Method of Timing
1 year ago
United States

Hello! Recently, there has been a major breakthrough for Kula speedrunning as I have discovered a ton of new glitches that can be used to lower the time. Most of these glitches are completely fine to me for being used, but there are a couple that I think are a little problematic.

The first one is "Hiro Skip", which is a glitch that upon exiting to main menu on the very last bonus level in Hiro, Hills will appear on the main menu instead. This mean that when a new Arcade game is started, it's started from Hills instead, which means the IGT shown at the end of the game will not take in account for Hiro, effectively "skipping" it.

The second one is "CopyCat Skip". It is executed by exiting to the main menu after saving progress, loading up CopyCat and exiting while doing the second input. Upon loading up your save game, it will instantly complete it. This allows you to "skip" every 5th level in the game, as you can only save on every 5th level. So since the level is completed upon loading it, the IGT does not take that into account.

I have mixed feelings about these glitches, as this makes the run look like it's just constantly going in and out of the save just to save in-game time. Now, the auto-splitter that was developed will still take the in-game times in account regardless if you leave the save and come back (I'm pretty sure), which makes it so that the in-game timer in-game isn't abusable as much. Better yet, it makes it so that not everyone can just leave their save if they make a mistake, and load it back up and the in-game timer resets.

So is that what the preferred method of timing should be now, not the in-game time shown at the end of the game but with the autosplitter? Just wanted to get an opinion on this :)

Yasahiro, LePalmypede, i RealRaptoReX podobało się to
Italy

The IGT calculated by the autosplitter matches the one displayed at the end of the game only if you don't perform those glitches. And that was fine before those glitches were considered.

I am looking at the submission I received and I'm kinda conflicted, because the internal game timer, as you correctly said, doesn't consider all those saves and reloads. Worse, it doesn't consider the fact people need do the entire Hiro normally before even "starting" the run. The point is doing all those glitches artificially lowers the game time, but RTA the run is considerably longer than the standard Any% for this reason.

Using the timer as displayed by the autosplitter can be fine (it supports every PS1 emulator) but excludes people who play on real hardware. Which is an issue simply because the current WR is still done on real hardware.

I need time to think about this. Pretty sure though glitched runs would probably be part of a separate category.

RealRaptoReX to się podoba
Italy

The run is accepted for now in the new category I created for it. For the time being, I think it's best to state both IGT and RTA for glitched categories.

Ukraine

To my mind there should be 4 categories:

  1. "Arcade Any% RTA (Real Time Attack)"
  2. "Arcade Any% IGT"
  3. "Arcade Any% Glitched" (only with IGT)
  4. "The Final" (only with RTA)

1-2. Why "Arcade Any% RTA" and "Arcade Any% IGT"? Because different strategies apply for RTA and IGT. If we speak about RTA, we should reach the exit as fast as possible without collecting fruits. In IGT we collect fruits (minus 3-4 seconds).

  1. In "Arcade Any% Glitched" we use glitches to save IGT time. RTA doesn't matter in this category because when we save or load game it takes time in RTA. For example, level 21. If we speak about RTA, we can complete level without saving after level 20, and reach the exit for 2-3 seconds. It's more quickly than we'll save game progress, enter the CopyCat and after that load the saving progress. It takes long time. And again, our goal is saving IGT time.

  2. And in "The Final" only RTA should be because IGT isn't shown in the game statistics after reaching the exit in level 170.

Okay since i got into an heavy discussion with Dima, I need to give my opinion on this whole situation as well. Fact is: SaturnKai changed Kula Speedrunning forever. The ammount of glitches he found is insane and it's up to everybody to judge that for themselves.

As for the categories and it's rules, this is how I would imagine it:

  1. Arcade Any%: This one needs to stay completely glitch-free, period. It would not make sense to allow anybody to use any of the found glitches. Because even using one of them would invalidate all the runs on there. Adeyblue holds the world record for a reason and i would find it shitty if anyone was allowed to use glitches in a "Non-glitch" category, that just doesn't make any sense to me. Should it be allowed to reload the save file? Hmm i'm not so sure. First of all, only if the save file menu is completely empty at the start of the run, so nobody can trick the viewer by reloading other save files. Reloading doesn't necesarily make you better at the game, it just makes your runs less riddled with errors. As long as it is a single segment, that would be fine. The IGT measures the run just fine and can't be tricked.

