I really think this is harmful for a research community. I know some security-privacy venues have even a lower acceptance rate. I believe increasing acceptance rates will also decrease some of the randomness.

Discuss. https://twitter.com/martinkl/status/1351934593293840385

Wanted to comment on "... decrease of randomness". In my experience, there are three equivalence classes of submissions [1] clear accepts [2] Ambiguous set [3] clear rejects. Most of the randomness comes from [2]. Often it comes down to luck (e.g. some reviewer fights for you).

Another addition. As we take more papers, the randomness in [2] decreases. Of course, this is anecdotal. This is based on my experience from chairing some major conferences.