TEB: Random Entry - Coin Toss Great with R.R.R. | Page 2 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Scripts for Limit entry, Stop entry, Market entry 47 replies

Is random really random? 231 replies

Coin-Toss Trading Strategy 152 replies

coin toss entry question 24 replies

Attachments: TEB: Random Entry - Coin Toss Great with R.R.R.
Exit Attachments

TEB: Random Entry - Coin Toss Great with R.R.R.

  • #21
  • Apr 18, 2014 5:58am Apr 18, 2014 5:58am
as long as you always do the same , in any condition always the same;
trading
- with fix spread , variable spread will change conditions on entry & exit
- always same hours of day
- always same pairs
- always using same exit strategy
- always using same RRR

Result = BE - Cost

if you can manage some of those trades then you can make profit.
My Best Trade is Next Trade!
  • #22
  • Apr 18, 2014 6:00am Apr 18, 2014 6:00am
Quoting dkrock
The Law of Large Numbers dictates that at the end of your experiment you will break even...minus the spreads
Van Tharp's backtesting of the S&P would indicate otherwise, of course that may just be luck, although it was conducted over several decades of data
  • #23
  • Apr 18, 2014 6:02am Apr 18, 2014 6:02am
Quoting dkrock
The Law of Large Numbers dictates that at the end of your experiment you will break even...minus the spreads

What laws? you mean the law of dismissing returns?
I think the laws of probability will have more impact- but lets see?
  • #24
  • Apr 18, 2014 6:04am Apr 18, 2014 6:04am
Quoting Andy B
{quote} Van Tharp's backtesting of the S&P would indicate otherwise, of course that may just be luck, although it was conducted over several decades of data

I believe it will work - I played with the idea of going for 10 baggers- But the law of dismissing returns also starts to kick hard after 4 to 1 RRR
  • #25
  • Apr 18, 2014 6:11am Apr 18, 2014 6:11am
Quoting dkrock
The Law of Large Numbers dictates that at the end of your experiment you will break even...minus the spreads
That's only valid with a true random game and 1:1 Risk:Reward
MT4/MT5 EAs/Indicators/Alerts coder
  • #26
  • Apr 18, 2014 6:15am Apr 18, 2014 6:15am
Having a higher reward:risk ratio reduces the probability of it being hit in a system where you simply use a 50:50 system to pick INITIAL direction, making it unprofitable in the long run. That's how the math works out, and that is why people don't use this system to trade. Sorry, have fun trying it by all means but it wont work.
  • #27
  • Apr 18, 2014 6:16am Apr 18, 2014 6:16am
Quoting FerruFx
{quote} That's only valid with a true random game and 1:1 Risk:Reward
Very true
1. This system would be a loser for sure!!! - Even with none random entries I would say 95% of all systems with a 1 to 1 RRR are loser
2. But I am not going for a "true random game"- rather I am going for a random entry - and a none random RRR and MM
  • #28
  • Apr 18, 2014 6:20am Apr 18, 2014 6:20am
Quoting AlanB
Having a higher reward:risk ratio reduces the probability of it being hit in a system where you simply use a 50:50 system to pick INITIAL direction, making it unprofitable in the long run. That's how the math works out, and that is why people don't use this system to trade. Sorry, have fun trying it by all means but it wont work.

I understand this but (and most times this is seen as a negative- though I do not understand why)
1. the market is not random and so the laws of math are altered.
2. Added to this you have a great variable that over comes most of the math models you speak of and that is .... trend.
  • #29
  • Apr 18, 2014 6:28am Apr 18, 2014 6:28am
Quoting TEB63
{quote} I understand this but (and most times this is seen as a negative- though I do not understand why) the market is not random and so the laws od math are altered. Added to this you have a great variable that over comes most of the math systems you speak of and that is .... trend.

By increasing the amount of pips you require the market to move in one direction, without it reversing and coming back to hit your stop loss which is less pips away than your target, you are reducing the success rate of your system. The initial 50:50 coin toss is not providing you with a 50% win rate, not even close, when your R:R is not 1:1.

Like I said, have fun with it, if you have money to spare there's all kinds of funky experiments and things you can do for fun in forex, but think about it .. if this system were even remotely profitable, don't you think people would use it considering how easy it is and how many people have come up with the idea before?
  • #30
  • Apr 18, 2014 6:28am Apr 18, 2014 6:28am
I did code a flip a coin strategy when I was debugging my simulator. The SL and TP are predefined at the beginning of the simulation.
Trades are taken at the open of a H1 bar if none is already running.

I played around with the SL and TP. Whatever SL and TP you choose the result is an edgeless system. The winrate converges to its theorical value 1/(R+1).
@TEB Law of Large Numbers => Central Limit Theorem

There are wild equity swings but the result is always in the negative because of the spread.

