I just noticed you submitted a run where you said you got way less viruses when you did Raccoon City last. I haven't tested that, but I will today and let you know if that's the same for me too. The '?' abilities can be some of the most powerful, as you have a reasonably high chance of getting your character's special ability (I don't know the numbers for it). So if it's possible to limit the amount of virus spawns for '?' abilities, that'd be a huge game changer.

Also, on character choices. For high combos I think Cosmi and Itazura must be two of the best, but Baron Von Proptop's ability to tie everyone up could have a similar utility in the get-to-500 stages AND be good for the battle stages too. I'm not sure about the other characters. Princess Cleo's ability to summon a UFO would be good during a casual playthrough, but unless we find a way to glitch it I don't think it's necessarily as good for speed as it could be. Same applies to Becky's ability. Magnetron's ability would be less useful in Cocotown but very useful in Raccoon City and also potentially Pranksylvania if the AI is off on the other side of the map while you're dealing with people on your side. I think Magnetron could be good but would require a different strategy to make good use of his ability compared to the obvious utility of Itazura and Cosmi's super-form abilities.

I've gotta believe that glitches are possible! Glitches that stop AI movement, or glitches that break game time would be perfect. I haven't found a way to clip out of stages yet, though I'm not sure that in itself would be helpful. It makes me wonder actually if maybe an emulator glitch% category would be a good idea? running the game at ridiculous fps to break it to the max might be an interesting and different way to play it fast.

thank you so much dude! :D

Hey JJ_Klooney, I was thinking it'd be cool to have some custom themes for the Under The Skin page. I did this for a game I run called Into The Vortex (https://www.speedrun.com/Into_The_Vortex/full_game), and I think it'd be great for this awesome game to have a snazzy theme of its own. I'm also considering uploading a whole bunch of runs for this game next week, now that I've sorted out some technical issues that got in the way recently. I know you check this page regularly, but i was wondering if you'd be willing to let me co-mod this game with you, so that I don't have to wait to update any records I may beat (this'll be particularly applicable for normal and hard mode, where the times might change quite rapidly), and so that I can edit the theme in detail myself. I understand if you'd rather be the only mod, but I'd be very grateful if you DID let me mod, as i feel there's a lot I can add to this game and it'd save us both the hassle of me pestering you to do stuff.

Thanks again,

Rainbow Lemon

Ran into a million and one technical issues while recording my run, but I'm gonna upload it today. I also discovered something that'll make battle stages even easier:

If you and the AI are running towards a UFO, even if you are BEHIND the AI, you will be the one that gets to refresh your disguise. As long as you are right next to the AI, you will always outprioritise them for UFOs.

Hope this helps someone!

Today I managed to beat the game in 33:38, but I didn't record the video cos I never expected to get so far below my previous time! I'm gonna record the run tomorrow and explain what I do, but here's the jist:

Just as JJ_Klooney discovered, you can reduce the battle stages well below 10minutes. I managed to finish one in 2 minutes and the rest (apart from Frontiersville) in under 5. The trick is this: it's not about attacking the cpu constantly - you just have to deny them UFO spots. If you systematically deny them the ability to disguise themselves, they will eventually be reduced to 0 coins just from humans beating them up (along with you beating them up of course). This actually makes Frontiersville the most difficult area in the game because you can't tell where a UFO is at any given time, though you can at least follow the AI around and try to cockblock it that way.

In my video I talk about routing, and I think that's even more important now we know how to get past the battle stages. You should do both Pharaoh Island and Frontiersville early on, though I do recommend you do Coco Town and Pranksylvania one after the other, to make best use of Cameras and Microphones. The reason I recommend Frontiersville and all that is because Bowling Balls from Pharaoh Island and the Rifles from Frontiersville are both excellent weapons for attacking the AI at a distance and stopping them from getting away. After that it's all down to personal preference but Raccoon City is another excellent choice due to the Missile Launcher you get there. I used to say leave Big Booty Bay till last, but the cannon could actually be useful for chasing an opponent.

TL;DR my recommendations are: stop the ai from getting to UFOs and do Pranksylvania straight after Coco Town.

Defining it by the bonus unlocks sounds like a good idea! On the 15 hour thing - I'd be happy to do a run like that, but as it stands now (and I've had to tweak the route a lot and am nowhere near being able to do a proper run yet as I'm realising there's a lot of optimisation I can do) it looks like such a run could easily be completed in under 10hrs, potentially under 5, and definitely in under 15. What I mean is, if we made it so people had to literally meet all the requirements, then the scoreboard would be a series of 15hr completion times. I mean it'd be interesting to see the bonuses attained efficiently anyway, but if there was room for improvement on the run time then I think that'd lead to more healthy competition and thus more creative ideas about how to do it as fast as possible. It'd only be a small deviation from the requirements of Ryu Book 1, and that'd be the only exception to 'unlock all the bonus stuff' that i'd suggest.

Use this to communicate with me about the game or to discuss strategies and stuff with any other runners that come along ^^

Hey guys! When I submitted my run I wasn't sure what was gonna happen cos this game had been quiet for so long, but holy cow we have two new runs already! That's absolutely awesome and it's given me motivation to wanna really work out how to get the time on this badboy down. i notice that JJ_Klooney actually managed to beat the battle stages in under 10, so I created this topic for us to share ideas and stuff!

