I thought I made a hard game and then speedrunners destroyed it
imgur.com
imgur.com
We may only be hearing about this story because this kind of language increased its chance of being shared.
[I don’t think the developer resented the speed runners. He sounded flattered and amazed to me. But it’s sometimes hard to judge tone of voice from just text.]
Why though, what does it add to the discussion? Why does it matter if a tiny indie developer gets a bit of exposure to his tiny little game? "It's just an ad" sounds needlessly cynical and dismissive.
To see people speedrunning as a negative is just...odd. People played it, and not only that, put the time and effort in to explore it and share with friends. What's the problem? How is spending literal hours collecting secrets and brute-forcing new tech "destroying" the game? Someone's clearly never run a game of D&D before, because I'd be ecstatic if people took literally anything I created and expanding it into a new form I couldn't have gotten on my own.
Imagine, instead of complaining about the exposure, the game's creator used the feedback for another level (LEAVING the exploits in) and leaning on them as intended tech instead of seeing it as some "look how they massacred my boy" pity party. Crazy, doesn't even have to rewrite any engine code, just design a new level that emphasizes the new tech for these speedrunners.
Average Joes can still play the "normal" version to practice, then ramp up to the new one. Why piss on a community that's shining a light on and celebrating your work? Author even ends with "It was difficult to watch them do this to my game but I learned to love watching them" - so is it just a clickbait title? Have I been played? Should I have listened to DJ Khaled?
> but I learned to love watching them. Their skill truly amazes me.
I got the impression that the game's author is seriously happy about all this. The timing of the article before a speedrunning event looks like it is intended for maximal publicity.
I'm not sure where you got the impression you did about the content of the article?
https://www.reddit.com/r/speedrun/comments/c2xapo/i_thought_...
"Totally joking with that post. I have no hard feelings. I really thought it was awesome how much some members of your community took to the game and made it so much better based on your feedback and play style."
It makes a bit more sense if you imagine the developer's intended way to play the game being a glass bridge across a chasm, and speedrunners literally shatter the bridge and glue the pieces into a ramp that they jump a dirtbike off.
Source: I've been speedrunning for 7 years and my full-time job is in the speedrunning community
I've never understood this need to take pot shots at the devs while running, but it happens all the time during events like GDQ. Even when the developers are on the line chatting while the game is being played (which just seems straight up rude.) It also ignores all the complexity involved in making and shipping a game, which is definitely non-trivial.
Several early 3D games increased player moment speed when holding strafe and forward at the same time. (Doom, EQ, etc.) That’s just incompetence and really should be described as such.
Now some games may have intentionally copied common bugs for style reasons etc, but more often titles are just rushed out with many known bugs and or vastly insufficient QA.
It's also one that doesn't impact core gameplay for most people. Heck, a lot of players will feel quite clever for discovering and using abilities like that in a game.
Fixing bugs like that require not fixing other bugs. They may have been aware of the issue and chose to fix other, more pressing bugs first. I'd strongly hesitate before labeling someone incompetent because their linear algebra didn't quite do vector math correctly in every case.
I'm also surprised to hear you describe the devs of Doom as 'incompetent', given their widespread praise in the engineering community, specifically for their code.
> Now some games may have intentionally copied common bugs for style reasons etc, but more often titles are just rushed out with many known bugs and or vastly insufficient QA.
So in some ways, those bugs become a part of the culture. That's an interesting avenue to explore, but it's an aside from my point that insulting the devs (who probably aren't the ones making decisions about QA or release schedules) undermines a relationship speedrunners could have with the developer community.
Unfortunately, QA costs money. Developers cost lots of money. Finding, root causing, and fixing bugs therefore costs lots and lots of money. You cannot just fix bugs forever, at some point you need to evaluate the current amount of money you've spent making the game and compare that to what you think you'll be able to make back on it. Does fixing that bug end up costing you more than you'll make back? Who knows, but it's a very real, complex prioritization problem that I don't think you are giving enough value or depth.
