Hyper Iria world record

[MarioNintendo]

Let's go

[Hakulyte]

Originally Posted by Hateandhatred

Hyper Iria news:

Haku dropped a run 3 seconds off any% World Record.

This statement is still correct, but the other way around now. (yeah, I just got WR on any%)

[MarioNintendo]

Good work! How much effort did that require?

[T-Force]

We got a scrap within a month of the race. Hot damn!

[Hakulyte]

Originally Posted by MarioNintendo

Good work! How much effort did that require?

Knowledge is power here. If you just try things blindly, you're unlikely to get a WR. Nearly every actions in this run has a clear intent and is trying to create a very specific outcome. Is the game mechanically difficult ? Well, for most of this run ? No. What if you want to do a "perfect" run ? Absolutely. What I'm saying is that it did requires a lot of effort because you need to constantly be ready for what's next until the end, but "pressing the buttons" is not more difficult than getting AAA on a difficult song on FFR. You don't need to vibrate a section or do some near impossible input combination under 1 second etc.

Boss fights have very specific strats. Both where I am during it and what I do is important. Too far away from a boss while shooting ? Well, I have bad news for you, your own bullets created lag and you're ruining your own run! So, yeah. It's mostly a battle of small adjustments. Unfortunately, I completely messed up the last boss in this run, but I did get the phase skip in Stage 4 so, yeah. (it involves shooting really fast before the game makes the boss invulnerable and create a 2nd phase. It's a run killer because you can lose 8~12 seconds here!)

Getting a time around 15:10 should still be reasonable with what's currently known. Anything under 15:10 with our current knowledge is going to become absurdly difficult. Me and hateandhatred are actively trying together to find ways to save more time to make the run easier. There's still many things we don't understand or simply overlooked. This speedrun is not on the same level as one with a leaderboard of 100+ players. I do believe someone could look at our times, improve incredibly fast and start getting runs close to us. The game is relatively short. We're not spamming frame perfect tricks over multiple hours. This should be a reasonable challenge for anyone as long as they break down what to do at every moment. Also, there's more than any%. There's segmented runs to practice each levels and there's also normal% and even extreme% where buying the right items between levels will change the outcome of every levels. Enemies have more HP and the meta for these 2 runs are completely different. You need a bazooka for normal% and you need specific grenades for extreme% or you will most likely not only lose time, but also game over. (there's currently no time for extreme% so, if any of you play it and beat it. You have a WR. Hateandhatred has beaten extreme% before tho)

So, RNG manipulation is a thing. The problem is that it's not fully understood and we're only scraping at the surface. There's no one who looked into this game's code and could explain in details what's happening. Lag in this game also exists. Enemy projectiles can create it (and yours as well!). There's strategies around choosing which enemy should be killed and how. Which ones should be left alive to optimize RNG etc. Which enemy should we knockback with a punch and then run away (because the death animation also loses time). If there's backtracking, killing an enemy could save time for the backtrack etc. (most of the important ones are the ones that shoots projectiles and add lag by simply existing)

A tricky mechanical part of the game is that some platforming involves doing either a low jump, medium jump or high jump. They're tricky to do because how long you hold the jump key quickly shift between each type of jump especially for low vs medium. How fast and how long you hold a key will decide instantly how high Iria jumps and this can lead to mistakes not to mention Iria also has a double jump. I personally failed a jump in Stage 4 in this WR because Iria couldn't reach her max height. I'm not even completely sure why it failed, but I think that it was because I just barely didn't hold the key long enough to make the game understand I wanted to do the highest possible jump. The turrets under me were shooting projectiles and I think it impacts how long you have to hold the jump key. So, yeah. A lot of things can go wrong.

So, it's a lot of small considerations like that which piles up on long term. The more you play the game, the more you can pick these things up. You just eventually get a feeling that something isn't right and you try new things until it does.

The game also has cancels. If you do a standing kick and press down like 0.3 seconds after, Iria will cancel her kick animation. If you do it too fast, Iria will cancel her entire kick and you will do nothing. So, this could be double edged. It took me a while to start doing them more consistently.

btw, the levels are also a maze. Most of them are actually 3-4 times bigger than what you see in our runs. You will have to memorize which door to take. You could spend 15min+ in level 1 and barely see the entire level. Fortunately, every door leads to a specific place so, there's no RNG for that. Also, you need to double tap forward to run. You also need to enter doors by aligning yourself with the center or Iria will lose time walking toward the center. There's a tons of doors so, this adds up. btw, every time you get out of a door, you need to double tap again to run. If an enemy hits you, you will stop running and walk. If you get hit while walking, you will start running. This can be troublesome when a lot is going on. Climbing is easier while walking so, I recently intentionally pressed down in a place in Stage 4 to make Iria intentionally walk. H&H also recently found that there was a door we didn't need to go into while doing Stage 2. We could just go around the level and save 2 seconds. As long as we keep finding small things like this, this game will remain interesting to run imo.

