What I mean is that at a setting of 0.80/-0.80 I find that it tends to get to miss the very small timescale(daily), get it wrong on the short timescale(weekly) but get it right on the medium (monthly) scale.

This is of course very general and it hasn't been tested enough. The short weekly timescale is a shame. For example in the log here I have:



Which of course is the very opposite of what you want right!

Now, come to think of it, I remember why it caught the 70 dip - it was because I manually intervened and then used the bot to buy if anything changes.

I was hoping that I could use the EMA to catch those very large changes and basically create a less speculative more stable level but it doesn't seem to do that and just holding USD would be better at that.

Anyone had any wins with EMA?
Even though it's made losses I somehow how feel safer thinking that something less human is watching some coins for me. I'm not sure if it's too early to say experiment failed.