Well, my opinion is that a movie should end when it's impossible to make it fail to reach the ending.
Suppose you're racing your car, and you're close to the finish line and going so fast that no matter how hard to step on the brakes you can't slow down fast enough to stop before the line. Suppose that you're time from starting the race to crossing the line was 1 minute, but after just 55 seconds you were close enough that you can't avoid finishing the race. What was your time for the race?
We are not in F1 racing here but a simple timeattack.
This site always show in-time emulator movie file. I am not cheating the system, I am with the system. Some guys doesn't seem to understand that.
Personally, I always thought that way and probably not change it. To respect rules, I have no choice to stop my movie later but I don't like that way.
P.S. I use the 1st person because people, between lines, are talking about that SMB1 submission.
One thing everyone should be looking at is that the point of this site is movies, not the keystroke files. The files easily tell us how long they are so we don't have to time it or watch the avi closely to get the exact time. Most visitors to the site only know the movies, so a special circumstance should be warranted here. One could argue that the Metroid movie doesn't technically end until Samus is up the elevator and the screen changes, but all the runs save probably 1 second by stopping the movie as soon as up is pressed. 1 second doesn't make all the difference in the world, so the time noted in the file is basically accurate. This is a smaller scale version of Monopoly - hit the button and the computer does the rest of the work. The only way to know which movie is truly faster is to stop one of them where the other stopped and compare.
Kyrsimys wrote:
I think the biggest issue is that the game isn't really completed in the reported time. Currently we say for example "NES Monopoly completed in 00:35 Acmlm", but this isn't really true since the game is not completed until the computer is bankrupt, which doesn't happen for another 18 seconds.
If we follow this logic, there are a bunch of movies with inaccurately reported times. Many movies end when no more user input is required and the computer does the rest of the work to get to the ending.
What is the end of a movie? When you fire the bullet to hit Red Falcon, when the bullet hits Red Falcon, or when Red Falcon explodes, the screen collapses and switches to the ending sequence. And what if you got that final shot in on Red Falcon and had to jump into a hole to escape. Would you wait to land in the hole or would you simply stop after jumping? Part of the fastest completion competition is stopping a movie as early as possible.
And, I agree with Phil in this respect - the viewers are not the players. If a movie ends right where no more user input is required, but the player can press start to pause indefinitely, I don't think it would be rejected (Metroid?).
In my opinion, if the Monopoly file ends at 0:36, but the movie ends at 0:57, and someone submits a movie with a file time of 0:54 and the movie ends at 0:54, it should take precedence. Similarly, if Phil and Genisto's movie is longer when hitting the axe than Mana's, it should be rejected.
The movies should count more than the time in the file, and once a movie is published, a player should look at stopping his movie at the same point the previous author did. But, in this circumstance, the current movie is faster than the published one using a new method to stop the movie, and currently the rules (understood or written?) state "no more user input required." I think Mana should change his movie and put it up against this recent submission, but as Phil noted, it wouldn't affect the time units, so I don't think it would be faster.
It's either this or the rules should be changed from "no more user input required" to "the user can no longer affect the outcome of the movie." But this wouldn't even help the Monopoly situation, as there can be no more user input.
No easy answers here, but remember when someone is downloading the torrent or fcm for Monopoly, right in their face is "This movie aims for the fastest time to get the CPU player bankrupt. It accomplishes it in 53 seconds in which only the first 35 seconds require player's input (to heavily manipulate the luck)." I think that basically covers it.
Now, for SMB, it would be silly to put 5:00 (Movie actually finished in 5:01). I think the standard of "no more user input required" is a solid one and if anyone out there is saying, "My run beats this but I stopped it later than he did" then he should go back and stop it sooner - no problems then.
The end of the movie is when the objective is completed. In the case of the Monopoly movie that is when the other player is bankrupted. Not 17 seconds before that.
Joined: 5/1/2004
Posts: 4096
Location: Rio, Brazil
ok, let me try to put it short:
IMO, the player should do everything possible to make the AVI shorter.
