Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
I'm replying to this thread because I once tried TASing this game, and gave up when I realized this game should be brute-forced for maximum efficiency and entertainment, similar to the old Lunar Pool movies. Peggle has massive potential to be a really entertaining TAS if someone wants to write a brute forcer. Every shot should be at least as impressive as a Zen Ball shot. As it stands, this movie is indistinguishable from a casual run that uses savestates.
Shot Angles: The angle at which each ball is shot is most efficiently controlled with the touchscreen. Since the touchscreen has a resolution of 256x192, it means there are about 588 possible angles each ball can be fired at. It's best to tap either the left/right (y=24 to 191) or bottom (x=1 to 255) edges of the screen to get the most precise angles; tapping further away from the edges tended to give the exact same results and is most likely arbitrary for angle finding anyways.
Did you know that you can subpixel-adjust your shot by tapping the D-pad left and right?
Here's something you did about 5 minutes into the run:
In this same position, here's a shot I found after about 2 minutes of attempts (though admittedly I wasn't holding the fast-forward button when it happened):
Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
FINALLY!!
I've always wanted to see this one done. Would have done it myself by now, but I really wasn't sure how to even go about C64 TASing or how to find the correct version of the game or anything. I thought about attempting a run of the GBA version, but that wouldn't be very ideal since the screen size is so small. And I even thought about running the Wii VC port. But now that a run actually exists, I'm very interested in copying the setup and examining this for improvements.
I see that you used a bot to manipulate the amoebas. I assume that the amoeba AI here is nothing compared to the NES version, which was eventually solved and found to be very deterministic (it had to do with objects touching the amoeba blob that changed its growth pattern, and number of free spaces around the blob that sped up its growth rate). So if I take a stab at this game, I'd be more than willing to try using your bot to deal with that cave.
I might be in the minority here, but I'd also love to see a full-game run. I know of the precedent set by the NES run of running only loop 4 as the "hardest difficulty", but there's a rather huge difference between the C64 and NES versions of the game: whereas the NES version increases the difficulty of each level on subsequent loops by loading the base level and running a script to make small tweaks (a blocked passage here, an extra enemy there, and so on), the different difficulty levels on C64 are practically entirely different levels from one another, requiring drastic changes to the routes and strategies. I estimate that such a TAS would take about 30 minutes more, but I don't think that fans of this game who have waited so long for a TAS would mind watching that.
About the ending input thing, personally I would consider it a perfectly valid speed/entertainment tradeoff if the movie hits the reset button (assuming one exists on the C64, which I'm not familiar with) to return to the title screen and leave the viewer with the music.
That all being said, I will now go and actually watch the video. Thank you for this, it's really made my day!
Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
If this site is going to keep a 1-level record, I'd much rather prefer it to be L20-HI, to obsolete the current movie.
Then again, perhaps the community will vote to replace the 1-level record with the full run.
Either way, while this is a cool run, I'm going to vote no.
Oh my god.
After watching the movie and reading the entire thread, I've concluded that this thread is undeniable proof that this community likes to argue over nothing.
https://xkcd.com/915/
As someone who hasn't seen every SMB TAS ever published (I haven't seen MrWint's TAS which is brought up several times in this thread), this looked like a quintessential TAS to me. It had a lot of tricks and glitches I've never seen before, and a 47-frame improvement is a nuclear bomb to a frame-war game such as this one. It's one of the most famous games in the world, and a monumental progression to one of the most culturally significant TAS games of all time. I see absolutely no reason this shouldn't be published.
But then again, maybe people are nitpicking because they know the movie is going to be published anyway, so they act contrarian because they know it will have no consequence and won't affect the outcome. Well, chasing skilled TASers like HappyLee away is a consequence. And it wouldn't be the first time, either. I've watched this community chase away dozens of good players over pure pedantry over this past decade over arguments that they didn't even start and weren't even their fault.
Scepheo wrote:
Because if you're bouncing around the level for 10 seconds, that looks like showing off, whereas slowing down for 10 frames looks like an oversight or mistake. At least, that's what it looks like to me in this TAS: I know it's intentional, I know it doesn't waste time, but feels slower.
I'm only speaking for myself, but many of the most memorable TAS moments I can recall are the moments that made me think the movie was desynching or that the player made some kind of obvious mistake, only to be revealed that it wasn't a mistake at all. A good example would be in the SMB2U TAS on world 4-3 where players like to make Luigi stall in the starting room, which clearly looks like a mistake, only for the viewer to realize later (or even better, not to realize and have to ask about it) that the Birdo's eggs are on a global timer and he needed to wait for it. To me, mastery and perfection are inherently entertaining, and making a fool of the viewer is one of the highest forms of demonstrating mastery.
