The recent improvement to the Super Mario 2 (U) Princess only has kicked up a bit of debate in the IRC channel --
What purpose does a princess only run serve?
Why is it that this improvement can be published, when better runs that don't obsolete old movies will be rejected?
Why should we allow the double-standard for improvements?
If the substandard runs are being kept because they show site history, they should be locked and no improvements accepted. If they're not, they should simply be removed, as they are not serving any useful purpose.
As it is they're just opening up the window for TASes that don't meet site standards to be published, and all the controversy that comes with it.
How fleeting are all human passions compared with the massive continuity of ducks.
Back then, peach was seen as the slowest character, and several runs with different choices were published every other day.
One day, someone (Bisqwit ?) dared anyone to beat the then current run, saying it was optimal already, adding that it'd be even impossible with peach only. A short time later, someone beat the record with peach only, and it was an instant publication, as it kinda taught a lesson to non-peach believers.
Of course the run did not survive long, obsoleted some time after, but it stayed in the list for those historical reasons. It reminds us to never pretend that a run is optimal, and never pretend that a choice is definately suboptimal.
But yeah time has passed and it should get removed from the list now. The run is not an exploit anymore. Historical reasons should be no exception, but it's up to the admins to strictly apply the rules.
I never sleep, 'cause sleep is the cousin of death - NAS
One day, someone (Bisqwit ?) dared anyone to beat the then current run, saying it was optimal already, adding that it'd be even impossible with peach only.
I don't think I have ever claimed any movie of mine as absolutely optimal.
Although I do not have a backup of the text of my sixth SMB2 publication, the fourth was: "My fourth attempt on SMB2, which is about 60 seconds faster than the second version and about 30 seconds faster than the third version. Uses warps."
I do remember that I mentioned something like "Princess has been entirely neglected in these runs ― perhaps there would be a need for a Princess-only run?"
And then it happened that Genisto submitted a SMB2 Princess-only movie, which also obsoleted my sixth SMB2 "any" movie: "Well this time its SMB 2 played with Princess Peach. Done in 09:07 by Genisto UNBELIEVABLE ;).
Genisto have a special request. He wants you to put that screenshot, included in the zip, on you're site." (by Phil)
I was quite humbled, and this Princess-only obviously obsoleted my old movie, because it was faster.
Then came Sleepz's Toad+Luigi movie, and created the second "any" branch, but that's history.
Joined: 8/1/2004
Posts: 2687
Location: Seattle, WA
I still don't see why we are perpetuating the SMB2 princess-only run whilst we reject the SMW walkathon submission. Both are quirky demonstrations of neat concepts, so why should one stay and the other be turned back?
I think, instead of deleting some of those runs, there should be categorization, dividing the site…
Something like:
― Good runs
― Other runs
― Hack runs
"Good" would be those that TASvideos wants to offer
"Other" would be those that satisfy a niche curiosity (movies of games that don't really work well as TAS (such as Mario Bros.) or niche goals like CCC in Mario64 and newgame+ in Chrono Trigger)
"Hack" would be runs of unlicensed games and hacks.
And possibly a "Premium" category for those that are now starred, but with different semantics more aiming towards awarding. Maybe not.
Joined: 8/1/2004
Posts: 2687
Location: Seattle, WA
And "other" is less vague than "concept demo" for a plethora of reasons, right? I don't mean to come off as condescending, but it seems like the same rehashed discussion we always attempt to have, and fail to conclude, every few months.
And "other" is less vague than "concept demo" for a plethora of reasons, right? I don't mean to come off as condescending, but it seems like the same rehashed discussion we always attempt to have, and fail to conclude, every few months.
Okay, maybe I was writing too quickly before having thought it through 100% with consistency and clarity in mind. Will try to make that mistake less often from now on. (Unfortunately, that may mean that I need to post less because these things just don't come off as God's word.)
A was thinking like this:
― A "hacks" category should be for "hacks" and nothing else ("concept demo" does not accomplish that)
― There should be a way to include TASes for every game, even the boring ones
―― but without reducing the signal to noise ratio in the site's main content
― In other words, I want to include Princess-only but like said, "concept demos" is something that should go away.
And now I'm thinking also these:
― There should be a way to include fancy demonstrations without "speedrun" / quick being an absolute goal.. and without having to put them all in the vague "concept demo" group
― There must be no "dumpster".
Shortly: Yes, this seems to be rehashing of old discussion, and no, I don't agree to deleting those movies unless as last resort.
Joined: 8/1/2004
Posts: 2687
Location: Seattle, WA
Mayhap we just need a distinction between Tool-Assisted Speedruns and Tool-Assisted Superplays? I'm against the idea, for now, but I'm very open to convincing, so long as there is good reason for it. As it stands, I accept the "concept demonstration" grouping in the sense that there are very entertaining concepts that are not speed-oriented, but only because of set site precedents. However, I can't accept that some of these demonstrations are accepted whilst others are not.
