Yesterday evening the third international Digital Games Research Conference closed in Tokyo. Professor Baba, the conference chair, presented some interesting figures; if I got them right from the translation, there had been 178 presentations in the conference, and 355 conference participants. Impressive figures for a conference of an emerging field, organised in (what to a Western-dominated researcher community is) a far-away Japan. There were much discussion about how to improve the conference and about various issues that could have gone a bit better, but the overall impression was that the conference was an obvious success. One of the future decision points for the new association board and the conference working group will be the focus and size, as well as the schedule of the conference. Many of those who voiced their opinion in the final discussion were taking the point that a broad, inclusive conference like the one that was organised this time in Tokyo serves the community best: basically letting in as much papers as possible, providing opportunities for also first time authors and students to gain the experience of presenting in front of international scholarly audience. I agree to a certain degree: the international, interdisciplinary community needs a meeting place. On the other hand, the downside of inclusiveness is that those who professionally work in a particular special focus area of game studies or (interdisciplinary) games research areas, will inevitably face papers in every conference that “start from the beginning”, written by newcomers to the field they discuss much of the same issues that have already been discussed in numerous papers, books, seminars and conferences in past years. Thus, we need also some tightly select and more focused venues, where a specialist can trust to hear something that will progress the state of the art in this particular subfield in all or most of the paper presentations. Thus: a double need — served best by some single event in hybrid form, or by several smaller conferences that are each differently profiled? More thought and discussion is needed.