Here is a screenshot of one of the realizations here http://www.forexfactory.com/showthre...74#post7115274
You may get this result and wrongly conclude that "it works".

Let's look a the chart of the result. Here I used E/U SL=50 TP=200. I zoomed on a down trend. Red lines are losers, green are winners.
We can see three things:
- First the strategy goes against the trend half of the time. That's expected.

- Second thing, counter trend trades can win and in-trend can lose. I thought that if the flip a coin was half of the time in the direction of the trend, using any MA should turn it slightly positive. It fails as well.

- Third thing is that a lot a losers were in the green before the SL got hit. Trailing the SL to BE when +1R or trailing the SL don't make things better. They make the result even worse in agreement with the equations I show in Expectancy Management.
Attached Image (click to enlarge)
Click to Enlarge Name: EURUSD-H4-coin-flip.png Size: 23 KB


The management of the trade is key but I'm still searching for the best way or at least a good way to do it.
No greed. No fear. Just maths.
  • #31
  • Apr 18, 2014 6:34am Apr 18, 2014 6:34am
Quoting AlanB
{quote} By increasing the amount of pips you require the market to move in one direction, without it reversing and coming back to hit your stop loss which is less pips away than your target, you are reducing the success rate of your system. The initial 50:50 coin toss is not providing you with a 50% win rate, not even close, when your R:R is not 1:1. Like I said, have fun with it, if you have money to spare there's all kinds of funky experiments and things you can do for fun in forex, but think about it .. if this system were even remotely profitable,...

of coarse that is the point of the whole thing - but its the number of times it will move for you vs move against you
  • #32
  • Apr 18, 2014 6:36am Apr 18, 2014 6:36am
Quoting PipMeUp
I did code a flip a coin strategy when I was debugging my simulator. The SL and TP are predefined at the beginning of the simulation. Trades are taken at the open of a H1 bar if none is already running. I played around with the SL and TP. Whatever SL and TP you choose the result is an edgeless system. The winrate converges to its theorical value 1/(R+1). @TEB Law of Large Numbers => Central Limit Theorem There are wild equity swings but the result is always in the negative because of the spread. Here is a screenshot of one of the realizations here...

1. thanks for doing this!!!

2. trailing stops (esp in the forex) have been know to be loses for a long time.

3. where you place your SL I think will be a real key and I do not think you can program that - I might be wrong!

4. would you mind try doing 10 to 1 RRR

thanks again
  • #33
  • Apr 18, 2014 6:49am Apr 18, 2014 6:49am
Quoting TEB63
{quote} of coarse that is the point of the whole thing - but its the number of times it will move for you vs move against you

Not sure I understand what you mean there, but you do understand that the 50:50 chance is only valid for a 1:1 R:R random based system right? Which in trading would mean you are unprofitable simply because of the spreads ..

The odds of a higher reward:risk ratio system with a coin toss as the initial starting position are nowhere near 50:50. That is the initial direction only, beyond that the math is much more complex and your system is 100% guaranteed to be unprofitable in the long run.

There are groups of very smart people with millions of $ at their disposal for research and computer systems you can only dream of that operate in these markets so if you genuinely think simply having a good R:R will be profitable and you just can pick a random entry point then you ought to let these guys know because they put a lot of effort into making complex systems to make money.
  • #34
  • Apr 18, 2014 6:58am Apr 18, 2014 6:58am
Quoting AlanB
{quote} Not sure I understand what you mean there, but you do understand that the 50:50 chance is only valid for a 1:1 R:R random based system right? Which in trading would mean you are unprofitable simply because of the spreads .. The odds of a higher reward:risk ratio system with a coin toss as the initial starting position are nowhere near 50:50. That is the initial direction only, beyond that the math is much more complex and your system is 100% guaranteed to be unprofitable in the long run. There are groups of very smart people with millions...

1. I am not going for a 1 to1 RRR
2. 1 to 1 RRR of coarse is a loser if just because of spread and slippage
3. I agree simple is best -
4. I think all complex system lose
5. I think the big boys win for two reasons and I do not believe its their systems.
a. experience b. they have information we do not have
  • #35
  • Edited 7:26am Apr 18, 2014 7:09am | Edited 7:26am
Quoting TEB63
{quote} 1. I am not going for a 1 to1 RRR 2. 1 to 1 RRR of coarse is a loser if just because of spread and slippage 3. I agree simple is best - 4. I think all complex system lose 5. I think the big boys win for two reasons and I do not believe its their systems. a. experience b. they have information we do not have

I know you are not going for a 1:1, however the reduction in your win probability with t his system does not alter in the same ratio as with the increase of your r:r ..

