ML has too many venues, and a flagship rejection says almost nothing about your paper. Some of the most cited papers of the decade got rejected somewhere first. Treat it as a scheduling event.
- Match the venue to the paper's shape: theory, empirical bake-offs, position papers, and dataset papers have different natural homes - Line up deadlines before falling in love with a venue; submissions cluster into a few windows each year - Use a deadline tracker that pings you; if you run a lab, a shared dashboard beats hoping students remember - Compare flagships, the strong second tier with better odds, and focused workshops where your subcommunity shows up - Workshops are not consolation prizes; half the collaborations I know started at one
VenueFit is free: it ranks ML venues by fit for your subfield and lays the deadlines out by month.