Very preliminary thoughts and questions on speedrunning

I’ve only recently discovered speedrunning. It’s a fairly obscure corner of gaming culture, so here’s a quick explanation that Kosmicd12, current holder of the Super Mario Bros. world record, gave to FiveThirtyEight:

The goal of speedrunning is to go fast.

Yup, that’s all there is to it: complete videogames as fast as possible.

Underlying that very simple explanation is a very intricately constructed house of cards, consisting of an incredibly deep knowledge of the game, an insanely high level of skill, an inordinate wealth of patience, and sheer dumb luck.

I’m not all that good at games, but I play them and think about them a lot (as you’ll discover soon). Many months ago, one of my favourite gamers, Quill18, gave a shoutout to Games Done Quick. I dropped in on the Summer Games Done Quick 2017 stream, and it so happened that the speedrun going on at the exact moment I joined the stream was Jet Set Radio Future (JSRF), one of my childhood games. I only managed to complete it once, and it took me all day. I can’t tell you now whether “all day” meant eight hours or 12, because I can’t remember. It wasn’t a very memorable achievement.

The JSRF speedrunner at SGDQ was FingerQuick, who completed the game in one hour, fifty-three minutes, and six seconds.

Months later, I was watching Quill18 play Pokémon Platinum, when he mentioned that he had seen Horn Drill used a lot in Pokémon speedrunning.

Horn Drill? I’ve played a lot of Gen 1 and 2 Pokémon, and Horn Drill is not a move I consider reliable. If it hits, it’s a one-hit knockout, but it is a fairly inaccurate move. I was curious why Pokémon speedrunners would choose to use it.

Off I went to watch a Pokémon speedrun. I found Exarion’s Pokémon Red world record run, and watched the whole thing:

I was stunned. I’d never seen Pokémon played like this before, with a total, relentless focus on speed. I didn’t fully understand what was going on, but even with what I did understand, it was clear that this way of playing required a completely different perspective on gameplay than I’d ever thought about.

Most Pokémon players will focus on building a well-rounded team of six Pokémon. Exarion and other Pokémon speedrunners choose one Pokémon to run with, and use it throughout their entire run. This cuts down on the time needed to train a full party of six, and on the time needed to switch Pokémon in and out of battle.

Most Pokémon players will plan movesets with a view to their Pokémon being able to handle any opponent. Pokémon speedrunners tailor their main Pokémon’s moveset to be able to take on specific opponents at specific stages of the game as time-efficiently as possible. For example, Exarion’s guide to speedrunning Pokémon Red has your main Nidoking with a moveset of Thunderbolt, Bubblebeam, Thrash and Mega Punch before Pokémon Tower in Lavender Town. He suggests teaching Rock Slide over Mega Punch before entering the tower, but has you teaching Horn Drill over Rock Slide before you even leave the tower. If you follow the guide, you use Rock Slide exactly four times before Nidoking is made to forget the move. As far as I can figure out, this is because otherwise, Nidoking would have to use Thunderbolt against four Ghost-type Pokémon, and Thunderbolt is not a one-hit against the Ghosts, while Rock Slide is. (I don’t actually know this for sure, it’s just the only logical explanation I’ve come up with!) Rock Slide actually has less power than Thunderbolt, but Nidoking’s base Attack stat is significantly higher than his base Special, which is enough to overcome the difference.

And there, I’ve gone and done it. I told myself I wasn’t going to get into the weeds, and here I am. Well, I guess that ties in nicely with my next few gameplay points: Pokémon speedrunners use in-battle special items (X Attack, X Special, X Speed, Guard Spec.) much more often than regular players, are far more judicious about using stat-boosting items (Protein, Iron, Carbos, Calcium, HP Up), are far more concerned about PP management (to minimise Pokémon Center visits), and pick their trainer and wild Pokémon fights very carefully (fight only if unavoidable or needed for XP). That’s what a casual player sees when watching a high-level speedrun.

Speedrunning is far more complex than that, however. When I first watched Exarion’s run, I noticed that the very first step he took into the tall grass on Route 22 netted him the Nidoran♂that he ran with, and that this statistical improbability passed uncommented on (Nidoran♂’s encounter rate on Route 22 is 35%). That’s when I realised that speedrunners must have some way to manipulate the random number generator. In Viridian Forest, he took a very specific route through the grass, and did not encounter any Pokémon in most of the grass — another statistical improbability.

I was curious enough to dig deeper, and that’s what led me to Shenanagans’s run of Pokémon Blue at SGDQ 2015:

That’s when I started to understand that the appeal of speedrunning wasn’t exclusively speed or skill; it was as much an intellectual exercise. It’s a self-imposed logic problem: given this piece of compiled code, what is the fastest way to get from A to B?

Seeing other highlights of speedrunning, such as Shenanagans’s Pokémon Glitch Exhibition at AGDQ 2016, Kosmicd12’s world record Super Mario Bros run, TriHex’s Yoshi’s Island run at AGDQ 2014, and the Tetris: The Grand Master Showcase at AGDQ 2015 made me start to ask myself: what is it about speedrunning that I find so compelling?

Take a look at this superb breakdown of Kosmicd12’s Super Mario Bros world record run:

When you look at the amount of planning, skill, and then the luck involved in trying to shave one frame off a 60fps game (i.e. 0.016s), that seems absolutely insane for something that has no evolutionary benefit that I can think of.

Let me make this absolutely clear: this isn’t a judgement about what speedrunners should be doing with their time. This is a question about me: why do I find speedrunning so compelling?

I spilled a lot of ink trying to figure this out, but I don’t have a lot of clear ideas. So here’s a list of questions I have for myself that I’d like to untangle:

  • What is the appeal of sport?
  • What is the appeal of beauty?
  • Must sport have an evolutionary basis to be appealing?
  • Must beauty have an evolutionary basis? (Yup, Richard Prum’s Evolution of Beauty is on my list)
  • Must a speedrun be a world record to make interesting viewing?
  • Watching a speedrunner stream a speedrun attempt is much less interesting (to me) than watching a completed speedrun after the fact. Why is that?
  • What is the function of entertainment?

I make no promises that I will approach these questions in order, or even try and answer them directly. Tomorrow I might sit down and write about something completely different, but the ink I spill on that might eventually lead back to the topic of speedrunning.