automatic chart pattern drawers for MT4? | Page 2 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

MT4 Enhanced chart cross-hairs with automatic chart-sync 83 replies

Automatic Notification of pattern detection 2 replies

MQL4 manual and automatic chart update on recompilation 4 replies

MT4 Automatic Trendlines Indicator 1 reply

MT4 Indicators-Automatic Trendlines 0 replies

Attachments: automatic chart pattern drawers for MT4?
Exit Attachments

automatic chart pattern drawers for MT4?

  • #21
  • Dec 20, 2012 10:45pm Dec 20, 2012 10:45pm
I re-read your post and I think a bit of beer and poor reading comprehension on my part is what you are reading


yes I agree with you 100%, you cannot have a random chart with out subjectively imposed chart patterns......so that DOES say something about the validity of the chart patterns on random data.


HAVING SAID that is not so much the chart patterns but the momentum loss in the impulse waves that might be the important thing because returns are not random, and there is something logical about price movement that aligns with systemic shocks and macro economic events......RANDOMLIKE but the daily log returns show that it is not entirely a random normal distribution.

perhaps we can get to something constructive with this thread for a change......you, yourself were able to notice the real chart...the question is HOW did you distinguish randomlike from random?
AVT INVENIAM VIAM AVT FACIAM
  • #22
  • Dec 20, 2012 11:07pm Dec 20, 2012 11:07pm
Quoting the redlion
I re-read your post and I think a bit of beer and poor reading comprehension on my part is what you are reading


yes I agree with you 100%, you cannot have a random chart with out subjectively imposed chart patterns......so that DOES say something about the validity of the chart patterns on random data.


HAVING SAID that is not so much the chart patterns but the momentum loss in the impulse waves that might be the important thing because returns are not random, and there is something logical about price movement that aligns with systemic shocks...
i'm not 'concerned' about your tone, it was more that i was 'confused' about what you were trying to say because it read like you agreed but sounded like you didn't. but you have cleared that up.

i didn't, and don't, expect this to be a big dialogue or whatever. i was only intending to toss out that first post and be done with it, and i'm merely procrastinating at the moment with these further posts.

what i was getting-at is that people attribute chart patterns as evidence that charts are not random. "look, a support line keeps being respected", "chart patterns are self-reinforcing" or the other version of this which is "history repeats itself", etc. after this someone like me will chime-in and retort with "randomly made charts will also show signs of the chart patterns you attribute to some expected 'order' in the charts, so your legitimate chart shows exactly the same patterns as a random chart".

but, is either really the point? does it really matter if a legit chart has patterns in it, or a random chart also has the same patterns in it, if in the meantime nobody can actually create a made-up chart that also DOESN'T show these very same patterns? i'm saying that it's a weak argument for people to claim that chart patterns are man-made, or chart patterns are self-reinforcing, or history repeats itself, etc etc etc, if you can't even find or make a chart that doesn't have these very same patterns! there's not really any point arguing with people with the stance that "random charts have these patterns", when really, ALL charts have these patterns. you can't even MAKE a chart that doesn't have these patterns. chart patterns are the weakest evidence of order within a chart when every chart MUST have these patterns!

what IS perhaps interesting to people like you and me is the randomness / non-randomness of volatility, not the 'chart patterns'.

as for how i deduced D as the genuine chart.... well, i had 1 in 5 (unless i counted wrong), so there's that. plus, i've looked at a lot... no, let me try that again. i've looked at A !@#$ING LOT of charts and spent more hours than i care to admit looking at charts, so perhaps that had some element to it.

in the end D just looked like a real chart to me. it was somewhat more random, in its own unique way, compared to the others. it looked like every other chart i've ever seen, whereas the others not so much.

now you'll probably say it wasn't actually D.
Congratulations ScalaFX for coming out of the closet!
  • #23
  • Dec 20, 2012 11:17pm Dec 20, 2012 11:17pm
Quoting nubcake
i'm not 'concerned' about your tone, it was more that i was 'confused' about what you were trying to say because it read like you agreed but sounded like you didn't. but you have cleared that up.

i didn't, and don't, expect this to be a big dialogue or whatever. i was only intending to toss out that first post and be done with it, and i'm merely procrastinating at the moment with these further posts.

what i was getting-at is that people attribute chart patterns as evidence that charts are not random. "look, a support line keeps being respected",...
come on nubs don't give up on me....I am finally getting to talk to some one that speaks my language


what do you make of the volatility of returns not being random at all......

and by volatility I mean the closing price and the squared variance from the drift......I have personally come to the conclusion that the most important thing to ascertain is the "drift", but I can just be grabbing at air here. some people might call this the "trend", "direction", "influx of capital"........ the way I have attempted to capture the drift (statistically I could just be lucky) is by doing linear regression, and to give me a perspective on the possible variance or probability I use a GARCH 1,1 with my own twist on the extrapolation on higher holding time.