- If the ending of a game needs you to hold a button for the text to scroll faster, that keypress should be included in the movie.
- If at the end of the game you need to input a high score name before seeing something further on, the player must input it.
i think everyone is with me here. but the problem seems to be about after-recording input that slows down or prevents the movie from ending. My opinion is that:
- the watcher shouldn't press buttons. the movie should be ended as soon as possible (as long as it doesnt make the AVI longer) without worrying about watcher input. Users who watch the movie on emulators are doing it the "wrong" way, the site is for AVIS, and the AVIS can't be changed by the viewer.
- the watcher could even hit PAUSE button, who knows... some games can be paused after the last boss is hit. so i dont think we should worry about that.
Joined: 5/4/2004
Posts: 90
Location: New Brunswick, Canada
I think once a precedent has been set for a given game, all subsequent movies should end at the same time.
For example, Super Mario Bros. The previous Super Mario Bros. videos ended when the axe was touched, correct? Then all future Super Mario Bros. should end when the axe is touched.
What if a glitch was discovered like, for example, land on Bowser's head and you're teleported to the axe? The movie would be stopped there, correct? Everyone would go for that.
Depending on how far away from the actuall ending a movie would be when record stops I would go "Woha! Thats an impressive early ending for the movie record"
Then you can start to bite your nails because you won't be able to tell if it still goes 100% well right. :D
Small pice of gambling eh?
Joined: 5/4/2004
Posts: 90
Location: New Brunswick, Canada
-ziplock- wrote:
What if a glitch was discovered like, for example, land on Bowser's head and you're teleported to the axe? The movie would be stopped there, correct? Everyone would go for that.
Sure, submit the run like that. If it was badass enough, I'm sure people would take it.
However stopping in mid jump instead of when you touch the axe when all other movies stop when you touch the axe is kind of cheap, in my humble opinion.
I think we need to change this thread from current heading into the following:
There are 2 lengths in question,
A: MOVIE FILE TIME (# of frames stored in the movie)
B: GAME COMPLETION TIME
IMO, the published time should be B.
The problem with B is that you need to apply common sense, as to when the game actually ends, and as someone above stated, does it end when you fire the last hit on the boss, when the last hit on the boss hits, or when the boss is done exploding, or the credits start rolling.
Common sense for SMB is that it ends when you hit the last hammer. Common sense for Monopoly is when computer accepts going bankrupt. Zelda when you touch the princess. Metroids when you touch the final lift.
Disclaimer: Common sense may differ wildly from individual to individual.
That sounds naughtier than intended...
Anywho, one potential problem I see with option B is that, if we're going to be measuring game completion time, it might not make as much sense to measure it from reset as opposed to when the game actually starts since that is how game time in general is measured. Although I suppose the reason for that in normal speed run sites could simply be because there's no consistent way to tell when reset happened, a luxury we have that normal speed running doesn't have.
Although in most cases option A might result in a shorter time, I think B will in some (admittedly fewer) cases do the same thing...where somebody records pressing keys in the ending (and we decide game completion is before that) then somebody else comes along and submits a run that doesn't press keys in the ending. Although I guess that could be sloppier and such.
Personally, I think that we should publish the length of input as opposed to the game time, and then if we think that something should be mentioned about the game time we can. I like the length-of-input option because it's less subjective and easier to automize. No need to make the world's best AI to tell us how long it is.
If we do it the other way around...where we publish the game time and mention player input ending early...I think we're circumventing automatic time calculations for no reason other than to try and seem like a normal speed run site (that publishes tool-assisted speed runs). Of course, a big part of that is that I'm a Computer Science student\fan and that I do prefer to use automization like that...hmmm...
Joined: 3/9/2004
Posts: 4588
Location: In his lab studying psychology to find new ways to torture TASers and forumers
-ziplock- wrote:
What if a glitch was discovered like, for example, land on Bowser's head and you're teleported to the axe? The movie would be stopped there, correct? Everyone would go for that.
It's funny you should mention that...
There is a glitch in SMB that if you land on bowser when he's too close to the ax, you'll both die, and you'll continue on despite that you died.