But, if we're comparing failure modes, can you at least agree that there's a pretty significant difference between:
* The tacoma narrows bridge, where there entire bridge was totally destroyed narrowly avoiding loss of life, AND * A bug in a video game that impacts the movement speed in a not-negative way for 90% of the players?
Like, if the strafe-bug was destroy millions of dollars in infrastructure and killing people I'd be on your side. But who cares if the 20th game has the same bug that doesn't affect most players in any way?
I agree it'd be better if games had no bugs, but if I had to choose between a game being 1% more fun, or a game not-having the strafe-bug I take the game that's 1% more fun every-time.
There's a nearly infinite set of tasks that sit in front of game-devs. The first thing you learn is that everything is an opportunity cost.
Nobody’s life is at stake in these games, but the gaming industry is both gigantic and competitive. So, a hypothetical 1% difference is enjoyment really can translate into 10’s of millions of dollars. Now, I don’t think this specific bug means a 1% drop, but it also does not take significant resources to fix.
Also, it’s not just about immersion this has real impact on game balance. Can players run away from enemy X often makes a huge difference in terms of fun. Balance assuming people are running sideways and people get stuck playing that way, balance assuming normal running and tension feels fake because they can easily escape.
Collectively obvious game bugs have killed off several game companies and represent billions in lost sales. As an industry that’s clearly extremely important even if we are just talking about entertainment.
This is where you are wrong. Even small fixes can destabilize builds. Every change needs code review, testing, a build, release (and possibly release notes). You need to schedule time to work on it, which means prioritization meetings and advocating why this bug should be fixed and not the other large pile of bugs.
Maybe you need to update some unit tests or automated feature tests. Maybe stop by the level designer's desk to make sure there weren't any baked in assumptions about player movement and space. Probably a good idea to chat with a gameplay designer as well to make sure it wasn't intentional (and also to validate that the fix doesn't actually make the game less fun.)
Wouldn't hurt to go to a playtest and mention that it's been changed, see if anyone on the team notices.
Actually, come to think of it, maybe you could put your fix behind a console command! Then you could quickly flip back and forth and see what you like better. But that also means adding a little more code and probably a little more documentation.
Sure, on the face of it you've just changed one arithmetic function. But there is a ton of hidden work and complexity that surrounds that.
As someone who made games for a living, there's a huge amount of unseen work that goes into even small changes. Players consistently underestimate the complexity of even minor fixes.
Not making the mistake in the first place should generally prevent the issue even if it was never fixed in any game ever. I am calling it incompetence specifically because it’s the kind of mistake you make when you don’t know what you’re doing.
PS: If detecting and fixing an issue like this is a significant issue for your team, that’s a much larger issue you need to deal with.
You seem to believe this notion that it is possible to write bug free code. Which suggests to me you’ve never written anything serious because you’re comments are textbook examples of the Dunning-Kruger effect.
Sure, some games are released buggy. But even there, blaming the developers just demonstrates how little you experience you actually have with professional software development. You think developers are really that stupid that they don’t know when their own games suck? Usually bad games get released because of financial and/or time restrictions rather than a lack of pride from developers. And often bugs will just be placed on a snag list and prioritised in order of estimated time to fix vs the severity of the bug. Which means many of the bugs you described would just never be financially beneficial to patch.
You keep talking about the monetary impact of bugs while ignoring the fact that software development is a business and thus saving will always be cut where needed. Having your development time prolonged by 12 months to fix bugs most people won’t even see is perhaps the single most expensive and unnecessary way to drive your development.
Frankly I’m astonished anyone would believe development happens the way you idealise, let alone dare to call other engineers incompetent for not writing perfect software.
Errors with collision detection are one of the most common speed running exploits and in excess can make games unplayable. Broken quest chains are similarly exploitable, but in excess result in an unplayable mess.
Anyway, cost is really not the driver for most of these issues it’s poorly scoping what can actually be finished in a given time and budget. I have no idea what your background is in game dev, but it sounds like you don’t understand the overall development process.