Also note that I still hate the term "TAS" and I wish that the forums could be reversed to a single order stack.
If I understand you correctly Bisqwit, you want one category for what we normally consider a standard TAS, playing through the game as fast as possible (in general, also some 100% runs etc), as well as a different set of runs with non-standard goals. Stuff like princess only SMB2, SMB/SMW walkathon, runs using passwords to access different modes etc (chrono trigger +). Also a third category for hacks, but that's not very relevant here.
I think this is a great idea, myself. I've always been a fan of non-traditional runs, and I've felt it's a shame there aren't more of them, because of the opposition to include them. Setting up a separate section dedicated for such runs sounds awesome, I hope it happens. One criticism I have of how the concept demo section has worked out so far is that excellent candidates for runs in that category have been rejected, I think partly because many (including judges) still hold them to the exact same standards as a normal run. Revamping the system and setting up a whole new section for them should help solidify in people's minds it's something else, something different and separate, and I think it will be helpful.
On a more personal note, I appreciate that you're responding in a more clear and less avoiding manner than in the past. You know I've criticized you for that (some star debates and talk about concept demos come to mind, I'm sure there have been other things too), and I'd like to think that maybe your Ask Bisqwit thread has something to do with this change. Whatever the reason, I think it's awesome of you.
Joined: 3/9/2004
Posts: 4588
Location: In his lab studying psychology to find new ways to torture TASers and forumers
Breaking into an more entertaining and more technical subsets sounds good to me.
Like people submit a perfect run of a game where they frame perfectly walk into a wall right at the start which triggers the end game sequence, we can agree it can't get any faster, but pointless to watch. This would go in the technical but not entertaining category.
On the flip side, we have a game where doing exciting things slow down the play a bit, in those cases, we should have some kind of agreeable trade off. Think of where in a game doing the really quick root keeps the screen black for five minutes, where the slow route takes six minutes, but you get to see a lot of action the whole time.
Warning: Opinions expressed by Nach or others in this post do not necessarily reflect the views, opinions, or position of Nach himself on the matter(s) being discussed therein.
How about these categories:
- "any%" TASes: The only goal here is minimizing the number of frames needed to complete the game. "Entertainment" (however you want to define it) is irrelevant. No voting needed: Whenever someone submits a run which uses less frames than the currently published one (and someone verifies that the run is valid) it automatically obsoletes the existing one. This is just the *absolute minimum* needed from game reset to see the end of the game, without any regard to anything else.
- "Entertainment" TASes: These would be runs which aim for a different goal than "any%" for entertainment purposes, while still completing the game as fast as possible while achieving that goal. The goal should still be rational and sensible, in the same way as currently (for example "uses no death-warp", "collects all items", "uses this alternative route" or "uses this character only").
- Superplays: The goal is not to complete the game as fast as possible, but to complete it while achieving some other goal. These obviously can be done with just some games where it's sensible to do so. Examples of such goals include getting the highest possible score as fast as possible, or demonstrating the largest variety of complex combos and finishes in a fighting game. Fixed-speed side-scrollers where the idea of the video is to show superhuman precision (instead of completing the game as fast as possible) could also be put into this category.
- Machinima videos: The aim of the video is something else than completing the game. A "just for fun" video made with the game.
The "uses a hacked/unlicenced game" and "starts from a savestate" could perhaps be additional tags instead of being their own categories.
I still don't see why we are perpetuating the SMB2 princess-only run whilst we reject the SMW walkathon submission. Both are quirky demonstrations of neat concepts, so why should one stay and the other be turned back?
Because the walkathon submission wasn't a walkathon- it was a jumpathon. Not to mention, unlike the SMB walkathon, the run button was pushed quite often (even if the arbitary "doesn't go over normal speed" limit is imposed). Most people thought it was annoying to watch due to all the jumping, even.
Also, why not just have a "Speedruns / Non-Speedruns / Hacks" division? Though things like the SMB2 princess run would still fall under a fuzzy line.
Thinking further how this would play together with Tub's idea of per-game entries where different categories for the same game could be mentioned as extras… I still this could be combined.
Each category page could basically get the same view; list of games that contain movies matching the category criteria.
The game information would contain brief links to other type movies of the same game.
Which movie it showcases in the game entry depends on the selected criteria.
Hmm…
I was thinking like this:
― A "hacks" category should be for "hacks" and nothing else ("concept demo" does not accomplish that)
― There should be a way to include TASes for every game, even the boring ones
―― but without reducing the signal to noise ratio in the site's main content
― In other words, I want to include Princess-only but like said, "concept demos" is something that should go away.
