Forum posts

Page: 1 2 » »|

It looks like there's a faster time on a japanese video hosting site -- contacting that guy to see if he would allow us to submit here.


There's no video for the only run?

I plan on running this in 2019 if I can find the time, but don't really want to compete with a mystery video. Can the video be replaced, or the run removed?


I could definitely see the "main" category requirement being a full game (meaning start -> ending), but my point was more the arbitrary nature of the various boards. Speedrunning is more or less fragmented at the moment, but it is growing. As people stop running (but remain inactive mods), certain runners become mods on dozens of boards, and new blood come in with new ideas, arbitrary can lead to unnecessary sluggishness or conflict.

Imaproshaman likes this. 

For the community deciding thing, this would be letting the community decide -- the people who run the game. If a category is a "meme" category it, by default, will have a small number of runs. I recently added a game to the site and am the only run, it is a rare game and is essentially a "meme" game -- but the Any% category is considered more legitimate than a misc category in others. By default misc categories are hidden on profiles and misc categories are hidden on the boards unless someone goes looking for them.

As an example, the "All Levels" category in SMB2 has less entrants than the Any% (Peach Only) in SMB2, but only 3 mods have submitted a time for that category. It's objectively more popular, but not necessarily with the gatekeepers. I don't see why an empirical metric shouldn't be used instead of an arbitrary metric decided by (mostly) arbitrary people.

You can see evidence of arbitrary decisions in multiple games -- Super Mario Bros. 3, for example, has its Any% as its last category while 100% is its first -- despite Any% having drastically more runners than 100%. Zelda II has Any% as its first category and 100% (all keys) as its second, but 100% has significantly more runs. They have the same situation on both games but with the opposite action -- its arbitrary.

Legend of Zelda "second quest -- Low%" has five entrants, and its one of the most popular games of all time. The Legend of Zelda "Traditional Dungeon Order -- No Glitches" has twenty two entrants. Assuming these were on the same board (the site's intended direction), why would the more popular category be listed as "misc" or "category extension"? What would the process be to change that and is that process the best way of going about it?

If the popularity of a run matters, that can be done automatically. If the decision is arbitrarily chosen by a select few, is that really the best outcome? I have the world record for Wolverine, but it doesn't mean I should decide whether or not a "Wolverine -- 100% All Burgers" category should be considered a 'main' category or not if others would want it to exist as a main category and actively run it.

I don't really care about SMB2 (just a visible example) and I'm not saying that certain categories shouldn't be manually chosen by mods to be considered main categories even if they don't have any runs; there's certainly room for arbitrary decisions, but in the opposite direction. Rather than arbitrary choices being exclusive, they should be inclusive -- e.g., mods saying "even though 100% is an unpopular run due to its length, it is a staple of the boards and a legitimate run that should be showcased".

I do think a system that is arbitrarily exclusive by nature and creates gatekeepers doesn't actually provide any benefit. Having one that allows for the natural, organic growth of a board without having to deal with gatekeepers seems to be a better long-term system, especially when this site doesn't have any direct forms of communication.

Imaproshaman likes this. 

Things I dislike about the site:

There are "main categories", "misc categories", and "game extension" categories. I think automatically making certain runs NOT a "misc" category based on number of entrants would be a good change. I ran SMB2 "Princess only Any%" and was a little surprised to see it listed as a "misc." category when it has more runs than many other NES games. Rather than leaving it up to a moderator to decide what is and isn't, an automated system based on the number of runs could be good.

I'd also say that the "category extenstions", "misc." and "main" categories should all be under the same title card. If I open "Legend of Zelda" I should be able to toggle between main, misc, and extension categories on that same page. Ideally a mod should be able to create multiple "boards" for the game -- like individual levels categories, full game, high score, oddball stuff, etc. Maybe a selection of pre-set boards, but multiple boards would be great.

There are no values other than "time" -- having custom values (time, score) and the ability to rename those values could help for different types of runs.

