Ventures of an ex indie game developer

Bad (manual) prediction and extrapolation

The mayhem I predicted last post is yet again postponed by yet another slow uptrend. This could go on another month, but I doubt it could last much longer.

Anyway. One thing I've been thinking a lot about is that some currencies have a much larger price variability than the commission. That means that it is easy to make a fortune if you're able to predict the stock prices just ever so slightly. The bank will make 90% of the money (the fuckers), but if you can shave ~10% off every time period, you're set for life.



Heikin Achi just makes every currency look so amazingly well-behaved. My tip to you is to run some correlation calculations on all price fields you can come up with, including Heikin Achi ones, and I'm sure you'll find some very interesting things to extrapolate from.



Python Pandas (an easy to use vector library) is soooo good at going from an idea into statistical evidence that it's ridiculous! I've thought some about the idea in the last couple of days, and implementing the calculations took less than on hour. I can't recommend it enough.

Anyway, let you know shortly exactly how good my prediction capability is. I'll almost eat one of the hats in my hat collection if this doesn't work. See, I've learned! I said "almost," not bad!

About the author

Mitt foto
Gothenburg, Sweden