I currently believe 15:00 is a wall that both me and H&H haven't figured how to overcome yet. 15:00 ~ 15:10 used to be unthinkable and now it's starting to be within reach. H&H was hoping to get a time under 15:00 before the live. I have no clue how that could be possible, but I'll try to do what I can to get closer.

This is starting to feel like a puzzle game tbh.

[Hateandhatred]

Originally Posted by Hakulyte

Knowledge is power here. If you just try things blindly, you're unlikely to get a WR. Nearly every actions in this run has a clear intent and is trying to create a very specific outcome. Is the game mechanically difficult ? Well, for most of this run ? No. What if you want to do a "perfect" run ? Absolutely. What I'm saying is that it did requires a lot of effort because you need to constantly be ready for what's next until the end, but "pressing the buttons" is not more difficult than getting AAA on a difficult song on FFR. You don't need to vibrate a section or do some near impossible input combination under 1 second etc.

Boss fights have very specific strats. Both where I am during it and what I do is important. Too far away from a boss while shooting ? Well, I have bad news for you, your own bullets created lag and you're ruining your own run! So, yeah. It's mostly a battle of small adjustments. Unfortunately, I completely messed up the last boss in this run, but I did get the phase skip in Stage 4 so, yeah. (it involves shooting really fast before the game makes the boss invulnerable and create a 2nd phase. It's a run killer because you can lose 8~12 seconds here!)

Getting a time around 15:10 should still be reasonable with what's currently known. Anything under 15:10 with our current knowledge is going to become absurdly difficult. Me and hateandhatred are actively trying together to find ways to save more time to make the run easier. There's still many things we don't understand or simply overlooked. This speedrun is not on the same level as one with a leaderboard of 100+ players. I do believe someone could look at our times, improve incredibly fast and start getting runs close to us. The game is relatively short. We're not spamming frame perfect tricks over multiple hours. This should be a reasonable challenge for anyone as long as they break down what to do at every moment. Also, there's more than any%. There's segmented runs to practice each levels and there's also normal% and even extreme% where buying the right items between levels will change the outcome of every levels. Enemies have more HP and the meta for these 2 runs are completely different. You need a bazooka for normal% and you need specific grenades for extreme% or you will most likely not only lose time, but also game over. (there's currently no time for extreme% so, if any of you play it and beat it. You have a WR. Hateandhatred has beaten extreme% before tho)

So, RNG manipulation is a thing. The problem is that it's not fully understood and we're only scraping at the surface. There's no one who looked into this game's code and could explain in details what's happening. Lag in this game also exists. Enemy projectiles can create it (and yours as well!). There's strategies around choosing which enemy should be killed and how. Which ones should be left alive to optimize RNG etc. Which enemy should we knockback with a punch and then run away (because the death animation also loses time). If there's backtracking, killing an enemy could save time for the backtrack etc. (most of the important ones are the ones that shoots projectiles and add lag by simply existing)

A tricky mechanical part of the game is that some platforming involves doing either a low jump, medium jump or high jump. They're tricky to do because how long you hold the jump key quickly shift between each type of jump especially for low vs medium. How fast and how long you hold a key will decide instantly how high Iria jumps and this can lead to mistakes not to mention Iria also has a double jump. I personally failed a jump in Stage 4 in this WR because Iria couldn't reach her max height. I'm not even completely sure why it failed, but I think that it was because I just barely didn't hold the key long enough to make the game understand I wanted to do the highest possible jump. The turrets under me were shooting projectiles and I think it impacts how long you have to hold the jump key. So, yeah. A lot of things can go wrong.

So, it's a lot of small considerations like that which piles up on long term. The more you play the game, the more you can pick these things up. You just eventually get a feeling that something isn't right and you try new things until it does.