For example you start with the coin toss entry of 50% and 1:1 R:R, by changing that to a 1:2 you don't then end up with a system that is 33.3% or better win rate and thus have a system with profitability (your wins are 2 times the loss, thus any win rate over 33.3% would be profitable right?)

By increasing the reward part, without imparting some sort of edge on your initial trade direction bias (ie you have stuck with the 50% coin toss) your win rate will be less than the 33.3% required for at least break even. The market is not going to move in certain directions long enough on a truly random entry for your higher reward rate to get hit more often than the risk.

Yeah, those big players have all sorts of ways to win, not all of them very fair either but if a system as simple as this were even remotely possible it would be used. It is not.

Anyway, I mistook this for trade discussion forum but it is infact trade systems so I wont keep banging away in your own thread, as I said before it is your choice and have fun but I guarantee this system is not profitable, it is not possible.


edit: I will add in this link as I was trying to find something after writing this reply http://www.forexfactory.com/showthread.php?t=146510 it's quite an interesting discussion on this very system, note one key point however is that even if the market were completely random (it is not) and thus your win %odds with different R:R could be calculated easily by using a coin toss system to trade you will inevitable end up a loser regardless because you have to take spreads into account. Throw in the fact that the market is not random, your win % will actually prove lower the higher your R:R, and this system is doomed to failure.

p.s. having just seen your replies I think what I am saying is actually flying right over your head, so good luck and have fun, but anyone else reading this check the thread I have linked if you want more actual information, it's one good place to start.
  • #36
  • Apr 18, 2014 7:13am Apr 18, 2014 7:13am
Quoting AlanB
{quote} I know you are not going for a 1:1, however the reduction in your win probability with t his system does not alter in the same ratio as with the increase of your r:r .. For example you start with the coin toss entry of 50% and 1:1 R:R, by changing that to a 1:2 you don't then end up with a system that is 33.3% or better win rate and thus have a system with profitability (your wins are 2 times the loss, thus any win rate over 33.3% would be profitable right?) By increasing the reward part, without imparting some sort of edge on your initial...

lol really !!!!! You really did not think I knew this???????????????????????? really !!!!!!!!!!!??????????????
Attached Image
  • #37
  • Apr 18, 2014 7:17am Apr 18, 2014 7:17am
Quoting AlanB
{quote} I know you are not going for a 1:1, however the reduction in your win probability with t his system does not alter in the same ratio as with the increase of your r:r .. For example you start with the coin toss entry of 50% and 1:1 R:R, by changing that to a 1:2 you don't then end up with a system that is 33.3% or better win rate and thus have a system with profitability (your wins are 2 times the loss, thus any win rate over 33.3% would be profitable right?) By increasing the reward part, without imparting some sort of edge on your initial...

I think you are missing a very important point - and that is my stop loses and targets and RRR are not random
Only the entry

and yes I do think a little something something added would increase the odds of winning - but that is not the point of what I am trying to do
And I am trying it - for fun
  • #38
  • Apr 18, 2014 7:25am Apr 18, 2014 7:25am
Quoting TEB63
{quote} I think you are missing a very important point - and that is my stop loses and targets and RRR are not random Only the entry and yes I do think a little something something added would increase the odds of winning - but that is not the point of what I am trying to do And I am trying it - for fun

So picking your s/l after entry seems straight forward enough...s/r, RN's, fibs, etc. But what is your plan on the T/P's...same thing s/r, RN's, fibs, etc.?
  • #39
  • Apr 18, 2014 7:33am Apr 18, 2014 7:33am
Quoting spyderman
{quote} So picking your s/l after entry seems straight forward enough...s/r, RN's, fibs, etc. But what is your plan on the T/P's...same thing s/r, RN's, fibs, etc.?
I am going to look for a SDZ at least 3 to 1 as compared to my stop
I will use RN in one of two ways
1. if there is no SDZ in the area I need , esp. in a break out into nothing or new highs or new lows
2. if we have a RN just before we hit the SDZ take that into account

I never use Fibs - other then the 50% which is not a fib


by the way I like swings for stops but not for targets
  • #40
  • Apr 18, 2014 7:37am Apr 18, 2014 7:37am
Quoting TEB63
{quote} I am going to look for a SDZ at least 3 to 1 as compared to my stop I will use RN in one of two ways 1. if there is no SDZ in the area I need , esp. in a break out into nothing or new highs or new lows 2. if we have a RN just before we hit the SDZ take that into account I never use Fibs - other then the 50% which is not a fib

So will you do that analysis before opening the position (and then not take it if it doesn't afford a reasonable opportunity)
Seems like that would help tip things in your favor
0 traders viewing now