I'm not much of a mean reversion guy.

My preference is momentum and trend systems.

But I understand the importance of having mean reversion in the portfolio.

Even if a mean reversion system isn't amazing on its own, its capital efficiency and uncorrelated return profile make it extremely valuable when combined with other system types.

The performance of the individual system doesn't matter in the grand scheme of things.

As long as it's uncorrelated and makes money during different times than your other systems, it can add to the portfolio.

Stop optimizing for the best standalone backtest.

Start optimizing for how each system improves the combined result.

That's the shift from system thinking to portfolio thinking.