Twitch Reflexes

Like a lot of older people, the very concept of Twitch has been somewhat baffling for me. Why would I spend my precious free time watching someone else play video games instead of, y'know, playing one myself?

Over the years I've come to understand it more. For certain games, like competitive player-versus-player ones, it can be really exciting to watch highly skilled players performing well; watching those, I realize one could just as easily ask "Why would I spend time watching televised basketball instead of, y'know, playing basketball myself?" Beyond the content of the game itself, much of the appeal in Twitch lies with the broadcasters, and watching a stream is often at least as much about their performance as with the game they're playing. That used to always turn me off when I would watch streams of games I liked, as it can seem very distracting and undercut the content of the game; but just as often it can enhance my enjoyment. Maybe the streamer finds something that I'd missed, or can offer some funny commentary, or does fun impromptu voice acting for a text-only game.

In the last couple of months I've come to realize that Twitch can be a phenomenal tool not only for game publishers, who notoriously use the platform to advertise and promote their existing games, but also for game developers, as it gives access to see how real people are actually playing your game. In my case, I've been able to watch people play through my Shadowrun mods. This is the first time I have ever seen another human being actually play my games, and it is a fascinating, illuminating experience. I get plenty of feedback from my players, via comments on my Steam Workshop pages or direct messages or blog posts, but they're inevitably condensing their experiences of playing 10+ hours into a couple of sentences. Watching what actually happens during those hours is a whole other level.

One impact is that I feel a lot better about giving players copious hints and directions. I provided more and more of these with each new campaign, and would often worry that I was "dumbing down" the game too much, over-communicating their goals and objectives. But, many players tackle these games in chunks: they're definitely not playing through the whole campaign in one sitting, and probably not over consecutive days. A Twitch streamer will typically play once a week for a few hours; in real life, people might go even longer between sessions. It's actually helpful for me to hear someone at the start of a stream say something like, "We're in San Francisco, but I forget why, I think we're supposed to meet someone or kill them or something?", and then over the next hour or so reconstruct the current state of the game. This is also an argument for keeping plots simple: not dumb or boring, but I think that if you make a few vivid things (colorful personalities, high-stakes problems, anything to stick in a player's mind), it really helps orient them.

One of the mortifying experiences I had came a few months ago when I was watching a player attempt to navigate the Sacramento map in CFiC, one of the more puzzle-and-exploration-heavy scenes in the game. After many agonizing minutes had passed, I started to yell at my monitor: "Click the icon! It's right there! It's glowing and blinking! The compass is pointing right at it!" I watched as he paused, walked his character over, see the mouse slooooowly move towards the icon... then he would turn around and wander over to the other side of the map. Stuff like that keeps me from wanting to stream myself: not being able to solve a puzzle feels kind of embarrassing to start with, I don't think I'd be able to handle knowing that other people are watching me not get something.

For the most part, though, it's been encouraging to see how rarely players get stuck. Most of the riddles and similar things work like I would hope, with the player mulling things over and maybe failing a couple of times before figuring it out. Other times they brute-force their way through, which is fine too; I'm sure that for both players and their viewers, it would be annoying to get stuck on a single part of the game for very long, and in general it seems like even the more obtuse puzzles in Antumbra Saga don't hold them up for very long.

Another unanticipated finding has been how many players almost exclusively use the default attacks: the basic single-bullet attack on a gun, Powerbolt on mages, and the basic deck attack in the Matrix, never using more powerful cooldown-limited abilities like Aimed Shot, Flamethrower, or Killer. In my own playthroughs I make heavy use of these, using Aimed Shot any time it's available and a given shot has a 50-80% chance of landing, and using Killer or Flamethrower any time an enemy has more health than my default attack. Not using the special attacks makes each fight longer and harder than I planned while designing the combat, though this is mitigated since the streamers seem to be playing on Easy mode. This is one of the few cases where I would genuinely like to interview the players and figure out why they so strongly prefer the basic attacks. Are they deliberately holding back the special attacks so they would be available if needed? Are they trying to keep things fast-paced for their viewers (a single click to attack is quicker than one click to select a mode and a second click to attack)? Do they think the special attacks give themselves an unfair advantage over enemies who eschew them? Are they unsure what they do? I'm now curious to see what those same streamers would do when unlocking custom attacks in the Caldecott Caper or CalFree in Chains: if there's a small plot and some dialogue introducing the attack, will they be any more likely to use it than if it just appears as an option in the UI?

There's a ton of useful information in these streams, and I've even pushed out a few new updates as the result of what I've seen: typos in the dialogue, clarifying some plot points that were confusing, noticing that some events wouldn't properly fire in all cases. I'm already thinking ahead to a hypothetical future (non-Shadowrun!) campaign and wondering if future beta testers would be willing to privately stream their runs; my existing system is great for identifying game-breaking bugs and the most important feedback, but there's so much more that I get from seeing people actually, y'know, play the game.

That said, the experience has also reminded me that, my best intentions aside, I can be kind of thin-skinned when it comes to criticism of my work. For the most part streamers have been very complimentary and excited, and it's a very fulfilling thrill to see others derive joy from something I put many, many hours into creating. But when I hear complaints, even made idly in passing, I instinctively want to put up a defense: "This part is supposed to be hard!" "It will all make sense in an hour!" "That's because you didn't talk to anyone back at the bar!" "You could have sided with the other faction!" "The engine doesn't support that!" Of course, I'd never actually pull a Brett Stephens over this; it seems petty and creepy for a creator to argue with players. And my focus should be on the experience my players are actually having, not the one I wanted them to have. I do think that, even for smaller dev studios, this is probably the sort of thing that could benefit a lot from some mediation: an intern or professional sifting through this raw footage, synthesizing the common, valid feedback, and propagating it up the studio as actionable feedback.

By far, my favorite part of dipping into Twitch has been watching players have emotional reactions to the stuff I've made. My entire motivation for creating these campaigns has been to instill a certain feeling, hit certain emotional beats, and I put a lot of thought and effort into reaching those points. The written feedback I've received so far has been really encouraging and validating, helping me recognize that at least some people are connecting with my story. Actually seeing it is even better, and makes all the long nights of work feel worthwhile.


It does feel a little narcissistic to focus on my own games like this, so I'll probably try and wean myself off of these particular Twitch streams in the near future. But it's been a very illuminating experience! I've gotten a ton of useful feedback for the games I've made, have made some discoveries about the different ways people play games that will help inform any future projects, and have some new tools that may help with future development and playtesting efforts. None of that is what Twitch was originally designed for, but it's a pretty darn good reason for it to exist.