I'm a non-speedrunning gamer looking to understand how speedrunning buffs view the world of gaming. In particular, I want to understand the major goals of the community.

Obviously one approach people follow is to just pick any game you enjoy and jump into it; but some gaming challenges may be considered especially important, either due to their association with a popular/historic game, or due to their known difficulty or even to uncertainty as to whether they're possible.

As a scientist by training, I could give you a list of the most important "open problems" (i.e., unsolved problems) in my area, and some sense of why they're important. I'm wondering if any of you would be willing to try to do the same for your field. Now this presumes that speedrunning has some degree of unity as a discipline, and is not just a collection of isolated challenges; I also welcome comments on this assumption.

I'm especially interested in qualitative challenges (e.g., no-death runs, or games that have never been beaten), even though I know the community's main focus is on bringing down the record times. I'm also interested in the role of tool-assisted runs since they are obviously a powerful device for "research".

Please help me and others get informed and become better fans of your efforts. Thanks!

I'll try to give you somewhat of an answer, but first things first: "Now this presumes that speedrunning has some degree of unity as a discipline, and is not just a collection of isolated challenges; I also welcome comments on this assumption." The "unity as a discipline" is not given imo. As games are widely different in genres, consoles, engines, programming styles and so on... you will find a lot of different challenges for different groups of games. You will find somewhat similarities in game series, especially if there it's a follow-up title and no big change in technology at that point. But as there are lot of standalone games it is more of a collection of isolated challenges.

Still there might be some general things that can be said: As the goal of speedrunning is (usually) to complete the game as fast as possible, you pretty much have an optimization problem with multiple functions. I'd define some functions like (as examples): character movement, items, menu movement, dialogues, battles, game RNG Depending on the game genre you might only have some of the functions there. Like in a simple platformer you might only do character movement and just care about getting the fastest from point A to point B. In an RPG you might have pretty much all of the above: you need to move to town X to meet a NPC to get item F and equip it (in the menu) to a character to be able to battle enemy Z to get past it...

There are usually constraints in runs like "you can't pass this door unless you have the key for it", which can refer to both functions movement & items as you need to find the minimal path to solve this. OR you find a way to remove the constraint, like using a glitch to go out of bounds and pass the door without the key. -> This may define different routes for different categories of a game (in speedrun terms). Another example would be game RNG -> if it is totally random you have to prepare for it, e.g. getting good equipment to win battles against certain enemies. If you can know / manipulate RNG and you know what you can do to avoid death you can pass the area earlier/differently.

Another think I'd like to bring up is the state of "open problems" as you'd like to know them: Not every game is explored in the same depth. There are games where speedrunning community members figured out "all" of the bugs/glitches of that game by various means (e.g. observations, looking at specific weaknesses, code is open source, brute forcing with a bot,...). Then there are games where there is no knowledge of bugs to date and finally there is everything in between. It limits or opens the possibilities of that game and the categories to speedrun. Also it defines the "open problems" as in some game one guy might have found a "new" glitch which could be useful, but nobody knows how yet (or it might never be useful). Other open problems might be more straight forward like optimal movement through a certain part of the game, like when there are multiple possibilites. Or battle strategies when there are different enemies in an area and you need to figure out how to equip your characters best for them (while still being time optimal).

Sorry for the mathematical terms if some ppl have troubles with them... natural scientist here. Man... you gave me quite the inspiration to sum up some of my thoughts here. :)