I haven't got anything cool to add yet cos I've only just now logged in to discover this awesomeness, but i'll let you know if i find anything fun in the future ;P

I totally forgot about that! It seems they actually did further tests and as you say, the lack of elemental hit in the NSTC-J version is why Killer Snake breaks later there than in NSTC-U - i.e. it requires a similar amount of damage, but elemental hit reduces the time required to achieve that breaking point. I have no idea exactly how iriirisu did these tests mind you, so I should probably ask them xD it looks as if they calculated the amount of damage racked up on an enemy before it protect broke, so that would only take account of damage required, and not how quickly you can dish out that damage to reach protect break.

I've been talking to iririsu on twitter, and they've been testing to see the differences between Japanese and North American NTSC versions. Apparently, the key data bugs in the any% speedrun actually protect break later in the JP version than in the NA version of the game. I speak Japanese (iririsu doesn't speak much English), but google translate should actually do an adequate job for non-Japanese speakers of clarifying which numbers in this tweet string represent what:

https://twitter.com/KyokoIrisu/status/865899677002022912

In short, Red Wyrm breaks at around the same time, Killer Snaker breaks much earlier in the NA version, and Skeith potentially data breaks slightly earlier in the JP version.

The areas visited component of Ryu Book 1 would still be required, but you're right that leaving out the time-based bonuses would render it not truly 100%. Would labelling it something like Ryu Books 99% or even Ryu Books% be more satisfactory? Honestly I'd be happy for it to be called whatever you think would be best - I'm just glad you like the idea of it! ^w^

In fact, leaving out the 100% is definitely a good idea because obviously, completing the Ryu Books doesn't require completing the goblin tag events or, I believe, the second Gardenia quest. It also would not require the bonus dungeon - though it wouldn't rule it out, assuming that we didn't make it a rule to finish the Ryu Books before you defeat Skeith.

Sorry to double post, but just to be clear, cos I think I wasn't really that clear in the first post, the kind of run I'm suggesting is one where you fulfil all the Ryu Book bonus requirements (e.g. as listed here https://www.gamefaqs.com/ps2/551240-hack-infection-part-1/faqs/22282) apart from the ones to do with total amount of time played. Basically it'd be a 100% run, but 100% as measured by the in-game mechanic of Ryu Books. The timesaving would come from careful routing, and by combining the goal of finding grunty food with the goal of visiting areas, opening chests etc. Also, sorry if the forum is the wrong place to suggest something like this. I'm not new to .hack but I'm very new to speedrun.com

Has anyone ever thought of attempting a No Disorder run? Obviously the disorder is the way to achieve the true fastest possible run of Knytt Underground, but going through the process of mapping out a no disorder route right now, it's looking like there'd be quite a few different routes to get all the items and get to the bells efficiently. If such a run were attempted, would it be better to be a full game run, or just a run of chapter 3?

I'm planning on recording a .hack//Infection run next week where the goal is to 1) finish the story as usual as well as 2) max out the Ryu Books (i.e. get a game '100%'). This would involve meeting all the potential party members for .hack//Infection, opening 300 chests, getting 15 Dungeon Clears, 10 Field Clears, and raising 3 types of grunty, among other things. I was wondering what you guys think of that as a potential category. Would people be interested in seeing a run of that kind? Should I tweak it before I go ahead and record it?

I think that's what I love most about LSD - it's like a huge maze that you accidentally shape around you. I'm definitely gonna have to experiment more with trying to link to useful points of Kyoto. Also, I read on the wiki a while back that Monument Part was meant to link to specific locations, but I've never really had much luck with manipulating the links there. What are other people's experiences?

It's not a consistent link inasmuch as you have to hit it at the right angle to get the correct warp, but in a run I'm about to submit to you now, I found a way to make it so the girl in the Black Space warps you right outside Temple Dojo. As Temple Dojo takes a lot of walking to get to normally, this could potentially be a very useful warp. I managed to warp right in front of Moonlight Tower as well but that one's definitely inconsistent.

the Under The Skin community has no runs submitted, and the moderator has not logged on in 3 months (the user account was made 4 months ago). I've submitted a run, and I attempted to contact the current moderator (there's only one) on discord and on twitch, but have received no reply. I would wait longer but I noticed your post about removing games with no runs or active mods, and I thought it best to post here before that happened, requesting that I be made a moderator of Under The Skin in addition to the current moderator, so that I can add my run and update the game categories etc.

thank you

I can confirm Kite runs incredibly slowly in NTSC-J haha. Also, one other good thing about the J versions of .hack infection, mutation, outbreak and quarantine is that they don't seem to have the weird control layout switch that the PAL versions have. So circle buttons stays as your confirm button throughout the series.

I think the .hack series is just one of those game series that doesn't get the attention it deserves in terms of like, the size of the fanbase. More runners equals more of the games done and all that kinda stuff. It could also come down to availability of the games themselves. I've actually found the available runs quite inspiring and plan to do a 100% run of Infection as well as a full series run of the first series of games (i.e. not GU). I'm not competitive, so I'm not gonna try and beat anyone's time, but if anything the lack of coverage of the games and of categories within those games is an opportunity, friend! You get to plan out those first routes for others to work with! That's why I really think you should go for the GU trilogy NG+, cos it'd add a lot to the available .hack speedrunning resources and vids and hey, you might even inspire someone else! I wish you good luck whatever you decide to do.