Monday, July 14, 2008


Now I am attending GAMES 2008, the third World Congress of the Game Theory Society being held at Northwestern. A real shame that GAMES will prevent me from visiting AAAI also in Chicago.

GAMES has about 800 participants much larger than any theoretical computer science conference I've ever attended (which was STOC 1987 in New York at about 500). Are there that many more game theorists than CS theorists? No, just that GAMES is held only every four years and has massively (up to 14) parallel sessions where most everyone who wants to talk can talk. This gets the whole community together, much like how Mitzenmacher describes ISIT, an information theory conference. The invited plenary and "semi-plenary" (5 parallel sessions) talks at GAMES are reasonably strong invited talks, though the regular sessions are more mixed.

So should we have a TCS Congress held every n years with theory broadly defined and massively parallel sessions to encourage participation to bring our community together at least once in a while? (And no, FCRC doesn't count.) Unless we have a corresponding reduction in emphasis of the other conferences, having one more conference to attend will not likely have the desired effect.


  1. Obviously, I vote yes -- theory needs a universal, flagship conference. (No, FOCS/STOC are not it. Heck, from what I've heard, arguably ICALP is playing that role currently!)

    We're currently far too split into small sub-communities that get together infrequently (or never). This hurts our sense of community, our ability to lobby for ourselves as a community, and one could argue long-term hurts our community's research output, by limiting our connections amongst ourselves.

    Never mind in an era of diminishing NSF budgets, increasing travel costs, greenhouse gas concerns, and a more family-friendly awareness in academia, having a big conference to reduce overall conference travel seems like a good idea to me.

    I just don't know how we get there, without a lot of upfront community buy-in. Maybe start by co-locating half a dozen or so conferences/workshops, like ICALP seems to be doing.

  2. Could someone elaborate as why "FCRC doesn't count?"

  3. I'd say because FCRC is a flagship conference for CS more broadly than theory -- I don't think of FCRC as a theory conference, although I appreciate that it includes several theory conferences.

  4. My guess would be that GAMES also attracts more "edge people" (to appropriate Bill's term) -- i.e., general economists, social scientists, etc. Whereas theory doesn't seem to want to reach out to people who want to apply theory (hence the sentiment that FCRC doesn't count).

  5. Maybe start by co-locating half a dozen or so conferences/workshops, like ICALP seems to be doing.

    How about collocating STOC, SoCG, COCOON, CCC, CPM, SPAA, WEA/SEA every five years or so?

    As an added bonus, since they are summer conferences we could hold them on a university campus and save even more money.

  6. A key observation in Lance's post is that "most everyone who wants to talk, can talk". That is absolutely necessary if you want to be inclusive and attract the community. Most technical talks only attract about 20 audience members, but that's OK.) But at the same time, GAMES has some big flagship talks, the plenary and "semi-plenary" ones, that are the real crowd-pullers. So, that's how to ensure that this putative TCS meeting should have some talks that most people want to go to.

    Best regards, Paul Goldberg (posting from GAMES 2008).

  7. Colocating all the theory conferences every five years might be a good plan. Although, after seeing the negotiations needed to produce FCRC (I wasn't involved, but was next door to one of the organizers), it's going to be tough to carry out.

    I don't think there's any way to introduce a new flagship conference; we lost that opportunity when we kept STOC/FOCS small and exclusive rather than increasing the attendance.