Candle trading pattern seems too good to be true | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Is the sky the limit or is it too good to be true? 27 replies

Why does this Co-integrating basket look too good to be true? 1 reply

Too good to be true? 17 replies

FXCM managed accounts - Too good to be true? 5 replies

Too good to be true 5 replies

Attachments: Candle trading pattern seems too good to be true
Exit Attachments

Candle trading pattern seems too good to be true

  • #1
  • First Post: Jun 1, 2010 3:10pm Jun 1, 2010 3:10pm
Hi, I'm actualy using a candle pattern indicator, I've not tested it on live yet.

It seems to good to be true, so, can someone tell me if it repain or not ?
Because if not, it's really well coded..

Thank you Factory
  • #2
  • Jun 1, 2010 3:11pm Jun 1, 2010 3:11pm
Here it is :
Attached File(s)
File Type: mq4 Candle Sticks reconize.mq4   8 KB | 2,078 downloads
  • #3
  • Jun 1, 2010 3:48pm Jun 1, 2010 3:48pm
I've seen questions like this posted occasionally throughout the forum. If you want to know if an indicator repaints, just put it on the M1 chart and watch it for 15-20 minutes. If you don't want to spend 15-20 minutes investigating it, then there is nothing in this forum that will do you any good.
  • #4
  • Jun 1, 2010 4:09pm Jun 1, 2010 4:09pm
The fact is, I don't exaclty know how to do that.
The indicator draws points on the last pattern candle to tell you when to buy/sell but I don't know if it's calculated on the previous candles or if it's a kind of repained thing doesn't work live.

Bytebodger, I spent about 1h on chart trying to understand.
In the code there is a "buffer".. I'm not a programmer, I can't tell.

Thank for you help
  • #5
  • Jun 1, 2010 4:50pm Jun 1, 2010 4:50pm
Quoting Walcott
Hi, I'm actualy using a candle pattern indicator, I've not tested it on live yet.

It seems to good to be true, so, can someone tell me if it repain or not ?
Because if not, it's really well coded..

Thank you Factory
Hi this indicator don't repaint it only show you some type of candle that can lead to a reverse, like a hammer candle a doji, an inverted hammer a spin etc..
  • #6
  • Jun 1, 2010 5:45pm Jun 1, 2010 5:45pm
Quoting Walcott
Hi, I'm actualy using a candle pattern indicator, I've not tested it on live yet.

It seems to good to be true, so, can someone tell me if it repain or not ?
Because if not, it's really well coded..

Thank you Factory
Open Strategy Tester
Put in any one of the sample ea's
Tick visual mode and when it starts add the indicator and watch it to see if repainting occurs
We are all One. Be present. Be the Observer
  • #7
  • Jun 1, 2010 7:23pm Jun 1, 2010 7:23pm
As far as I can tell by looking at the program code, its calculations are based only on prior candles. Since the data (OHLC) on these candles remains fixed, the indy shouldn't repaint.
  • #8
  • Edited 11:42pm Jun 1, 2010 7:33pm | Edited 11:42pm
to me candle pattern indicators are not really that good because they are based on mathematical calculation whereas we identify the patterns' quality by subjective mind. Programmes are good at calculating values but that's not the case while it come to determing shapes by comparing values.
  • #9
  • Jun 1, 2010 9:46pm Jun 1, 2010 9:46pm
You can see on this pic that the text does appears on what candle it refer
I didn't writte these tag it is the indicator that write it. So you can see it is just like the indicator says a candle pattern indicator.

I don't see in this indicator to good to be true. candle pattern are good when using at resistance area. But you can not rely only on this
Attached Image (click to enlarge)
Click to Enlarge Name: candle indicator.gif Size: 23 KB
  • #10
  • Jun 1, 2010 11:01pm Jun 1, 2010 11:01pm
Quoting lmyyyks
to be candle pattern indicators are not really that good because they are based on mathematical calculation whereas we identify the patterns' quality by subjective mind. Programmes are good at calculating values but that's not the case while it come to determing shapes by comparing values.

As long as the computer is aiding the trader and not taking over for the trader, it can handle things like candle pattern recognition just fine. I have writtern several myself. It will probably pick a few invalid ones, but a good trader can handle that. Better to have a computer do the grunt work than the trader.
  • #11
  • Edited Jun 2, 2010 12:11am Jun 1, 2010 11:48pm | Edited Jun 2, 2010 12:11am
Quoting CodeMeister
As long as the computer is aiding the trader and not taking over for the trader, it can handle things like candle pattern recognition just fine. I have writtern several myself. It will probably pick a few invalid ones, but a good trader can handle that. Better to have a computer do the grunt work than the trader.
What i think is if you do not totally trust the indicator you're using and you frequently need lots of other indicators to justify its signal, you should either modify it or simply throw it away. It's just like they are generating random signals to make a bet on future movement. When you find it generates too many wrong signals that you start to look for confirmation from other indicators, then you'll know that you can live without it because the others are already good enough to do the job.
  • #12
  • Jun 2, 2010 12:15am Jun 2, 2010 12:15am
Just a cursory look and I'm not sold on it. The marked candle patterns are not up to the criteria defined by Steve Nison, but maybe I'm a Steve Nison snob. . The dark cloud cover patterns are not really accurate (real body's are not 50% or more into previous candle real body). In more than one instance there is a huge bearish engulfing bar and it's calling in Dark Cloud cover. Also, no 3 bar patterns, I mean the evening/morning star?!?! Great setup in the right place.

