It's something that is constantly being done on this site; someone makes a run of a game, then someone else (or the same person) takes the same game and does it faster. The newer run replaces the older run and the older run goes to another dimension... or something.
My question is this: if someone submits a run that can be improved, why is it accepted? I've seen runs here (Rockman/Megaman series for example) that have been improved 2-3 times and are still being improved. Why not make the run as good as it can be before accepting it? I love a fast run as much as the next guy but eventually it gets boring seeing the same game over and over again no matter how fast you go through it.
I'm asking this because I'm confused about some things. Mainly the Amagon run I recently submitted then cancelled because I was told it needed to be improved. Don't get me wrong, I'm going to improve it, but I'm sorta dumbfounded that a run (my run) that could be improved by 22 seconds or so is declined when there are runs that are accepted then improved by minutes months later.
All you need to do is look at the front page to see my point. Megaman 6 and Little Nemo were both improved by minutes. If such improvements were possible, why were the runs accepted before being improved?
I dunno, maybe I'm wrong. Maybe my run wasn't as good as I thought it was and that's why it was turned down. I didn't make any mistakes, I just didn't use the edgeboost as much as I could in the first few levels. I got great feedback on the Amagon thread so I thought it was ok. Either way I wanted to bring this point up because I'm confused and I'm hoping someone can clear this up for me. Thanks for letting me rant.
Guybrush: "I'm selling these fine leather jackets."
Wally: "Really?"
Guybrush: "No. I 'm lying."
Wally: "In that case, I don't want one!"
Currently working on:
Nothing at the moment.
It's not just about whethe the movie IS improvable or not, it's also about if it LOOKS improvable. I don't think my Megaman 6 run looked all that bad, it's just that more creative solutions which make the movie even faster are constantly found.
Another thing is that standards change. When Sleepz made his Little Nemo run a little less than two years ago there was no frame advance or even FCEU. Hell, I'm not sure if Famtasia had the 60fps patch back then. The movie was considered to be pretty good back then. Also, there was no huge submission queue back then, which made it easier for Bisqwit to notice and review (and of course publish) every single movie more carefully.
For me the question isn't "is it improvable?" That's a question for future speedrunners, yes, but not directly related to submissions. For me the question is "Is it well thought out and well executed?" That is, it's not necessary to analyze every minor nuance of a game's physics to make a good movie. For the most popular games (Rockman, Mario), that's what happens anyway, but that doesn't mean the previous versions aren't both well thought out and well executed.
I'd say some ninety percent of reduced time in already-published runs are strategy improvements, new bugs, and new insights into the game's physics.
Also, if you were to not submit runs until you were 100% positive there would never be any improvement, there would never be new movies. There are always new things to discover, even in the most basic of games (SMB's been obsoleted and redone how many times?).
A lot of runs are just "I've played this several times, and this is utilizing all time-savers I currently know of." It also challenges other people to best that movie, to find new techniques the original author may have missed.
Of course, any movie that has obvious flaws, or flaws the author already knows about and isn't correcting should not be published, unless the mistakes add up to an nelegible* amount of time lost, such as in the MM6 TAS with the 1-3 frame ladder trick.
* incorrect spelling
Joined: 4/20/2005
Posts: 2161
Location: Norrköping, Sweden
I basically agree with what everyone else has said already. Making a totally perfect run is practically impossible, but the goal is to make them as good as we can. If neither the author or any other user on the forum can find one single flaw in a movie (or at least practiacally no flaws), then of course it will get published. Then, somtime later, perhaps someone finds a new trick or glitch, or just uses new techniques in a new run, thus making it even faster.
The main goal on this site is really to make entertaining movies... And a run becomes entertaining when the viewers is amazed by the time it's completed in, if the viewer can spot obvious mistakes in the run, then it hasn't really reached up to it's goals of being "perfect". But as Xkeeper said, these timesavers should add up to a noticable amuont of time...
On a more personal note, I'm sorry about your Amagon movie... I look forward to your next version, if that's any comfort. :) Have you started on it yet or when are you planning on doing it?
