Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

Price action trading - Trade like the pros! 964 replies

Retail Trade Gains vs Pros 6 replies

newbie questions on daily pivot points 18 replies

Retail Sales? Trade Balance? What it means? Newbies ask here. Pros answer here. 9 replies

Attachments: Questions for real pros: How to trade with reference points only?
Exit Attachments

Questions for real pros: How to trade with reference points only?

Quoting nubcake
{quote} Nobody is going to hand you their homework. I know where you are wrong, but have no reason whatsoever to tell you where to look instead. You are, after all, the one complaining that nobody is contributing... there's nothing to contribute. This topic isn't a winner, and what you want out of it isn't likely to happen. What you should take from this thread is that you are probably on the wrong path.
Why would this be a bad path? Why this topic isn't a winner? Have you performed research in depth as OP? Do you at least understood what OP is talking about?
It would be better for everyone if, instead of labelling something as "winner topic", you shared what you researched and what where your findings and why you got to the conclusion this was a dead-end topic. Maybe you find some flaws on the researches you've done on this topic, maybe you can help readers to find eventual errors by OP.

Other than that I don't see any kind of reason for you to post on this thread.
Regards.
Thanks apprendista,

since I know which argument will come next, I'll give something away from what I've done. Only a little for the crowd ... This is what I've done and what I've asked for in post #69.

First: How do AVERAGE bullish candlesticks look like?
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 8 KB


2nd: What about the AVERAGE bearish candlesticks?
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 8 KB

Same but switched!

And last but not least the AVERAGE dojis:
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 7 KB


Which trends can you see? Well, the bigger the bodies become, the smaller the wicks get. Moreover, the range of dojis is relatively small in comparison to the AVERAGE bullish or bearish candlestick. There are other observations that have been mentioned earlier in this thread. If you just look at the AVERAGE bullish or bearish candlestick, you can see that the "close wick" (the extremum hit before the candle closed) occurs about 25 % before the candlestick closes. Moreover, the wick of the AVERAGE candlestick is ~25 % of the AVERAGE candlestick's range.

If you like to measure the market performance (which can be seen as rolling candlestick), then how could you act? If you see the close minus open as the performance of the whole candle, then which measure tells you how strong the candlestick performed during the last part of the candlestick's lifetime? Which measure tells you, how much the market is in a range or trend? Which tells you how strong the current retracement, pullback or trend of the market is?

Marry your answers and see if you could withstand when the market behaves in more realistic ways ... To do this I suggested to have a look at bullish candlesticks in which the lower wick is build before/after the upper wick ....
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 7 KB

Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 7 KB


Yes, I've done the same for bearish candles - but because my sample size was quite big (> 27.2M rolling candlesticks), the forms are almost identical (but of course switched).

How far could you have come? During which market formations would you like to stay in your trade, when would you've like got out? How should your measures look like to achieve these goals? Things to ponder on for those who like to participate/stay engaged in the thread.
Quoting apprendista
{quote} Why would this be a bad path? Why this topic isn't a winner? Have you performed research in depth as OP? Do you at least understood what OP is talking about? It would be better for everyone if, instead of labelling something as "winner topic", you shared what you researched and what where your findings and why you got to the conclusion this was a dead-end topic. Maybe you find some flaws on the researches you've done on this topic, maybe you can help readers to find eventual errors by OP. Other than that I don't see any kind of reason for...
Just assume I know nothing, and continue on in this thread looking at a coin flip. If all OP is going to do is keep measuring candlestick sizes, relationships between candlesticks and OHLC and so on... it's going nowhere. If you can't see that then that's your journey. Don't expect anyone to tell you where to look, though.

OP, from what I can see so far, is literally just looking at randomness. Trends start just as fast as they stop, until they don't. Bars go in one direction, until they don't. OP is trying to see the yellow part of an egg after it's been whipped to death with the whites and is rightly concluding that there's very little to gain from this endeavor. OP wants to figure out how to tease out that yellow part, and that'll be for OP to figure out and discover.

Prices behavior is random, except when it's not. If you look at it all in total, you'll only see the total randomness.

