Today I made an experiment/hunt glitch to try to find a setup to get both keys at the beginning of the level. I expected that this setup existed because there's also glitches for Rapid and Laser shots, but just never did the proper testing, until now.

Here's a video with the full experiment (very long and I forgot to put my keyboard display, but I uploaded anyways as a documentation of the test):

And here's the short version with just the results and some explanations:

With this we can avoid A LOT of almost every level.

I'll be working on re-routing the different categories.

TL,DR: just type 'L' 12 times and you'll get both keys

Ok, here are my final thoughts in the subject:

• I think it's better to not add more categories since the comunity is quite small to have so many of them (currently there are 3 difficulties times 4 episodes, so 12 individual levels. Also 3 difficulties for full-game. With the new glitch we'll almost duplicate the # of categories).

• In my opinion, we should just maintain the same name categories and just change the rules to allow the level skip but ban the usage of savefiles, cheats, and accidental bonus of both keys. With that, in individual levels: Episode 1 will be the same, and the others will be open for new runs with the level skip (yeah no more level 1 of episode 2, 3 and 4)

• If you agree with the points above, we still need to discuss one last thing for all-episodes: do we need to 'see' the credits of all episodes? or just 'beat' every level? This makes a huge difference, because if we decide the second option, then we can use the last level of Eps 1,2 and 3 to set up the glitch for the next episodes, just quitting right before the credits of the episode. While in the first option, we need to set up the glitch on level 1 ep 1 after every episode.

I tried the glitch and couldn't find a way to get out of bounds.

effectively, The character auto-stands up in narrow places. I scan the whole area 1 and 2 for exploits, found nothing. I'll probably explore the other areas during the weekend to see if we have something promising.

great discovery!

This could be a first step for setup out of bounds stuff. Johny look quite stuck there anyways xD. I'll try to look for more places where this works.

btw, I know that your character auto-stands up when is going backguards while crouching and the legs aproach a ledge (if you go foward, it performs a flip animation in the ledge). So maybe the explanation to this is that the clone man pushed your character legs out of bounds (so, into a 'ledge') making him auto-stand up.

With two Categories "Any%" would allow the glitch in both individual episode and all episodes, while "no level skip" would be like the runs I already have. And yeah, for episode 1 I think isn't necessary to make another category (but we can do the NG+ if someone wants to run that).

About the off-topic (thanks for bring that up, it's super important). The Random Bonus Glitch (RBG) gives you any bonus item at ANY random moment, this includes rapid, laser, full heal and both keys (no super jump sadly). The only way to "manipulate" this is at the beginning of the level with the hold up or two up inputs for rapid and laser, I tried many ways to manipulate getting keys but I never succeded -__-. I got both keys randomly at some runs but is like 1 out of 30 runs and at a complete random moment (when might not be useful). So for me at least, if think it would be unfair to allow it while we don't have a consistent way to do it (like rapid/lasers at the beginning of the levels).

And of course, if there is away to do it consistenly, that's several minutes choped from ALL categories xD (specially ep. 4)

little update:

I rewatched all my videos of glitches and, effectively, every level skip was at the beginning of a run (i.e. at level 1 of some episode, so I probably reset the previous run right at the end of other level and never noticed xD). I quite remember having a random level skip in level 4 of episode 2 (right after beating level 3, just at the beginning of level 4, no saves, etc), the problem is that I have no proof because I don't see it on my videos (probably was during practice x_X). So I guess I can't prove there is a random level skip =/.

I think is ok to just ban savefiles and cheats, allowing the new glitch without using savefiles for all categories

normal play, and on multiple levels, I'll upload a video in my channel with some of them today

Shoutouts to Melonax for the discover!

I personally prefer the second option, because banning level skip also tackles the problem of accidental level skip (sometimes when you enter a new level you just get warped to the next one inmediately). The accidental level skip is absolutely random as far as I know, and rare enought to just broke a category (I have some "PB's" with minutes ahead of my normal PB's even with bad execution, because of a random level skip).

