EUR/USD patterns

good pattern, but I feel dizzy when looking your chart

Already noted in post 14 and replied to in 15.

The method'll take a while to get right. You noted one problem, and there'll be plenty of others. Once the method's reasonably done, I can explain it, because I think it'll give an effective approach to organizing the trendlines one draws toward the next move.

(thanks to the various coders)

Hiekin Ashi.............. rl balance (current: http://s5.tinypic.com/140i792.jpg ;
.............................next vsn in draft)
CDRW Spread...........highest: hi-lo candle channels
bettervolume............low (yellow) high (red) medium, etc. (lt blue)
CDRW volume...........bettervolume confirms (matched colors)
recentcandles hist.....5', 15'; 125 bars

http://i40.tinypic.com/2mzjfk5.jpg

Longterm indicator for smallest and highest price-bar spreads. Possible uses:

1. On 15', highlight lightly biggest spreads on chart; on 1', find and boldly highlight biggest spread at each area. Then check H1 (has nice pattern repetition): 1' chart's biggest spreads turn/explain each longterm trend. http://i40.tinypic.com/im8pl5.jpg , http://i43.tinypic.com/x5nko2.jpg .
2. Also do the same for smallest spreads on the H1; they show stability price returns to later before a major turn. One can trade above/below the trendline extrending from the price trend the smallest-spread bar is on. http://i42.tinypic.com/20hvmva.jpg , http://i40.tinypic.com/2mq193p.jpg .
3. Draw trendline connecting highs of two biggest-spread bars and another for their lows. Extend the two lines into future. One can trade with given channel intersection in mind. (This on http://s5.tinypic.com/30nhpbk.jpg .

(continued)

A special example of the 1' highest-spread bar is multiple 1' hsb's together. One draws a line in the middle of the high and low trendlines. http://i44.tinypic.com/2ldwupx.jpg http://i40.tinypic.com/2jcyhbo.jpg http://i43.tinypic.com/xc4y90.jpg

There's also a 1' lowest-spread area. Find a balance pattern amidst one, and its shape reflects the larger waves pre-,post- of it. http://i42.tinypic.com/wiu2b8.jpg .

A volume-spread description (CDRW Volume, bettervolume) of a few to several 1' bars at a time, a pattern-description method, http://i44.tinypic.com/2hcnh8w.jpg (with price-spread) http://i39.tinypic.com/2uivrb4.jpg to get portions of a trend, and the noted uses of CDRW Spread (1H lowest-spread bars, balance patterns in 1' lowest-spread areas, 1' highest-spread bars, and 1' multiple-hsb areas) for the overall trend complete the trading approach.

This use of the 1H lowest-spread bars, 1' multiple-hsb areas and 1' highest-spread bars are fairly already explained. The pattern-description method, with volume-spread description included, can be left for elsewhere. That leaves this thread to deal with balance patterns in 1' lowest-spread areas.

---
UPDATE:

I'm not asking you to believe yet, but I can read the MT4 tick's patterns with my pattern description sketch, AND trade with it. (Sunday open update) That's worth finishing (elsewhere, as noted.)

Well, over three months have gone, and I still haven't gotten an approach. I have learned a lot about the last post's subject, and too little about mm. So, I continue from the last post.

Here's where I am currently: http://i30.tinypic.com/1z70ot3.jpg

Yes, it's messy. I'm at a stage where trendlines of the last post's type have turned valuable, and returned to this thread to have patterns reduce the number of trendlines, however effective, I need to trade.

This is where I first used patterns exactly, consecutively and only to trade (posts 1-4), and I've returned to this thread to figure how patterns can reduce the number of trendlines, however effective, I need to trade.

a one-pip-spread-bar inter-1psb approach, this version still exclusive of consecutive-1psb, or pip, lines and other spread trendlines, and this version also excluding few-1psb areas and thus language may change a little: first version


After using One to One scale on MT4 1-minute charts for a while, I found there's a parabolic effect, in which the eyes see more patterns when a parabola curves. A parabola curves a limited amount of chart to see it better or for other reason, such as with the given scale in MT4 charts. There, I made roller language and the pattern recognition table (prt).
There are patterns in 1*1 scale, only they're highly 1psb-to-1psb dependent, with a range of extra 1psb's adjusting pattern in hyper-1psb areas, to interpivot and 1psb-pivot trendlines and a mixed pattern, that is, line1 in line1-line2-line1, in hypo-1psb areas.
Thus, patterns are random and/or line-/1psb dependent in 1*1 scale, and purposeful, that is, the prt and its rl, with disjointed groups of patterns, such as stringing all of the classic chart patterns and a few others together, also possible.

Interest in patterns usually centers around isolation for a simple pivot or their being taken for granted to register the price-time line's direction, so I gathered patterns' attributes to create a baseline of understanding.



