Footon is right. If you compare the short trade of the first strategy and the short trade if the "short only" strategy you may see that the trades are different. The execution is not the same by two reasons:
- some long trades may disturb eventual short entries and vice versa.
- "Reduce" rule will not work in "one direction only" strategies.

...

Leon, you have a great example of an over-optimized strategy.
I bet the pessimistic method is not the worst one in that case. Can you post a Comparator screenshot? I assume it will look like a peacock's tail.

You have two separate strategies with non-matching trades, care to explain the logic behind having the same number of ambig. bars?

No problems, sorry if I sound strong, I'm not shouting or anything. I try to keep the contents high with less words, seems more constructive.

You can achieve two-side strategy if it closes at bar closing or it reverses at opposite signal. Essentially every signal has to be used in the original strategy, then if you separate it to long and short strategies, you have mirrored sides.

@Leon: Only when seperating long and short you have also the possibility of running a long and a short trade at the same time. This opens completely different strategie possibilities. See a clear example (from Topic: Long and Short independent at the same time) attached.

@ hannahis
The idea is simple and based on following main hypothesis:
- no one knows which direction the next move will be, but it will fluctuate up and down
- all indicators are based on the past and cannot predict the future for sure
- often a trend in one direction starts only after going first in the opposite direction
- looking at any point of the chart you will often find a long as well as a short target which can be reached, but not at the same time.
- Clever entry points for long and short are independent from each other and the trades may run completely different time span.
So the idea is to independently trade both directions.

Hello Bru1,

It is very interesting idea to separate the long and the short trades. The main reason for that is that the trades in the one direction will not prevent opening in the other direction.

I think that we can develop two symmetrical strategies - one for long trades and one for short trades. The only difference between them to be the direction of the "Long or Short" indicator.

This is what we can do:
- set "Long and Short" indicator in the Editor page for "Open long positions only".
- lock "Long and Short" in the Generator and generate
- return to the Editor and switch to "Open short positions only"

Our goal is to find a strategy that works in both direction. A side benefit from such approach is a reduced danger of overptimization because we search a "long" strategy and confirm the results with a "short" one without any manipulations.

I tried that for 2 minutes and found one example. It is not perfect, but shows that there are chances.

Attached images for long only, short only, and combined trades.

Bru, indexes are different than the fx rates.
For an index we have a product with price. We have clear separation between bears and bulls. It is especially true for the commodity market.

On the other hand, the fx symbols are symmetrical. We don't have bulls and bears because  a bull for the base currency is a bear for the underling currency. We don't have a "side".
I think a fx strategy must be absolutely symmetrical.