As for the sentiment about discouraging playaround with spare framerule time for the sake of making things easier on future TASers: I find that a bit silly and irrelevant, as I imagine that future TASers will have no trouble finding the resources they need elsewhere, and if not, a skilled TASer who understands a game can generally look at a movie and see where all the time/entertainment tradeoffs are anyway.
And as a side note:
Memory wrote:
I'd like to say that requiring everyone who criticizes a work to make something better themselves is not being receptive to criticism. You don't have to be the cook to criticize the food.
There's a huge difference between criticism and opinion. Criticism (at least in the sense of artistic criticism) isn't about liking or disliking the end product, it's about showing that you understand what the creator intended, and offering advice about how they could have reached their intentions more effectively. In other words, you can dislike chicken, and you can criticize a cook for cooking chicken badly, but you can't criticize a cook for cooking chicken because you wanted steak instead. In the case of HappyLee, his artistic decisions were fully premeditated and intentional, and you can dislike it if you want, but it's unproductive to tell him that his artistic direction was a mistake.
Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
Even memorization is a skill, something which many people can't do well.
I hear this a lot, when talking about hard accomplishments in games, that "well, anyone could do that if they spent enough time, and I just don't want to waste the time, therefore your accomplishment is relative and meaningless."
Firstly, I don't find that to be true. I feel like some things are are simply out of the reach of some people regardless of practice, similar to how a person in a wheelchair can't play rugby. Secondly, even if it were true, it might take one person a year of practice, and another person 50 years to develop the same skill, and that is the difference.
Whenever someone says "get a life," I reply, "If getting a life means not doing anything novel, then I don't want one"
Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
Some games will never be accepted, but I think we'll still make the best TAS we can for them.
Ideally, once all the publishable games are published, we can simply begin counting movies that are made but not published, and perhaps get a wiki page up to list TASed games that are unpublishable but part of the project.
Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
My speedrunning contributions...
-The hyperspeed carpet glitch in SMB2U
-The "boost ball instantly ends Ridley's first form" glitch in Metroid Prime
Not much, but they're pretty significant glitches, and I always feel happy seeing them performed every year in GDQ.
Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
Mothrayas wrote:
This whole argument rests on the idea that because categories are humanly defined, they're subject to change. In practice though, that doesn't really happen - at least not to any relevant capacity. Name a single publication on this site for a Vaultable category that ended up having its goalposts moved for what constitutes any% completion or full completion.
The stage -3 ending SMB run. It was first submitted as an any% run, then there was some drama over whether it should have obsoleted the warps (any%) run, and in the end it was put in its own category.
Note that you said "vaultable category", not vaulted movie, and any% is a vaultable category. Besides, I don't think that "any%" and "100%" should change definitions depending on whether the run is being considered for vault or not.
Mothrayas wrote:
This violates our Vault rule directive of being minimally subjective, on top of being practically impossible to define. Where do you put the limit on what's acceptable or not? Also, I don't think "we reserve the right to randomly reject movies from the Vault that meet all requirements for publication" is a very favorable rule.
Sometimes, there is little difference between a rule, and a principle on which the rule is based.
Let me ask you this: what, exactly, makes Pinball and Desert Bus so bad? What constitutes an undesirable game, worthy of molding the vault rules specifically to keep them out? I'm not saying I want to see these movies published; I agree that publishing these movies would be a ridiculous violation of common sense. I'm saying that every objective rule has a subjective sentiment behind it, and it's better to simply describe the sentiment, rather than building rules to enforce and obfuscate it. This is transparency. Is Desert Bus unpublishable because it is too long? Then why not state that outright?
I like the vault and I'm a huge proponent for its existence, but edge cases with the vault bring out some of the worst of bureaucracy of this site. It's so much simpler when we can just collectively say something is fun to watch or not.
Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
I had an idea for an interesting "magic trick" that could be presented at a TAS block.
