{quote} But if you're asking me for permission wouldn't I first have to get permission from you so that I could give you permission to remove me or you or whoever we are?
Is forex education misleading or everyone is stupid?
- #1,121
- Feb 9, 2022 5:26am Feb 9, 2022 5:26am
- Joined May 2010 | Status: Straight line Fest | 3,196 Posts
- #1,123
- Feb 9, 2022 5:37am Feb 9, 2022 5:37am
{quote} Man only Gepards could work that one out, would have to be a super sleuth…
- #1,124
- Feb 9, 2022 5:43am Feb 9, 2022 5:43am
- Joined May 2010 | Status: Straight line Fest | 3,196 Posts
{quote} You know, this gepards fellow strongly reminds me of ryuryu. In fact, he came on the scene at exactly the moment that ryuryu was mocked as RickM's waterboy. I'm thinking that ryuryu switched to his alternate alias (gepards) so he could continue posting stupid comments -- this time as joyny's waterboy. It's hard to find a good waterboy these days so the two must be the same.
- #1,125
- Feb 9, 2022 5:50am Feb 9, 2022 5:50am
{quote} Now that’s interesting, might spare the time tomorrow to look into it....
Okay, moodybot, thanks for humoring me. I'm done. Have a good rest of your day, mate...
{quote} Page 56, 1110 posts, Lets get to the next level - markov chain.
I'm not going to start talking to myself about math.P(Xn | Xn-1. Xn-2, ....X2,X1) = P(Xn | Xn-1)
Someone comment on that equation then we Go! Comment anything that will suggest you know the topic.
And I'm not going to start talking in numbers and equations, I want to explain that equation to a 7 year old. I'm not here so sound smart, but to have a normal discussion with everyone.
- #1,128
- Feb 9, 2022 6:26am Feb 9, 2022 6:26am
- Joined Jan 2020 | Status: Membership revoke | 2,332 Posts
{quote} That guy is not back yet
P = probability,
X = independent variable,
n = sample space converging to limit
Know what you trade
{quote} That guy is not back yet
- #1,130
- Edited 9:13am Feb 9, 2022 8:58am | Edited 9:13am
- Joined May 2017 | Status: Trader | 377 Posts
Yes it has to do with Markov chains. But that was not the kind of reply I was hoping for.
I was hoping for comment that'll will make that equations talk...like in English so that anyone would understand.
You see mathemathician equations are just simplified statements. We simplify everything. Make it easy, it's the whole point of equations.
Now let me try and convert that equations to a statement...
I have a problem, I have random events happening, I want to find out the probability of this particular event, Xn, of happening in the future.
P(Xn...
so I come up with a condition, if these other events...
P(Xn | (other random events) )
... happened, what are the probability of Xn to happen in future?
That's : P(Xn | Xn-1. Xn-2,.....X2,X1) in simplified form.
Now....
Markov comes in and say... Listen! You don't have to calculate that other useless events...
P(Xn| (useless events))
...you only need to calculate the most recent random event.
That's: P(Xn | Xn-1)
So this is a over simplified explanation of what Markov model is. See how lovely mathematic equations are
I was hoping for comment that'll will make that equations talk...like in English so that anyone would understand.
You see mathemathician equations are just simplified statements. We simplify everything. Make it easy, it's the whole point of equations.
Now let me try and convert that equations to a statement...
I have a problem, I have random events happening, I want to find out the probability of this particular event, Xn, of happening in the future.
P(Xn...
so I come up with a condition, if these other events...
P(Xn | (other random events) )
... happened, what are the probability of Xn to happen in future?
That's : P(Xn | Xn-1. Xn-2,.....X2,X1) in simplified form.
Now....
Markov comes in and say... Listen! You don't have to calculate that other useless events...
P(Xn| (useless events))
...you only need to calculate the most recent random event.
That's: P(Xn | Xn-1)
So this is a over simplified explanation of what Markov model is. See how lovely mathematic equations are
Fun Fact
If you would buy SP500 right now... Now now.
There 99.99% probability you will be in profit in 10 years time
If you would buy EURUSD pair right now
There's 33.6% probability you will be in profit in 10 years time
If you would buy SP500 right now... Now now.