I’d already addressed that when I said the kind of bugs you’re complaining about wouldn’t have an effect on sales.
> Anyway, cost is really not the driver for most of these issues it’s poorly scoping what can actually be finished in a given time and budget
That is exactly what I’d just said except you’ve overlooked that development time is still an expense.
> I have no idea what your background is in game dev, but it sounds like you don’t understand the overall development process.
Not my current role but my previous job was a lead developer. So I was not only hands on but I also had to manage workflows and budgets. Basically all the stuff you’ve just claimed I don’t understand ;)
Comparing a minor bug with player speed calculations in an edge case scenario with the total collapse of a bridge. They are on totally different scales. One cost lives. The other is a minor inconvenience at worst. When bugs are found you triage them, and if the expected cost to fix is greater than the expected benefit gained, you don't fix it. The benefit gained from fixing an edge case movement bug like this is minimal, so it will often be left unfixed. This isn't incompetence, it's just prioritisation.
velocity.x = (RIGHTKEY-LEFTKEY)*speed;
velocity.y = (BACKKEY-FORWARDKEY)*speed;
Compared to building even a 2d side scroller let alone 3D worlds it’s not meaningfully more complex or slower.
PS: For a spaceship to move like that it would need equally huge engines on all sides. A design like the space shuttle with big engines on one sides and little ones for rotation seems much more logical outside of possibility low speed zero g docking.
For a spaceship to accelerate like that...
No, it alters gameplay balance, but it doesn’t break it any more than shooting your film at 24 FPS breaks the cinematic experience. At this point, it has such a storied history in video games that “fixing” the problem may make the game less fun to many people, despite being more accurate.
That is one way to see it, another is to add extra depth to gameplay by introducing secret areas only accessible through gaps that you can only cross if you move diagonally.
Designers often acknowledged these bugs, especially in older games. The classic Tomb Raider games, for example, has several secrets only accessible through outright breaking the engine. One very obvious example is in TR1 a health pickup is only accessible if you keep jumping upwards at a corner of a pillar which causes Lara to move forward a single unit and eventually, due to some collision bug, she teleports at the top of the pillar which is otherwise inaccessible... and the only reason you could know about this health pickup is because due to the painter's algorithm used to draw the level, sometimes under some angles the pickup is drawn in front of the pillar even though it should be invisible.
Also strafe-jumping has been a common tactic in FPS games (especially multiplayer games) to the point where games reimplement it in engines that otherwise work 'properly'.
It’s like blaming knife manufacturers that they’re incompetent because you accidentally slice your finger while cutting vegetables. Except the kind of gameplay bugs you’re describing aren’t security vulnerabilities and can’t be used to cheat in multiplayer games (outside of speed running where these kinds of cheats are often allowed). At some point you have to accept responsibility for your own actions; you guys are deliberately trying to break games to get faster speeds so it’s hardly a surprise that you’ll find ways to break games.
Whenever this gets under my skin, I simply remember the classic: "I could place chunks of fat in a bathtub and call it art!" "Maybe, but you did not, Beuys did."
Haters gonna hate, it just shows their ignorance.
Additionally, there are constant charity events that are moving millions of dollars annually. For example, a week long event just closed that raised 3 million US dollars for Doctors Without Borders. Those events need staff full time to support their mission, plan for future events, etc.
Perhaps a way to inject some visceral excitement into an eSport. Also makes it feel like more of an accomplishment.
https://www.youtube.com/watch?v=cRX-kxiNfes
With Freecell, once you know what you're doing you know at a glance what your options are: how deep a stack you can move and which cards are key to clearing out and in which order.
As well as playing way too much Freecell back in the 90s I played Minesweeper to death. Probably a better example of how you can progress from needing to consciously work out where you could safely click, to knowing at a glance, to being able to click the one cell that would open up the area. If you were actually flagging bombs you were going too slow.
[0] To be clear, I'm talking about multiplayer as it's unpredictable.
Yeah, I try to snipe someone, and they respond by building a 5-star hotel with WiFi.