With Super Mario World, use arbitrary code execution to make the game play itself, similar to the "Ends Input Early" movie from last year. Upload the contents of the TAS to the SNES itself, then unplug TASBot, and the SNES will complete the speedrun on its own with no input. This would be and interesting an unexpected trick that could easily be played up for comedy, e.g., "Super Mario World 11-exit speedrun played by the Super Nintendo Entertainment System" or "played by nobody at all"
Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
Mothrayas wrote:
This was answered in the Punch-Out submission thread - you're comparing apples and oranges, or to be precise, you're comparing optimization level (which cannot be precisely defined) with completion criteria (which can be precisely defined, and which have to be precisely defined for any movie to be publishable - any movie that does not reach a clearly defined goal or end-point is rejected). Remember that the Vault only allows any% and full completion runs - if a max score run isn't full completion nor can prove itself to be such, it does not count as full completion, nor any%, so it cannot be accepted for the Vault.
I don't think this is an unfair comparison. What if a vault submission aims for 100% completion, but it is later discovered that 101% completion is possible? It implies that the first submission's goal wasn't precisely defined. I seem to remember a "105%" Donkey Kong run recently which begged this question.
I think that "full completion" is just as subjective and fluid to define as any other goal -- for instance, Zelda 1 "full completion" doesn't toggle every save flag such as bombing open all the secret doors. The community came to a consensus that collecting all items counts as 100% for the game. Every game is going to involve this kind of consensus at some level.
Even "beating the game" can be controversial, such as in the SMB -3 stage ending, which "beats" the game simply by setting a flag (Beaten castle level in world 8 or higher) rather than beating Bowser. This could be considered the same kind of memory corruption we're trying to avoid in score-attack runs.
Vault rules wrote:
Maximum points or score is allowed as a full-completion category, provided that:
There is no better way of defining full completion in the game.
The maximum score is easily defined and absolute - it must not be possible to gain a higher score, even theoretically. It must be definable without being dependent on precise time, speed, or similar requirements.
The maximum score is limited by not being able to gain any more points, not by hitting a score cap or overflow. If it is possible to score points infinitely, score cannot be used to define full completion.
Can you give an example of a published movie on this site which falls under these rules? Personally, I feel that this wording is very poor and does not convey the spirit of what it was intended to mean. If we have no movies which fall under these rules, I'd advocate changing this wording to something more clear, such as "Score attacks are not allowed in the vault"
Mothrayas wrote:
Sorry, but I am going to instantly reject any proposal that would force us to publish a 74-hour pinball run, or a 792-hour long Desert Bus run. This is one of the main reasons we require entertainment voting support for publishing movies with goals like this.
Add a disclaimer: "We reserve the right to reject any movie which is unreasonably long, even if it meets all other requirements for publication, on the basis of reducing stress on our server and our publishers. Such movies generally include those which are over ten hours long and largely consist of no meaningful content."
Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
marder1991 wrote:
maybe I'm not completely sure what i did wrong or what I'm missing. what can i do better in the future. i thought the point of tas was to make cool looking play through as for times i got hit it was intentional if i did not want to get hit i could have just rewind.
It helps if you think of this site as a leaderboard that only accepts submissions for world records. Unlike speedrun.com where anyone can submit a run and get put on the leaderboard, here you will only get put on the site if you beat the existing #1 record.
There is a gray area for playaround runs, but the vast majority of the movies on this site were made to break records. Also, we already have a SMW movie that aims for entertainment, so if you wanted to submit a SMW playaround, your movie would have to surpass the existing movie in entertainment.
Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
I feel that it should be published to the Vault. It is a relevant run to the SMB community, which I feel that this site should host.
However, judging from just the way it looks and plays, and from comments in this thread, it does seem like more of a port than a revision of SMB. Saying this should obsolete the SMB run because it's faster seems similar to saying that the run of GBC Super Mario Bros. Deluxe should obsolete the NES one.
Currently, the definitive version of SMB is the U version, simply by overwhelming popular consensus. Because of that consensus, it should be observed on this site as the definitive version as well. Relevance sets precedence, not the other way around.
Now if this gets published, does this mean we're setting a new precedent and the whole site should go crazy publishing the PAL version of every game with a NTSC publication? No, because relevance is a key factor. If there are other games where PAL is a popular separate category? If so, they should be considered for publication individually, but not every single game.
Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
I have solid plans to study Sokoban someday, and having a reference movie like this with known perfect solutions for so many puzzles is an awesome idea and something I would definitely want.
Yes vote! This is a movie not meant to be watched from beginning to end, but it definitely deserves to exist.
Experienced Forum User, Published Author, Active player
(325)
Joined: 2/23/2005
Posts: 786
One of my favorite old puzzle games! I greatly prefer the music on the PC version.
I'll watch this in multiple sittings, but what I've seen so far is very satisfying.
In this same position, here's a shot I found after about 2 minutes of attempts (though admittedly I wasn't holding the fast-forward button when it happened):
