this seems really useful. the region at the top is obviously the largest/greenest (probably the most robust compared to the other regions).
- but what makes it robust?
- what if the largest region was a single square?
A region has to be a grouping of squares or it is not a region. If you're looking at a single green square in a sea of red, you're looking at a system which happened to catch a single large price movement at the perfect time - which the neighboring cells did not catch. By choosing this square, you would be "cherry-picking" a combination based upon a single trade which will never occur again.
You're looking for a profitable system which captures the general characteristics of the market. Specific enough to give direction; vague enough to avoid curve-fitting.
thanks for the reply.
the general idea seems to make sense. but, the issue for me is in considering a green square in a sea of red...
if you:
- use a smaller scale (ex: hours instead of days)
- or exclude the points where the red region occurs
you can make the region really big (a scaling problem).
You're looking for a profitable system which captures the general characteristics of the market. Specific enough to give direction; vague enough to avoid curve-fitting.
i assume you mean something like finding a formula for part of a line. i can see that being dangerous
- but would this include sine waves?
greetings