if that was difficult to parse, the idea is that the greater the number of points of contact, the more robust the trendline is.
this is because, as frozen pointed out, any line at all can be drawn across one point, two points is enough to define an arbitrary unique line, but three points is a 'coincidence', or in statistics terms, a correlation.
i admit that the lines above only have 2 points of contact. the evidence for their robustness is in the price movement right now.
Not difficult to parse. Only arbitrary.
The 260ies marked the on-trend inevitable climax (which had to happen somewhere, and happened to happen there) of of the trend, not an outlier of that trend.
Discarding the last, crucial day in favour of the day before (which was not an outlier? if so, why?) seems disingenous.
As Frozenlock points out:
But why aren't you taking the high of the $265 candle?
You can make many, many shapes if you can choose any length of candle.
that high is very obviously an outlier.
the process of drawing the 'correct' triangle is as follows: identify the possible top and bottom trendlines with the greatest number of points of contact, disregarding outliers.
we can even disregard the highs that break robust (multi-point) trendlines because that itself is evidence that they are outliers, based on the theories of triangle consolidation patterns.
...
as you can see, there are many possible triangles. determining which is the best fit is a messy process that i'm still refining.
You are saying you can use any wick and/or candle, without your method providing guidance.
And proceed to arbitrarily discard any inconvenient data, after you've found a pattern you like.
It is this need to discard data you don't like for the pattern to hold that I find worrying.
(I've been asleep, good morning) 
