Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
Patashu wrote:
I'd love to see a category that takes on Deluge mode, since that's where the real challenge is, but I'm not sure what the category would be - max score, 32 planets sounds intense. 'fastest max score' might be more swallowable, and entertaining enough for moons.
Maxing every planet is clearly the "correct" category (perhaps also including unlocking every planet along the way, rather than using SRAM? That'd add interest from managing the rare metals). I'd be entertained, at least; the way you reach a maximum score for the planet can vary somewhat from planet to planet (the exception would be if there are any planets which require a very slow strategy – Hevendor is an obvious possibility for that due to its excessively simple mechanics – but more commonly you're scoring so fast that the score counter can't keep up). I'm also curious as to whether there are any planets where maxing out the score counter is impossible; if so, then trying to score as much as possible could be a very difficult optimisation problem. The resulting run might end up in Vault, but I wouldn't care; it's an obvious 100% definition, so it'd be acceptable, and seeing the strategies would be valuable.
Deluge is likely to be more interesting than Star Trip for someone who knows the game, because Star Trip is dominated by RNG about how badly the AI decides to screw up. When you're playing the game as a human, it's basically about trying to survive until the AI doesn't, which you don't have much control over (when the AI dies, it's typically for some silly/pointless reason rather than because you filled their screen, something which the AI can often survive!).
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
I believe a 100% run would pick up all the cosmetic upgrades. Many of them are placed in awkward places to serve as optional goals to aim for, which is sort-of the point of a 100%. Sure, they don't do anything useful towards a run, but that's fairly common with upgrades-that-contribute-towards-a-100%.
I'm very much in favour of categories which lead to a mix of techniques; that's why I think IGT-including-damage is a good category. Some levels will prefer to be completed directly to avoid the 3s penalties, others would want the wall clips, and it'd serve as a good "main category" for people who want to know what the game is like. Damageless IGT is fairly pointless; you'd want to use real time for a damageless run.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
BTW, in-game time is a category missing from the list above. I think it might be a good compromise between "phase through the wall and run to the goal every level" and the potential waiting times of damageless.
FWIW, I think a varied category list that showed off most of what the game had to offer might be any%-no-training, IGT-with-training, 100%.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
Voted no for entertainment, for the same reasons as on the other run. I do, however, believe that one of the runs should be published.
I think it makes more sense to publish the run without the training levels, because I believe this game can easily support multiple categories (and that alternative category choices would be more entertaining than the any%), and placing the training levels in a pure any% seems weird compared to placing them in a longer game. That said, if the alternative publication is one which can't zip along the walls of the training levels for whatever reason, that might be a reason to include them in the any%. (That said, given that I believe the any% to be unentertaining to watch – although very interesting to read about – leaving out the training levels may be required by Vault rules.)
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
I was the first No vote.
I think this run is an amazing technical achievement and should be published on that basis. I also found it very repetitive to watch; the intended gameplay is almost entirely missing (with the exception of Cloud 3, which is done in something approximating the intended fashion), and "clip through wall, go to the closest (intended or OoB goal area), clip back into goal area" gets tedious pretty quickly. As such, a pure any% seems like the wrong category for this game.
There's definitely space for a no-damage run alongside this one, I think; it'd look pretty much entirely different, but would likely show off TAS-level precision in an entirely different way. (It would, of course, be much longer). Other possibilities include an in-game time run (taking damage costs +3 seconds, so this would reduce the number of clips that were useful), or a 100% pickups run (potentially too similar to this one?)
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
Matslo123 wrote:
ais523 wrote:
Matslo123 wrote:
P.S In Iji 1.6 only the first sector can be done on reallyjoel's dad difficulty. After that there is a barrier blocking the way. I have TASed the first sector https://www.youtube.com/watch?v=NCGjinbDSL0
Isn't there some rule in the description of the difficulty that says you're supposed to kill every enemy to take the barrier down? (IIRC that isn't actually implemented because the developer knew it was impossible, but I'd expect a TAS to at least try.) You can see the same message on the screen, "100% kills required", when you reach the barrier.