The game also has cancels. If you do a standing kick and press down like 0.3 seconds after, Iria will cancel her kick animation. If you do it too fast, Iria will cancel her entire kick and you will do nothing. So, this could be double edged. It took me a while to start doing them more consistently.

btw, the levels are also a maze. Most of them are actually 3-4 times bigger than what you see in our runs. You will have to memorize which door to take. You could spend 15min+ in level 1 and barely see the entire level. Fortunately, every door leads to a specific place so, there's no RNG for that. Also, you need to double tap forward to run. You also need to enter doors by aligning yourself with the center or Iria will lose time walking toward the center. There's a tons of doors so, this adds up. btw, every time you get out of a door, you need to double tap again to run. If an enemy hits you, you will stop running and walk. If you get hit while walking, you will start running. This can be troublesome when a lot is going on. Climbing is easier while walking so, I recently intentionally pressed down in a place in Stage 4 to make Iria intentionally walk. H&H also recently found that there was a door we didn't need to go into while doing Stage 2. We could just go around the level and save 2 seconds. As long as we keep finding small things like this, this game will remain interesting to run imo.

I currently believe 15:00 is a wall that both me and H&H haven't figured how to overcome yet. 15:00 ~ 15:10 used to be unthinkable and now it's starting to be within reach. H&H was hoping to get a time under 15:00 before the live. I have no clue how that could be possible, but I'll try to do what I can to get closer.

This is starting to feel like a puzzle game tbh.

I definitely 100% don't believe in RNG manipulation. There are a few things we can influence but the randomness, as far as I can tell from practice, there's just nothing we can do about it but plan for every outcome, there's some stuff you just can't make consistent with human methods. The flyers in stage 1, we literally hold right until they spawn, and they consistently do whatever they feel like doing based on some obscure intangibles like run seed and global timer.

I dunno about grenades, even in extreme%, I beat it with them but I only did it a few times for the memes, not as a speedrun. There's just too much risk involved with middle stages, and frankly I think we can just find a way to be a bit safer and take out everything without eating damage. I don't think there's any damage we take intentionally that we can't avoid by slowing down, and the only differences with extreme% is how much health enemies have and how much damage they deal. They don't change beyond this.

I think I have a good idea already on how to reach sub 15 personally. The issue is more with execution and it has to be unreasonably tight for it to happen, but I see it. I'm 100% convinced a sub 15 sum of best between us is all but guaranteed. Also, I personally have been believing in sub 15 since around AN last year, no joke. I just haven't been playing Iria. It's not something I think I can just drop for free, I think with the current knowledge, it could happen if *everything* went right

Otherwise, I agree! And we're entering the anti lag era of the game for sure. There's a lot to study when it comes to that, and I really need to go back into practice mode and do a lot of testing. I think I can improve stages 2-3-4 a good deal, stage 1 will take a lot for me to do anything like what Haku does in it, stage 5 I think I can make a few things happen, but there's so little to squeeze out of that one.

It's kind of fun how the focus has evolved. For me, there was the cancelling stuff, then the door entry stuff, then there was the jump drop stuff, and now it's all about lag in my mind, and I think we haven't even scraped the surface of that.

That being said, congrats on beating the any% WR! It was a long time coming and it makes me extremely aroused to get into Iria again. Now, if only GPW3 can end...

Edit: Oh yeah, the levels. 3-4 times bigger, that's right on average, but stage 1 is... I have no idea, at least 10x bigger. It's enormous!

[Hateandhatred]

Well, I'm a Kaizo Mario expert now, so it's time for me to go back to Iria, took me a bit over 6 months to go from noob to amaze, I'm really hyped.

How hype has this weekend been, seriously?

[Hakulyte]

Improved to 15m12s today even though I should really fix sound soon.
https://www.speedrun.com/hyper_iria/runs/m7lq38ey

Anime Speedrun Festival is happening next weekend. Both their twitch/youtube are @animespeedruns.

Should be happening at 1:05am EST according to schedule.

[Zyxel]

Originally Posted by Zyxel

i cant wait for the video essay that comes out 3 years from now detailing the incredibly competitive speedrun ladder of "Hyper Iria" that features this thread for one of the games world records

i am manifesting this line
@summoningsalt

[handpanda]

Originally Posted by Hakulyte

I'm surprised there's not some sort of FFR speedrunning group.

Hyper Iria is a great game to learn speedruns.

H&H made me try it out and I'm getting close to under 17 min now.

I was just thinking that if I can do it, others probably can too.

You've come so far~

[MarioNintendo]

It's happening tonight!

[Hateandhatred]

https://www.twitch.tv/speedgaming in 3 hours!

[MarioNintendo]

that was awesome

[Hateandhatred]

We had a crowd of about 1000 people watching, it was hype!

[ShinyStardust]

We were preparing a new strat to achieve unlimited ammo. Throwable grenades are related to that (the onion one is the only one that doesn't work, the other two does).
Btw, I'm trying %any category right now. I still have to do more runs.

[Hateandhatred]

The first ever Extreme% speedrun as a proof of concept:

This hilariously bad, never seen before glitch: