Update: I'll have a trajectory for next week posted before Monday, along with the new odds table.
I've been meaning to address what Peter mentioned in the comments about bringing back market trajectories, but in a smarter way than before. The way I started doing it was flawed in that it produced random results that were hard to decipher and told no clear story.
The new & revised trajectory method uses an expanded odds table to pick "gradient" moves in the market (just a simple mid-point of the expected range) with bonuses for expected returns over 1% (adding 1.5% etc). Just to endorse the method a little, this is what my trajectory method showed for 11/2, right before the rally exploded.
Your first thought might be, wait a second, this didn't accurately predict the path. Well, not entirely, but it does show a clear trend upwards for the next few days, which would have been a good clue to be long.
I did this by expanding the odds table to investigate the 2-5 period returns after all the analogs I'm testing and then using the mid-point of the expected range to come up with a trajectory. Kind of a sloppy approach, but actually much better than using averages because averages get way too distorted by outliers.