I'd say by Steve Nison rules, this indicator is pretty weak. Just my opinion though and since I can't program to save my life, it's not like I could do better...



P.S. Using Aud/Usd 1hr.
Don't wish it were easier, wish you were better. ~ Jim Rohn
  • #13
  • Jun 2, 2010 3:05am Jun 2, 2010 3:05am
Quoting lmyyyks
What i think is if you do not totally trust the indicator you're using and you frequently need lots of other indicators to justify its signal, you should either modify it or simply throw it away. It's just like they are generating random signals to make a bet on future movement. When you find it generates too many wrong signals that you start to look for confirmation from other indicators, then you'll know that you can live without it because the others are already good enough to do the job.

Well I'm not sure what criteria you use to judge the usefulness of an indicator. I figure 50% accuracy is a pretty good indicator. I know of some that were 25% accurate and were still useful because I didn't let the computer make the trading decisions and tried to not be influenced by the computer was telling me to do. As far as using confirmatory indicators, thats a bad trap. Its better to learn and accept the limitations of one or two indicators than to pile indicator upon indicator.

The computer is an assistant that helps by screening potential setups and allows me to focus on the important things like market conditions.
  • #14
  • Jun 2, 2010 3:39am Jun 2, 2010 3:39am
Well, according to the code, it doesn't repaint

By following your advices, I tried a system :

I put a HMA(60) as a trend filter (on GBP/USD 15M).
I take only Pattern in the same direction of HMA, if the HMA slope is UP I take Long patterns (blue arrows).
I don't take pattern on the middle of the night, no volatility.

(I'm actualy trying to associate this system with Pivot Points using specific rules, on M5, interesting).

I use HMA because it stick well to the price but you can even try a EMA with.

If the dark cloud cover patterns are not really accurate, you can delete it easily in the code and just use reliable patterns.

Now, I realize that maybe a program can't give better candle pattern interpretation than human, but by watching the chart, it seems not this bad to me.
  • #15
  • Jun 2, 2010 4:28am Jun 2, 2010 4:28am
Quoting CodeMeister
Well I'm not sure what criteria you use to judge the usefulness of an indicator. I figure 50% accuracy is a pretty good indicator. I know of some that were 25% accurate and were still useful because I didn't let the computer make the trading decisions and tried to not be influenced by the computer was telling me to do. As far as using confirmatory indicators, thats a bad trap. Its better to learn and accept the limitations of one or two indicators than to pile indicator upon indicator.

The computer is an assistant that helps by screening potential...
then I would say the criteria is that the signal itself is persuasive and correct enough for you to ignore override opposite signals by certain indicators you're using at the same time. This is why we need multiple indicators to work complimentarily to prevent you from entering the wrong direction. I've tried to read bar-by-bar to seek explanations for the coming movements and wrong signals of other indicators but didn't get a satisfactory result. maybe i'm just too biased against it.
  • #16
  • Jun 2, 2010 4:36am Jun 2, 2010 4:36am
On Strategy tester, the arrow appears only after the next two candlesticks appears.. What does it mean ?
  • #17
  • Jun 2, 2010 4:52am Jun 2, 2010 4:52am
Quoting Walcott
On Strategy tester, the arrow appears only after the next two candlesticks appears.. What does it mean ?
you mean fractals?
  • #18
  • Jun 2, 2010 5:03am Jun 2, 2010 5:03am
I'm just asking my self if the arrow over the pattern is right.

The arrow to indicate the new pattern over the candle had to appear maximum on the very next candle opening (or the close of the candle pattern), am I right ? Not two candles after like it is on strategy tester.

I'm confused.
  • #19
  • Jun 2, 2010 5:15am Jun 2, 2010 5:15am
Quoting Walcott
I'm just asking my self if the arrow over the pattern is right.

The arrow to indicate the new pattern over the candle had to appear maximum on the very next candle opening (or the close of the candle pattern), am I right ? Not two candles after like it is on strategy tester.

I'm confused.
it would be better if you can give a screenshot
  • #20
  • Jun 2, 2010 5:39am Jun 2, 2010 5:39am
It's easier if you try yourself in the strategy tester to see it.
0 traders viewing now