Not sure how you would use that to your advantage. I think when you exploit that glitch the bridge animation doesn't play, so that's faster, and it effects your powerups later on in the game which probably only makes a difference in an all level run.
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.
Anywho, one potential problem I see with option B is that, if we're going to be measuring game completion time, it might not make as much sense to measure it from reset as opposed to when the game actually starts since that is how game time in general is measured.
A solid point, and worth considering. If the only thing that matters is the time you're actually playing, then the opening title sequences for every game shouldn't be counted either. But, as Bisqwit pointed out, they're out of our hands - the ending point isn't.
TNSe wrote:
Common sense for SMB is that it ends when you hit the last hammer. Common sense for Monopoly is when computer accepts going bankrupt. Zelda when you touch the princess. Metroids when you touch the final lift.
Isn't it just easier to say, "end all your movies as early as possible" than to layout a stopping point for each game? That way everyone's competing on the same level and working with the same set of guidelines.
Again, movies are the goal. Monopoly says "0:35, but actually 0:53" in the description, and, for special circumstances such as these, I think that basically covers all the bases.
Any frequent visitor understands the times listed don't correllate with the actual movies, as the endings aren't counted, so I see one of two options:
-Change the time in the descriptions of each movie from the keystroke file's definition to the actual movie time. This will cause problems with the endings, as a SMB time in 5:14 doesn't sound too impressive.
-Define the starting and ending points of every game submitted. This is a serious amount of legwork that I wouldn't imagine anyone wants to get involved in.
Obviously, there are varying opinions on the subject, but I'd still recommend going with the standard for every movie - stop when no more user input is required.
Joined: 3/11/2004
Posts: 1058
Location: Reykjavík, Ísland
I'm going to change my mind on this. I say the rule should be: stop as soon as no more input is required, as long as it doesn't slow down the speed of the actual completion.
So if in the end of a game you're supposed to fall down a hole, and there's NO way to speed it up with input, the movie should end as soon as the character starts falling. However, if the movie ends early and makes the character slowly drift towards the final goal, that would not be OK (if it was possible to speed it up).
Of course, that's just MY opinion, ... well, have fun debating...
Joined: 3/9/2004
Posts: 4588
Location: In his lab studying psychology to find new ways to torture TASers and forumers
I agree with Blublu seems to be the most logical side in this debate.
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.
I'll give my input on this, care how much you want about what I have to say:
If you fall unconscious in a race 5 seconds before the goal (and it is noticed by the judges and the audience) but your car still crosses the goal line, does it count as a valid result?
If the car collapses 10 meters before the goal but a wheel of it crosses the goal line, does it count as a valid result?
I like the reasoning that this quote represents, and the variations that followed. I will use this theme for a (fictitious) scenario of my own.
I may regurgitate what others have said, but at the risk of doing so, I say:
Say that the driver is unconscious before the finish, and won't be affecting the outcome of the race from that point on. Does that make him reach the finish line faster?
The winner is announced. One of the judges comes running, gasping for air to deliver the breaking news:
We just discovered the driver in 2nd place, who finished 1 metre behind the first racer crossing the finish line - went unconscious one minute before he reached the finish line. (The ending straight was long, or something, it doesn't matter in this fictitious example.) As a result of this, his race time is considered a minute shorter. He wins the race by a margin of 59.XX seconds.
I may have taken som values to the extreme to easier illustrate, but the situation, and resoning itself appears strange to me, even with more sensible values.
Kyrsimys has said most things for me.
I also agree that for all improvements of existing runs, the movie(s)should be stopped at the same point for all attempts.
Andreas
We all realize that there the movie length isn't necessarily the game length. No one is saying that the game of monopoly was won at 35 seconds, just that the movie is completed then.
There isn't really an parallel in the race car scenerio, other than saying that the driver was only racing for x minutes, rather than x + one minutes, which doesn't really apply (but does in our case)
Earlier or later, it doesn't change anything but the fact that the rest of the video is so well played. Everyones still got their idea pretty anchored. So it's hard to convince people as well as convincing me.
