Why does Google Scholar's "Sort by Date" feature give you a completely different number of papers? I'm not speaking about approximation, like here, I'm speaking of real numbers being 1000 times different.
If I search "genome sequence homomorphic encryption" (any time, sort by relevance, all the words,) I obtain about 5,030 papers. If I change to sort by date I obtain... 3 papers. Literally 3. Google Scholar changes my query to "genome sequence "homomorphic encryption" " (I mean, it puts "" before and after homomorphic encryption, searching for the whole expression instead of separate words). I can manually change the query to be the same as before, by either changing it directly or crafting the URL, but the number of papers remains 3.
I can change to "Everything" instead of "Only abstract" (I don't get why this option appears only when I set "sort by date", but whatever) and I get 724 results (or 596 if I use the modified query it automatically does).
I mean, it is not an approximation, it is simply a different set. Completely. As far as I can tell it simply takes the papers from the last 364 days, but I don't know for sure.
Moreover, the results aren't accurate. With sort by relevance, by adding the constraint "Since 2021", I get 281 results. Among the 724 results I obtain using the "sort by date" search, around 215 (and surely less than 220) are from less than 112 days ago, meaning that I have lost about 60 results in this way (it is more than 25%, it is huge).
Why? Like, it is completely idiotic. It should be very simple, the normal research contains the date, it is a very easy sorting to do. It is not like it does not know the date or it is unable to find it. And, more importantly is there a way to change it/use a workaround?