OP already knows they are on the wrong path, they just don't know where else to look. Step one is to look at something different instead of beating the same dead horse.
Congratulations ScalaFX for coming out of the closet!
Quoting nubcake
{quote} Just assume I know nothing, and continue on in this thread looking at a coin flip. If all OP is going to do is keep measuring candlestick sizes, relationships between candlesticks and OHLC and so on... it's going nowhere. If you can't see that then that's your journey. Don't expect anyone to tell you where to look, though. OP, from what I can see so far, is literally just looking at randomness. Trends start just as fast as they stop, until they don't. Bars go in one direction, until they don't. OP is trying to see the yellow part of an egg...
Well, maybe you're right. I cannot deny it. But I've already seen some potential in these stats and made some profits (of course not multiple years, but it's a first step to me). Just to repeat this: It's not about beating in the sense of making a prediction. It's about observing what the market is doing every moment (every new incoming tick). I mean, if most traders see a trend that retraces more than 50 %, they would exit. However, this study tells me to exit earlier. I mean: Yes, candlesticks are symmetric if you average it. However, one side is worthless to me - it's just the past. In a bullish candlestick with high time > low time it is the low of this candlestick.
Quoting AntiCre
{quote} Well, maybe you're right. I cannot deny it. But I've already seen some potential in these stats and made some profits (of course not multiple years, but it's a first step to me). Just to repeat this: It's not about beating in the sense of making a prediction. It's about observing what the market is doing every moment (every new incoming tick). I mean, if most traders see a trend that retraces more than 50 %, they would exit. However, this study tells me to exit earlier. I mean: Yes, candlesticks are symmetric if you average it. However,...
You still need to know when anything you discover applies, and when it doesn't. That's what will begin to make a difference. You find a meaningful measure that approximates over some swath of data... now go find when it's above that value, and when it's below, and what might influence that.

But this OHLC nonsense isn't going to get you far. I'm only skimming, but I haven't noticed a reference to what timeframe you are even getting this bar data from, which will make a substantial difference with just that alone.

Anyway I'm out. Do your research, but don't expect to have a roaring discussion about a dead-end topic.
Congratulations ScalaFX for coming out of the closet!
Quoting nubcake
{quote} You still need to know when anything you discover applies, and when it doesn't. That's what will begin to make a difference. You find a meaningful measure that approximates over some swath of data... now go find when it's above that value, and when it's below, and what might influence that. But this OHLC nonsense isn't going to get you far. I'm only skimming, but I haven't noticed a reference to what timeframe you are even getting this bar data from, which will make a substantial difference with just that alone. Anyway I'm out. Do your research,...
In post #69 I recommended to use 1000-tick candles on a rolling tick base. However, for some reason I've used 1040-tick candles from 2018 GU. Nevertheless, the general findings will be the same. Bye
Quoting AntiCre
{quote} In post #69 I recommended to use 1000-tick candles on a rolling tick base. However, for some reason I've used 1040-tick candles from 2018 GU. Nevertheless, the general findings will be the same. Bye
Ok, I'll try and save you some time and I wish I'd listened many years ago.

A pattern you see is only there because human beings are programmed to see patterns.

A pattern is only there because it has formed from another pattern.

This other pattern (forming your pattern) has "failed" in a fractal form (smaller or larger)to create your "pattern" If this (your pattern you are waiting for) fails to form it is because it is likely forming the same shape which will materialise in another size.

Anything more complicated than this and you invite insanity with the rest of the 99%.
If you are a Pro you will understand real quick that reference points alone don't provide an edge.

What i've found works best though is to spot the imbalance in the market (spotting the (hidden) aggressor & direction for a period of time) and then use reference points for potential targets.
Then you can add/place bigger orders in the same direction and make some nice money.

If you also understand how they execute their orders over the time it's also a big plus.
As if there are any "pros" here
What most people struggle to understand is that there are many ways to skin a cat. Just because it doesn't work for you, or it isn't the way you trade, doesn't make the method a lousy one. The amount of close-mindedness in this forum is amazing. Here you have AntiCre dropping golden nuggets, saving you hundreds of hours that would otherwise be dedicated to research, and we have people shitting on his work. If it doesn't work for you, then don't read the thread. No need for unconstructive feedback.

