A page with fewer backlinks outranks one with more because links are a single input among several, and the page with fewer simply won on the inputs that mattered more for that query. Backlink count is not the ranking, it is one contributor to it, so a smaller link total can be outweighed by stronger relevance, a better match to what searchers want, higher content quality, or links that are fewer but far more topically relevant. The upset only looks like an upset if you mistook the count for the whole score.
Relevance and intent fit do a lot of the work here. If the page with fewer links answers the query more precisely, covers what searchers actually came to find, and matches the form they expect, it serves the search better, and serving the search better is closer to the actual goal than accumulating links. A page can have a long list of links pointing at it and still miss the question, and missing the question is hard to outrank your way past with link count alone.
Link quality compounds the gap. Ten links from weak or unrelated sources carry less weight than a few from authoritative, on-topic ones, so the page with fewer links can hold more genuine link value despite the smaller number. Counting links treats them as identical units, but they are not, and the page that looks behind on quantity can be ahead on the quality and relevance of the links it does have.
So the reframe is to stop reading backlink count as a ranking score. It is one signal feeding the result, and a strong showing on relevance, intent, quality, or link relevance can move a page past a rival that leads only on raw numbers. The page with more links lost because it was ahead on the one input you were watching and behind on several you were not.
Next time a competitor’s higher link count puzzles you, set the numbers aside and compare the two pages on relevance, intent match, and content quality instead, and judge your own position by those rather than by reading link totals as predictions.