Keyword tracking vs. keyword research
Two different questions, built on the same scoring engine, and why the same term can show a slightly different difficulty in each one.
Two different questions
Keyword tracking answers "where does my app rank for this term, over time". Terms you track are checked by the daily collector, kept as a history, and charted. Keyword research (the finder) answers a different question — "is this term worth targeting at all" — with a single live search, and saves no history: it is for comparing opportunities before you commit to one, not for watching a term over time.
The same formula underneath
Both read difficulty from the same scoring function: the median rating volume of the ten apps actually holding the top positions, log-scaled, blended with how full the result field is. Nothing about the formula changes depending on which page you are on.