The tactical opportunities in a fight that are available to a team with a gibraltar and pathfinder vs a gibraltar and pathfinder.. huge. Where does the shield go? Is claiming the high ground worth exposing yourself to the airstrike?
What? Two of the top four~ most played games on earth right now are battle royales.
You might like my game. Bomberman meets RTS. https://youtu.be/Zr0A8VcBnYk
If you just moved around using the obvious controls then you would generally get destroyed by opponents who had mastered that unintuitive way to move fast.
I've seen the same thing in people I know who can pwn me at things like COD. It's 95+% knowing where to look and where to have your gun pointed no matter which part of the map you're on, so you can shoot before the pure-reflexes n00b even knows you're there.
A master swordman does not fear novices precisely because they know their movements before they do. They look at (imperceptible for amateurs) body language hints and know exactly what you are going to do before you do it.
In reality there are physics constraints. Before you do a movement, you need to put weight in some part of the body, tense a muscle....e.g it is impossible to jump without a specific movement of going down and preloading.
Not only that , they know exactly what is the best response to it to neutralize whatever comes to them.
>"But don't you know, there are some things that can beat smartness and foresight? Awkwardness and stupidity can. The best swordsman in the world doesn't need to fear the second best swordsman in the world; no, the person for him to be afraid of is some ignorant antagonist who has never had a sword in his hand before; he doesn't do the thing he ought to do, and so the expert isn't prepared for him; he does the thing he ought not to do: and often it catches the expert out and ends him on the spot."
I played SSBB (Super Smash Bros. Brawl) competitively. I'd occasionally drop games to the worst players because they would choose one of several of the worst possible options available to them at any given time. I keep expecting them to react the "proper" way and so it would throw off my reads. Thankfully sets are best of 3.
SSBB is a game entirely based on reads due to the characteristic lack of combos and "neutral position" being popular since there aren't many scenarios where you can build momentum in a favored position (unlike the more popular SSBM (Super Smash Bros. Melee) which has combos that allow you to create favored positions to keep momentum.
It obviously doesn't eliminate luck; but the person who doesn't understand strategy isn't going to do well either.
Also, that's all well and good for heads up play, but if you're at a full table, playing tighter for one person at the table is just going to see you winning a lot of nothing. (Players paying attention will just skip hands where you bite.)
CSGO is a fantastic example where someone with poor aiming skills but good gamesense (holding angles, using echo-location where you move your view to triangulate opponents based on how the sound moves and what type of footstep sounds are generated, timings, grenade throws, etc) can wreck opponents with poor gamesense and better aiming skills.
I have gotten reported for walling but people just need to realize how much damn noise they make.
The Shift key became infinitely more useful once my ears acclimated to the setup change.
In many ways it's not so different than billiards/snooker/pool. Being a good shot-maker helps but just about all the consistently performing pros focus first and foremost on position-play and thinking several shots ahead.
Always amusing when this happens.
I've noticed people who perform best are the ones who pick a strategy and stick with it rather than trying to get the glory.
Another thing is lower level players use nades poorly and are always caught off guard.
I'd say in most of my games I've got more playtime, trying various strategies, and learning from mistakes, than the other 9 people in the game put together. But at some point it all just comes down to who wins a trade, and I very rarely win them. Game's still fun though, somehow.
(I’m not suggesting aimbots, mind you; the bot layer wouldn’t have more complete information to base its decisions on than the player does.)
I remember feeling guilty when I remembered what was on the next screen, because it felt like cheating. It kind of broke my heart when I realized that was the best way to get good and it was pointlessly stubborn for me to resist a technique that games were designed to cater to.
wikipedia.org/wiki/Mornington_Crescent_(game)
This is what is nice about research. Real problem solving matters.
But also what is frustrating about research: real problem solving is hard!
My completion of Age of Engineering is this small countryside feeling open air base, Direwolf20 (famous YouTube Minecraft player) beat it with a space station. The only obvious commonalty is we both have huge floating energy spheres from Draconic Evolution which looks kinda out of place next to a greenhouse compared to inside a space station. "There's more than one way to do it" is a driving ethos in most of these packs.