There 99.99% probability you will be in profit in 10 years time
If you would buy EURUSD pair right now
There's 33.6% probability you will be in profit in 10 years time
And Markov chains or model, is NOT some magical mathematical formula discovered by Markov that solves any random problems. No, its Just like any other breakthrough mathematics, it's something that occur naturally that was observed and documented by Markov. Take for an example coin flips. Naturally, what do you think are the probabilities of heads occuring twice in a row over tails also occuring twice in a row, when flipping a coin 1 billion times? 50 50?
here we enter deals when previous X deals (period for moving average - in this example it is 40 - so we take a look at previous 40 deals win %) win % is under 35%.
When re-run random numbers (in Excel) get this:
here if we would enter at 35% would get some losing deals anyway. But still majority would be profitable.
when re-run random numbers - again different oscillating but still we see extremes exists about 30-35%.
I implemented this approach and run now 1 account more than 7 months. seems for EURUSD this works perfect. but did with NQ100 index and that failed.. because of strong trends. EURUSD and maybe GBPUSD, EURGBP is more choppy and suitable for that "random" idea. But stocks and indexes are not so "random" - they move with stronger trends and fluctuations therefore seems not working there.
here you see some last trades for EURUSD (overall in 7 months 23% profits and 3% dd).
as you see i audit fluctuation numbers in comments and Magic means EA instance with different pattern set.
you are on the right path and believe that is working. still i wonder you have quite a big number of trades (88 in 3 monts) for your EA..seems you do not enter on extremes (they are rare) but somehow different. you mentioned ticks... you get some few last ticks and calculate probabilities... hmm... ok i was thinking more about previous win rate % like it is with coin flip. you do something different maybe.
in forex there is no 50/50. it is 49/49 becuase of spreads/slippages. but this can be adjusted in the model so no problem.
the beauty of idea i tried to explain is that you can use any patterns for probabilities calculation (therefore i freely speak about this.. anyone can build his own patterns sets - imagination is only limit, and this "strategy" wont be excausted because all would use that - because entries still would be "random" from traders. who would pay for that? banks...
or big corps who exchange currencies for their business deals. but we small traders all can benefit here - everyone do various patterns and no impact on charts - all still would be "random"). even weird patterns like "enter after 3 h1 bars up" or "enter 18 minutes after London opening" or "enter reversal when price run X pips after particular clock time" - you see infinite combinations to generate patterns because win % probabilities does not care - different extremes still exists in case if asset moves "random". some SL will be hit anyway if some "non-random" event like brexit or any other super strong trend which sets bias for market (like it is with stocks). This system still will have losing trades but overall those extremes will give positive outcome in longterm.hope my input might help. tnx for not blocking me - like your ideas
{quote} here is coinflip 50/50 chart based on winning % moving average. {image} here we enter deals when previous X deals (period for moving average - in this example it is 40 - so we take a look at previous 40 deals win %) win % is under 35%. When re-run random numbers (in Excel) get this: {image} here if we would enter at 35% would get some losing deals anyway. But still majority would be profitable. when re-run random numbers - again different oscillating but still we see extremes exists about 30-35%. {image} I implemented this approach and run...
I read this in a hurry, interesting.
Why do you think random startegy didn't work well on stocks or commodities?
Will reply when I get time
{quote} Beautiful Gepards I read this in a hurry, interesting. Why do you think random startegy didn't work well on stocks or commodities?
What is the difference between Chaos and Randomness? Randomness is - broadly speaking - just the absence of any pattern. Chaos refers specifically to a process or system that's sensitive to initial conditions. Perhaps the best example here would be the weather.
think stocks more move as "chaos" but currencies as "random".. still any can at some moment switch movement behavior but majority movements for choppy currencies pairs like GBPUSD, EURUSD, EURGBP - will be "random" most of the time. But for stocks more "chaos" with "initial conditions" and yes I saw with stocks that there do exists patterns. You can see this my FB trading results - swings. Just simple patterns...
Stock in deed can be traded with patterns and correct risk management. Currencies not. no patterns in "standard approach" can be used. But... for probabilities yes - currencies patterns in combinations with win % extremes can be used for the most "random" (choppy) pairs. I speak non scientific... maybe there exist precise definition what i mention here..