Having multiple names for a game is important. If you search for "The Peace Keepers" you'll not find it on the site because it is listed as "Rushing Beat". This is kind of bonkers, as the image actually shows "The Peace Keepers" and the runs are in the US version of the game.

Things that would be cool:

Having a "cobweb" game selection -- games that haven't gotten a run in a while, as a search option.

Having the ability to search games by oldest WR.

Having the ability to search for MULTIPLE user histories would be neat. "How do I compare against so-and-so" could be answered by searching for both and only showing games where both runners have submitted something.

Having a % score based off of the WR could be good too. 1st, 2nd, 3rd, etc., is great, but even if it isn't public it'd be nice to see the % difference in time. If I have the WR in a game at 4:59.95 and 2nd place is 5:00.15 its a different story than if 2nd was 7:00. Just be a neat metric to see how far ahead/behind you were.

I also think a built-in "speedrun score" could be something neat -- having a ranking system that gives point values based off of your placement (or time) compared to others in the same game could be a neat leaderboard, especially if it had different categories (e.g., "All games", "NES", "SNES", "Zelda games", "Mario games", etc.). Making it a limit to only count games with X runners to be calculated could make for a pretty neat leaderboard.


So I read this thread after watching Arcus play TMNT for fresh five day.

I program games as a hobby and from looking at what you guys are discovering, it looks like you should be able to predict the technodrome spawn if

A) the randomized seed is used for other things
😎 you see those other things during normal gameplay
C) this seed doesn't change during gameplay*

An example would be Redheaded's Bebop/Rocksteady having their AI manipulated by the random seed. This should be easy enough to check by creating a save state post-spawn (after the seed is set), then moving through the game and re-checking the things you think are set by that seed.

This could be things like enemy spawns, drops, etc., etc. You could find things like "oh, if this enemy drops a single ninja star it is the bad star" as a consistent thing. You'd just need to play through twice, know what the randomized seed is set to, and then see what lines up as consistent.

You'd need to do things consistently between the two playthroughs to guarantee it just in case things altered by the seed could also be altered by something else.

*It is possible this seed DOES change during gameplay -- specifically, using a continue -- and this gives you the bad spawn. If this is the case you could test whether you have THAT seed pretty easily by killing yourself immediately at start, using a continue, and then playing through the game and looking for consistencies. Then if you don't see those consistencies, you know you have a 50% shot at a good spawn (instead of 25%).


I didn't check the other frames, so that might be the case -- I figured I'd give the benefit of the doubt because I couldn't tell 😃


@Oh_DeeR I think the last pixel of Sabretooth in WhiteHat's video is an artifact -- I thought I had a one frame advantage too at first, but looking closer you can see other artifacts appear on the same frame and the size of the pixels seems off compared to everything else. Unless the resolution of all the sprites in the game was far greater than it appears, I don't think it's possible for those little black spots to be part of Sabretooth.


So... frame counting, my time is 23,372 frames. WH's is 23,372 or 23,373 frames -- I think on the very last frame of my run you can see a pixel of Sabretooth at the bottom, but that may be an artifact done via the video capturing.

This may be an actual to-the-frame tie which is quite possibly the coolest thing that could ever happen.


When asked about how retiming runs worked I was told to make a forum post, so here it is!

Current runs between WH and myself are within .01, from my timing I'm ahead by a bit over .01; framecounting may be needed. Above are images showing the timing changes based on our split timing. Very close!


Neither the megaman boards or have the legacy collection's challenges (boss rushes, boss rushes buster only, wily battles) as speed run categories -- they're designed to be run as a speed run and have a built-in timer for easy verification.

Would it be possible to have the Mega Man Legacy Collection challenges added on If Mods are needed for verification I'm more than willing.


I keep wanting to run this game -- I got a 1:04 at one point with a wild warp run, but its slower than darc's no-warp time! 😃 I'd love to see what better runners could do with a wild warp run. More interesting to me than a standard non-100% run.

Can be a frustrating game to run.


Page: 1 2 » »|