We've spent the last fortnight checking the VCP model book by eye. Every setup the rules had ever found was measured against the textbook — a flat top, successive tighter contractions, a pivot under real resistance — and more than a hundred were then judged one at a time, with the chart cut off on the day the rating peaked so the outcome couldn't colour the call.

Two things changed as a result. First, V-shaped recoveries are out. A base whose later swing highs never come back near its high isn't a VCP, and nine in ten of them failed the eye test, so the model book and the scanner now hold flat-top bases only. It's a smaller book — a little over half the size — and a more honest one. Second, a 5 rating now has to be earned by the structure: a flat top within a percent of the base high, or three or more contractions. Rating 5 is now the cohort most likely to be a real VCP, and the most likely to clear its pivot.

One finding is worth knowing as a trader. How textbook a base looks predicts whether it breaks out, not how far it runs afterwards. What predicts the run is relative strength, an actionable pivot and how extended the stock already is — which is what the rating has always weighed most. Every setup now also carries an E badge once the right edge has turned up off its last low: the point at which a pivot is buyable rather than still falling towards you.

The model book and scanner are part of the Pro plan. The outcomes-by-rating table on the model book's analysis page has been re-run against the new book.