Pure Risk to reward trade System - 1 to 3- Totally random entries

So happy to see at least one person understood. I don't think I could have made the math any clearer. Somehow I got into an argument about beliefs while discussing math..

no, you are at BE if you hit that 3 one quarter of the time - 25% wins/75% losses.

3L + 1W = BE

You're right...of course...thanks for the correction.
However, I still think you have missed the point, but thanks for your contribution to the thread none-the-less.

dont think I missed any point as I only wanted to point out your error for the sake of other readers.

Trade 11 Sun, July 29, 8:33 Am UsdCad Buy 1.0046
dice rolls where 3 and 5



( I had to use the UsdCad not the GpbUsd as it would have killed on other trade)

So I was kinda bored this weekend and whipped up and EA for this

I simulated trading this method for the month of july, and the results are posted in the CSV file. I had to generate the backtests seperately for the different currencies because strategy tester doesn't support multi currency backtests.

The EA does everything the way TEB does except use a different currency to support a trade in the opposite direction. Since my broker does indeed support hedging trades, and I'm not looking to trade this method live, I can't really be bothered to double the lines of code in the EA just to support this. Besides, this is more in the spirit of the experiment anyway. FYI, I used a gomarkets demo.

If u wanna play around and generate results yourself, here are the inputs:

extern int hourentry1 = 10;
extern int hourentry2 = 22;

*enter the server hour you wish you enter the trades. I estimated the time teb uses by looking at his posts.



extern int TP = 1580;
extern int SL = 500;

*default TP and SL used in this experiment by teb



extern int magicnumber = 10825342;

*just make sure it's a unique number



extern int decimal = 5;

*broker decimal


extern bool managerisk = FALSE;
extern int riskmultiplier = 1;
extern double lots = 0.1;

*risk in percentage. uses what's listed in lots unless managerisk is set to true



extern int random_seed_supplement = 0;

*dice roll will be exactly same no matter how many times u backtest. change this number everytime you try it again to get different dice rolls.


If you look at the csv file, you'll see that while the number of data points isn't statistically significant, the curve is pretty much what you would expect of a 0 EV model.

What you mean?
The paramenters were set by you at 158 pips for TP and 50pips for SL. The EA takes in inputs as fractional pips.

The accuracy of mt4 backtesting increases as your TP and SL become bigger. This "strategy" is actually pretty ideal for backtesting due to it's big TP and SL.

I don't know why you would lie about the results when you obviously read into detail the trades in my csv file, but the net balance was close to BE, not +12800 pips like you claim.

hey guys, such a niceeee experiment. In theory you should BE using random entries but using settings of RR 1:3 then with a 50% of success rate then it should be a winner scenario, now the problem maybe the first RANDOM variable (Selecting random pair) Now what i did was just selecting the EURUSD profit/loss and now we see some profit. (its like 445 am here so if i did something stupid in the xls. sorry) just my idea. Now the next step should be testing random entries on 1 pair using 3:1 rr and see what happens

Not sure what you are saying here but you are of course right that a system with R/R 1:3 and a 50% hitrate will be a winner. That is what you are saying right?
Anyway that have nothing to do with this experiment because it is not 50% hitrate here. This experiment will probably end up around break even minus the spread if you give it a large enough number of trades.

Arctic4x

Yeah, lets see ill be testing this later now im cooking my breaKfast then i g2g to class "investment theory" soo later guys

Trade 12 Mon, July 30, 7:58Am Audusd sell 1.0495
dice rolls where 3 and 5

so to summarise, you looked at my csv file in detail, decided to spice up the results a bit and declare the method made +12800 pips, even though anyone could easily download the file for himself and see your words for the steaming pile of BS that they were.

I caught you doing that ridiculous claim and called you an outright liar.

You responded by changing your post from one that declares "mission accomplished, epic win" to "i guess we're losing money with this, but it's summer!"

And THEN you responded to me with that unreadable garbage quoted above, telling me that "lie" is too strong a word to be used here, quickly following it up with ANOTHER lie, that you changed your post before I made mine.

The time stamp is right there for everyone to see! Why would you continue to make lies that are so easily found out?

Alright had to register to comment on this atrocity to basic game theory. Just for a little background I'm brand new to forex and came over from online poker after the US DoJ got they're panties in a twist over online gambling. One of the things that really interested me in forex is how similar it is to zero sum games such as poker. Anyways on to the thread.

Let's establish some common ground first:

1. Forex trading allows you to manipulate your own risk reward ratio through the placement of Tp's and Sl's. This everyone can agree on (I hope). So we can set up a 1:1, 100:1 or a 1:100 trade.

2. The Forex market in itself is not entirely random. The degree of randomness is up for debate but just the fact the market can jump 2% when some European dude in a suit gives a speech is proof that it's not a 100% true random system.