My question is this: if someone submits a run that can be improved, why is it accepted?
That's the thing ― we don't see the future. We can't make a better movie than what we know the means how to. Records are records until they're beaten, and until then, they're indeed, records.
People who haven't made these movies often think that it's easy to make the movies with tools, "cheating", but it's not so.
You really need to have the creativity in order to break the accomplished standards and establish a new record. Such creativity doesn't appear in an instant. Nobody of us is enlightened.
You should see the history of Super Mario Bros 2 speedruns as an example. It started with a mostly glitchless run. The double-jumping glitch was discovered. The route changed from worlds 1,4,5,7 to 1,4,6,7. New ways of handling 6-2 were discovered, more glitches were discovered, etc.
I have 7 different SMB2 runs which obsolete each other. Maybe more exist. Should the first 6 have never been published? Do we have proof that Phil's current run is unbeatable?
I agree... most of the movies that are published were published because there was no knowledge at that time on how to improve them further.
Of course there are some movies published that were not known to be not perfect when they were published... but I think it depends on several things...
- If a movie takes 2 hours, it's not as terrible if it loses a second somewhere compaired to a movie that's 5 minutes...
- If a movie involves lots of different possible paths, strategies, weapon switches, difficult weapon usage, it's also more likely the movie won't be perfect, and it will probably not count as hard, if it loses a little time to the perfect game compaired to a very linear game...
BagOfMagicFood wrote:
See, this is why I never finish movies anymore.
I do agree with this a bit... I think my first movie, the teenage mutant ninja turtles 2 movie, provides perfect entertainment, but it's played far from perfect (famtasia + no experience)... I don't know whether the same movie would be published again nowadays, which is too bad...
Tailz wrote:
I'm sorta dumbfounded that a run (my run) that could be improved by 22 seconds or so is declined when there are runs that are accepted then improved by minutes months later.
You might be interested to see these quotes from one of my lolo submissions:
Baxter wrote:
I know a level where I could improve one square
One square equals 16 frames...
The reply on it:
Bisqwit wrote:
Ok. Since you already know how to improve this movie, I'll reject this submission, but I will publish the improved version when it's submitted.
This is basically why I think it's not that unreasonable if you would redo your run, if you could save ~22 seconds...
That makes a lot of sense. I'm still new to this community so I didn't realize that the tools used have come a long way from what they were. I see now why runs are improved and why some runs are rejected. It's not about how perfect you can make it (well, maybe it is a little bit) but whether it's entertaining and looks really good.
Thanks everyone for clearing this up for me.
Guybrush: "I'm selling these fine leather jackets."
Wally: "Really?"
Guybrush: "No. I 'm lying."
Wally: "In that case, I don't want one!"
Currently working on:
Nothing at the moment.
I think it also depends on how much strategy is involved. People expect nothing but perfection with runs like SMB or other linear platformers where the objective is simple: go right. However, if you're talking about an RPG with insanely hard luck manipulation, several different paths to an objective, and very tough choices to make, like whether or not to get into a battle and steal weapon A from an enemy right now or just wait until you can buy it and possibly save time, a little leeway is almost expected.
Essentially, how "hard" the run was (hard in quotations because that can be interpreted to include trail-and-error) helps determine if it's necessary to go back and improve a few things or be rejected.
I don't know about others, but if a game is fun to watch, I like to watch it. For example, if someone were to redo the Yoshi's Island movie (My favorite movie on the site so far) claiming to be 5 minutes faster, I'd gladly watch the whole thing over again. In the same way, if someone were to take a 6-minute movie and improve it by 30 frames, I'd watch that over again too, just because I like to watch tool-assisted movies.
As far as absolute perfection goes: I think it's okay to drop one or two frames here and there, as long as they're not easily noticed by somebody not trying to break your record. For example, a movie which makes someone say "Whoa, that last move looked like it might not have been as fast as possible, but maybe it was" is okay, but something that makes the viewer say "Whoa, that definitely was not as fast as possible" is not, and makes the run lose a little of its entertainment value.
