The similarity system | Page 491 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Similarity: A Practical Approach 859 replies

Similarity Trading - The ZZBB Method 215 replies

Rainbow similarity 256 replies

Similarity Stuff 17 replies

Directional Similarity 0 replies

Attachments: The similarity system
Exit Attachments

The similarity system

  • #9,801
  • Oct 2, 2014 11:59pm Oct 2, 2014 11:59pm
  •  Neio
  • Joined Oct 2012 | Status: Can't think outside?Enlarge the box | 688 Posts
Quoting Barbaresco
Thanks Neio, yes it seems rather straightforward to come up with a strategy with an edge based on the principles in the thread. However, I'm still searching for something more exact. When computing the probabilities using the premise laid out by EURUSDD and coded into the indi by K, the misleading part seems to come from the bars where the movement was too small to be tradeable, i.e. <= Spread + a tiny profit, say 1 pt. Essentially all of these bars are recurrent, and become a large part of the data set. So, your 97% isn't 97% at all. What if bars...
I see what you are saying and I like your thinking...I would be interested to see the results of that. A range bar chart with a TZ indicator?
An expert is a man who has made all the mistakes which can be made
  • #9,802
  • Edited 12:42am Oct 3, 2014 12:31am | Edited 12:42am
  •  Neio
  • Joined Oct 2012 | Status: Can't think outside?Enlarge the box | 688 Posts
Quoting Barbaresco
Thanks Neio, yes it seems rather straightforward to come up with a strategy with an edge based on the principles in the thread. However, I'm still searching for something more exact. When computing the probabilities using the premise laid out by EURUSDD and coded into the indi by K, the misleading part seems to come from the bars where the movement was too small to be tradeable, i.e. <= Spread + a tiny profit, say 1 pt. Essentially all of these bars are recurrent, and become a large part of the data set. So, your 97% isn't 97% at all. What if bars...

I would imagine that a large amount of the recurrence happens within 5 bars. What if we were just to exclude these from the stats?

So what is the probability of a potential TZ being cleared after the first X bars.

Just thinking out loud...
An expert is a man who has made all the mistakes which can be made
  • #9,803
  • Oct 3, 2014 12:57am Oct 3, 2014 12:57am
Quoting Neio
{quote} I would imagine that a large amount of the recurrence happens within 5 bars. What if we were just to exclude these from the stats? So what is the probability of a potential TZ being cleared after the first X bars. Just thinking out loud...
With regard to regular candles, wouldn't we want a recurrent candle that was right next to the transient, if it were say 10 pts? I thought of removing just the small candles from the data set, but working off of Range Bars would be more exact. Will see if I can get something going in the morning (CA time here).

Could probably work (differently of course) to provide tradable probabilities on Renkos as well...
  • #9,804
  • Oct 3, 2014 1:48am Oct 3, 2014 1:48am
Quoting killerno
{quote} Hi MikiMax, nice result, well done!! Would you share a bit more about your "technic"? It is based on TZ or Similarity too? Thanks Regards
will be back in late NY
let me check myself today...
Mind is the King
  • #9,805
  • Oct 3, 2014 1:56am Oct 3, 2014 1:56am
Quoting Barbaresco
First, confusion re: the 97%, 3%. Using the tester written by K, find approx. 3% frequency of a true problem bar using H=28 on M15. The other bars are 97%, but only ~15% become potential TZ's, i.e. have satisfied H on the left side. Secondly, trying to trade against the probability of 3% twice, many of the PTZ bars have negligible range. In fact, many of the PTZ bars with sufficient range to trade, that also extend further beyond the PTZ end up becoming TZ's themselves. All of this can be visualized in any major price move, where often 4 bars in...
Hi there,

what occasionally fails are not the probabilities, it is the Markov property. Market does have a memory, and this shows itself in strong trends. This is why you can have several transient zones in succession. You would be wise not to trade against strong trends. There are ways to either completely avoid doing that or to even use it to your advantage.

It is an interesting idea to try to use range bars.

k
  • #9,806
  • Oct 3, 2014 2:10am Oct 3, 2014 2:10am
Quoting Eurusdd
{quote} Power is not in special cases - pictures - . Power lies in the general case, statement. "At any give time, the current price or price range wants to be recurrent(97%) rather than transient(3%)" How I apply this idea exactly, I cannot share. I expect everyone to find her/his own level.
got a spare minute...
That answer to my earlier post by The Founder and Generous Friend EurUsdd helped me a lot
I've also read a few times the Valdy's pdf.
I'd also read comments of other Clever Brains.
Mind is the King
  • #9,807
  • Oct 3, 2014 2:10am Oct 3, 2014 2:10am
Quoting evrynet
{quote} 1.26 seems to be quite a bottom for 2 years...
well, dosnt mean 1.26 cant be violated. time will tell
  • #9,808
  • Oct 3, 2014 2:46am Oct 3, 2014 2:46am
I've gotten a couple of emails about similarity. Congrats to vlady some of the other Originals who have kept this idea going.

My public comment is that similarity is about finding ways to tell when traders are going to try to revisit a price. I myself used the RSI(14) and the CII, or the Composite Index.

Simply put, the RSI and the CII will be similar during a good trend, but a some point the trend begins to weaken internally. If the trend is up, the higher high but lower momentum is your dissimilarity. Price will at some point try to come back and test that.

Another way to see this is with the AO and AC oscillators of Bill Williams.