But that's how it is, if you try to crowdsource ideas and build quality discussion around potentially good concepts - It rarely happens.
Keep at it, OP. Don't let that deter you.
ask questions other wont or dont. basic game of trading. if website folk say its garbage then you know its worthy. if you do research and know what you have seen then why worry about online nothing nobodies. they empty fluff. mediocre minds with big mouths. they on carousel never to get off. ur premise is good start. keep asking questions. 99% of info on here is not 100% field tested. so start there.
protein for breakfast !
Quoting AntiCre
Over the past few years, I've been hanging out here at FF. I've particularly analyzed two systems (or theories of systems), collected statistics, researched how prices behave under certain situations, and so on. Despite conducting sometimes intensive research, I haven't been able to consistently make profits. As a result, I've taken a step back and considered starting anew. Throughout the years, I've consistently observed that the way prices move is not significantly different from random walks. For example, the expected profit is strongly correlated...
there is a good thread which has been very active for the last few years, but unfortunately one or the other has already died or withdrawn.
https://www.forexfactory.com/thread/...ppy-trading-ej

we have only traded horizontal lines here. the "line in the sand" works best, as the market also moves along the y-axis.

and don't forget the x-axis;-)
trading is like dancing, don't think and just listen to the music!
Quoting carddard
What most people struggle to understand is that there are many ways to skin a cat. Just because it doesn't work for you, or it isn't the way you trade, doesn't make the method a lousy one. The amount of close-mindedness in this forum is amazing. Here you have AntiCre dropping golden nuggets, saving you hundreds of hours that would otherwise be dedicated to research, and we have people shitting on his work. If it doesn't work for you, then don't read the thread. No need for unconstructive feedback. But that's how it is, if you try to crowdsource...
Well, you can also take your horse to the water but you can't make him drink..

Good luck all, yet another closed minded thread.
Hello AntiCre,

This idea might direct you to another direction (maybe it is just a stupid idea, but it is an idea)

This is one quote that is haunting me every time when I look at the candles and I switch from one time frame to another : @mzvega - "There is no way to see how prices are distributed throughout the time interval using OHLC data (candlesticks) The data is just not there."

This is just a little experiment:
If I take the the 4H candle: High and Low and divide by 2 to find out the MidPoint I will get a value
If I take all the 1H candles: High and Low divide by 2 and all 4 values divided by 4 - I should get the same value as the 4H Midpoint
If I do the same for 1H and 15Min - there is always () a different value

@ Pipalicious :
Mathematics: "2x3=6" , "3x2=6"
Market: "2x3=6", 3x2=5"

Trei
Attached Image(s) (click to enlarge)
Click to Enlarge Name: 1H Median.png Size: 4 KB
Click to Enlarge Name: 4H Median.png Size: 3 KB
Click to Enlarge Name: 15Min Median.png Size: 5 KB
Quoting treigaini
Hello AntiCre, This idea might direct you to another direction (maybe it is just a stupid idea, but it is an idea) This is one quote that is haunting me every time when I look at the candles and I switch from one time frame to another : @mzvega - "There is no way to see how prices are distributed throughout the time interval using OHLC data (candlesticks) The data is just not there." This is just a little experiment: If I take the the 4H candle: High and Low and divide by 2 to find out the MidPoint I will get a value If I take...
Hello, the values are different because you're averaging multiple high/lows giving each one the same weights. Think of it like so: you've a 1H candle and you calculate the midpoint, then you go to 1m timeframe and calculate the midpoint of all candles and then you do sum of all midpoints / 60. They will differ, ok, but why and based on what? Think if the 60 1m candles were ranging for 58 minutes and then in the last two minutes you got a 40 pips spike that created new high and low for the 1H candle, now where would the "cumualtive" midpoint be? It will be very near the first 58 minutes range, because that range lasted more and hence it has more weight in the final result.

From what I understand this measure is heavily related to time, basically it averages a series of time points giving each one the same weight.