For speedrunners though any sources of randomness are mostly just annoying. If it's 50 seconds faster to start on a hill, and your personal best is 40 seconds slower than the world record, you're not going to play runs where you didn't start on a hill. You're just going to reload.
I totally relate to that guilty feeling. I like having to think on my feet, and never saw the point in memorizing a sequence of controls. It's like it ceases to be a game at that point!
It's just that some games have this strange curve in the learning curve, for instance Starcraft, where beginners think the game is all about strategy. But as you learn to play, you find out that it's more about macro, cause having more units normally beats any strategy. But then if both players have top notch macro, they have to begin doing tradeoffs in what they build, maybe cut some units to get a technology faster, or be greedier but more vulnerable for a while for a later payoff. In other words, the game becomes about strategy.
Another set of games my girlfriend recently turned me on to are rhythm games, particularly VOEZ[1]. Again, as I don't think I said in the first comment but said elsewhere, it's not like I don't like the "classical music" style and I'm all "jazz music," but my "jazz" style is definitely my default mode. If you asked me afterwards to remember the notes, I would be able to tell you the gist but I certainly don't remember details, a lot of it is focusing when it gets complicated and responding on beat. And, as I said above, my absolute favorite is playing a song for the first time and trying to get a good score, because I really have no idea what's coming and it pushes my "respond quickly" buttons.
[0] not that you don't learn quickly/during the fight without dying, and as many games often are, there is an obvious hint of what to hit, the rhythm to a boss, etc.
There's something un-fun to me that is hard to explain about all this. I don't like spending hours retrying a single challenge when I know there's a super precise series of inputs elite players or a TAS could use to get through in 5 seconds. I realize that's a very subjective thing and this is still fun to some players.
I've moved to playing mostly puzzle games like The Witness, puzzle platformers like Braid, and challenging "simple" platformers like VVVVVV. What I like about VVVVVV is that the inputs are incredibly easy to learn. You will die a lot, and there are people who can make it through the whole game in deathless mode, but I'm okay with that because it just means I need to get better at the game, not get better at memorizing a precise series of inputs and strategies.
I will never be a chess grandmaster. That doesn't make chess unfun.
But at least for me, the existence of this second game (and third) does not in any way detract from the quality or enjoyability of the first one. The speedrunners are basically playing a different game, and unlike most other video games, in Celeste this is by design.
These are the things I had in mind. It's quite astonishing to see a TAS or even some runners skip straight through an obstacle or jump off of a spike.
Seriously though, what does this even mean? I love platformers, but strategy and muscle memory pretty much sums up most action oriented games.
> You can jump straight through some obstacles, get additional dashes, etc. that the developers didn't originally intend.
Most of these techniques were intended; the game was designed to have them in. The game actually teaches you most of this tech extremely late in the game. It's a replay-ability thing. There's an extremely good talk at GDC a few years back about Celeste level design, incidentally...
There have been a few situations where runners did find an actual glitch, and the devs patched them out, because they weren't intended. There have also been one or two where they've left it in, because it's healthy for the running community overall.
I haven't played much VVVVVVV, but isn't it basically the same as celeste? Or is it Celeste's lack of randomness that makes it worse for you; I thought VVVVVV didn't have any randomness either, but it's been a few years.
As another reply reminded me, I'm thinking of stuff like demodashes that definitely weren't intended. In any case, the comment I'm replying to is also criticizing design styles that benefit speedrunners over the average player, so my comment is kind of a reaction to that.
> I haven't played much VVVVVVV, but isn't it basically the same as celeste?
The difference with VVVVVV is that the controls are simple, instead of having runner-friendly combos that get extremely long or involve glitches. The game is played with arrow keys: up / down flip the direction of gravity, and left / right move the player character at a constant speed to the left or right. That's all there is to the controls. Similarly, Super Hexagon has exactly two inputs: left and right, which move an arrow around in a circle.
