I ran a Kolmogorov-Smirnov two sample test. The result was a dissimilarity of 47%. If you normally trade on Wednesdays, then you know it has a reputation for being "whipsaw Wednesday" with choppy action. Today's response to several weeks of lower volume and a high CPI led to a huge spike in price and volume. The Standard deviation in the two sets shows that the model is typically flat (CV - .3%) whereas today's actual data has a CV of 1.5%, which means the data points are further spread apart (which represents the huge price action we saw today.
Because the Dissimilarity was so high, it is likely I will exclude this data in the model and likely hold onto it to build a separate pattern. Or in this case....big news Wednesday pattern -- CPI data/etc.
This day's match was a bust, but that was fairly obvious after the 1000 AM swing point. The morning open was a good match though with a strong move up, which shows the model continues to have good morning reliability, and the chop zone held true to its word as well.