1. Harmonic patterns are often early, But seldom Wrong

2. How's the testing of stats regards failed and successful patterns? All I've seen on this is Bulkowski

3. Originally Posted by jago25_98
How's the testing of stats regards failed and successful patterns? All I've seen on this is Bulkowski
Stats can only make sense if you test the same type of harmonic with the same fixed SL and TP, ... however we all manage harmonic trades differently.

4. Originally Posted by jago25_98
also:
Lack of backtested stats for each pattern completion vs failed vs stop hunted patterns
That is the most important of all! I was also just curious why so many traders buy a strategy if they hear "high probability" without doubt.
I mean it is your job as trader that every information is valid, so there must be a a concrete check to replace the word "high" with a non nebulous hard number.
Also important the probability can change over time so, the correct way to display a probability is like this.

1.Succsessrate 40%
2.proceeded Cases/Patterns= 100
3.Timeframe D1
4.Date window= 0.01.2014-01.01.2015

If only 1 of these 4 facts are missing the probability given is useless. Also there could be fanboys, throwing fakenumbers around, so you must see the probability only has hint for a maybe worth study, to check the probability FOR YOUR SELF !

greetings

5. hi agora,
thank you for your commont. I agree. The numbers that are often published together with a system are worthless, especially when system is not explained, entry, stop loss and take profit levels are clear.
So I would add to your list another point:
5. entry, stop loss, take profit rules provided

greetings

6. Originally Posted by kor4x
hi agora,
thank you for your commont. I agree. The numbers that are often published together with a system are worthless, especially when system is not explained, entry, stop loss and take profit levels are clear.
So I would add to your list another point:
5. entry, stop loss, take profit rules provided

greetings
Yes , i know this would be suggested. But The Problem here is also Volatility changes over time. (I mean the average Daily Range Movement for lets say 10-20 Periods of D1). I call it the D1-factor (divide or multiply by case).
So a better way is instead of naming pips, to use Reward-Risk relations. Like this you can always adapt even if the volatility doubles.
Example.
Reward vs Risk is 2 to 1
actual Avg. D1 Range (20 Periods) (2015_03-2015_04) is 100 pips of Pair X
So if SL is for example 25 Pips, TP must be 50 pips.
Entry Rules are needed offcourse.

greetings

7. yes, if stoploss and take profits are additionally expressed as dependent on volatility, is even better.
nice contribution, Agora.

ortodox harmonic trading as commontly presented on www is just fibonacci price ratios based. Since I created korHarmonics / AMP and DaVinci and while doing it I spend time with harmonics, I drifted away from the ortodox approach. Actually I am much closer to the original Gartley book approach than to Pesavento's Fibo add-ons, but it's already another topic.

Page 2 of 3 First 123 Last