Even walking to all of the areas of sector 1, let alone killing the enemies would take more than 2 minutes. You only have 2 minutes to do it so doing a legit TAS would be impossible. I don't see a point in only killing like the first 10 enemies and then dying to the time limit.
Well, it would serve as a good explanation of why the difficulty is impossible, and it might be interesting to see how far you could get. I guess the category would be "maximum kills within the time limit".
Walking to the barrier is less interesting because it doesn't look all that different from a regular run of the game.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
Matslo123 wrote:
P.S In Iji 1.6 only the first sector can be done on reallyjoel's dad difficulty. After that there is a barrier blocking the way. I have TASed the first sector https://www.youtube.com/watch?v=NCGjinbDSL0
Isn't there some rule in the description of the difficulty that says you're supposed to kill every enemy to take the barrier down? (IIRC that isn't actually implemented because the developer knew it was impossible, but I'd expect a TAS to at least try.) You can see the same message on the screen, "100% kills required", when you reach the barrier.
The very highest difficulty level of this game is a joke, and believed to be impossible even with TAS. (Someone investigated it using cheats to make it through the first level, and IIRC discovered that the author hadn't bothered programming the second, because the first was literally impossible.)
The highest non-joke difficulty level would likely be worth seeing, though.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
This entertained me, so I voted Yes.
One concern I had watching it is that it's unclear how this would differ from an unassisted speedrun (but I haven't seen one of those for this game). It feels like it could look the same, but maybe not? However, given the cube bounces don't seem to be an intentional part of the game, presumably they'd be too hard to get consistently without tools, providing the difference we need. (They should be part of the game, though! It's a very fluid form of movement.)
The levels also started to get a bit repetitive near the end of the run, but not quite by enough to bore me.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
dwangoAC wrote:
This is the wrong place for this but I haven't made a dedicated thread just yet - I would like to solicit a Celeste TAS for SGDQ 2019 and I think an All Berries or 114% or whatever category is ideal for this. I know there's some questions about what a "more complete" run might look like but I'm just throwing it out there that I'd love to see it happen. TGH is on board with attending to provide commentary. Thoughts on the viability of that?
The two most viable full-completion categories are All Red Berries and All Stages / All Hearts. All Red Berries has less overlap with the categories already shown at GDQ, though, so would likely be the best choice. 100% is basically All Red Berries + completing the B-sides (which you have to do most of anyway for ARB) and C-sides (already shown at a GDQ), so it'd be very comparable to All Red Berries but a little longer. 114% is probably a bad idea, because it's basically "complete the game, then do it again deathless", and deathless isn't that interesting in a TAS (but would be rather time-consuming). That said, you could probably add the secret gold berry as a donation incentive; that's the only really new content in 114% over 100%.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
When doing 114%, does anything force you to complete the A-sides on the first run through? You need the heart and cassette, but I can't see any reason to actually complete the stage until after the gold berries have been unlocked.
That said, you'd be playing all the B-sides twice, and some hearts come so late in their levels that you'd be playing most of those twice, too.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
I think Vaultability is only relevant when actually planning to accept a submission to Vault.
I agree that this submission cannot be accepted unless it meets the entertainment standard for Moons. (I haven't watched it, and don't know how entertaining it is.) If it does meet that standard, though, I don't see why similarity to the any% would be a reason to pick that run over this one; as long as a run is being published to Moons, it should be the most entertaining of the possible options (as that's what Moons is for, maximising entertainment.)
I believe that TASvideos should also host all TAS records for any% and 100%, and thus if this movie is published to Moons, the less entertaining any% should be published to Vault alongside it (showing the speed record).
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
I asked about a similar situation in another game recently. The opinions posted on the forum seemed to be in favour of allowing the all-levels run to obsolete the any% run if it was more entertaining.
(For what it's worth, you (ThunderAxe31) have pointed out here that any%s have been rejected in the past for being too similar to a full-completion submission, despite being a vaultable goal!)