There's more to this though since dima also wants a "Arcade any% RTA" category. I don't see any point in that either, because it is technically there. Just click on "Time" in the leaderboard. Just maybe renaming time to RTA would be enough. At the same time, I fully agree that the RTA should exclude any loading screens. For some reason, even though my pc is good, my load times are terrible (look at my glitched run) and it loads longer than real hardware. So to make it fair for everybody, RTA runs should exclude those. Because I agree with dima on this, that an RTA run is completely different from an IGT run. Collecting fruit in IGT is beneficial but not in RTA.

  1. The new category "Arcade any% glitched" is fine as it is. This one should allow any glitch with save reloading, again as long as we can make sure the save file screen is empty before starting the run. The IGT once again is enough for this and I don't think RTA runs make sense in this category.

  2. The Final. We also have this category now, which can only be measured as RTA, since there's no IGT. This should once again exclude loading times. But this category should remain for legit runs - just like the arcade leaderboard.

  3. The Final Glitched. I swear this is the last one. You can use all glitches you use in arcade in this one as well and you can skip 1 level, the 11th one by using the copycat glitch. SaturnKai demonstrated the possible time in this category with his TAS. If we would add this one, i think it's complete.

I think we don't need the "Broken IGT" category we were talking about, since even if we would exclude the hiro skip and copycat glitch, the times we can achieve are still ludacris and we should make it even more awesome how much we can shorten the timer. I hope we can find a proper solution for everybody.

Ukraine

I somewhat agree with RealRaptoReX. But I suppose there must be as "Arcade Any% IGT" as "Arcade Any% RTA". Because different strategies apply for RTA and IGT. If we speak about RTA, we should reach the exit as fast as possible without collecting fruits. In IGT we collect fruits (minus 3-4 seconds). It depends on what the speedrunners want to do: showing their best RTA or showing their best IGT. Because if your goal is showing the best RTA, then what's the point in collecting fruit? You're only wasting extra seconds for your RTA. And if your goal is showing the best IGT, then it makes sense to collect fruits (minus 3-4 seconds from your IGT). And here it doesn’t matter what RTA you have (50 minutes, at least 2 hours, anyway). And vice versa. If your goal is to show the best RTA, then it doesn't matter what IGT you have.

For example, my speedrun in Arcade Any%. My IGT is 34m 58s, my RTA is 43m 30s. But my RTA can be better than 43m 30s. Because there are levels which speedrunner can complete faster (without collecting fruits) and my RTA will be better. But at the same time my IGT will be worse than 34m 58s because I don't collect fruits. That's why there should be two categories. As you can see, these are two different strategies. And accordingly, these are two different videos, that is, two different speedruns, that's why there should be two leaderboards: "Arcade Any% IGT" and "Arcade Any% RTA" if we speak about "Arcade Any%". In the "Arcade Any% IGT" leaderboard there should be shown only our best IGT, without RTA. In the "Arcade Any% RTA" leaderboard there should be shown only our best RTA, without IGT. I don't know if you understood me or not, just don't know how else to describe this situation.

And RealRaptoReX is right. If we speak about RTA, RTA should exclude any loading screens because everything here depends on the hardware.

Italy

I thought for a while about this and this is my proposal:

  1. Arcade Any% stays as it is, with IGT as its main timing method. As I previously said, the main point is the route keeps into consideration the fact you can collect the fruits to reduce the game timer.

  2. Arcade glitched% should be considered in both IGT and RTA/Loadless. My main issue here is that the Hiro skip essentially skips 1/10th of the run before the IGT even starts ticking... and it's debatable whether you'd call it a skip as you need to do the entire Hiro anyway first. Which is exactly why a different timing method should ALSO be included alongside IGT. My idea here, however, is that Loadless should be its main timing method.

