I was on Roguelike Radio recently, talking about game jams. I have a bit more to say on the topic.
I have a lot of ideas. Most of them are bad. Most of the bad ones I can throw away just by thinking about them for a minute, but that still leaves many I can't tell without trying them out. So this is one advantage of jamming: the quicker I can make a working prototype to test out an idea, the quicker I can throw away the ones that don't actually work and keep the good ones. And the more games I make, the more ideas I have and the higher a proportion are good, so it becomes increasingly important to work quickly.
Glitch Tank is the best thing I've made. I made it in half a day. I'd made another game already that morning. It took a few more weeks to polish and debug it, get it tuned perfectly - and some sporadic updates over the next year improving it further - but I had the core game playable in just a few hours. Zaga-33 and VESPER.5 are the games I'm best known for, that have gotten me the most attention (and attention is a valuable currency for a lone game developer). Each of them was made in a week (with a few extra weeks on Zaga-33 porting it to iOS and polishing the final version).
Couple of years ago Chris Hecker ranted about "please finish your game". He criticised the shallowness of game jams; ideas being dealt with quickly rather than explored in depth. Hecker and Jonathan Blow are partially responsible for the prominence of game jams in the first place, but they seem concerned that jams are not fulfilling the particular role they had in mind for them: prototyping new ideas and "pushing game design forward". There's a different role that jams have taken on - a community spirit, a welcoming atmosphere for newbies - but Blow contends that this environment has become unhealthy for experienced developers.
They're missing a few things.
First, I think they're making a fundamental error of measurement. Lists of games released are misleading. Large games take disproportionately more time to make than small ones, so a disproportionate number of small games released does not imply a disproportionate amount of time and effort spent on them. Most people making quick jam games have "main projects" they're working on, and giving the attention they deserve. We don't have enough time in our lives to work on every good idea we have for a couple of years.
Second, there's the aesthetic virtue of things made quickly. Jackson Pollock's "action paintings". This relates somewhat to my previous post.
Third, making a monolithic game is not the only way to explore an idea in depth. Hecker's prescribed cure for the shallowness of jams was to rent a beach house for a few days to work on long-term projects, and still not finish them. That's not a depth jam. Kompendium was a depth jam: I spent a couple of months making a game every day or two, digging deep into a focused set of ideas, mostly not finishing my games and throwing them away. My life is a depth jam: I come back to the same ideas over and over again, understanding them better with each game I make. So is Stephen Lavelle's series of Sokoban-inspired puzzle games; they climax in English Country tune but he still hasn't shaken the obssession. Bennett Foddy's endless fascination with physics simulations. Reiner Knizia's study of auctions. Hecker's dismissal of jam games treats them as though they're isolated barely-considered throwaway things; this is a strawman, practitioners like Sos Sosowski and these others have developed a consistent ethos around it.
Making things quickly is not an easy task, it takes practise. As I said in the podcast, I was initially completely baffled at how people could make games so quickly, but now it's something I've learnt to do myself. To quote Franz Kline: "spontaneity is practiced". Hecker mentioned Cactus, who is well known for making lots of games quickly, and more recently Hotline Miami - perhaps having taking the rant to heart and working on something larger. I don't think he'd be the artist he is today and have been able to make Hotline Miami what it is without having spend so much time jamming deeply. I always come back to this apocryphal story about a pottery class, comparing striving for perfection in a single piece against improving yourself by producing a substantial body of work.
I don't dismiss the value of working on large projects. The ideal approach for me is a balance between short- and long-term. Some games do just end up needing more time, especially ones that are content-heavy. Some games can be made into something more profitable by spending more time on them, and we need to generate an income somehow; but let's not make the error of conflating how much money something makes with quality. (Hecker brought up the example of Braid, and I disagree: it has some good ideas in it, but spending years polishing it to maximise its commercial potential is the single least artistically interesting thing Blow could have done. I'd rather see the dozen other games he could have made in the same period.) Also, not every game is financially worth spending more time on - even if it's interesting and deep; the sad truth is that many games don't pay for the time spent on them, and spending longer on one thing just creates a higher concentration of risk, especially when you're exploring weird ideas on the fringe.
What's the "normal" rate to do art at anyway? Videogames are an anomaly with multi-year projects being typical; a novelist writes about one book a year (plus a few short stories), a painter produces dozens of paintings a year, a musician records an album of a dozen songs every year or two. I haven't found multi-year projects to be the healthiest approach myself, regularly releasing smaller things is much more productive and satisfying - maybe several pieces a year is the "right" approach to art in general (if there could be such a thing).