What patterns have in common:


symbol set -- a particular trading symbol tends to have only a certain type of pattern, such as a stock symbol's rl set. This is for a price-time line, such as heiken-ashi; an indicator would have its own set of patterns. For instance, a stock with a simple story that turns its fate repeats its rl set exactly at critical junctures in its existence, with only slight variation to explain the new version of its story.

morph-copy rule -- the price-time line flows, pattern-wise, from copying an unusual pattern that just formed, whether plain, flipped or reversed, to morphing an unusual pattern that just formed, and also from morphing with a simple copying. The turn point for each side of the rule's range is morphing with a simple copying, such as copying a large area that morphed and settled, and copying with a simple morphing, such as morphing a small pattern that copied and settled.

parabolic effect -- Filling the space of one's attention, such as a platform's chart window, increases the chance there is a pattern, due to the curving of the presented data to make this happen.

pattern-line formula -- Patterns arrive at much of their content when they tend to flatten, jump or drone on according to the spread trendline(s) (or other just-as-good indicator) currently valid to it. A spread trendline seems to midline or touch the price-time line; the proportion of doing and how doing this seems to carry through the lifetime of its use, and some do so with a special pattern.

recognition table -- Any pattern or set of patterns can be broken down into a representation of other patterns or of another set of patterns.




Another way to look at it is to range patterns to given indicator (such as spread trendlines).
Patterns-only get morph/copy, patterns with lines get touch/midline around a line, lines with patterns get a particular mix of touch/midline and sometimes even a special pattern, and lines-only get (the price-time line going) to/from parallel line / greater or less of an effective substitute for the sometimes-parallel line. Then just add line, pattern descriptions.

Temporary patterns may better serve bar spread trendlines. In the example below that introduces volume in with bar spread trendlines, I found this copy of patterns within the same bar channel!
I was working on a notion of small price bars with large relative volume. In the picture below, such a bar starts on the left; then, a medium-high price spread aligns with a large relative volume; finally, on the right, an obviously high price spread, over two bars, has an incredibly high volume on both bars, the latter having less of a price bar.
(I decided to draw line segments, instead of the usual rays, to reflect medium-spread price bars, which I haven't found to be especially useful unless related to a high-impact event such as news, in the formulation.)
I've done very little research into volume-price types of bar spread trendlines, especially via temporary patterns, so this is just kick-start a new path in my studies.







Small, same-size volume-and-price bars also make decent trendlines before jumps, as in the very low relative volume of each two-pip bar below. The trendlines are in ray form, because sometimes 2-pip-spread bars do above average for a while. (I didn't do any pattern study with these trendlines, although the proportions after the fact are obvious, so it's incomplete to trade with.)

I'm finding it's easier just to observe the price/volume spread, generalized-candlestick-pattern experience; curved- or straight-weighted vs. mid-lines, directional slants and pattern-capturing horizontal lines; and, morph/copy/derivative of patterns. This is to be, rather than to deal with sole patterns completely described. Sometimes, however, the 'pattern' is just a single candle, in which case, sure.
A two-pip-spread, no-shadow candle can be seen below to determine the trend's flow and major turn (ignoring all trendlines, unless written about on picture):

Wow..so many lines on one chart... How do you use it with so much lines on it? How do you know which one is more important?

Welcome to my thread. I'm still trying to get together a trading method.
Just ignore the lines not described in the pictures.

Yeah, I'd like to figure which ones are more important; it's on my to-do list. Right now, it's going with the ones with the most recent and big-candle-based (the gold line being based on a 120- or so pip spread bar, trendlines attached to top and bottom of it, crossed and the crosspoint carried to a higher-TF-significant pivot (see 'balance lines' thread in Trading Discussion)). Since the charts are in 1*1 scale, it's fairly easy to scroll along a trendline to find its origin, once they're on there of course, which helps a lot in deciding if the line is important or not.

I guess it was kind of rude of me to post such messy charts. I have been warned before. I was just kind of happy I'd found an approach to read easily-exchangeable patterns and bar spread trendlines.

Alright, fixed.

Just wondering aloud if any of my work will be more than "interesting" or "messy". Ah well...

You saw patterns a and b a few posts ago. Well, look at this. The same pattern c and d counterbalanced and flipped upon a central two-pip spread bar, to baseline the entire turn (yes, the upper level held after this chart, levelling for a while and staying within a range just above pattern d):



Alright, I'm gonna move this discussion over to my journal, where this method is summarized and actively rewritten in post one.