But again I could just be using mathematics like people use chart patterns..................lucky guess that is worth a shit.
AVT INVENIAM VIAM AVT FACIAM
  • #24
  • Dec 20, 2012 11:31pm Dec 20, 2012 11:31pm
Quoting the redlion
come on nubs don't give up on me....I am finally getting to talk to some one that speaks my language


what do you make of the volatility of returns not being random at all......

and by volatility I mean the closing price and the squared variance from the drift......I have personally come to the conclusion that the most important thing to ascertain is the "drift", but I can just be grabbing at air here. some people might call this the "trend", "direction", "influx of capital"........ the way I have attempted to capture the...
i might be speaking your language, but you aren't speaking mine (though i do have a tentative grasp on what you are saying if i don't go and spend 30 mins in wikipedia). the area of stats and whatnot was the part of maths i grasped the least in school... and this is someone who did the 'big boy' maths classes reserved for those who are smarter than the average bear.

and yes, the "drift" is surely important, otherwise you are just swimming against the tide. without trend and volatility you have nothing. you can't make pips from a flat line. so how do you get to grips with either? well, that's probably where some sound fundie knowledge gives a head start... but now we are venturing ever so slightly into grail territory.

i don't have all of the answers, but i think some people maybe have the wrong questions. that's what my first post was about.
Congratulations ScalaFX for coming out of the closet!
  • #25
  • Dec 20, 2012 11:51pm Dec 20, 2012 11:51pm
well I don't quite know what to say.....because for every macro economic analysis you can throw at me....I can probably justify a position the other way....that is not as clear cut as FA traders might like to admit.
AVT INVENIAM VIAM AVT FACIAM
  • #26
  • Dec 21, 2012 12:59am Dec 21, 2012 12:59am
Quoting nubcake
that forex guy has been drudging up all manner of old threads and starting new, bad, threads. this thread is oooooooooooooo

ooooo

ooooooooooold. old.

but, since it's up here's something i'd like to see. i want to see someone give a rough chart-like drawing that does NOT show any typical technical analysis 'shapes'. they say that charts are random, and you'll find all manner of geometry within even a genuinely randomly generated chart, so let's take it from another angle. let's make a chart that is 100% manually created, or you could maybe...
So I suppose what you are saying is that at least some chart patterns which appear to have significance are random and have no meaning, even if they appear to do so? That is possible and even probable.

A truly random chart would not have a time element and would not have a logical sequence, just random dots over a two dimensional area. To me, that underlines the importance of chart patterns and what they reflect in the market.
Gone to a better place
  • #27
  • Dec 21, 2012 1:11am Dec 21, 2012 1:11am
Quoting smikester
So I suppose what you are saying is that at least some chart patterns which appear to have significance are random and have no meaning, even if they appear to do so? That is possible and even probable.

A truly random chart would not have a time element and would not have a logical sequence, just random dots over a two dimensional area. To me, that underlines the importance of chart patterns and what they reflect in the market.
you missed the whole point


go back to when I SIMULATED RANDOM charts
AVT INVENIAM VIAM AVT FACIAM
  • #28
  • Dec 21, 2012 1:23am Dec 21, 2012 1:23am
Quoting the redlion
you missed the whole point


go back to when I SIMULATED RANDOM charts
You're right.
Gone to a better place
  • #29
  • Dec 21, 2012 1:27am Dec 21, 2012 1:27am
Quoting smikester
So I suppose what you are saying is that at least some chart patterns which appear to have significance are random and have no meaning, even if they appear to do so? That is possible and even probable.
almost. i'm not saying it's possible, or probable. i'm saying it is unable to be any other way. i'm saying that you can't even deliberately make a chart that doesn't have these very same patterns. unless you rig the chart by only plotting one or two points, for example, then you can't escape patterns forming of one kind or another, or some mystical 'tech analysis' manifesting itself.

some people expect the existence of different patterns and whatnot to be evidence of their significance. really, i'm just taking a dig at the people who believe "chart patterns 'work' because.... just because they do".

legitimate charts have a wealth of patterns. random charts have a wealth of patterns. deliberately created charts made with the intention of having no patterns.... still have the patterns. if it's not some trend line, or s/r, it'll be some fib bounce or astrologically derived turn-point. bogus.