  3. The Final can be measured with IGT, but the issue is the game doesn't show it clearly (you'd need to use the autosplitter or read the PS1 memory, so that needlessly overcomplicates things). Which is why I propose simply timing it as Loadless.

Keep in mind Loadless timing is fairly easy to do with the autosplitter, and keeps into consideration the fact load times are dependent on the emulator you're playing on (loads, for the most part, don't depend on your PC hardware). If you play without livesplit, or if you play on official hardware, runs would need to be retimed manually, which is not a great thing to do for glitched% but basically free on the other categories.

Edytowane przez autor 1 year ago
Italy

So essentially I propose this:

  1. Arcade Any% (the current one) --> IGT, glitched banned
  2. Arcade glitched% --> Loadless & IGT, loadless timing starts at the beginning of Hiro 1
  3. The Final --> Loadless
  4. The Final glitched --> Loadless

(My presence on the internet might be spotty for this weekend, but I'll be sure to read everything by the end of Sunday)

Edytowane przez autor 1 year ago
Ukraine

I think it's a little difficult. Everything can be done much easier.

  1. Arcade Any% IGT (IGT is main timing method). Glitches like avoiding lasers, avoiding arrows, jumping 7-8 times should be allowed, as they can be used in each level. It depends more on the version of the game (Kula World, Roll Away or Kula Quest) and the skills of the speedrunner, how well he can use these glitches. Hiro Skip and CopyCat are banned.
  2. Arcade Any% RTA (RTA is main timing method, with loads or without, admin decides). Glitches like avoiding lasers, avoiding arrows, jumping 7-8 times should be allowed, as they can be used in each level. It depends more on the version of the game (Kula World, Roll Away or Kula Quest) and the skills of the speedrunner, how well he can use these glitches. Hiro Skip and CopyCat are banned.

If we speak about Arcade Any%, you need to play ALL 150 levels. For me, this is the most important thing in Arcade Any%.

  1. Arcade Any% Glitched. All glitches should be allowed. Hiro Skip and CopyCat too. It makes no sense to add RTA here, since these glitches (Hiro Skip and CopyCat) are designed to reduce IGT time.

But I'm not sure about the Final. Yes, maybe there should be "The Final" and "The Final Glitched". But only RTA should be there (with loads or without, admin decides). IGT isn't showed after the Final.

And I suppose we don't have to use livesplit, I think it's optional.

Edytowane przez autor 1 year ago

@Jujstme

In general, load times should never be counted. But let's look at the categories.

  1. Arcade Any% should stay as it is, with IGT as its main timing method and collecting fruit to reduce it. NOT as dima proposed using glitches, that makes the glitched category pointless. Keep the leaderboard as it is with RTA and IGT present

  2. Arcade glitched% - I'm fine with both ideas. Not loading any savefile and doing all glitches or with loading savefiles. If we choose to exclude runs with loading savefiles, that would make my run disqualified, but that's okay, since it isn't a good run anyway, i did that one for fun. RTA isn't really something we need in this category, IGT is fine.

  3. The Final loadless and 4. The Final glitched loadless is fine to me as you said.

Autosplit is okay for me if it excludes the load times, that's a good solution to me. Real hardware runs need to be calculated manually then, but if that's okay for you, i'm okay with it.

Edytowane przez autor 1 year ago
Ukraine

Agree with RealRaptoReX, that RTA shouldn't be in Arcade Any% Glitched. RTA is simply mindless in this category. RTA as IGT should be in Arcade Any%. Admin decides, with loads or without (I mean loading time).

And I suppose we can load savefiles. It's CopyCat glitch. And if we use this glitch, then we can use it when we want. But no quicksave 😁

Edytowane przez autor 1 year ago
LePalmypede to się podoba
Finland

Hi, sorry for reviving an old thread. I've recently taken up Kula World speedrunning and have been confused about the categories.

I have to say I side with Dima here with separating IGT and RTA properly as different categories. In essence they are different categories and you employ different strategies depending which timing method you focus on, Dima already explained the reasons so I won't repeat those here again.

