Tuesday, February 11, 2014

Six Things I Learned From Speedruns

(I think I've done a list like this before. Don't judge me)

There are certain principles of behavior, learning, and efficiency that are nearly universal. If you have the right mindset towards altering your behavior, then on average you are going to find success more quickly than people without those mindsets. I think there’s a lot of useful lessons to be drawn from gaming that can be applied to life.

One of my favorite aspects of gaming are watching speedruns, because they illustrate principles of efficiency, they reward perseverance, and they encourage players to dig deep and collaborate to explore and exploit as much of a game as possible. They represent many of the skills, behaviors, and principles that lead to better learning and efficiency. These are some of the lessons I draw from trying to play games quickly, and watching others do so.

1) Efficiency is decided mostly by what you don’t do.

Watch a speedrun, and in many cases, most of the game gets ignored. You run past enemies and obstacles, because fighting them takes time. People who love the game, funnily enough, will try and figure out ways to skip more of it so they can beat it more quickly.

The first goal of doing something quickly and effectively is not figuring out how to do everything, but figuring out what is unnecessary and can be skipped. You want to solve for ways to generate the same output with less effort. Your goal is not to increase action, but reduce it, and get more for each action. That’s the very basis of efficiency; finding ways to do less.

2) Most time we spend doing things is figuring out how.

Many video games boast about their long play times for the new player. This makes sense, because the person who hasn’t played before needs an advertisement that the money they spend will get them a long way. Seventy hours of gameplay, with multiple endings and numerous class trees and playstyles and talents (and DLC and an identical sequel next year)! Sounds good to me.

But then again, most games can be beaten rather quickly. They don’t actually take all that long if you know what to do. We spend a lot of time standing around going, “well, what next?” If you already know what next, then you’re on your way there. You aren’t going back and forth talking to every NPC and checking every nook for some item you need. You aren’t losing twenty minutes of progress to a silly death and you aren’t attempting an impossible jump for thirty minutes. Half--or more--of the game that you purchase is actually somebody selling you your own ignorance.

If you ever get discouraged by a project or task that you think will take a long time? Most of it will be spent figuring out how to do it. Once that happens, they tend to go by much more quickly. Also, if you remember something taking forever, odds are it won’t now that you know more about it.

3) Tiny optimizations matter most to the best.

If you aren’t at the point where you’re seeking absolute mastery of something, you don’t want to spend your time obsessing over things that give you a small amount of benefit relative to the effort invested. The biggest amounts of time saved, on average, are from changing your route. You do things in a different order, you find a different path entirely that renders old optimizations moot.

Generally speaking, better decision making will trump minor perfections. Mastering nuance will bear you fruit once you have the route planned out well; beforehand, it will just be a time-sink and give you little to no reward. Start with your route before you try and optimize intensely.

4) The line between amazing and awful can be very thin.

Every now and then you may watch somebody play a game of some kind, and attempt something very difficult, risky, crazy, or all three (with a heavy dose of crazy). And when they fail, all the laypeople watching may think, “wow what are you doing, that was extremely stupid.” Either they don’t realize what the potential payoff of success was, or they’re being ignorant to a simple fact: the more difficult something is, the higher the odds of failure. Every corner you take tightly puts you a little closer to a wall you might crash into.

Part of expertise is how it makes the near-impossible seem easy, but it’s still near-impossible. With each step you take closer to the line, you run the risk of crossing it. When you do, you will make mistakes more costly than just playing it safe. And when you want to push into realms of absolute mastery, you will end up having to take those chances.

5) We are poor judges of what is possible.

I find it really funny that in many speedruns, the runners and the communities will talk about what times are possible. They will say “I think with perfect play, we might get this game under two hours.” These people aren’t even ignorant about the game; odds are, they know more about it than many of the people who made it! They aren’t even overconfident that they’ve reached perfection.

Somebody just happens to find an interesting glitch or trick, some re-routing is done, and suddenly you are beating the game in an hour and a half. It usually takes a long time before they really find most of what there is to know, and even then they don’t account for the power of optimization and perfection when judging best possible times. Some people will make a TAS (tool-assisted speedrun, using a program to manipulate the game perfectly) and then a year later somebody will find strategies to beat that “perfect” time.

It’s not even hubris. It’s not even cynicism. The simple fact is that we don’t just lack knowledge, we lack understanding about the depth of our ignorance. How are we supposed to know what we don’t know? When you ask somebody “what do you think the fastest time can be?” there is always an implicit prefix of “Well, given what we know…” in front of the answer.

Life, as it turns out, is full of some ridiculous possibilities. Science and knowledge are like digging into the game’s code, and technology is just finding ways to exploit life’s physics engine. Both in games and life, we slowly find ways to apply our knowledge in more esoteric, complicated, and obscure ways. Sometimes we use a new piece of knowledge to obsolete huge quantities of old data, and prove all the old guesses wrong. It happens surprisingly often, often enough that you should raise an eyebrow when even an expert declares almost anything “impossible.”

6) There is no telling what you will end up loving.

There are some speedrun games that I watch and think, “wow, that is intense. That is amazing. I would love to play that game.” Then there are some games that I watch and think, “why on earth would you ever play this game more than once, let alone hundreds and thousands of times to perfect it?” Certain strategies and segments of games I might find boring and tedious, but somebody out there is practicing them relentlessly to shave off five seconds from their personal best.

Some games have runs that are five (or more) hours long and partially dependant on luck. Some games are very short and the records are decided by saving one to two frames. I watch some people play their games and every streaming session they come away frustrated and angry at some part of the game that screwed them.

If you ask somebody why they play a game, and lay down the proof of why their game is objectively bull****, and try to convince them… they might say “yes, you are right.” They might even emphatically agree with you on the dumber parts. They might nod and say, “yes, this game really is extremely stupid.” They might get mad and take breaks and get burnt out and quit and come back and quit again.

Then you will ask them why they are running it, why they keep loading it up day after day, and they will respond with, “Because it’s my game,” and at that point the conversation is basically over, because there really is no telling.

Thanks for reading. See you Friday.


  1. I wish I could do a TAS of real life lol. Great read as always, Rob.