JapanSaw this post several weeks ago, and I thought I’d give shell hyperspeed a try. Turns out that you only need two frame perfect inputs after initiating the glitch to finish 5-3 with 268 IGT remaining. Minuscule time save, I know, but it’s RTA viable in my eyes. One benefit of this strat is that it gives you a mid to high 268, meaning that even if you only get one frame-perfect jump after setup, you can still finish with 267. Here’s the setup:
Minor update on what I found:
As I stated previously, the toad pattern in the cutscene(will be referred to as TP for convenience) is a visual indicator of what seed you hit. As such, the same TP will always give you the same red block RNG so long as you play 1-1 the exact same way. Note that multiple "good" TP exist even during manipulation due to slight differences in the timing of your game selection input from the DS menu.
Now for a not-so-easy and very convoluted solution to inconsistent RNG manip: once you’ve found at least one setup for your own cartridge, keep note of the TP in the cutscene (i.e. film your gameplay/take a pic of the top screen right as the cutscene ends) for the successful attempt. From that point on, keep observing the TP at the end of the cutscene for every attempt and take notes (again, filming the process makes your life much easier)— chances are, you will occasionally hit a different seed with different TP, which will sometimes have an undesired outcome (will be referred to as bad RNG with bad TP). Once you’ve sorted most of the TP into good TP and bad TP, start playing differently for bad TP—change the number of double/triple jumps, hit a different number of coin blocks or do both. With enough experimentation, you could have a 100% consistent RNG manip so long as you know the 1-1 playstyle for both good TP and for every bad TP. I do not recommend doing this since the process itself takes way too much time, and you’d have to memorize which TP corresponds with which 1-1 method. In theory, this is the best way to do the manipulation but it’s also very impractical. Until someone finds a 100% consistent method for manip setup, I’d like to think that this is the best that we could do.
Console: DSi
Region: J
Date set: 01/01/2000
Time set: 00:00
Brief overview: set the date, then set the time to 00:00. As soon as the "made changes to settings" text pops up, I hit the power button to go back to the DSi menu. The colon on the digital clock (upper right on top screen) should be on the "on" state. Counting this initial "on" state as the first, I select the game on the third "on." I find that it's more consitent if you select the game a tad later than the third "on." I usually mash A to select the new file. 1-1 is very basic movement
Video:
Upon testing my setup, I noticed that the roaming toadstools at the end of the opening cutscene seem to have very few patterns (about eight or so) when I manip. I then figured out that the same pattern of toadstools will always give you the same red block RNG. While I know two of the "bad" patterns of the toadstools and one "good" pattern (shown in the video), I have yet to figure out all the "good" patterns
I’m extremely iffy when it comes to emu/VC runs because I have no experience in the technical side of this game, and so I can’t say anything tech-related things for certain. However, what I understood from the discussion so far is that emulator runs are unreliable for accurate timing, even with the built-in recording feature. No matter the player's skill level, it’s unfair to both the player and the rest of the runners to have their times be inflated/deflated just because of their PC capabilities. VC, I find, is perfectly acceptable since it’s an official hardware with mostly the same controls from the original DS (correct me if I’m wrong) with uniform load times. We can time the load differences for all categories at some point. Another issue with emulator, like lakitu said, is that it allows for a series of inputs that are otherwise impossible on original game. Overall, I personally think that there aren’t many legitimate reasons for future runs to be done in emu because 1) webcam recordings are perfectly fine under our current rules(not just on OG, but even on VC!), 2) load time differences fluctuate way too much, and 3) getting your hands on either OG or VC isn’t exactly the most difficult task. (It’s one of the best-selling games of all time, after all. Not just in terms of DS games)
Of course, I strongly disagree with deleting emu runs from the past.
As for separating the leaderboards, I don’t have much issue with it so long as we create tabs for "all", "OG" and "VC" with the "all" being the default. I don’t know if you can have tabs for "all" AND its subcategories simultaneously though; we're gonna have to work that out. Main issue I have, like Wal said, is that the "VC" tab might not have enough runners to accurately indicate the relative placement of participants.
tl;dr: future emu runs should be banned, VC are fine. Splitting the leaderboards is ok, so long as we have the "all runs" tab along with it
I saw Cheese05's Non-stop 120 star run the other day, and thought that "NG any%" seems like a good idea. You basically just go through all the levels in the usual any% run, but it’s on a completed file so you don’t have to wait for cutscenes and such
Like I said, I made a thread specifically for this debate. There’s no point in repeatedly posting on the main boards anymore
I will make a thread about this in the extensions page. Let’s discuss there, DiamondcrafterA
KingCoolgadgets,
Talking down to the community won’t make your point any stronger. I advise you not to do that if you want to be taken seriously and not like in your other thread.
