It basically back tests (checks against historical data) various versions of popular evergreen portfolios. These are portfolios like the Permanent Portfolio that are designed to be put into place regardless of risk tolerance and not be adjusted. What I found most interesting about these is that a bunch of these portfolios are VERY different. But despite some of them being that different in make-up, the actual long-run performance doesn't change that much. Sure, there are some objectively better and worse portfolios in here (worse means the same or lower average return for the same or higher standard deviation than another). There are some that not better or worse, just different (appropriate risk and return trade-off, but different volatility levels).
It seems to me that this tends to reinforce the idea that if you just get good, low cost, relatively diversified investments and then leave them alone you are probably going to do pretty well. That spending a lot time trying to pick the perfect stocks has nominal impact on your total returns. Spend the time billing another couple of hours and you'll likely be better off.
Most of you know I don't track market performance on a daily basis or anything, but I still like to keep track of where the market is going in general. For the most part I use my good friend Matt's site Model Investing to do that. He uses an algorithm he built that is based on the ideas found in momentum investing. His site can explain it better than I can, but it has served me pretty well in understand where the market is heading.