Ranking Academic Conferences Solely by Buffet Quality (2025 Edition)

Every year, the academic computer science community engages in a ritual that reveals more about our collective psychology than any study we have published: the ranking of conference buffets. We claim to evaluate papers on originality, rigor, and significance. In reality, we evaluate conferences on the quality of their lunch spread.

I have attended 23 major AI/ML conferences over the past decade. I have reviewed 847 papers. I have eaten 23 conference lunches. Of these, three were memorable. The rest were a beige continuum of unidentifiable protein sources, wilted salad, and bread rolls with the structural integrity of peer-reviewed software documentation.

A Proposed Taxonomy

After extensive empirical observation (and four years of a PhD that permitted me to call eating “field research”), I propose the following classification scheme:

Tier 1 — The Grant-Funded Experience: Linen tablecloths. Three courses. A brief moment in which you forget you are attending a conference about gradients. These venues exist primarily in Europe, where conference organizers have not yet internalized the American belief that intellectual stimulation is sufficient nourishment.

Tier 2 — The Adequate Spread: A reasonable buffet. Enough variety that dietary restrictions are technically accommodated, though the vegan option is clearly an afterthought, assembled in the final hour by someone who Googled “what is vegan” immediately before service.

Tier 3 — The Workshop Experience: Sandwiches in plastic wrap, assembled before dawn and left in a hallway adjacent to the registration desk. Temperature ambiguous. Structural integrity of the bread an active research question.

Tier 4 — The Virtual Conference: You ate whatever was in your refrigerator, alone, while listening to a presenter whose audio quality suggested they were speaking from inside a running dishwasher.

Findings

My analysis reveals a statistically significant inverse correlation between the prestige of the conference venue and the quality of its catering (p < 0.05, but I’m not telling you what p-value I started with). The most selective conferences, held in the most expensive cities, reliably provide the most austere meals, on the grounds that the honor of attending should be sufficient nutrition.

This is consistent with a broader pattern in academic life: the greater the privilege required to participate, the less materially that participation rewards you. I leave the implications as an exercise for the reader, the reviewer, and whoever approved the catering budget.

The author declares no conflicts of interest with any buffet operators, though he has accepted free coffee at several industry-sponsored workshops and considers this spiritually equivalent.