3. EV (expected value) is a term from probability theory and is the single most important thing with regards to trying to make money off of any system. (I think they're is some confusion on this one so I'll elaborate). EV = (chance of event occurring)(Payout)+(chance of event not occurring)(Loss). Loss would be a negative off course. So if a have a 150pip:50pip tp to sl trade with a 50/50 chance of hitting tp or sl. Then our ev would be, ev= (.5)(150)+(.5)(-50) which is ev=75-25 or ev=50. Meaning we would expect on average to 50 pips. (Hopefully we now all understand the most basic concept in gambling or investing. if not my buddy google knows this dude named wikipedia etc etc I'm sure you get the gist of the joke)

Now onto the retarded assumptions being made (I'm not gonna direct quote cause this is already taking too long).

1. "The market isn't random so assuming that it has a 50/50 chance of going up or down is wrong." - True but because the choice of buy or sell (in this system) is random the chance that you will be going with the trend or against it is exactly 50/50.

2. "Well I'm going to make money because if it's a 50/50 chance with a 3:1 RRR so my ev is +50pips" - This assumes that the market is a 50/50 to range to 150 in your direction before ranging 50 in the opposite direction. Which is false it's only 50/50 that it will be moving with or against you when the order is placed. This should be obviously for most, but to amplify the point if you changed your system to a rrr of 1000:1 people would call you an idiot, why? Because the market will most likely never range up 1000pips without dropping at least 1 pip below your EP. The fact that it's 3:1 just means it's more likely to occur then the 1000:1 example but in no ways a 50/50. Determining the actual chance would require a lot of back testing etc etc, and would change depending on pair and currency etc etc, but luckily the next point makes this all unnecessary.

3. "You can never be anything but BE minus spread." This isn't a quote of a fallacy stated in the previous pages but rather a fact, and here is why. In your system/experiment your playing both sides of the coin. Because the sum of all probabilities must always equal 1 (refer to 5th grade math) and because in this system there are only 2 outcomes (+150pips or -50pips) the probably of 1 is the inverse of the other ie if you have a 75% chance to hit -50 before hitting 150 you have a 25% chance of hitting 150 before hitting -50. Add in the fact that we are using RNG (random number generator, in this case a die) to determine which one of these two outcomes occur. Remember the market may not be random so it can go +150 or -150 due to external factors however your choice of direction is random therefore making the system random.

Still with me? let's make up some numbers to hammer this home. I'll even throw you a bone and give you favorable odds (at least according to the way you claim the system works). Let's say you have a 40% chance of hitting your Tp of 150pips before hitting your SL of -50pips. If we throw all this along with your random direction pick (ie 50/50) into an eq equation we get:
ev= (.4)(150)+(.6)(-50)=30pips, great we made money, except we left one major variable out... the buy sell choice which is a 50/50. If we add that in we get ev= ((.4)(150)+(.6)(-50))+(-1(.4)(150)+(.6)(-50))=0 . So the first half of the equation is when you rolled a 1-3 and went long, and the second half is when you rolled a 4-6 and went short the second half is multiplied by negative 1 because your are doing the exact opposite of going long and therefore would get the exact opposite results. Subtract spread from the result and you get your expect value of pips per trade.

I realize the last couple jumps in logic here may be a little difficult for some and maybe I could have explained them better but regardless they are correct.

Btw lol at the guy saying the die roll wasnt random cause the guy rolling it had knowledge of the market.

[quote=Donko;5887886]Alright had to register to comment on this atrocity to basic game theory. Just for a little background I'm brand new to forex and came over from online poker after the US DoJ got they're panties in a twist over online gambling. One of the things that really interested me in forex is how similar it is to zero sum games such as poker. Anyways on to the thread.


Its your first post on FF and your new to Forex..... your going to make a success of forex, no doubt of that. Excellent post.

I like this point and have never thought of it this way.
Let me see if I can use an analogy that would be correct.

You're standing on top of a building. If you jump off you have a 100% chance of death...but if you stay put you have a zero % chance. So before making your decision you have a 50/50 or break even chance of surviving.

Nice one Donko. Hopefully that will put this to bed.

ya i guess if you were on a ledge flipping a coin on whether or not to jump your ev of being alive would be .5 every time you flipped. Or while the coin was in mid air flipping you would be half dead half alive at the same time....not exactly the same b/c life and death are a finite options as far as i know (Schrodinger's cat would prob disagree though if anyone's familiar).

ok you are playing nice

so let me say again great math.
but
for a lot of reasons this does not work the reasons are many a few of which are
1. The game is rigged!!!!
2. you have a trend which you do not have in games
3. you have outside powers in FX that you do not see in cards

so this changes all this, and your math is not accounting for these

cheers

But aren't you negating the force of all those by the fact that from the beginning you are randomizing your entry.

It may be true that you have a trend, but if you enter on a coin flip your odds are still 50/50 of entering the right way. Same with the other reasons I would think.

Let's say we have a pair that always goes one way 158 pips then reverses and goes the other way 50 pips. Let's say further that we know exactly when those swings start and finish. But if we flip a coin we're still going to lose half the time.

half the time is 50Pips Loss/ half the time is 150pips profit.

To my understanding: TEB63 is randomizing the entry in a "none random market".

If approx. 200 pips are the average of a move in one direction than he could be profitable.

Not so sure how it works out in action: actually in such case one might try to open a trade long and one short of the same pair and just let them run