----

Might as well show the results. I'd midlined some horizontal lines around the basing patterns (forgot basing ray extension in first pic), which held all night long:





-----

This makes me wonder if patterns a and b also had useful basing lines. I found them problematic. Price kind of rushed in on this area, so that a lot of localized midlines formed to making the one true basing pattern, b. Only two lines looked trustworthy:



Let's see together if these two lines amounted to anything, working from, really, only one basing pattern (b).
As I went over and erased lines unrelated to this venture, I found I'd discovered, late last night, a minor basing pattern not too long after pattern b (increasing the width of the only line that really fit pattern b and the minor basing pattern):



Now that I go over the lines, I see that I'd found and now find more probable basing patterns, since I had to subdivide with smaller midlines those xy-symmetric patterns and, implicitly, their areas. Patterns a and b's lines, which, as you recall, had price rushing in on them, held through the five + basing patterns I found, through which price took it easy and went sideways until patterns c and d
This is a bit of a story. The second picture below shows more minor basing patterns. Price held onto the example of a minor basing pattern, shown in picture above, with the one line that was thickened to highlight. The last minor basing pattern was one that wrapped around this line, climaxing the thickened line that had begun to act as a centering device for all of the basing patterns it went through. This last minor basing pattern, especially designed to wrap around the thickened line, gave birth to a large jump upward.
After price settled at the new level, line 3 in pic, I found I had to return to patterns a and b and highlight their upper line 2, as it midlined with the now-lower thickened line 1 to give the new level. The new level didn't last long, serving only to be disproved with another basing pattern; 1 was the jump-off, of course, so it isn't really around in this last picture; and, 2 is the base line for everything here. I think 2 is so because this area stayed sideways through the picture.







So now this is a method worth considering. Logic all the way through. Begs for higher-TF knowledge, though -- How are you going to see the sideways market or the push into patterns a and b's area? The method could also apply to higher TF's. Finally, I drew awfully rough lines, so practice and experience help.

A first guess would be super lines, i.e. the gold line described on a chart in the above examples, which can go about anywhere; news lines, which have major and minor versions, and especially have extra lines that, as well as taking care of their specific area, have those lines extend into regular lines' areas to help super lines denote major turns and such for the regular lines' area(s); and, regular lines, which, like news and super types, go with a pattern, only the pattern is surrounded with a complex nature of lesser patterns.
This, at least, would be easier to draw up from scratch, if you went over the news of several years enough to choose the minute of each major news. Then, you'd just find super lines, which is obvious, and the regular lines happening right now, whip up the batch with appropriate bar spread trendlines, and you're good.

LINE PATTERNS

I've got carry-thru. Consecutive. Anywhere. I added all doji, looking for counterweights representing larger, isolated areas. Line segments. It works.

candle types, by pip spread: 1; solid; doji; high; other
pattern types, by #bars: 1; 1 far from 1; 1 close to 1; pattern; patterns

1,solid candle types strongest with pattern(s);
high pattern is logarithm, and they get a pip line after each and, if two close bars, a balance line;
doji's are useful temporarily at close range, and, if big, at far range;
regular, basing patterns.

Right now, I've a loose collection of line patterns:

TYPES --
a group of low-spread bars naturally copying within;
doji's or same-low-spread solid candles, each set highs, lows connected with lines, and eyed for proportion within other bars more solid or shadowed;
an entire-group-copied group of average bars, with its small trends midlining into loose patterns;
a typical logarithmic decline of a high- (or two) to a low- (or two) spread bar, with a pip (or balance) line after it

POSITIONS --
in a small lull before a jump;
two one-spread bars on a group of sideways bars;
midway between two new items,
or a row of sets of midlines logicizing between each other for a logarithmic conclusion making a pivot, between news items;
with a pattern either end of the line;
fitting the length of a trend

my next idea of getting a pattern method to work:

Temporary, trend-change / setting patterns found regularly may define each trend area, thru the area somehow copying the pattern for its area a bunch of times, and, with bar spread trendlines from, I guess, the more pivotal of them (especially news ones), guide the price-time line and thus where and how often the copies appear. The patterns are both obvious and vague; at the same time, the idea repeats indefinitely, so it may work.
The typical response to these patterns is to define and identify them. Yet, as a lot of 'random walk' traders note, this is self-defeating. Instead, one draws lines denoting the influence originating or unusual, usually small patterns have, to put between these temporary patterns.

Working on the pix. Here's a sample of line and trendsetting-pattern mixing, I hope about as detailed as it'll get:



Next, I'll include the bar spread TL's real influence, beginning with its origins (other trendsetting-patterns along the way shown; key: dark green, trendsetting; bright green, addition to trendsetting pattern's area; white, exceptional pattern, not copying or more often not, and also used for midlines and other measuring or pointing lines):

1, 2, 3, 4, 5.

Finally, the initial picture, with bar spread TL's entire influence:

Here's how I usually make my lines; I found more continuity through multiple patterns connecting with each other. I'd already suspected about news evening with prior technicals.

