Rhythm and Readability: Why Bubsy: Paws on Fire! is the Best Bit.Trip Runner

This is kind of two articles in one. On the surface it’s about readability in rhythm games but the seed behind it is the confusing arc taken by the follow-ups to Bit.Trip Runner. So I’m gonna ramble about those games for a bit here.

The first Runner was an experiment - as far as I know, it was the first notable rhythm-based auto-runner. But it did a good job establishing the formula and in particular it didn’t have the readability problems that plagued its sequels. Its biggest crime was that some of its levels got quite long and had no checkpoints.

Runner2 was clearly an attempt to expand and improve the formula. Some of these changes were fine - more mechanics, more level environments, more playable characters and outfits, etc. - but some were problematic. I made a whole video essay about why I disliked the bonus cannon, but there were several other flashy additions that I think harmed the game. For example, there were a few types of branching paths and alternate exits in many levels, which created an obnoxious need to replay levels. They felt out of place to me - you’d never find forking paths in a song in Rock Band, for example - and I think it’s because they have basically the same problem as the ambiguous cues - the player has to choose their actions based on factors other than those immediately in front of them, which breaks them out of the moment.

(As a tangent, you can see this problem with smaller-scale choices as well - for example, in Senran Kagura Bon Appétit!, the cues are a Rock Band-like series of button prompts. Under some circumstances, one button prompt per song is replaced with a heart icon which treats any button as the correct button. Whenever I saw this, my automatic process didn’t know which button to press and I had to loop in my conscious process. I ended up solving the problem by just telling myself that the heart prompt meant a specific button that I hit every time it came up, so it could be automatic again.)

I still really liked Runner2 overall, but its design philosophy seemed weirdly inconsistent. For example - in the article’s animation of the rail segment where the player has to jump over the fireball instead of going under it, the other obstacles next to the rail are these round things with boxing gloves. By the time the game introduces rails, it already has obstacles that are established as things you go over and others established as things you go under. But for the rails, they created this brand new obstacle to mean “go to the other side of the rail” - these boxing orb things never occur in any other context in the game. This seems to imply that the designers understood the value of consistent and unambiguous cues enough to spend resources on specialized rail-only obstacles - but then they still did things like throw in a fireball you need to jump over. It’s weird and I don’t know what to make of it. I’d love to know what the conversations were like at the developer - whether the design inconsistencies were due to internal disagreements and compromises, for example.

Runner3 has inconsistencies as well, but they lean further on the side of unreadability. When reviewing the footage I took for the article, I noticed there actually is what looks like a very subtle cue hinting which wooden structures should be jumped over versus slid under. At least in the levels I played, the structures you are supposed to jump over have vertical supports in both the foreground and the background, while the ones you are supposed to slide under only have supports in the background. (You can see the difference in the animations in the article.) This seems like an attempt to provide readable cues - but before you are taught the double-jump you go through several levels sliding under wooden structures that vary somewhat in appearance. Thus you are taught to ignore minor visual differences in the wooden structures before they actually become relevant. I didn’t even notice the difference until reviewing the footage - it doesn’t seem fair to expect the player to notice while playing.

Runner3 also does things like throw sudden vehicle segments at you in the middle of a level that change up the gameplay in unpredictable ways and aren’t bookended by checkpoints. I very much felt like the game was trying to justify the sequel by being brighter and flashier and I can’t deny that the results are visually engaging. But they also make the game more frustrating than flow-inducing. I know it’s arrogant to assume you know better than the developer, but the failure to prioritize readability seems like an obvious bad call to me.

I didn’t decide to write an article about this until I happened to try Aaero and found it to be another promising rhythm hybrid that (for me) was ruined by a failure to prioritize readability. (If the moving barriers mentioned in the article were the only problem, I could have gotten over it. I can forgive anything once. But the game wants you to keep an eye out for red lights that indicate danger or attack targets, and then also uses too-similar orange lights as decoration, which I found distracting and flow-breaking as well.) And due to a happy accident of timing, Paws on Fire! (which I had completely forgotten about) happened to come out while I was working on the article.

I was thrilled that it avoided every bad decision that Runner2 had flirted with and that Runner3 had fallen into. It’s the most readable rhythm platformer I’ve ever played and a new entry on my list of all-time favorite games. It doesn’t have the branching paths or the bonus cannon. It does have the gameplay variety Runner3 tried to create with its vehicles - but distributes it amongst the four playable characters, so that the player controls when they get a change of pace.

Without knowing why the Runner games went the way they did, though, I’m a little worried that if Paws on Fire! gets a higher-budget sequel it’ll follow the same trajectory as Runner. I suppose time will tell.

But enough about Runner and Bubsy! Let’s talk about the article we ended up with.

I expected it to be easy to find quotes or references to back up my claim that rhythm games are about flow. It really wasn’t. I suppose rhythm games are somewhat niche, but I couldn’t find any serious look into why people play them. Thankfully, flow has been studied in actual musicians, so that gave me enough of a connection to work with.

Speaking of which, I thought of mentioning Clive Wearing when discussing music and flow. Wearing is an accomplished musician who contracted a virus that gave him anterograde amnesia - he’s unable to form new long-term episodic memories (similar to the condition depicted in the film Memento). As a result, he is incapable of following a conversation or holding onto complex thoughts and is often confused or distressed. But he can still play the piano, sight-read music, and conduct choirs, and while he does so he’s calm and in control for the entire duration. I do think this is quite interesting and a powerful image, but only tangentially related to what I wanted to discuss and quite distracting. So I left him out of the article.

I also came across a tangentially related article that argues that maximizing flow is not how you improve performance of actual pieces of music: Flow is the Opiate of the Mediocre: Advice on Getting Better from an Accomplished Piano Player. There’s also some interesting discussion in the Hacker News thread, including questioning whether the article is using a correct definition of “flow”, but either way the point of my article is that rhythm games are about maximizing flow to have fun and not to improve performance, so this still felt like a tangent at best and I didn’t get into it.

There are a few other games I considered discussing. Rhythm Heaven is a rhythm game franchise I haven’t been able to get into because it seems to mostly be a series of minigames that all require learning new cues but don’t have a ton of depth, so there’s much less potential for flow. I’m told that they’re mostly re-skins of the same five or six types, however, so maybe it’s better than it looks. Regardless, they didn’t seem like they’d add much to the article, so I didn’t dig in.

Also, people seem to really like Thumper - I watched two videos about it by Writing on Games (which mentions flow about four minutes in) and Jake Butineau, and while the game is intriguing it didn’t seem like it would add much to this particular article’s thesis either, so I didn’t delay the article to play it.

Somewhat more interesting is osu! which is an open-source rhythm game inspired by games like Elite Beat Agents. As an open-source game, a lot of its content is community-driven. Users can create playable songs as well as UI skins to go with them, and unsurprisingly a lot of these skins are not readable. Some do things like make the targets you’re supposed to click blend in with the background. However, due to the open nature of the game, players can turn off these aspects just as easily. You can play songs created by other players but use your default UI skin. So there’s room for fun surprises from the community, but you can also enforce readability.

And that’s about it for this article. Hope you enjoyed!