If you have a lot of duplicate or very similar content on the same site, without using proper canonicals to indicate which version Google should index, Google will default to the shortest URL over the longer URL.

I asked Gary Illyes from Google about how Google selects what content is canonical when dealing with duplicate or similar content on the same site, when there aren’t canonical or other signals to do so.  Normally, Google will choose the originating source to display, but when the duplicate content is on the same site, it gets a bit murkier over what Google would choose to be the URL to display.

And he confirmed it in his response.

I raised the question after noticing Rand Fishkin commenting on Twitter about an issue that Google was choosing one version of a very similar category (https://moz.com/ugc/category/whiteboard-friday; part of the YouMoz section) to display over another version (https://moz.com/blog/category/whiteboard-friday; part of the main blog), the latter being the page that Fishkin felt was the better option for Google to display.

Both pages had links pointing to them, and each page had the canonical listed as itself.

That said, if you have a similar issue and are not clear why Google is choosing one duplicate or similar page over another, the URL could provide the clue.

But the problem is solvable if you find yourself having the same issue.  Simply use the correct canonical to select one page over the other (which is what Moz is doing for the individual Whiteboard Friday pages) or employ a redirect.