Fox Studios logo Fox StudiosGame projects, devlogs and Scratch builds
Community

How to Host a Small Game Jam for Your Own Community

The first jam I organised had 34 people sign up and four submissions. I had picked a 48 hour window that started Friday evening, posted it eight days ahead, chosen a theme I thought was clever, and then spent most of Sunday watching a page...

A game jam page on a laptop showing a countdown and a gallery of small submitted game thumbnails

The first jam I organised had 34 people sign up and four submissions. I had picked a 48 hour window that started Friday evening, posted it eight days ahead, chosen a theme I thought was clever, and then spent most of Sunday watching a page that stayed almost empty. Two of the four entries came from people I already talked to every day. One came in an hour after the deadline and I had to decide, on the spot, whether rules mattered more than kindness.

Four jams later the numbers look different: 61 sign ups, 23 submissions, and a feedback thread that ran for nine days afterwards. Almost none of that improvement came from better promotion. It came from fixing the parts of the format that quietly told people not to bother. If you are thinking about running one for your own group, here is what actually moves the submission count, and the one piece of standard advice I would now ignore.

Sign ups lie, submissions are the only number

A jam page with 200 interested people feels like a success before anything is built. Across the four jams I have run, the share of sign ups who submitted anything went 12 percent, 19 percent, 31 percent and 38 percent. The industry wide figure for open jams is often somewhere near 10 to 20 percent, so a small community jam can do much better, but only if you design against the three reasons people drop.

The first reason is scope panic. Somebody opens the editor, imagines what the other entries will look like, and closes it again. The second is the lonely middle, usually the Saturday afternoon of a weekend jam, when the project is ugly and nothing works. The third is submission friction: a form that wants a cover image, a trailer, a build for two platforms and a description, all at 11 at night.

Everything I changed between jam one and jam four targeted one of those three. The theme matters far less than any of them, which was humbling, because I spent two weeks on themes for the first one.

Length, and why a weekend is usually the wrong answer

Here is where I disagree with most jam advice. The standard recommendation is 48 hours, because that is what the famous jams do, and because a short window creates urgency. For a small community with school, jobs and timezones, a weekend is the format most likely to produce four entries.

A 48 hour jam demands a specific weekend be free. In a group of 60 people, maybe 15 have a clear Saturday. My third jam ran for nine days, Friday to the following Sunday, and submissions tripled. People worked in two or three sittings of a couple of hours, which is how anybody with a life actually builds anything.

The counter argument is that long jams let scope grow until nothing ships. That is real, and the fix is not a shorter clock. It is a size limit stated in the rules: one screen, one mechanic, or a stated target of about three minutes of play. I now write a line into the jam page saying that anything over five minutes long is probably too ambitious for this jam, and three people have told me that single sentence is why they finished.

A timeline that has worked four times

  1. Three weeks out, pick the dates and say them publicly before the rules are finished. Dates spread, details do not.
  2. Two weeks out, post the full page: length, theme reveal time, size guidance, what the submission needs, and whether late entries are accepted. Answer that last question in writing, because you will be asked at the deadline.
  3. One week out, open a sign up thread and ask one question with it, something like what are you hoping to learn. Reading those answers tells you what the feedback thread should focus on afterwards.
  4. Reveal the theme at a time that suits the most people, not at midnight. I use 10 in the morning on the Friday, which lands in the evening for the furthest members of my group.
  5. On day two or three, post a visible progress check and share your own ugly screenshot first. This is the single highest value hour of the whole jam.
  6. Twelve hours before the deadline, post a reminder with the submission link and a one line checklist. My late submissions dropped to zero the first time I did this.
  7. Keep submissions open for a 24 hour grace window and say so in advance. Writing the grace into the rules removed the awkward judgement call I had to make in jam one.
  8. Run the feedback round for at least a week, and ask every entrant to play at least three other entries. Make that an expectation stated up front, not a plea afterwards.

Step five deserves the emphasis. In my second jam I posted a screenshot of a character falling through the floor with the caption that this was the state of things, and seven people replied with their own broken builds. Submissions from that jam were nearly double the first, and the only structural change was that mid jam thread. People quit when they think theirs is the only mess.