In this case I'll stick with "Any% No Level Skip" and the broken "Any%" for single-episodes.

And for all episodes I think it's a great idea to just ban savefiles and accidental level skip, so that way the first levels can be skipped and we don't need to create more categories for full-game

pd: If it sounds too inconsistent, then I'll stick with option 1, just ban savefiles and accidental level skip from single-episodes and all-episodes

It actually works with every level regardless of episode. Like: every episode will take the same amount of time separately with the new glitch (just doing level 1 Ep 1, 9 times).

Also Moderate and Easy have the same strats on Lvl1 ep1, just hard will make a little difference.

The "All Episodes" full-game category with the glitch will be doing lvl1 ep1 36 times (if the objective is to reach the credits of every episode)

With the brand new gamebreaking glitch, we need to redefine the rules or categories of the game. I see two options:

1: Just merge everything, and allow everything except cheats

2: Make an Any% category which allow everything except cheats, and convert the old categories to "All Stages" in which is not allowed to have accidental wrong warp or the new save warp (and of course, no cheats)

wow, I'm glad Looper Land has the highest encounter rate. Also Aika can kill most loopers with Psyclone having high magic, but I don't know if in 100% speedrun she'll have that much at that point (Focus Robe helps a lot). And with all that leveling up soon or later Rain of Swords on first turn will be a thing.

On the other hand, I think I can do some time test on grand fortress. I think the big question will be what saves more time, faster 2500 battles or having more level in late game so the bosses are even faster (also a LOT of money, meaning more paranta seeds for vyse)

That's a good point. And yeah, late game it's pretty broken in general xD

In regards to the 2.500 battles (I guess that's what takes the most time), you think it's just better to wait for the end of the game to grind? I guess on a high rate place with a black map (soltis?).

I was thinking how optimal would be grinding at a fixed encounters place, like the flashlights zone in the grand fortress before the delphinus, one can just sit at the spawning point of the light and get encounters pretty fast. The downside is that we'll be less OP at the end but maybe it's still worth.

I totally agree about Elcian and the upgrades of crescent island.

And about the Three Secrets: Do you consider just "unlocking" them (fulfil the requirements), or also "getting" them (buy the Sky Fang, "discover" the Golden Hamachou, and beat Vigoro) ?

Quick question:

For 100%, do you mean just get "Vyse the legend" title and then beating the game?

or: do you also include "getting" the "Three Legends"? (Sky Fang, Golden Hamachou, beating Sky pirate Vigoro)

also: what about Elcian?

Update:

• Project64 2.x is effectively broken (framerate goes randomly faster), but is the only one without graphics and sound issues.
• Project64 1.6 randomly crashes with no reason if you play more than an hour (some kind of sound glitch).
• Most non-Project64 emulators have graphics and sound issues and also inconsistent framerate.
• Project64 1.7 is the only one that emulates roughly at the same speed of console. (Essentially the whole game, except it manages some laggy parts different). The problem is that it has graphics problems (blue shinning on mines instead of red and a weird graphic glitch when a body part is injured).

Conclusion:

• I'll recomend to use only Project64 2.x and 1.7 (if someone don't care about graphics issues).
• Emulators are still allowed but runs on emulator cannot have real time on the leaderboard, this way console is favored because it can submit real time.
• Of course my word it's not rule. So if there is a larger community in this game in the future, and that community wants emulators to have a separate category, so be it.

Thank you for posting this, I actually didn't knew about these issues.

I'll remove the emulator suggestion for now and I'll do several tests with lower versions and maybe other emulators (I tried some others, but most have problems with graphics and colors for this rom). As soon as I get back my N64 I'll also do a test like the one shown in the video to see the differences in framerate.

Those results will tell whether or not runs might be separated. Also (just for the discussion) the important measure is In-Game Time (because the game timer never stops, even at pause and cutscenes) and is independent of the frame rate.