I believe the current consensus is that only one of the two runs can be published, but that there is no consensus on which. (I personally disagree with this precedent, and think that both should be published if the any% run is the less entertaining of the two – the any% in the vault, and the similar longer run in Moons – but suspect I am outvoted on this.)
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
That's not directly comparable; doing "all levels" in Kuru Kuru Kururin requires a no-damage completion of most of them, in order to unlock the Last Land. So this would be "all initially unlocked levels" or the like, which is a somewhat arbitrary category.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
Suppose a game has multiple sets of skippable levels. Some are unlocked from the start, whereas some are only unlocked when you 100% the rest of the game (which includes optional goals in the nonskippable levels).
In a non-full-completion category, what would be preferred, completing only the minimum number of levels, or completing as many levels as possible while keeping to a fastest completion of each level that's actually done (thus skipping the initially locked levels)? The former seems less arbitrary, but the latter shows more content.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
The tutorial is skippable, so I think it'd be interpreted as not being needed in an any%.
An "all levels" run will be great, but it should probably play the postgame levels too (which would in turn means you have to unlock them, making a rather different category from any%).
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
I like the little entertainment tricks in the times when you have to wait.
This run wastes quite a bit of realtime to improve the in-game time. That's noted in the submission notes, but needs to be noted in the movie/publication notes too.
For anyone who doesn't know the rules of this game: most types of block will "glue"/"stick" to other blocks of the same type; any group of 4 or more blocks of the same type will disappear after a short time period when disturbed (i.e. it's just been newly formed, or it falls or a new block sticks to it). Silver X blocks are an exception, they can't be destroyed by normal means and don't interact with anything (thus serve to get in the way). The playable character can destroy a block they're adjacent to (in any direction), move left or right, or climb diagonally up/left or up/right if there's a block to support them; and falls if there's no block beneath them.
Brown X blocks take five hits to destroy and come with huge penalties when you do so (varying by game mode; in this mode, I believe it's an in-game time penalty), which is why this run doesn't destroy any of them (even when it looks like it would be faster to do so), except by allowing them to form a disturbed group of 4. Crystal blocks disappear after a short time regardless of whether they're disturbed or whether they're in a group of 4 or more (their main purpose is to cause a player who's trying to play more cautiously than this TAS is to react fast). There are also a few special blocks that do things like flipping or rotating the playfield, or making some blocks into crystal blocks; and timer blocks which directly subtract from your in-game time (and are necessary to get a good score on many levels).
In casual play, there's two main ways to play this game. The "cautious" mode is to stay on top of the collapsing pile, trying to drill down through it carefully and letting the pile stabilise before digging further. You can also do what this TAS does, and try to outrace the falling blocks; of course, in Time Attack mode, that's pretty much the only viable way to play.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
I'm not sure if it applies here, but: does it make a difference if you don't enter the password manually, but rather the game automatically enters the password from your previous save file after a reset? Several games will automatically fill in the password for the state they think you've reached for you, and continuing from that state feels to me more like reloading a save file than entering a password by hand.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
fmp wrote:
It's not fair in the slightest to say it "looks massively suboptimal", especially when I made sure the submission text covered anything and everything that would be useful for the audience to know. It's a lot, yes, but that's what you're going to get for a complex run. Plenty of runs do things that aren't understandable just by watching. Part of reviewing a TAS for the quality of its optimization is doing research on how the game works. The onus is on the author to explain things of course. I did my part, so you should do yours.
I'm not saying that it is suboptimal, just that it looks like it if you don't have a lot more knowledge of the game than most viewers will have. As it happens, I did read the submission text, but many viewers won't. (In such cases, it's common to use subtitles or audio commentary to explain what's going on, although other visual aids could also work.) Even so, though, I had to give up watching the encode halfway through because it didn't really add anything to the commentary; what's actually going on has sufficiently little correspondence with what's on screen that you can't tell anything about how good or bad the run is, nor try to work out what actions the run is taking and think up alternative strategies, or the like. (And even when I understood from the commentary that, say, you must be loading a screen adjacent to the one you're on with no visible change, having no guide as to how the screens are laid out or what the new screen looks like means that that's very similar to TASing a level entirely in the dark; the entertainment level of that in a TAS is low because you can't appreciate all the smaller things that the runner is doing.)