Quote
A truly random chart would not have a time element and would not have a logical sequence, just random dots over a two dimensional area. To me, that underlines the importance of chart patterns and what they reflect in the market.
tbh i'm not really understanding what you are trying to say here. if you have some reason to 'believe' in chart patterns that is better than "just because" then all the power to you. but for anyone who just wants to see pretty pictures within a chart and takes that as evidence of these patterns holding some inherent power within them then maybe take a moment and think a little harder about it... or try deliberately making a chart that doesn't have any patterns in it. i'd be interested in seeing that unicorn.

this wasn't meant to be anything 'deep'. i only meant for the one post and nothing more, but it's gone on a bit further now. this is in the 'rookie section' no less.
Congratulations ScalaFX for coming out of the closet!
  • #30
  • Dec 21, 2012 2:17am Dec 21, 2012 2:17am
Quoting nubcake
almost. i'm not saying it's possible, or probable. i'm saying it is unable to be any other way. i'm saying that you can't even deliberately make a chart that doesn't have these very same patterns. unless you rig the chart by only plotting one or two points, for example, then you can't escape patterns forming of one kind or another, or some mystical 'tech analysis' manifesting itself.

some people expect the existence of different patterns and whatnot to be evidence of their significance. really, i'm just taking a dig at the people who believe...
Well, I'm not trying to be clever here, but for a chart to be truly random it would have no time element and would have infinite range. Completely useless for a trading strategy. I suppose after a ranging year such as we have had this type of debate is useful. I don't have a problem with it being in the rookie section because these kind of studies are important in my opinion.

I agree that some types of chart pattern are more useful than others and it does depend on how it is defined. I don't believe that they work just because they do. The clearer the pattern, the better it works, in general.

Take one of the simplest and most common chart patterns, which is consolidation after a sustained move. It's very easy to envisage what the market is doing here. It's a simple pattern which occurs because there is a market. It could appear in a randomly generated chart with time as the x and a random value as the y but most unlikely, even though time is such a consistent parameter. The longer a pair consolidates, the more sustained the move is following that consolidation. This is a high probability. I did the work. So, that leads me to believe that flags, pennants and triangles all work, more often than not. I find it quite difficult to understand why some other chart patterns (and I'm including candlesticks in this definition) work, but again, the simpler ones such as engulfing, pin bars and so on are pretty obvious, even though the context is as important as the pattern.

Market psychology produces these patterns, bubbles, exhaustion spikes, what have you, and you can even see it on these threads when people say something like "It's like watching paint dry, I'm closing my longs and going to..." Guess what? Bull flag. I'll be watching that, even if you aren't, and I'll wait till it breaks, preferably with an engulfing bullish candle.
Gone to a better place
  • #31
  • Dec 21, 2012 2:44am Dec 21, 2012 2:44am
Quoting smikester
Well, I'm not trying to be clever here, but for a chart to be truly random it would have no time element and would have infinite range.
hmm, i think maybe you are trying to be clever. a coin doesn't have a 'time element' and certainly doesn't have an 'infinite range', and yet a coin flip is still random. why doesn't it have these properties? because it has clearly defined constraints.... as do markets. let's suppose that the markets truly are 'random'. this doesn't require what you said. a market has constraints, and a random market has the same constraints. you can't say a market is NOT random just because it doesn't have a bid/ask of infinity / infinity.

Quote
Completely useless for a trading strategy.
i never touched on trading strategies, and don't intend to. so the discussion of strategies is beside the point and a different discussion altogether.

Quote
I agree that some types of chart pattern are more useful than others and it does depend on how it is defined. I don't believe that they work just because they do. The clearer the pattern, the better it works, in general.
if you have sufficient evidence to prove this then great. most people don't, and you also might not. i simply don't know or even care if you do because again it's not the point.

Quote
Take one of the simplest and most common chart patterns, which is consolidation after a sustained move. It's very easy to envisage what the market is doing here. It's a simple pattern which occurs because there is a market. It could appear in a randomly generated chart with time as the x and a random value as the y but most unlikely, even though time is such a consistent parameter.
not at all unlikely.

Quote
The longer a pair consolidates, the more sustained the move is following that consolidation. This is a high probability. I did the work. So, that leads me to believe that flags, pennants and triangles all work, more often than not. I find it quite difficult to understand why some other chart patterns (and I'm including candlesticks in this definition) work, but again, the simpler ones such as engulfing, pin bars and so on are pretty obvious, even though the context is as important as the pattern.
the context.... so not just the 'chart pattern'. this is also sounding like a subjective observation. i've seen plenty of 'consolidations' that broke-out.... to nowhere. it works both ways. maybe, just maybe, more one way than other, but that's a different conversation.

