know where it comes from)

I also believe EA can help this, with more focus, enhancement and correct setting, we can have more robust strategy in an already very strong method.

Off course we have to monitor always the chart and EA, but we could put an early warning EA before the EA really takes an action.

I don't want to talk about EA in this thread, but since our goal is to be mechanical, it is not a bad idea to have an assistance such as EA, but not to be confused as a reason if the method is profitable or not. EA is just as dumb as rock. (not to mean harsh)
  • #498
  • Feb 24, 2007 3:58am Feb 24, 2007 3:58am
Here is an unfinished indicator to detect price ranging, I think this could be useful if price break the range against BAT signal.

I hope someone could finish it to be better, in the mean time, we can find reliable setting as usual.

I am thinking of drawing rectangle object and placing price label at resistance and support, instead of as an oscillator.

Duty calls, whoever interested.

Rules : if more than 4 bars in histogram, price did range, we find highest high and lowest low and draw rectangle.

This indicator is also based on MA, it has been a very powerful in most of the time.

Attached Image
Attached File(s)
File Type: mq4 MAwithHiLo.mq4   3 KB | 587 downloads
  • #499
  • Feb 25, 2007 10:53am Feb 25, 2007 10:53am
Hi all

Thank you for your great system. I trade almost a year and I find your system very comfortable .

I thought abaout whipsaw filter : all the rules as you wrote except entry must be the first candle that his open price is above/under the bat line. what do you think about it?

Simon
  • #500
  • Feb 25, 2007 5:09pm Feb 25, 2007 5:09pm
You can filter out some of losing trades by using your entry filter, but, on the other hand, you will sacrifice some pips due to late entry. Therefore, we recommend you to do a proper backtest to make a comparison.


Quoting simonr
Hi all

Thank you for your great system. I trade almost a year and I find your system very comfortable .

I thought abaout whipsaw filter : all the rules as you wrote except entry must be the first candle that his open price is above/under the bat line. what do you think about it?

Simon
1 trader viewing now