I have collected the opinions of a bunch of other devs on the ideal patch frequency into a new blogpost. Here it is, including opinons from the devs of Prison Architect and Don't Starve:

Other developers on the ideal patching frequency

I have also added Orymus3's quote on the word "patch" since I found that one highly interesting. :) Orymus3: if you have made a game or want your real name or company mentioned in the post, then let me know and I will edit it. :)

My understanding with iOS apps is you have to download the entire app every time there is a patch, personally I find it annoying to have to do a big download frequently just for little bug-fixes.

One thing I've seen work is staggered development. I've had to resort to that on a project I did (white label). Basically, we had two teams, each of which were on 2 weeks sprints, meaning we ended up delivering the equivalent of 2 weeks worth every week. A lot of this hinged on getting the build approved through very efficient QA testing (which we had to supplement with 3rd party QA-ing in a different timezone).

The end-result was astonishing, but I don't know how long this strategy can be employed before the build eventually breaks, and you miss a deadline.

Staggered development only works for as long as your player base knows it can trust you to deliver on time. The minute you break that bond, (most/all) added gains are lost.

I have also added Orymus3's quote on the word "patch" since I found that one highly interesting.

Orymus3: if you have made a game or want your real name or company mentioned in the post, then let me know and I will edit it.

I'm good with anonymity, although that's a moot point since my username has been tied with at least 2 articles on game dev (which means my "real identity" is fairly easy to uncover). Given that the bulk of my experience comes from the industry, and not from being an indie (I have yet to score in that regard), putting my name up there would hardly give the thought more credibility ;)

If you can, you should really inquire to Keen Software House regarding the success they've had patching Space Engineers weekly!

EDIT: Eh... I replied with my google account anyway, so my name is written in bold in the comments section. Ah well... :P

Staggered development like that is an interesting idea, I hadn't heard of it before.

It does sound too rigid though. What if you want to make features that are bigger than two weeks of work? New characters we make for Awesomenauts take several months to develop.

I think if we felt the need to do weekly patches it would work better to just let devs develop their stuff in however much time is needed. New things would by default be excluded from the release build and then every week we would pick what we want to release that week. With a good overview of what is ready and how far along other things are I think it would be possible to select features in such a way that you have weekly new content consistently. I think this approach would work better with a larger team and various development times on new features.

This is all purely theoretical though because my main point was that bigger patches less often is better for publicity reasons.

Google: Perfect imbalance in game design

http://extra-credits.net/episodes/perfect-imbalance/

Pretty good video on the topic.

It does sound too rigid though. What if you want to make features that are bigger than two weeks of work? New characters we make for Awesomenauts take several months to develop.

Split your team in 4, and work 4 week cycles. Still deliver each week, but 4 teams have 4 different ETAs.

I think if we felt the need to do weekly patches it would work better to just let devs develop their stuff in however much time is needed. New things would by default be excluded from the release build and then every week we would pick what we want to release that week.

That seems to be what they do for Space Engineers. They just deliver that which is "done done". Sometimes, they happen to release something that is clearly a stepping stone to something else too because they are halfway through and can cook up something that emulated the final feature. Also gives them a chance of evaluating the issues with it before making a full release.

Bottom line is, you've probably tried many strategies. If you stuck with yours, then it's because it works for you, and ultimately, you need to "do what works".

Do you feel your strategy is optimal for your team, project and player base?