In contrast, games like Celeste put an enormous amount of complexity behind input combos. The air dash mechanic alone took me hours to learn with any precision. The complaint I have (and maybe the parent comment too) is that as a casual player it's hard to shake the feeling that these games were designed for someone who has a lot more time to train finger memory on a bunch of precise input combinations than you do.
So it's not the existence of speedrunning that bothers me (I watch a bunch of GDQ, actually), it's the experience of playing games that are easy to exploit and "beat" if you have time to practice frame-perfect inputs, but aren't so rewarding casually. As I said in my comment you replied to, I realize this is all very subjective and probably won't bother most people (and I actually, I like Celeste quite a lot). But it's also the reason I find games like Smash very frustrating as a casual.
I find Smash difficult too. Wave dash in Celeste? I’ll do it all day. In smash? Dunno, just doesn’t work. Humans are funny :)
On the other hand, though, I've seen friends of mine who are less experienced with video game controls struggle with all of the above actions. I could watch my friend manipulate the controller and see that she was only able to handle one button input at a time, so for example in order to jump right, she had to break it down into the discrete actions of: hold down the right button, hold down the jump button, release the jump button, release the right button, with maybe half a second to a full second in between each action. So for this friend, pressing any two buttons at the same time clearly was an "input combo" that was a challenge to execute.
The interesting thing to me is that in my mind, the basic controls of Celeste feel very simple: it's "only" run, jump climb, dash (until 7B and 8C teach you wallbounces and hypers). But all of those basic actions have so many subtleties. For example, jumps can also be wall jumps, and there are even two kinds of wall jumps: straight up the wall if you hold climb, which consumes climbing stamina, or away from the wall if you don't hold climb, which doesn't consume stamina (not counting the other wall jump type, neutral wall jumps, since those aren't taught or required to complete the game). Somehow all of this complexity felt very intuitive to me, even at the beginning when I still hadn't learned to execute it all consistently. So I never felt during my first playthrough that Celeste was somehow catering to speedrunners at the expense of the casual experience. In fact, my impression was that Celeste does a very good job of providing the smoothest possible ramp up from casual play to speedrunning.
Anyway, if there's a point to this meandering comment, it's that variation in the way that people perceive the complexity of video game controls is very interesting. The same friend I mentioned above, who had trouble stringing together two inputs in Celeste, can give me a run for my money in Mario Kart, managing steering, mini turbos, and using items all at once.
I call those "do over games", because you have to just do them over and over and over in order to win.
I don't find them fun.
I love Hollow Knight to bits.
It is one of these "hard but fair" games. My first playthrough 1 year or 2 ago has seen me die a lot of times. And it was indeed because I needed to learn how each boss (there are so many of them .. maybe 30 or 40 ?? ) moves and react accordingly. I quickly started to see it as a dance or a dialogue with the boss though. You have a limited but versatile skill set at your disposal, so for each boss, I had to think about how to adequately counter a move.
The end result is a beautiful dance of action and counter reactions.
I had a second play through a couple of months ago. Since then, the devs have made the game change A LOT. Some bosses had their entire move set rewritten. New areas were added, etc.
So I was not playing on memory.
I had still acquired a wealth of knowledge on how to control that little bug, even with its limited initial toolset.
So I beat most of the game without dying just by reacting quickly to whatever it threw at me, it was an awesome experience.
I think what you say is probably true of most platformers and other games like Sekiro (a successor to dark souls which was one of the original non-platformer games like this) but definitely not all games.
It depends on the exact context and goal. To an extent, I would consider "map painting" the equivalent of a speedrun. It's usually measured in elapsed game time rather than play time, but there's a similar effect.
Especially with some of the more difficult OPMs, there can certainly be an element of memorization and optimizing starts. This falls off as the game goes further, but it definitely can be present. I recall some years back reading an EU3 byzantine guide that described many of the nations as being every bit as much puzzle game as strategy, where figuring out the tricks and traps of the opening years can mean the difference between a thriving growth and early defeat.