Last night:

1


2


3


4


Get it?
In 1, level catches.
In 2, a basing / average-spread pattern shows.
In 3, another one shows. Notice the large double-spiked average-spread pattern between them; the level sits atop this, and midlines with the first average-spread pattern's lowest point.
In 4, technicals having created room for news, news shows and adjusts to the level again.

This deals with high-spread news reactions better than just a set of hi-spread trendlines, and goes with balancing news logic performs.

Here's an update 0927est. Note the 00/50 levels as center lines ( that is, price either centers on it, with midlined trend on either side, or logarithmically approaches or regresses from it, or both, much more often the former .). I don't have any new lines in place, may have some from news; just noticed the lines above working:






-----


The above leads me to believe a revision of the old prt, which only connected stocks' rl to a sample of forex's balance patterns, is in order. Forex, I now see, has its own balance patterns and their results.

One range of examples, to begin evidence of this, follows:
There was a pattern I saw near the start of my forex trading (7 or 8 / 08) that has perplexed me ever since. (It's not just because the pattern ruined one of my (highly overmargined, for news) accounts and gave me an opportunity to make up for the margin call (a third of my account, where I was) less than five minutes later, which I didn't take, stunned it was even there.)
The last big news item before the EUR/USD fell from its all-time high offered had a really particular shape, one rarely seen: It was a corkscrew shape that was very wide on the left and, tapering the whole way, very narrow on the right, all upon a very horizontal line segment. (for comparison's sake, around the noted time: A very wide balancing pattern was in play around this center at the all-time high of EUR/USD. A simpler fall, based on a Trichet speech, happened on one or the other side.) In other words, it was both a logarithmic pattern, which hi-spread trendlines apply to, and a balance pattern, a very condensed balance pattern which encourages, with a sharply horizontal line, bigger falls/jumps afterward. It was a big, directional pattern based on news, and I'm only now grasping its meaning.

This example suggests I want to return to labelling patterns alone. No, it's a waste. I'm staying with patterns with viable trendlines. It's just that I'm slowly breaking into a comprehensive view, and maybe a table, of how these line patterns interrelate for a continuous understanding of the (1') chart.

I worked a general description of all of the spread patterns, one that avoids the less precise small- / average- / high-spread description. I want to identify the shape of the area the spread pattern TL's. So here's my first attempt, fairly untested:


A pattern is meant as two or more bars in a row or connected with a TL, usually at least three and often dozens. All spread patterns have at least some perfection, or balance, to them that distinguishes them enough to be worth drawing TL's from them. After this, they divide into two types, more perfect or more average:

Some patterns depend on being perfect, and tend to be very small or very big. They tend to be of very few bars.
The lines they project are more irregular, any repetition of the patterns of little consequence.
They depend on contrast within their pattern: Wide up and down vs. close side to side, or narrow up and down vs. far side to side.
The pattern of perfect bars is usually lens-like, that is, a small, oval pattern resembling a lens ; often, a large, frame-like pattern will form around the perfect pattern.

Other patterns are made of average bars that gather in somewhat average ways. They tend to be of lots of bars.
They project more-regular/parallel lines to elsewhere; any repetition of the patterns is highly consequential.
They depend on agreement within their pattern: Wide up and down vs. far side to side, or narrow up and down vs. close side to side.
The pattern of average bars is usually frame-like; often, a small, oval pattern resembling a lens will form within an average pattern.


A direction from either type is all in one direction or in conflicting directions. The shape of the pattern itself will show which it is.
Also, very wide lines may form from one portion of the perfect pattern, which is a predecessor of more important TL's from the entire perfect pattern: Just pick an area with a lot of one-pip-spread bars and draw a TL ray between consecutive 1psb's.



----------------------



I also worked a general description of how spread patterns fit in with other types of patterns.

In ex.:
BALANCE-MAJORITY
GU
--spread patterns, with rl build in isolated spots
EU
--some spread patterns, some rl build patterns
-----
ROLLER LANGUAGE-MAJORITY
MSFT, other conglomerate-like single symbol
--some rl, w/ rl build as energy
various one-product,other P/L-close single symbols
--rl, with rl build as energy in isolated spots

With caution on considering the below and thus viewing it only as filling in what's relevant to forex in the above table, below is an untested, unexplained organization of spread TL's (which depend on 'spread' patterns).

Spread patterns divide nicely into dot (logic dot),
short line (balance lines),
middle-length line (positioned sides),
longer-length lines (hyperdot),
and longest-length lines (balance areas).

a momo-line view of patterns:




After settle (see pic), switched to connecting highest-spread bars and putting short horizontal lines at base and top of them, while still connecting with a line the bodies of the tiniest doji/single-pip bars and including midlines everywhere to follow flow.