I've skimmed through Grimes' book a little and liked what I saw. I've independently taken a similar approach to TA and it has mostly worked out well for me. My

I've skimmed through Grimes' book a little and liked what I saw. I've independently taken a similar approach to TA and it has mostly worked out well for me. My personal take on TA is that Markets are psychology and TA is a slowly evolved form of feature engineering that has been finely tuned to be used by the most advanced visual pattern recognition algorithm we have, our brains. Many of these features can be tested and proven statistically, but they require additional feature engineering work to convert them from their visual pattern recognition purpose into something more suited to statistical inference. And just like any form of feature engineering, a lot of it doesn't work, but when you find what does work, it's pretty awesome.