My suggestion for a good visual aid would be to add a couple of extra screens beside the main screen. One would show the game map as the game sees it, including a dot showing Link's position and a rectangle showing the camera position. The other would show the collision around Link (and perhaps use a different color to show collision on other layers of the same area). (Obviously, you can't change/remove the main screen because that's important too.) Those would make it clear what was happening when a glitch was used (e.g. if you see Link teleporting a single screen width in one direction and the camera staying still, the effect of the glitch you just used would be clear). It'd also likely mean that there would be no need to add additional commentary.
It's obvious that a lot of work went into this TAS, so it's a pity that it's presented in a way that means that it's impossible to appreciate that work. Because this game allows so many forms of movement, and ones that often have unusual effects, TASing this category means that the game is no longer about the intended gameplay; what the run focuses on, and what you're optimising, is the "wrong warps route" where you're moving around the game blending intended and unintended methods to reach the points at which you gain the various items that make up 100%. It's that movement that's the focus of the run, and the entertaining part, and yet as the run is presented you can't really follow any of that.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
This run could really benefit from an encode that shows information showing how the game fits together from the point of view of the glitches used in this TAS. I'm finding it hard to enjoy watching from the direct encode, and the reason is because there's basically no information visible that lets you see why the TAS does anything; it's basically "run around a bit in a room, now you're somewhere else, which may or may not be visible onscreen". In order to be able to follow the run, you'd need to show not only the view as the game shows it, but a map showing where Link currently is, and a picture of the collision that's loaded and that Link can collide with. The run just looks massively suboptimal when there's no information available as to what constraints the run is facing that determine where Link can and can't go; the run gives the appearance of being able to go anywhere and yet sometimes going around in circles instead, and although I know that's not what's actually happening, I can't see what is.
I'm not currently voting; although I'd vote for No based on the present misleadingness of the encode, this seems like a fixable problem and I couldn't change my vote once it was cast. So consider my vote a No up until the point where this sort of issue is fixed.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
Variation in RTA timing method is addressed by linking to the game-specific rules.
Wait, are you assuming that all speedrunning communities for each specific game use the same rules as each other (e.g. by following the rules on speedrun.com)?
speedrun.com isn't the end of the Internet, and the decisions made there might or might not be representative of other communities for the games in question. I'd be opposed to giving it any sort of special status, especially as it has no consistent global rules (with the rules for each game being up to a moderator who might or might not have any strong connection to the community of the game in question).
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
For high-profile runs, we sometimes write the time using other timing methods in the movie description. That seems like a better place to put it than having a centralised page listing every run.
Bear in mind that you have to specify the exact timing method used, because timing methods vary among the unassisted speedrunning community. (The term "RTA" used to refer to a timing method that went from power on to end of credits, always giving a higher time than TAS timing, although it's become more generic since because the timing method in question is rarely used nowadays. Other timing methods used for unassisted running, such as SDA real-time timing and in-game timing, can give different results.)
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
It needs to be listed, somewhere in the category name or in the submission post's extended description of the category, that resetting during a save is categorized out from this run. Otherwise the run is just suboptimal, as faster game end glitches than the one here exist (specifically, resetting during a save can also be used for a game end glitch).
Saying "game end glitch" in the title is all very well, but when multiple such glitches exist and the run isn't using the fastest, you need to explain why.
Editor, Experienced Forum User, Published Author, Player
(44)
Joined: 7/11/2010
Posts: 1022
I think it's reasonable to treat a microphone like a controller; any sound that the microphone is capable of picking up seems like valid input to me (as a user sitting at the console could input it via making the right sounds). In fact, it's considerably more legitimate than things like left+right (which couldn't be input without damage to the console).
This is different from, say, expansion port input, as creating your own expansion cards is not part of normal play and not an intended input method.