Brain teaser 1: Is it possible to turn a losing system into profitable one? | Page 2 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Very Profitable System - Roi system 355 replies

MA, Gann and Ichimoku system (profitable system!) 60 replies

Can someone code this profitable system into an EA? 52 replies

Change a losing system into a Winning system 68 replies

Brain teaser 2: Who made the most? 9 replies

Attachments: Brain teaser 1: Is it possible to turn a losing system into profitable one?
Exit Attachments

Brain teaser 1: Is it possible to turn a losing system into profitable one?

  • #21
  • Dec 20, 2010 11:48pm Dec 20, 2010 11:48pm
Quoting nhct
Yes, it's definitely...
Congratulation!!! You have won the jackpot!!! The keyword to solve the puzzle is conditional probability!

Math is weird right? With a little trick an original strategy with expectancy of -2.3, 0.8, 2 (Before spread) can turn into a system that looks like new system (But it is not) with expectancy of 7.3 pips (before spread). This is not a magic and it is completely logically sounds!

Anyway I think you already explained everything in simple but detail enough. So I guess my explanation is no longer required.

btw I believe that TraderHotch managed to find the solution too, just his explanation is a bit messy and lack of the keyword that I required.
  • #22
  • Dec 21, 2010 1:49am Dec 21, 2010 1:49am
I'm a bit confused as my calculations leads me elsewhere...

If you do what nhct wrote, then out of 100 trades you'd get this:
1) 41 trades hits stoploss @ -22
2) Out of the remaining 59 trades:

 

  1. a) 15 trades will hit stoploss @ -32
  2. b) 44 trades will hit TP3 @ +28

That will give: -22*41 - 32*15 + 44*28 = -150 or -1.5 on each trade

What am I missing?

  • #23
  • Edited 2:15am Dec 21, 2010 2:00am | Edited 2:15am
I just made a Monte Carlo simulation for this scenario. Have fun!
Attached Image (click to enlarge)
Click to Enlarge Name: simulation.PNG Size: 15 KB
Attached File(s)
File Type: xls MonteCarlo.xls   857 KB | 467 downloads
  • #24
  • Dec 21, 2010 2:13am Dec 21, 2010 2:13am
Quoting NowAndLater
I'm a bit confused as my calculations leads me elsewhere...

If you do what nhct wrote, then out of 100 trades you'd get this:
1) 41 trades hits stoploss @ -22
2) Out of the remaining 59 trades:

 

  1. a) 15 trades will hit stoploss @ -32
  2. b) 44 trades will hit TP3 @ +28

That will give: -22*41 - 32*15 + 44*28 = -150 or -1.5 on each trade

What am I missing?

41 trades will never hit TP1 therefore you got no entry right? so -22*41 is not required because it never happened.

Since we have entered at TP1, therefore the distance from TP1 to TP3 is 30 - 10 = 20 but the distance from TP1 to SL = 10 + 20 = 30

So after considering spread, Net P/L = -32*15+18*44 = 312, expectancy = 312 / 59 = 5.3pips. Why divided by 100 since you got only 59 trades?

Got it?
  • #25
  • Dec 21, 2010 2:22am Dec 21, 2010 2:22am
Of course, I didn't understand that we only enter at TP1. Tnx!
  • #26
  • Dec 21, 2010 3:00am Dec 21, 2010 3:00am
Is it possible that we did something unnoticed by ourself where we actually reduce the profitability of our system, or even turn it into losses??

Plausible.
Attached Image (click to enlarge)
Click to Enlarge Name: simulation.PNG Size: 19 KB
  • #27
  • Dec 23, 2010 12:56am Dec 23, 2010 12:56am
1. is there a question in your last post? i'm actually not sure, or if you are trying to point something out or hint at something?

2. fquant and nhct and so on... did y'all go to college and learn math there? or to put it bluntly, how the hell do you guys know this shit and furthermore, understand it?

i grasp the foggy concept of it, but just can't focus it into a solid thing that i can grasp with both hands. :/
Congratulations ScalaFX for coming out of the closet!
  • #28
  • Dec 23, 2010 2:42am Dec 23, 2010 2:42am
Quoting nubcake
1. is there a question in your last post? i'm actually not sure, or if you are trying to point something out or hint at something?

2. fquant and nhct and so on... did y'all go to college and learn math there? or to put it bluntly, how the hell do you guys know this shit and furthermore, understand it?

i grasp the foggy concept of it, but just can't focus it into a solid thing that i can grasp with both hands. :/
1. Not really a question. It is just another scenario with the same concept.

2. I dont know the others but I learnt it during my young age at secondary school. Conditional probability is not something that difficult to understand, the hardest part is how you applied it.

It may not known to many traders but you are actually make a trading decision everyday without knowing what is conditional probability.
  • #29
  • Dec 23, 2010 2:45am Dec 23, 2010 2:45am
Quoting CindyXXXX
Funny thread I think people have a hard enough time turning a profitable system into a profitable one. So I can only guess whoever is interested in this garbage is just rich and bored of how easy the forex is.