Regards.
Quoting treigaini
Hello AntiCre, This idea might direct you to another direction (maybe it is just a stupid idea, but it is an idea) This is one quote that is haunting me every time when I look at the candles and I switch from one time frame to another : @mzvega - "There is no way to see how prices are distributed throughout the time interval using OHLC data (candlesticks) The data is just not there." This is just a little experiment: If I take the the 4H candle: High and Low and divide by 2 to find out the MidPoint I will get a value If I take...
Dear treigaini,

sorry for the late reply. I think apprendista already gave a good explanation and I agree. Nevertheless, I still think these differences could also be used to tell you something about the market itself. I think it was strongtrader who used MAs (not medians) on different TFs and analyzes whether they converge or diverge. Sry, was not able to find these posts of him. Nevertheless, you can probably find it in some CP-related threads ...

Cheers,
AC
Quoting apprendista
{quote} Hello, the values are different because you're averaging multiple high/lows giving each one the same weights. Think of it like so: you've a 1H candle and you calculate the midpoint, then you go to 1m timeframe and calculate the midpoint of all candles and then you do sum of all midpoints / 60. They will differ, ok, but why and based on what? Think if the 60 1m candles were ranging for 58 minutes and then in the last two minutes you got a 40 pips spike that created new high and low for the 1H candle, now where would the "cumualtive" midpoint...
The other way to explain these differences is that the 1H candles highs/lows overlap each other hence the median differences.
Quoting apprendista
{quote} Hello, the values are different because you're averaging multiple high/lows giving each one the same weights. Think of it like so: you've a 1H candle and you calculate the midpoint, then you go to 1m timeframe and calculate the midpoint of all candles and then you do sum of all midpoints / 60. They will differ, ok, but why and based on what? Think if the 60 1m candles were ranging for 58 minutes and then in the last two minutes you got a 40 pips spike that created new high and low for the 1H candle, now where would the "cumualtive" midpoint...
Your explanation is spot on, and this is confirming the following facts:
- Candles do not show the reality
- Changing timeframes is not going to give an "better" understanding of price
- Removing time will give us price on tick chart (tick bars, number of ticks calculations ... anything else?)

I'm trying to understand the weight, because you say that the same weight is applied for each time series (time series : 1min, 5min ... etc).
Could you please define weight and how can it be different for each time frame.

Regards,
Trei
Quoting AntiCre
{quote} Dear treigaini, sorry for the late reply. I think apprendista already gave a good explanation and I agree. Nevertheless, I still think these differences could also be used to tell you something about the market itself. I think it was strongtrader who used MAs (not medians) on different TFs and analyzes whether they converge or diverge. Sry, was not able to find these posts of him. Nevertheless, you can probably find it in some CP-related threads ... Cheers, AC
I could not find the post from strongtrader , but I think you are talking about the post by forrestang : Higher Edge within a Single Candlestick | Page 119 | Forex Factory

It is a good idea.

Regards,
Trei
Quoting AntiCre
Thanks apprendista, since I know which argument will come next, I'll give something away from what I've done. Only a little for the crowd ... This is what I've done and what I've asked for in post #69. First: How do AVERAGE bullish candlesticks look like? {image} 2nd: What about the AVERAGE bearish candlesticks? {image} Same but switched! And last but not least the AVERAGE dojis: {image} Which trends can you see? Well, the bigger the bodies become, the smaller the wicks get. Moreover, the range of dojis is relatively small in comparison to the AVERAGE...
Hello AntiCre,

Have you ever considered to run a study for candles that will fit into the following criteria: Open=High, Open=Low, Close=High, Close=Low
How many candles of this type can you see on the 4H timeframe, 1H timeframe, 15 min timeframe ... (the lower the timeframe, more occurrences of this types of candles but when you get a tick ... what could those candle mean in the tick timeframe)

I might sound like there is a secret in the statement above, but this are just ideas that one might want to explore and maybe find an edge.

Regards,

Trei
Attached Image(s) (click to enlarge)
Click to Enlarge Name: 1H Candles.png Size: 4 KB
Click to Enlarge Name: 4H Candles.png Size: 5 KB
Click to Enlarge Name: 15Min Candles.png Size: 8 KB
0 traders viewing now