When I designed my 3D puzzle game 'O' I had two main intentions: to learn the webGL technologies and to learn about the broader social environment in which HTML game development exists. In service of this second goal the game has comprehensive google analytics functionality that was designed to tell me exactly who played what sections and how long it took them.
My basic idea was to actively try to promote the game across a few different channels and see if I could:
1. Just get people to play the thing I worked so hard on
2. Get useful, usable, intelligent feedback on various features of the game
Doing this wasn't going to be fun for a number of reasons: self promotion is culturally and personally loathsome to me; I have no interest, experience, knowledge or skills in social media and therefore come across as a 10 year old with a precociously impressive vocabulary when communicating online; also I'm a introvert mostly, anti-social mostly and misanthropic often.
So I made a list of sites and decided to work one each day for a week or so and see which groups paid attention to me.
About 30% of the hits on the game came with out referral information suggesting that they had linked directly to game from an emailed link - these users I suspect are mostly my friends, family and business associates to whom I sent the link. I didn't ask them for feedback of any sort and so didn't get any. A further 7% directly from this blog which suggests people from the small following that blog had recently gathered through js13k.
My plan was to provide a brief description of the game a screen shot or two and ask for comments and feedback on four specific points: the difficulty of the game, the learning curve when you start, the style of the rendering and the audio components.
I started with the social programmer places with which I am (slightly) familiar; specialist in these technologies is http://www.html5gamedevs.com/ a place for browser game developers to ask, show, help etc. In all I got 23 test plays from posting in their game showcase section, almost all of these players actually played the game over half way through - I got two comments on the game, both were constructive and extremely useful.
Locally the organization that seemed relevant was http://makegamessa.com/ a mixed group of amateurs and professionals in the South African scene whose meetups I have occasionally attended in the past. Although the site indicates that my pretty screen shots and links got 150+ views only 10 people actually played the game, most of them for a long time. I got three comments from this forum, all of these were useful and interesting, one was from an expert in audio who warmly, correctly (imho) and concisely pointed out my flaws - basically the best feedback you could ask for.
My next point of call was the sites that are specifically intended for indie game devs to showcase their work; these sites are not (mostly) communities but they exist to promote indie games and get players to look at them.
http://indievideogames.com/ has an elegant form that you fill in with all you details that makes a pretty page on their side and is promoted by a complex and widely followed social media community. I got 23 referrals out of this site from a profoundly international audience which pleased me.
http://gamejolt.com/ is a small but active community of developers and associated game artists who aggressively promote indie work. According to ther site the page only got 46 views - however I got some 20 players from this source; two of these rated the game and someone left a useful and constructive comment.
Finally on my list of things to try where the big places - Facebook and Reddit. I found these communities extremely difficult for an outsider to understand and I suspect someone with a deeper understanding of these worlds might have found some more appropriate places and have extensively cross posted.
On facebook I made a post on the groups 'Gamedev show and test' and 'Indie Game promo' - together these garnered 41 players and the 'show and test' group included two accurate, explicit bug reports (which is great).
On reddit I used this sub-reddit https://www.reddit.com/r/playmygame/ and got around 10 plays from this source. The playmygame folk gave me three great and detailed responses, specifically responding to my request for feedback (and also an accurate bug report).
I did all this because I wanted to understand and get some real-life experience with promotion on the internet, I learned some things:
1. These communities are supportive and sweet places - I expected a lot of trolling or ignorant, mean-hearted comments - the comments I got were the opposite: thoughtful, intelligent and elequent .
2. I thought is would be harder to get players - I got more than 200 people to play the game (most of whom played with it for more than 15 minutes) by just writing a few posts in communities where I am not an active participant/
3. I thought it would be easier to get comments and - Perhaps my tone was not right on the introduction texts; perhaps it's always like this; certainly I need to learn a lot more in this area. If I were to release a commercial game I would need about 10 times the amount of players to give me good solid feedback before release - this experiment gives me little hope of achieving that through public media sites.
4. People liked my game, and many played it through to the end - the comments I received indicated that the wider audience agrees with my design sense about what worked and what didn't in the game.
That was my social media experiment - I hope all the folk that played the game had fun with it and sincerely thank those who took the time to comment.