And now I'm thinking also these:
― There should be a way to include fancy demonstrations without "speedrun" / quick being an absolute goal.. and without having to put them all in the vague "concept demo" group
― There must be no "dumpster".
I agree that this is a good idea, but it doesn't address the issue of double-standards for improvements to old runs.
There are surely movies up today that wouldn't be accepted, under either tier of the new system, if they had been first submitted today.
Should improvements of these movies still be accepted, if the improvement movie wouldn't pass submission standards if judged as a movie for a new category?
How fleeting are all human passions compared with the massive continuity of ducks.
I think the site would benefit from having a "technical" section as well, perhaps including only emulator movies and not avi's in some cases. Searching through the Gruefood I've found a few enjoyable movies (to me) that were deemed too boring (ok, they kind of are) but people who have played the game would like seeing. Off the top of my head, Golgo 13, Yo! Noid, Princess Tomato... again, not the most exciting movies (not even close really) but all fairly optimized and in my opinion should be easier to find by browsing through the site.
How about these categories:
- "any%" TASes: The only goal here is minimizing the number of frames needed to complete the game. "Entertainment" (however you want to define it) is irrelevant. No voting needed: Whenever someone submits a run which uses less frames than the currently published one (and someone verifies that the run is valid) it automatically obsoletes the existing one. This is just the *absolute minimum* needed from game reset to see the end of the game, without any regard to anything else.
.
This is correct in 95% of cases but what if the new run just uses a new strategy or trick(s) to improve the current one and doesn't have a good overall playing standard compared to the previous run, say for example the recent OoT run. We still need to have a quality assurance in this case.
In that case I feel both movies should be available through the main site. The nicely done one would be in the "good" section and the new shorter one would be in the technical section until it is obsoleted (maybe with disclaimers... though I still think that most casual viewers would not consider the faster one that poorly played)
Joined: 11/18/2006
Posts: 2426
Location: Back where I belong
I like the idea of further segregation of games, but I have a feeling it will simply breed further arguments about why one movie is published under a certain section, or the idea of dual obsoletion. Say that a movie comes along that accomplishes all of the goals of the "entertainment" run, but also does it faster that the current published "any%" run. Does the movie show up in both sections now, or would the entertainment run be deleted until some other, newer goal is accomplished?
I also have a feeling that the current movies would be split up into 70% speed (duh), 20% entertainment (several of the games in the Concept demos section right now, along with most fighting games), and 10% that would kick up quite a bit of discussion (100% games, especially Super Metroid, come to mind... do they primarily aim for entertainment because they go for more items, or is it primarily speed with a secondary goal?)
Again, I'd like to see a little bit of overhaul to the classification, but I don't know if it will cause more problems than it solves.
Each category page could basically get the same view; list of games that contain movies matching the category criteria.
The game information would contain brief links to other type movies of the same game.
I like it, but I'd also suggest listing the Gruefood submissions("There are ?? rejected movies for this game"). The page could be separated from the normal content(or listed at the bottom of the game's page) and simple(no screenshots, descriptions, AVIs, etc; just the category, who made it, a link to the movie, and a link to the submission thread).
How about these categories:
- "any%" TASes: The only goal here is minimizing the number of frames needed to complete the game. "Entertainment" (however you want to define it) is irrelevant. No voting needed: Whenever someone submits a run which uses less frames than the currently published one (and someone verifies that the run is valid) it automatically obsoletes the existing one. This is just the *absolute minimum* needed from game reset to see the end of the game, without any regard to anything else.
.
This is correct in 95% of cases but what if the new run just uses a new strategy or trick(s) to improve the current one and doesn't have a good overall playing standard compared to the previous run, say for example the recent OoT run. We still need to have a quality assurance in this case.
My idea was that this category would be a purely technical category, with no quality nor entertainment goals (other than what automatically is generated because of completing the game as fast as possible). With this I don't mean the videos couldn't be entertaining or of high quality. I just mean that entertainment or quality are not qualifying factors. It would be a pure "world record" category.
It would make many things easier. No voting needed, no discussion needed whether it's ok to publish it or not. Just pure frame count and nothing else.
(Of course there may be still room for rejection, for example if it's known that the submission is clearly suboptimal and not eligible for the "world record" category.)
Then the second category I suggested would be much more concerned about the quality and entertainment of the run.
It's just an idea to think about. I'm not saying it *must* be exactly like this.
It would be nice to have a pure speed, those video would be hard to argue if they are better or not. Or maybe have a Category (like any%) but now it would be the fastest.
Joined: 5/2/2006
Posts: 1020
Location: Boulder, CO
If it were pure speed, all you would need is the frame count to know if its better.
That was the whole idea of warps suggestion, that the speed only catagory would take judging out of the equation.