To me it looks like people generally agree that they are different categories but because the categories are sort of there already there's no need to change anything. However I'd say that in practice having IGT and RTA listed in the same category creates some annoying problems:

  • By looking at the scoreboard I can't tell whether the runner aimed for a good IGT or a good RTA time
  • Related to the previous one: when I submit a time to the scoreboard I don't really know how well I did because the other times on the scoreboard may or may not have tried to achieve the same thing I tried to achieve with my run, I can't really tell
  • Because each runner can have only one time listed on the scoreboard of a category, runners cannot have a good IGT time and a good RTA time (there's no one run that achieves both)

Hello, I also have to revive this old thread. Are we really allowing glitched runs in "The Final" normal runs list? Isn't that a bit unfair to those who want a glitchless competition. In general, no glitch should be allowed in any run unless stated that it is allowed. These glitches kinda ruined this games run in my opinion.

Finland

(Sorry for the gigantic post, had to split it up because went past max length :D)

I'm not going to be standing in the way if we decide to have a glitchless and glitchy category for The Final, and I'm fine if the run I did is transferred to a category something like "The Final glitched". However I think having a "glitchless" category comes with many issues...

So let's go down the route that we have a "glitchless" category. That means we have to define what are the banned glitches in this game. This results in a lot of tough calls which are ultimately subjective opinions on what you feel like was maybe a developer error. Some examples of what I think are some of those tough calls:

  • At 2:30 ( ) I jump through three heater blocks. This clearly breaks the logic of the puzzle, the point probably was that going through that way should be impossible and the player is supposed to find a way to the exit through the teleport. But all I did was simply move fast... In particular, I do the corners quickly by turning and immediately jumping after that, which is what people do in their runs all the time. If this is a glitch is the turn + quick jump a glitch everywhere and should not be done anywhere in any run? If it's a glitch only in certain situations we need to define all the situations where it is considered a glitch, which is a rabbit hole of it's own...
  • At 2:07 ( ) I again break the the point of the puzzle with some jumps in between the buttons that activate the lasers, as well as picking up the key while jumping off the trampoline even though they're on different blocks. I'm just jumping a little off-center of the blocks, that's all, and again that's something everybody are doing all the time, in fact it's really difficult to jump dead-center of the block every time. But it did kill what I think was the point of the level that you need to go back and forth to activate the lasers in a complicated way that ultimately lets you go through. Glitch or simply precise movement? And similarly to the point above, if this is a glitch is the very basic movement strategy of jumping off-center of a block a glitch everywhere or only in certain situations?
  • At 4:29 ( ) I do a series of quick movement on crumbling blocks, which lets me finish the level using a shorter route that probably was intended. Looks a bit more to the glitchy side to me but a case could be made that I'm again simply just moving fast enough not to fall with the crumbling block. Also, if this is a glitch, is this a glitch when adeyblue in their Arcade Any% WR does a 180 on a crumbling block? At 3:22: ( ) Looks like both tricks work on the same principle that a crumbling block lets you do about two actions if you're quick enough.
  • This is from SaturnKai's TAS, 2:46: ( ) SaturnKai jumps precisely over the pill, saving about half a second. I've managed to do this myself outside runs too so it's not a TAS-only trick. Here the impact is small, but should be considered anyway to be consistent, and maybe somewhere in the game a jump like that could skip say the pill that makes you jump two block instead of one, breaking the game more seriously. Simply a precise jump or a glitch?
  • Pause skip is IMO pretty clearly a glitch but causes another problem: if you manage to pause the game accidentally during a run, is the run immediately invalidated? You could potentially gain unfair benefit...
  • There are glitches we don't know about and might be doing accidentally all the time. Who knows if dying once in level 12 makes the ball go slightly faster for the next 5 levels? How do we deal with the situation when somebody ultimately finds this out and we realize some old runs have accidentally performed the glitch and gained benefit from it?
Edytowane przez autor 4 months ago
Finland

And so on and so on, you can probably see that this gets exhausting quickly. Even if you might have a clear opinion which ones of those are glitches other peoples opinions will differ and the discussions will drag on and any potential glitch like that requires month(s) long discussion here on the forum to reach a concensus if that particular trick can be included in the glitchless category or not. Finally the result is an ever-growing list of pretty arbitrary rules, which will be annoying in many ways:

