The past six months I've been conducting an experiment. I've written about it before but think I've kept fairly silent on the whole thing in "this" life. Then again, if you follow me at all you probably couldn't have missed it. The experiment was about recording games I played, while commenting on them, and posting them online. Not a new concept at all, don't get me wrong. It's just something I've been wanting to try for a long time and now I have. And concluded it.
The Experiment
It was fun, don't get me wrong. But it's just not worth it. Between August 2015 and January 2016 I've posted at least one video every day to a
fresh youtube channel. Most of the time two. The first three months were all about covering different games in the "Let's Play" category. Or well, more like the "First Look" category for most of them. The second three months were pretty much exclusively Isaac videos. Still are, I suppose. Just without comments now.
Looking back on it the First Look type of videos were sub par. I mean I'm not sure what I expected from that anyways. That's why I wanted to try them. Sure as hell got a lot of respect for people who do these videos and manage to be entertaining at the same time. It may look/sound easy but it certainly isn't. I had no idea what kind of level of videos I'd produce or what would be involved.
Time
I think I underestimated the time taken though. You may think a video of 30 minutes takes that; 30 minutes. But there's setup involved. First I need to pick a game I think will be fine to record. Then setup the recording software, mic, and everything. Some games are a real PITA to even record at all. That may in part be due to the million different ways you can do stuff in Linux. On that level I'm amazed by how well OBS does this. But the setup is there and it takes time. Especially for different games you have to calibrate the mic and volume levels every time.
Then there's recording, of course, which usually takes as much time as the video runs for. Sometimes more if you mess up or whatever. Theeen.. add the small review to the website (
played.today) in the backend, regenerate the site (in nodejs) and upload the statics, upload the video to youtube, setup all the fields because youtube doesn't feel like helping you much in that area. Seriously youtube, what's wrong with you anyways. Eh, so fill in all the fields that need filling in, schedule the video to be released at a certain time (I tried to keep this at a fixed time, not that it mattered much), and wait. Then I'd cross my fingers while the video is being processed and hope there's no content ID match. The matches were usually "right", in that it's indeed that piece of music. I'm pretty sure that whatever music plays in these games can be shoveled on the big pile of FUP, but nobody gives a shit about that. Certainly not youtube.
Content ID Scrubbing
When there's a content ID match I'd have to scrub the video a bit. I record mic and background in separate tracks so that I can use various linux tools to erase part of the background track while maintaining my mic track. That worked out pretty well and even though it's a bit tedious at times, it's worked out for most videos. I only had to fight a claim once, backed up by the devs, and prevailed there. From what I hear it's really just a crapshoot and if the protest is rejected, you're shit out of luck and marked for life. THAT SUCKS, YOUTUBE. So yeah, some videos cost a bit more work but I believe keeping the account clean is worth that trouble.
Syndication
Then once the video is released you post it to various networks. Youtube can automatically post to google plus, twitter, and facebook. It's kind of meh I suppose but I set it up to post to
twitter and google plus. Not sure whether that was worth anything. Initially I also posted them to a subreddit but that just got drowned there and I stopped bothering to do so. I would also post them to
a steam curator group and as a review on the steam store page for each game. I think that ended up getting the most views, though still nothing to write home about.
So yeah, all in all I would say it takes about 2 hours to post and spread a 30 minute video. More if there's a content ID match. And that's without any kind of post production involved.
Analytics
According to the analytics provided by youtube the first three months got about 6000 1600 views, with 5800 minutes watched. Though about 769 of those views (and about 2150 minutes) were a timely posted, though very bad, steam controller video. The second most viewed video in that timeframe has 40 views and 270 minutes. So you know, shared between 160 videos I'd say it was not quite worth the effort.
The second three months, consisting on nearly exclusively videos from "The Binding Of Isaac: Afterbirth", has about 17000 views and 22000 minutes watched. Noticed the low ratio? These are 30-60 minute videos each, yet on average they viewed less than two minutes. The reason is that the majority (14500 views, 12400 minutes) originated from a funny event in one of my Isaac videos that went viral on the subreddit. Kind of incredible and shows clearly that there's plenty of audience (that's just the tip of the ice berg) but that I'm not going to be drilling into that any time soon. If you leave that video out we end up on some 420 videos amounting to 3000 views and 10000 minutes. A slightly better ratio, considering. But still bad I guess.
Revenue
It should go without saying that the amount of money earned over the entire period literally doesn't even buy me a cup of coffee :p But that's okay, no problem at all. It wasn't meant that way. I mean hey, if it had led to a 7 figure thing then that'd be allright too. But I've got a day job and it's fiiine. Just an interesting insight into this shared revenue stream thing. I should add that I disabled all ads as much as was allowed, I figured ads would "scare" away potential viewers more than they're worth. I still think that was the right choice.
So now what?
Well I'll keep on producing videos. Just no longer on a tight schedule. Or any schedule, honestly. I'll do Isaac dailies when I can. But I won't comment on them anymore. I'll rant about that in a different post but it comes down to feeling that the devs are neglecting the community and not taking the daily as serious as they could be. Ought to be. Man I had plans for that game. So I'm giving up on that pipe dream and are just playing dailies for fun now, posting them for the record and without further comments. That also means I can record them at any time without having to clear the house to use the mic.
First Play to Let's Play
I'll stop with the First Plays and move to a Let's Play in review mode. That means playing the game first, getting to know it, and then doing a review play. Some of the PT videos I've posted so far were of those kinds and I think they were of a better level than the "uhhh, ok what is going on in here oh hey there is the exit and why is..." level the Firs Plays tended to end up like.
In all fairness, the top ten of PT in terms of stats was that viral Isaac video, the steam controller review, Vangers (a first play of a weird game from 1998), Tomb of Tyrants (a review), Bridge Constructor Medieval (a review because I had botched the recording of the first play), another Isaac video (because reddit), Siralim (first play), Last Warlock (first play), Isaac (the first special daily, IIRC), and A Wizard's Lizard (a review). I conclude that virals are great
(d'oh) and reviews are better received than random first plays.
So the frequency of review videos goes down and hopefully the quality goes up a little. For Isaac it probably is what it is now; as long as the community gets no attention I feel no need to put more effort into that either.
Conclusion
In conclusion, recording game streams is fun but kinda exhausting. If you really want to break through you either have to offer something significant in terms of entertainment, put more work in post production (to get said entertainment through alternative ways), or be more engaging with your audience through live streaming platforms like twitch. Again, the keyword is entertaining although I feel less so there since the engagement is also very important. I simply chose not to spend time in that area which I'm sure (but still assuming) is a big reason why view count is what it is.
For what it's worth I've persisted in posting one or two videos every day from the start to the (arbitrary) end over a period of six months. Mind you, this includes being abroad for two weeks for work and becoming a dad in the same period. No regrets, I had fun and something to show for it. PS. see
website and
youtube.
Now it's time to move on. Spend a little more time in personal programming projects because there's one or two that have been wanting my attention for a while now. There's also the JS1k competition which still has about two weeks to go and which takes up its own slice of time. Oh and I started writing paid blog posts, as you may have noticed. Finally there's a conference talk in June I have to prepare. But with this streaming biz out of the way I'll have more time for all of that. Well, and just work and spend time with my family too, of course.
Busy busy, but life is good :)