Quote
Market psychology produces these patterns, bubbles, exhaustion spikes, what have you, and you can even see it on these threads when people say something like "It's like watching paint dry, I'm closing my longs and going to..." Guess what? Bull flag. I'll be watching that, even if you aren't, and I'll wait till it breaks, preferably with an engulfing bullish candle.
randomness also produces these patterns, as does a deliberately made chart.

it's all well and good to say you have evidence of whatever. my post wasn't for you then. you also seem to have grossly different ideas of what constitutes random and what random can, and does, produce.

all i particularly care about right now is for someone to show me a chart that doesn't have 'chart patterns'. everything else is just going to result in the typical back and forth arguing that i care not to spend time on at the moment, and is all just not the point i was making.
Congratulations ScalaFX for coming out of the closet!
  • #32
  • Dec 21, 2012 3:04am Dec 21, 2012 3:04am
Quoting nubcake
hmm, i think maybe you are trying to be clever. a coin doesn't have a 'time element' and certainly doesn't have an 'infinite range', and yet a coin flip is still random. why doesn't it have these properties? because it has clearly defined constraints.... as do markets. let's suppose that the markets truly are 'random'. this doesn't require what you said. a market has constraints, and a random market has the same constraints. you can't say a market is NOT random just because it doesn't have a bid/ask of infinity / infinity.

i never touched...
OK, well thank you for your time and stimulating debate and you are both right, I am missing the point. Probably deliberately, although truthfully I don't know why. I do see your point about what "random" can produce, but since I don't see a value there, other than drawdown in the market context. All the best to you and have a good Christmas
Gone to a better place
  • #33
  • Dec 21, 2012 3:28am Dec 21, 2012 3:28am
Quoting smikester
OK, well thank you for your time and stimulating debate and you are both right, I am missing the point. Probably deliberately, although truthfully I don't know why. I do see your point about what "random" can produce, but since I don't see a value there, other than drawdown in the market context. All the best to you and have a good Christmas
i'm just throwing a wrench in the works of some chart pattern beliefs. nothing more. you have a certain perspective and that's cool.
Congratulations ScalaFX for coming out of the closet!
  • #34
  • Dec 21, 2012 3:42am Dec 21, 2012 3:42am
The markets arent random , chaotic yes but def. Not random

Do you think that there is a random outcome on price if there are more buyers then sellers ?
  • #35
  • Dec 21, 2012 8:31am Dec 21, 2012 8:31am
Quoting nubcake

the context.... so not just the 'chart pattern'. this is also sounding like a subjective observation. i've seen plenty of 'consolidations' that broke-out.... to nowhere. it works both ways. maybe, just maybe, more one way than other, but that's a different conversation.
.
a `fakeout` most probably happens due to to the fact that more often then not institutions `fade` breakouts..
as they need the liquidity to offset a positon .. or similar ..

but the law of cause and effect holds true to the markets , price doesent move randomly there is always a cause for a change in price


for me ChartPatterns are a by product of the markets action ... nothing more nothing less ..
  • #36
  • Dec 21, 2012 9:50am Dec 21, 2012 9:50am
people are missing the point, and that's fine.

i do understand that markets don't just move themselves. i'm not new to this. i understand the underlying mechanics of what makes a market, and what makes a market move.

this is a little in the realm of off-point, but if i gave you 1 input leading to some output then is the output random? no, of course not. you can figure-out what the output will be from the 1 input. what about 2 inputs? no, it is still probably fairly easy to determine the outcome. what about 3, or 10, or 100, or 1000? at what point do you decide that just because you know what the inputs are you are still dealing with something that is, for all intents and purposes, random? no really, because some of you have some very loose views on what is, and what is not, deterministic versus essentially random. if you get down to the nuts and bolts then nothing, anywhere, is truly random... that doesn't mean it's not virtually random for our purposes and perspective.

if charts were not as random as you guys want to try to impart on to me then y'all must be hitting a very nice high win-rate with a matching risk:reward, yes?

a lot of 'but but but' tonight on the forum. 'but charts are not random'. 'but chart patterns are not random'. 'but chart patterns are derived from human input'. 'but random would mean infinite range'. 'but......'.

anyway, it's all good.
Congratulations ScalaFX for coming out of the closet!
  • #37
  • Last Post: Dec 21, 2012 10:24am Dec 21, 2012 10:24am
Iam all good

Agree ,

Its all about the nuts and bolts in the end , puttinh as manny odds as possible in your favour and stick to your plan then the only random thing that can happen to you is...
a random distribution on winning and losing trades
0 traders viewing now