I think what happened is that a lot of links changed (the link text includes a count of matches in the link destination, and the counts have changed since new documents were published). Some ideas for being more nuanced here:
Rank changes in link URLs higher than changes in link text
Rank changes in numbers lowly (e.g. if the change was in a span of text matching the regex \b\d+\b)
Rank link changes lowly if just a small proportion of the link text changed (in most of these cases, it was one or two characters of a >25 character string)
This page got a relatively high priority (0.388), but clearly should not have: https://monitoring.envirodatagov.org/page/acbec676-e2e8-480a-a80a-79f3fc6f146f/4d2d7b8e-8770-4a17-a54b-b7399ed932a0..84729f0b-8627-4a8e-b23c-f09c81f064b2
I think what happened is that a lot of links changed (the link text includes a count of matches in the link destination, and the counts have changed since new documents were published). Some ideas for being more nuanced here:
\b\d+\b
)@danielballan any thoughts on these?