FinlandAll my submitted runs so far have been in the VGA floppy version; but I'm getting the feeling the EGA version may be optimal, due to the fact it's the version that supports Tandy and DOSBox always runs fastest with Tandy emulation. So the EGA version has faster loading times. (For some reason it also has extra slowdowns with the slime at sewer entry, and on flushing the brain; but I think it's still just a mite faster overall.)
The problem is simply with the barrier of entry in acquiring the EGA floppy version. Basically, you have to either own it from 30+ years ago, or buy it from someone else who does, or download a pirate copy (and the vast majority of pirate copies are CD or VGA floppy too).
What do runners/mods think - should we allow or disallow EGA floppy?
Badly executed sample run.
And there I was about to ask if we could have an IL leaderboard for this game :)
I can be a mod, though I have to acknowledge most of the run backlog is likely my own uploads, would feel a bit uncomfortable verifying the lot of it by myself if there's another active runner/mod who could do it.
Yeah, no debug mode or cheat codes. (One runner did do a 1-7 collection run in debug mode recently though; so if you want to run the game(s) that way, there may be some interest.)
The AGI version has its own leaderboard: https://www.speedrun.com/kings_quest_iv_the_perils_of_rosella_agi
David can give a more authoritative answer; but if the only major difference between the Spanish and English versions is the language of the text, using Spanish is certainly okay; and even if there's a substantial difference that makes the Spanish version faster, it's probably still okay.
The scorpazoid does not follow Roger across screen boundaries, and since it only generates on screen transitions, whenever it appears you will always have a screen boundary handy to escape to. You will also always get a warning message the first time the scorpazoid generates; if you get it, assess which boundary to run for and then dismiss the message. (Back the way you came is the safest, but if the scorpazoid's out of the way, you can just keep going in your intended direction.) Subsequent appearances from the scorpazoid are more dangerous since there's no warning message, so good eyes and quick reaction times are needed; but you need to be unlucky to generate the scorpazoid more than once or twice.
If you look at the current Any% WR you'll notice it hugs screen boundaries even when not transitioning, largely to stay safe from the scorpazoid. The WR actually has quite bad luck with scorpazoid generation, and almost dies to it 3:30 in (3:38 in the video, as the run starts 8 seconds in), but is saved by the warning message. A couple seconds later, the scorpazoid puts in another appearance, but Roger's placement near the corner keeps him relatively safe. At 3:35/3:43, there's a weird zigzag caused by scorpazoid paranoia. Even with the bad RNG and the unnecessary zigzag, the run still lived and lost no more than about 2 seconds.
There is a non-zero chance that you will die through no fault of your own to extremely unlucky scorpazoid RNG; but if you die to it more than about 5% of the time you're either doing something wrong or taking extra risks to save time (or playing some unusual game version with different mechanics I'm not aware of). If you start really pushing for an optimized time, the main run-killer on Phleebhut will be losing time to bad Arnoid RNG on the screen with the pod creatures.
There was that time when topkekshrek and I were battling each other for the King's Quest IV SCI Any% record, and we were both held back by our inability to type "open door"... he kept typoing "door" and I kept typoing "open".
I also had some difficulty with spelling the word "tidy" correctly:
But yes, Abney's runs are full of unexpected hilarity...
My keyboard battery ran out 21 minutes into a 25-minute run during my WR push in Manhunter: New York. It was the highlight of an epic fail trail that also included twice losing the WR in the last three seconds of the run, and once losing it to a bug I have never encountered before or since. Got so silly I had to laugh at my own misfortune.
Like most things in speedrunning, bouncing back from a painful miss tends to get easier with experience. Keep pushing, keep improving, lose faster and faster runs in stupider and more ridiculous ways; with any luck you'll get accustomed to it, and even if you don't, you'll still forget all about the original lost run that was so disheartening at the time. Eventually the good run will happen; and the more attempts it took, the faster it probably ends up being.
If it doesn't work that way for you personally... just get the good run in that first hour. This isn't an entirely flippant answer; there are incredibly good speedrunners who've consistently played a game at a high level for years, still lose heart after major setbacks, and just work around the problem by basically being better at the game than everybody else. You don't need an unflappable and permanently upbeat attitude to be good; sometimes, champions are champions precisely because of how intensely they hate losing.
That said, both of those approaches work best if even under all the discouragement, the prospect of that one good run still motivates you. If you stop wanting the good run at all I cannot help there, beyond saying "take a break and do something that does motivate you."
Most likely Flogus01 just split somewhat late while playing and submitted the original time to splits.io, but then retimed the run as a 3:06. Nothing particularly unusual or suspicious there, looking at the video of the run it's clearly faster than 3:08 and probably is a high 3:06.
Many players actually prefer games with bugs in; exploiting bugs to help a speedrun is part of the fun. No glitches, and even no major glitches, are fairly common speedrun categories. It also isn't unusual for each platform to have its own subcategory in cases where a game plays very differently on separate platforms.
Of course, if the glitches on a particular platform outright prevent the run from being completed, using that platform likely isn't recommended...
Because the run is quite short and you're running a tutorial, it's very easy to assume it's the tutorial to some much longer game. Unless you're already familiar with the tutorial, it's also quite hard to tell from the video which parts happen automatically and which parts are actual gameplay you can influence, which is particularly problematic due to the short length of the run. This makes a detailed explanation even more important than it would be for a conventional game.
Explain to the mods that it's just a tutorial and there is no longer game behind it; then explain why the tutorial still counts as a video game in its own right, what you are doing to make it go faster, how much skill the run takes, and how it might be further optimized. Link to other speedruns to show that you're not the only one running it. None of that will guarantee acceptance, but if it's still rejected it will be rejected by mods who understand what they're rejecting.
Time trials in racing games are just IL speedruns, and some of them are pretty popular. The popular ones generally don't have a time-trial leaderboard here simply because they already had a well-established one elsewhere by the time speedrun.com got big; but many more obscure racing games do have time-trial leaderboards here.
The rules currently state that generic typing games will not be added to the site, but there are games with a heavy typing element; many old adventure games relied on text parsers, and some are pure text adventures like https://www.speedrun.com/the_hitchhikers_guide_to_the_galaxy/full_game.
There's one major RNG spot on Phleebhut about 4 minutes in (with Arnoid) that costs maybe 10 seconds per unlucky roll. The random scorpazoid, also on Phleebhut, can be annoying but seldom costly unless you're super-unlucky (there's a very small chance, maybe 2%, that it spawns right on top of you and insta-kills Roger; but if it doesn't do that you very seldom lose more than 1 or 2 seconds). The enemy ships at the end are partly RNG but mostly skill. I think most of the earthquakes on Ortega are also RNG (the first ones are guaranteed); but bad earthquake RNG doesn't cost very much time, especially outside ScummVM/QuanVM where they seem to be slightly longer for some strange reason. Overall, a lot less RNG than Space Quest 1 (and most other early Sierra games) but slightly more than Space Quest 2.
What happens if I dismiss the closing ceremonies so quickly that they're never visible in the video? Does that invalidate the run since the literal condition for stopping time (closing ceremonies become visible) is never verifiably met, or should I just figure out when the ceremony exit happened and time by that?