Similarity is a concept that one can build a trading system around. But you, the trader have to decide what you are willing to do at these decision points.

I hope this helps someone.
If you are good at something, never do it for free--Joker
  • #9,809
  • Oct 3, 2014 3:16am Oct 3, 2014 3:16am
let's make some live check calls...
eur/usd tp at 2665
Mind is the King
  • #9,810
  • Oct 3, 2014 3:17am Oct 3, 2014 3:17am
eur/jpy 137.66 tp
Mind is the King
  • #9,811
  • Oct 3, 2014 3:18am Oct 3, 2014 3:18am
gbp/jpy tp 175.60
Mind is the King
  • #9,812
  • Oct 3, 2014 3:19am Oct 3, 2014 3:19am
gbp/usd 6127 tp
Mind is the King
  • #9,813
  • Oct 3, 2014 3:26am Oct 3, 2014 3:26am
  •  Neio
  • Joined Oct 2012 | Status: Can't think outside?Enlarge the box | 688 Posts
Quoting kprsa
{quote} Hi there, what occasionally fails are not the probabilities, it is the Markov property. Market does have a memory, and this shows itself in strong trends. This is why you can have several transient zones in succession. You would be wise not to trade against strong trends. There are ways to either completely avoid doing that or to even use it to your advantage. It is an interesting idea to try to use range bars. k

I think this is a significant point. The original assumption is that the market has no memory.
An expert is a man who has made all the mistakes which can be made
  • #9,814
  • Oct 3, 2014 4:03am Oct 3, 2014 4:03am
MikiMax - don't know what time levels your calls are made on but some of them look rather brave. Time (how much time) will tell.
  • #9,815
  • Oct 3, 2014 4:12am Oct 3, 2014 4:12am
Quoting handy148
MikiMax - don't know what time levels your calls are made on but some of them look rather brave. Time (how much time) will tell.
sure, that is just a live test as I stated
Mind is the King
  • #9,816
  • Oct 3, 2014 4:55am Oct 3, 2014 4:55am
Quoting Neio
{quote} Welcome Barbaresco, I have come to the same conclusions. I have experimented a lot and tried high and low h values on different TFs. You would think that a higher h value might reduce the number of successive TZs but when there is a powerful move into an area where price has not been for a while it makes no difference. So some of the probabilities presented do not always seem to stack up based on our interpretation of what Eurusdd has presented. Eurusdd although he makes good calls most of the time but occasionally he has got it wrong (so...
Have a little faith Neio I could post results like this every day ,and only posted this because I read your doubtfulness , (is that a word ?) but don't see the point , And I think I am prob one of the poorer traders here ?lol

Trust me when I say if I can do this every day , anyone can , ( that last trade was a wrong entry ) got out as soon as I could no excuse though.
no offence meant ,old chap
Regards Peter
Attached Image (click to enlarge)
Click to Enlarge Name: proof.jpg Size: 708 KB
  • #9,817
  • Oct 3, 2014 5:02am Oct 3, 2014 5:02am
Quoting ephisi
missed my swings. looking to take a swing short 1.2660/70 ish targeting below 1.2576
Slowly but surely. We should get there? Nfp will decide if it happens today
  • #9,818
  • Edited 5:24am Oct 3, 2014 5:05am | Edited 5:24am
Hello,

I think the most interesting TZ are the one in blue/purple in the chart bellow. They seems to appear in a nowhere lands of price (during an acceleration of price?) and showing clearly a trend market.

The others one are only showing simply High and Low confirmated x H later... they can be interesting only in a market with no clearly tendance for scalping.

Attached Image (click to enlarge)
Click to Enlarge Name: XAUUSDM5.png Size: 75 KB


Can someone adapt the indi to filter and showing only these blue/purple zone ? And maybe like the first indicator schowing these TZ in formation and hen they are confirmed?

PS: i dont really understand why they appears?

Ps2: i discovered something very good with only theses zones. Maybe showing ti all with the new indicator

Green pips,
Bottomless wonders spring from simple rules, which are repeated without end
  • #9,819
  • Oct 3, 2014 5:07am Oct 3, 2014 5:07am
  •  Neio
  • Joined Oct 2012 | Status: Can't think outside?Enlarge the box | 688 Posts
Quoting orid
{quote} Have a little faith Neio I could post results like this every day ,and only posted this because I read your doubtfulness , (is that a word ?) but don't see the point , And I think I am prob one of the poorer traders here ?lol Trust me when I say if I can do this every day , anyone can , ( that last trade was a wrong entry ) got out as soon as I could no excuse though. no offence meant ,old chap Regards Peter {image}
Congratulations mate. No offence taken. I never said that this is not successfully being traded. Just that nobody is 100% successful with their calls. Not even Eurusdd.
An expert is a man who has made all the mistakes which can be made
  • #9,820
  • Oct 3, 2014 5:19am Oct 3, 2014 5:19am
Quoting Neio
{quote} Congratulations mate. No offence taken. I never said that this is not successfully being traded. Just that nobody is 100% successful with their calls. Not even Eurusdd.
I must admit I 'm pretty sure I know how some people predict where price will prob be in so so time , but I wouldn't risk the house on it , the way it is done I don't think its that reliable , so I'm in and out quicksmart but I do miss a lot of pips .
heres a tip for yas be good to your mother hehe
Back to entertain the visitors ( they go back to the uk next week ) then I can get some serious reading the thread again
regards Orid peter
1 trader viewing now