Not that anyone ever has to play the games that way, but that's true of everything that gets put through a traditional speedrun as well.
What recent games are you thinking of?
Competitive Quake players knew the maps down to the pixel, along with the location and respawn times of every resource. Lots of modern 'hardcore' platformers have randomization and don't depend on pure memorization.
Then you have people like this, who've gotten so good at the game that they're able to beat the entire thing in 1 minute 17 seconds: https://www.youtube.com/watch?v=wnzTObVRwF0
I used to work nights doing production support (lots of downtime). To beat 88-4 [1], I exported the level, re-edited it to have 4 starting points (each a quarter of the way through [0/4, 1/4, 2/4, 3/4]). Then I practiced each of those in turn until I could beat them. After I could be each with about an 80% success rate, I tried the whole level again.
There are probably only 5 hard "gates" in 88-4. That is, 5 places where you are likely to fail (say a 15% success rate before extensive practice). The success rate after 5 gates is unexpectedly, phenomenally low: 0.008% (0.15^5). If you practice and get your individual success rate up to 80%, then the overall challenge is a doable 32.7% overall. I think my success rate was more like 40% on three gates, though, because even after practicing I think I tried about 30 times.
I still think about gates like this. I think Yatzee explained this in a Zero Punctuation episode, but I don't remember which one.
0. http://www.thewayoftheninja.org/ 1. https://www.youtube.com/watch?v=t6hxHi234AM
The game was built with a timer which counts up. It's intended to be speedrun...
Or are there also hardcore older gamers among them, who started playing in the 80s, 90s and they have jobs, families and still spend their free time spedrunning games?
There exist something called TAS (tool-assisted speedrun) where you use an emulator with save states and either slow motion or the ability to advance frame-by-frame in order to construct the "ideal" run, but runs constructed these ways generally aren't actually playable by hand because they rely too much on frame-exact timing.
If preprogrammed inputs or whatever are used, the term TAS (tool-assisted speedrun) is used. Calling a TAS a speedrun would be a faux pas.
A good person to watch for overviews of the speedrunning community is SummoningSalt on Youtube.
Here is a good video on the speedrunning of 1987's "Mike Tyson's Punch Out!!" on the NES while blindfolded, an incredible feat for humans regardless of the medium: https://www.youtube.com/watch?v=iZT6JEOC3D8
Here is a good video on the speedrunning of 1999's 'PokeMon Snap' for the N64, a game thought impossible to speedrun early on. Nonetheless, it has been speedran: https://www.youtube.com/watch?v=WYcWQP1tWfQ
In general, the community is non-normal and seems a bit, well, pointless, to a lot of others. But it's something that a lot of people enjoy doing to some degree or another and like everything else on the internet there is a thriving fanbase.
I watched a speed run of Half Life 2. It took them 4 years. That's equivalent to a university degree. Some things I'll never understand...
A couple of years ago, I played Hollow Knight on PC. I probably spent 60 hours getting to the final ending and finding the majority of the secrets. I enjoyed it immensely for a month or two. Sure there are speed runners who can beat it in under an hour, but that doesn't diminish the many hours of enjoyment I got from it.
If you haven't, I would recommend watching Summoning Salt's World Record Progression videos, each of which takes a game and talks about the series of exploits that allowed the world record speedruns to exist.
And if not, all the more impressive.
They had a speedrunner running their game in front of the development team which is hilarious.
But remember it doesn't matter if they finish it quickly if your audience is the average player
A game that's too difficult is not fun for a lot of people. So unless your target audience is specialized in that, I wouldn't worry too much
Then I realized all the little things I was discovering were actually known by everybody else. Specially, they were made by someone else.
As I grew old I kept thinking about how to make a game that would have secrets no one knew, not even the game maker. But that's impossible, not even Dwarf Fortress or similar very complex games where you can do basically almost anything are like that.
Every game is just a treasure hunt made by some adult for kids to play -- and that is very depressing.