Themes, prizes and judging

Good themes are open enough to allow wildly different reads and narrow enough to mean something. Two words beats one. Broken clock produced a platformer about rewinding, a typing game about a repair shop, and an unsettling thing about a town where time stopped. A one word theme like water produced mostly swimming.

Prizes change behaviour, and not always the way you want. The one jam where I offered a small cash prize, about 25 dollars, drew in four entries from people who had never spoken in the community and vanished afterwards. It also made two regulars tell me they did not enter because they did not want to compete. Now I give nothing but attention: every entry gets a written paragraph from me, and the top three get their game featured for a month. The engagement afterwards was better than the money ever bought.

On judging, let the community rate but publish the criteria beforehand. I use three: does it work, is one idea clearly carried through, and does it feel finished. Note what is missing from that list, which is graphics. Judging on polish punishes people learning to code, and a jam for a mixed skill community is the wrong place for that. Our notes on giving useful feedback on other people's games are worth sharing in the feedback thread, because ratings without comments teach nobody anything.

What goes wrong, from someone who caused it

I once set a deadline without naming a timezone. Three people submitted after what they believed was the cutoff. Write the timezone, and write the hour as a number, not as midnight, since midnight on the 14th is genuinely ambiguous to half the people reading it.

I also underestimated how much work the week after the jam is. Playing 23 entries properly and writing a paragraph each took me about six hours spread over four evenings. If your group is larger, recruit two people to help before the jam starts rather than appealing for help when you are already behind. The burnout pattern there is the same one that hits anybody keeping a chat community alive, which we wrote about in running a Discord server without burnout.

And if nobody has done a jam before, point them at the participant side first. Half my eventual regulars entered their first jam somewhere else before they trusted mine. Our guide on finishing your first game jam covers that side, and the rest of what we publish about groups and threads sits in community.

FAQ

How many people do you need before a jam is worth running?

I would run one with a dozen active members. My first useful jam had 34 sign ups and produced four entries, which was still four games that did not exist before. Below about ten active people, a shared build week with no theme and no deadline works better.

Should the organiser enter their own jam?

I do, and I exclude myself from ratings. Building alongside everybody gives you something honest to post in the mid jam thread. The risk is that you end up with no time to answer questions, so keep your own entry deliberately tiny.

Where should submissions actually go?

Wherever your people already are. A hosted jam page on a game site gives you ratings and a tidy gallery. A thread with links works fine for under a dozen entries. Do not make people create an account on a service they have never used in order to hand you a file.

What about teams?

Allow them, cap them at three, and ask teams to say who did what. Uncapped teams produce one entry that outclasses everything else and makes solo entrants feel pointless, which costs you more than the good entry gains you.

DF
Diego Fenn

Diego covers finished builds and the community around them, and runs the feedback threads that shape what the studio makes next.

More posts by Diego

More in Community

01doorganico 084goad 3hot8 4465456 after5internprogram ahandfulofblessings ahmadpc ahomcybersecurity allodialresources annodex antiquesandperiodcollectibles aotc asianworldkashmir beedlectrics bloggertemplatefree bluefeatheragency bodyweightsolution campaign08blog carfixadvisor catstreetcollective cbstvsports cnfanspurchasing committeeof500yearsofdignityandresistance covertopsdjs cs-kaiyuangame cuestacomunicaciontotal cxoutlet dancebetgame deskgearhub destinyvillastay detectmagazine di33 diamondsteelbuildings donmo dyanadropship emarsysfinance emimattl escapechan eskitarih evautoplay everythingweddings-orlando farmzone feedviewerapp gamevui123 gleasongrade greetingsday hacklabasu haskell iw-kaiyun jagoanproperti jilidskripsi joltbusinessfinance juegodepokergratis juegospucca kaeru-festival latestmehndidesigns lestanzediottaviano litemagazines localcommunityhero lustrsazan mariashebets menswheels newwebstechs1 nikkit ootheday oxigenvpn plurfitness prefetial probudgetcarpetcleaning psychicmrsmartin pyaten qvantor rejoignezlogis rexzbot riselayouts ruchin rwxqsgd5 sakaltime shibiaowang shop-bot skrtips spotlesshaven strcoding tetapatrium topselller uneplanteartificielle vipertechcr webtech1223411 wo-kygaming yamaha-mekar yeknava yoybuysheets zenunwa yamaha-mekar jilidskripsi ankertoto ankertoto ankertoto