Conference Record

Against the Listicle

Search for conferences in any technical field and the results are lists: the ten best, the twenty unmissable, the essential events of the year. Where the same decision also depends on how work is measured, fireable offenses at work is another practical point of comparison.

The format cannot answer the question it appears to answer, and its structural incentives are worth understanding before using one. For wider background, see Nature .

Why the format fails

"Best" is not a property of a conference. It is a relation between a conference and a person — their field, their problem, their budget, their location. A list ranks something that only exists in combination with a reader.

The list must be positive. An article titled "the ten best" cannot conclude that eight are not worth attending, so every entry gets a favourable description regardless.

It has no cost information. The total cost is the decision, varies by where the reader is, and appears in no list.

And it cannot read the programme, because the programmes are not published when these articles are written — most appear in January for a year of events that have not announced their sessions.

Where the entries come from

Search volume, largely. The conferences named are the ones people search for, which is a measure of existing prominence rather than of value.

Existing lists. These articles cite each other, and a conference that appeared in last year's list appears in this year's.

Sponsorship and affiliate arrangements, in some cases. Disclosed sometimes, and the disclosure changes nothing about the selection.

And whatever the writer has heard of, which for a general technology list is the largest and most-covered events by construction.

So a listicle reliably names the conferences you already know about, and that is its actual function.

What it is good for

Being fair, because the format is not worthless.

Discovering that an event exists. For somebody entering a field, a list is a starting inventory, and starting inventories are useful.

Names to check. Take the names, discard the ranking, and run the checks yourself — the same advice as for any list produced by somebody with different constraints from yours.

And spotting the annual rhythm of a field, which is genuinely hard to see from outside.

What to use instead

Your field's professional body calendar. Curated, maintained, and it covers events too small for a listicle.

Last year's programmes of two or three events, read properly. The programme is the product and last year's predicts this year's better than any description.

Where the people whose work you follow presented. Check three of them. This is the highest-signal method available and it takes ten minutes.

A subject librarian, which remains the most underused resource in this subject.

And the reference list of a paper you admire, which shows you where that community actually publishes.

The version of this site's own limitation

This site does not list conferences either, and that is deliberate rather than modest.

A maintained list would need constant updating, would age badly, and would put this site in the position of implicitly endorsing events it cannot evaluate for a reader it does not know. The checks transfer; a list would not.

The short version