  • 100% of the runners will be dissatisfied with some decisions on the list
  • Whenever a runner discovers a new trick it needs to be brought to the forum to be decided whether that is a glitch or not, resulting in new long discussions and potentially changes to the rules -> this doesn't really encourage people to find ways to beat the games faster
  • the rules are not constant and might change, so runners need to stay on top of the ever-changing rules
  • new runners are met with a long, discouraging list of arbitrary rules to memorize, and might end up never trying because of the effort feels so dull
  • good runs will be invalidated because the runner didn't remember the subpoint three of glitch 19
  • good runs will be invalidated because a runner accidentally performed a glitch that is easy to reproduce accidentally
  • a new glitch is found, the community decides (after another month long discussion) it is banned. Later somebody realizes the glitch was actually used already in a 5 year old run. Do we invalidate the run that we've been okay with for 5 years or then randomly allow this one glitch?

All in all, we have to spend tons of effort to maintain a separate layer of rules on top of what the code lets and doesn't let us do, and it will lead to lot of ambiguous situations.

The alternate viewpoint is so much simpler: we all have the same code in our hands that with same input produces the same output. Let's just embrace that and see who gets to the end fastest. This way the rules stay simple, runners are encouraged to find faster ways to beat the game, and (in my opinion) running is more fun and less memorizing which strategies are banned and endless discussions of which tricks to ban.

Most games seem to have come to this conclusion and allow glitches. See any Spyro run for example, within 5 seconds of the start of Spyro 2 Any% WR we're out of bounds ( ). Basically all games I've ever taken a look at seem to be fine with glitches, including other popular games of the era like Spyros, Crash Bandicoots, Crocs and Tony Hawk's Pro Skaters.

All this said, ultimately the point is of course to have categories that make speedrunning the game fun, so if people want to run and maintain the rules of glitchless category then of course go for it. I'm just writing this to make sure you know the implications and to defend the point of view that there should at the very least be a category that allows glitches :)

@haiseva_iljetys I need to thank you for your very broad view of the whole situation. You asked questions that should be asked and even provided some examples of stuff you talked about. For me, there is a clear situation here though.

I don't know about every game on the planet and how the speedrun rules are, I can only talk about metal gear solid on the PS1 which has so many damn categories that i could loose count. At the same time, even I did a run and that felt good, to be on the list of runners. The only thing I needed to do is to look at the emulator settings, rules and then I went for it. It can be the same for this game as well.

I'm for a glitched and glitched category in both the main game and "the final", because SaturnKai did a lot of research on the glitches that are available and I think we found most of them. And yeah the copycat glitch, pause glitch and so on are that - glitches, clearly. If a run accidently pauses by hitting the button, you can just look at it individually, because that doesn't mean you automatically performed a glitch.

As for the movement - all your mentions are, in my book at least, using the games mechanics to its limits. Even the game we develop at the time - RollingCube - uses the same movement system as Kula World so these things will be possible there as well. No run should be invalided because you turned 2 times on a breaking block or you jumped halfway between blocks or even skipped an item or block type because of an early jump. That's what all speedruns do in a way - break the game or at least do things that weren't intended by the developers. Crash Bandicoot comes to mind, slide-spin-jump comes to mind. Or even Metal Gear Solid where you weight-shift and avoid lasers of Metal Gear.

So the only thing i think we really need is clear rules. And if we have that, everyone can just go from there.

Rules could be like this for glitchless run:

-Any Glitch providing a faster time or outright skipping of levels or hazards is prophibited -Timer starts when you hit the arcade mode button and ends when reaching the last exit in level 150 -Movement mechanics like jumping between blocks and skipping blocks by early jumping are allowed as these are considered non-glitches

Statystyki gry
Obserwujący
34
Przebiegi
26
Gracze
9
Najnowsze wątki
Opublikowano 4 months ago
games:thread_reply_count
Opublikowano 4 years ago
games:thread_reply_count
Opublikowano 9 years ago
1 odpowiedź
Moderatorzy