Since you’re so hellbent on making that an actual category, why not post it on the category extensions page? Others have suggested it too, so why won’t you just do that? If you’re just going to continue this meaningless argument and not take action, you’re no better than an internet troll. I will not pursue this any further.
All the best,
Andre
Dear KingCoolgadgets,
Since you were able to bring this topic up, there is always a possibility of it being an actual category so long as more people agree with this and think any% glitchless should be one of the main categories. Thank you for suggesting the idea. I, for one, disagree though. Let me make my stance clear on that.
I’m not speaking for every single NSMB DS runners when I say this, but these restrictions you have set for "glitchless" are very difficult to call actual glitches. In my opinion, any% itself is already pretty friendly in terms of "glitches" in that none of them are hardly ever forced onto a runner. Boss quick-kills and death warps, even in a very broad sense, are far from actual glitches as far as I can see it. I would even go so far as to say that it’s just an attempt at making an easier any%.
Yes, I’m very well aware that glitch categorization isn’t your concern. However, that’s the only thing that truly matters when creating that category. You can propose your ideas on what’s allowed and what’s not, but as I said, people have different ideas for what a glitch is. You don’t get to decide the rules on your own, especially when glitchless isn’t a brand new idea. We'll have to wait and see.
Kind regards, Andre
Dear KingCoolgadgets,
About what’s intended by developers and what’s not... How do you know? The Bowser jump in the end, let's start there. Ok, certainly not a glitch, but that’s besides the point. How do you know the developers didn’t mean that to be possible? Any clever guy bored out of his mind could stumble upon 8-BC and think "hmm, how about I jump over them?" and actually do it, albeit it'd take many tries. And death warps too, how would you know? A really rash guy who knows his way through the level can go and say "alright, I don’t wanna wait for autoscroller so I’ll take a death" and skip boring parts. Mind you that two of these things I just mentioned doesn’t necessarily take a coding genius to discover, nor does it take intensive experimentation. Hell, anyone that doesn’t even speedrun this game could discover them individually. Bowser Jr. quick-kill? I guess that’s your strongest point. Surely the developers expected you to knock him out with a shell and stomp him. RNG manipulation is hardly a valid argument, since no one has figured out a 100% accurate RNG for any% yet, and people playing on OG wouldn’t even attempt it anyway.
What I’d like you to understand is that I’m not saying these things to demean you; the semantics of the words "glitch" "exploit" etc., are very vague even to this day, and making a "glitchless" category as you mentioned causes headaches in that what’s a glitch and what’s not isn’t intuitive at all. This is where the beauty of any% shines through; as long as there isn’t an external device, you can finish the game however you want! Sure, not everyone is comfortable with using minor exploits in a game, but understand that the spirit of speedrunning isn’t necessarily getting WR, but instead beating your own PB; you can set restrictions for yourself (e.g. no glitches) and improve upon it even if there isn’t a category.
Have you submitted any runs, KingCoolgadgets? I'd love to see it! I'd love to help you out if you’re just starting out, and I’m sure a lot of others do too.
Respectfully, Andre
No one except for a certain somebody who just loves SMB d:^)
That and the elevator skip, which would still be around 22:38 based on Wah's SOB
Must be some ridiculous trick he found if it’s 22:30 with loads
Wow my guide is obsolete now I guess, nice job!
Does anyone have an alternative for this? I have X/Y for run and A/B for jump, and I used to do the alternating A-B mash with my thumbs. It’s fast, sure, but my thumbs aren’t big or wide enough to keep it going(they usually cramp up either in the first room of 8-2 or the vertical section in 8-3) so I have to hold it differently in those sections
By second part, do you mean the one I referred to in the guide? Waaaaaaay harder, yes. The reason for this is having to deal with multiple pixel stops(the last one being completely luck dependent) as well as quick turnaround, making optimal jumps and trying to maintain horizontal momentum to slip through the gaps. I can probably do the first part blindfolded, but I’m not as consistent on the second part
@PainPita good luck! TAS cycle is nothing but grind. I strongly suggest that you practice other "difficult" strats in the any% route(cycle skip in 8-T, optimal 8-6, 70x in 8-BC etc.) before going for this, but I guess it won’t hurt since it’s very early in the run. I think that myself and a few others could help you out if you have any questions with the execution
I dunno, v1be seems to be getting the hang of it from what I’ve seen
