“Understand your own buying and selling policy well. Past verifications should be minimal.” (From Pan Rolling’s “The Psychology of Magicians”)

A phrase written under the title "Biases that affect system verification."

A longing for freedom. I want to optimize the system. And the more I tweak data to fit past conditions, the more it feels like I understand trading well. More than that, it's important to thoroughly understand my own trading approach. Past verifications can be minimal.”

I develop EA, and I feel that the above words apply well.

In long-term backtests, I optimize to achieve a clean upward trend, but

I forget the original principles and just try to make it rise to the right,

tweaking the indicator parameters over and over.

As a result, I fall into the trap of curve fitting, and performance does not carry over to forward testing…

Backtest results look as good as possible due to human nature, doesn’t it?

However, in exchange for that, I may be losing something.

If the basic policy for the EA I build starts to waver, that’s not good.

With the fundamental principle of “truly making a profitable EA,”

I will continue development without wavering.

Well then, until next time.

[Promotion] We are selling a single-position EA. Three years of forward performance.

If you have a creed of leaving things alone, it's fine (laugh).