Its just hindsight curve fitting BTW. You think these percentages will repeat exactly?

and yes this post will be deleted
Many people blame that quants are selfish and not sharing. But when they share some concept that really outside their box, they talk shit. I guess that is one of the reason why quants seldom share.

Why should I need these percentages to repeat exactly to make this concept works? Oh yes maybe your box is too big until you have a difficulty to reach the wall.
  • #30
  • Dec 23, 2010 3:14am Dec 23, 2010 3:14am
Quoting ForexQuant
1. Not really a question. It is just another scenario with the same concept.

2. I dont know the others but I learnt it during my young age at secondary school. Conditional probability is not something that difficult to understand, the hardest part is how you applied it.

It may not known to many traders but you are actually make a trading decision everyday without knowing what is conditional probability.
to my understanding this is a spin on monty hall... which i don't recall encountering until watching "21". i did the 'hard' maths at school and really don't recall anything like this.
Congratulations ScalaFX for coming out of the closet!
  • #31
  • Edited 5:45am Dec 23, 2010 5:20am | Edited 5:45am
Quoting ForexQuant
Congratulation!!! You have won the jackpot!!! The keyword to solve the puzzle is conditional probability!

Math is weird right? With a little trick an original strategy with expectancy of -2.3, 0.8, 2 (Before spread) can turn into a system that looks like new system (But it is not) with expectancy of 7.3 pips (before spread). This is not a magic and it is completely logically sounds!

Anyway I think you already explained everything in simple but detail enough. So I guess my explanation is no longer required.

btw I believe that TraderHotch...
Thanks for the interesting puzzle; and thanks to nhct for his answer. I actually was wondering if the answer was conditional probability, but I foolishly rejected it, because the only way I could make it work was to have the stop loss be not in one of the sets, the same way that nhct and you did. I am still not clear why you were able to make a larger stop loss and still remain in the rules in your post one, but that point is trivial, because the larger picture is what is important that one can watch the TP1 trade and wait for a hit of the TP1 and then take the TP2 trade from that point, regardless of the definition of the stop loss.

Thanks again for a stimulating thread. Gave me some great ideas.
  • #32
  • Dec 23, 2010 5:25am Dec 23, 2010 5:25am
Quoting ForexQuant
Many people blame that quants are selfish and not sharing. But when they share some concept that really outside their box, they talk shit. I guess that is one of the reason why quants seldom share.
OH and BTW Quants don't have feelings. Do they? LOL; just a little joke.
  • #33
  • Dec 23, 2010 7:00am Dec 23, 2010 7:00am
Quoting smjones
Thanks...
You are welcome. I hope you and anyone who interested apply this idea on your own system. If so I believe we will start another round of new discussion because you will find another puzzle.
  • #34
  • Dec 23, 2010 7:02am Dec 23, 2010 7:02am
Quoting smjones
OH and BTW Quants don't have feelings. Do they? LOL; just a little joke.
Quants do have feelings, Algorithm doesnt.
  • #35
  • Dec 23, 2010 12:33pm Dec 23, 2010 12:33pm
Awesome thread and puzzle, FQ! Thanks for the fun reminder of an important concept I hadn't applied to trading.
  • #36
  • Dec 25, 2010 6:31am Dec 25, 2010 6:31am
Quoting ForexQuant
Congratulation!!! You have won the jackpot!!! The keyword to solve the puzzle is conditional probability!

Math is weird right? With a little trick an original strategy with expectancy of -2.3, 0.8, 2 (Before spread) can turn into a system that looks like new system (But it is not) with expectancy of 7.3 pips (before spread). This is not a magic and it is completely logically sounds!

Anyway I think you already explained everything in simple but detail enough. So I guess my explanation is no longer required.

btw I believe that TraderHotch...
Thank ForexQuant for this tips !! but i don't understand how translate this conditional probalility in mql4 code .. it's possible ?

Regards and Merry Christmas
  • #37
  • Dec 25, 2010 6:43am Dec 25, 2010 6:43am
Quoting surfeur
Thank ForexQuant for this tips !! but i don't understand how translate this conditional probalility in mql4 code .. it's possible ?

Regards and Merry Christmas
Did you mean that you want to apply calculation directly into trading and let your EA decides when and where to place your entry?
  • #38
  • Dec 25, 2010 7:53am Dec 25, 2010 7:53am
Quoting ForexQuant
Did you mean that you want to apply calculation directly into trading and let your EA decides when and where to place your entry?
Yes it's the idea..
  • #39
  • Dec 25, 2010 9:39am Dec 25, 2010 9:39am
I dont think so. You are the one who suppose to decide the 'condition'. This is the hardest part because I gave you the condition in this puzzle but in real life you got to figure it by yourself.
  • #40
  • Dec 25, 2010 11:58am Dec 25, 2010 11:58am
Quoting nhct
Yes, it's definitely...
Great demonstration !
I and other people here do that but we call it "Order Management", "Money Management", "add to your winners", among other things. Condtionnal Probability sounds way more nerdy It looks like maths could discover many little tricks used by profitable traders. I don't like talking about them because I wasn't smart enough to discover them other than by trial & error but it's really nice to see them nicely formulated and in plain sight


Thanks for your input to the forum ForexQuant
0 traders viewing now