Say a cutscene last a minute, and you implement a technique/glitch that takes 30 seconds, and consenquently save 30 seconds of non-gameplay.
Did you save 30 seconds or lost 30 seconds of "gameplay"?
I think you would go for saving 30 seconds gameplay, even if you lose it in cutscene. Cutscenes usually are skipped/removed so it is pointless to ruin actually gameplay over skipping a cutscene in-game.
Cutscene are definitly not gameplay, does that mean theres no merit to avoid a cutscene in the end?
Ending input sooner has merit as long as the game still ends properly eventually. Another traditional approach is making that proper game end happen as soon as possible even at the cost of longer input. Cutscenes or not, doesn't matter what exactly you optimize and how, it's just up to you which ending metric to use.
arukAdo wrote:
Taking this by the letter sound like just watching the cutscene would obsolete skipping it, which I suppose is not what you want for counting frames/differences.
If your new movie is shorter or it makes the ending happen sooner, then it's a legitimate improvement regardless of cutscenes. But taking someone else's movie that ends input early and switching it to fastest ending or vice versa is not an improvement. Improvement must apply using the same ending metric as before. Then if it's faster that way, you may change your ending metric on top of that legit improvement.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Cutscene are definitly not gameplay, does that mean theres no merit to avoid a cutscene in the end?
arukAdo wrote:
Yeah sorry, I did not mean a cutscene at the end of the movie, I meant in the middle of playtrough.
Please decide.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
In this movie [2541] PSX Castlevania: Symphony of the Night "luck mode" by ForgoneMoose in 18:40.08, there is a couple cutscene that you have to go trough, because its baseline it does not use sram, which mean you cannot skip cutscenes by pressing start.
The first cutscene is unavoidable (maria).
The second cutscene is cleverly skip by going outside bounds with glitches and completly skip the trigger (richter cutscene), it saves some seconds compared to watch the cutscene.
-it is possible to go trough the trigger, and still keep going trough the second part of the game, wanted to make sure doing that mean loosing time, not saving, since its longer than going outside bounds. (and technically its not "gameplay").
The third cutscene happens right before dracula with shaft boss; I wanted to make sure that if I manage to skip that cutscene trough a glitch (like I did in the recent pacifist), it saves time compared to just wait the cutscene is over.
Now on a different matter and where it starts to be complicated is, the glitch is for jp version only, in us version you get to skip the FMV at the game start, in jp you cannot skip that, but, you can skip the cutscene right before you switch to alucard, while in usa you cannot skip that one.
Cutscenes ingame in jp are also ironically, slightly longer than their us counterpart, at the very least for prologue and shaft.
Bottom line, what I would hope is, the intro and fmv doesnt matter, but skipping the dialogue for shaft does really save time/valid for obsoletion (its at least 40 seconds compared to wait the cutscene to end).
Edit: "in the end" in the context of the quote meant "after all"
what I would hope is, the intro and fmv doesnt matter, but skipping the dialogue for shaft does really save time/valid for obsoletion (its at least 40 seconds compared to wait the cutscene to end).
So switching to a different version allows you to save time in one part of the game but lose it in another?
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
So switching to a different version allows you to save time in one part of the game but lose it in another?
It save time in a gameplay part (skipping a boss cutscene) and could/should/might loose time in non-gameplay unskipable part (fmv, dialogs).
Its pretty difficult to quantify this without going to the end, since well the boss cutscene is really close to the end of the movie.
If your movie is faster on the same game version and emulator, you can obsolete the current publication. I believe the rule you're referencing is only there to prevent people from using the same gameplay on a version or emulator with faster loads, text, cutscenes, etc. If skipping a cutscene is faster than watching it, skip it.
Heh, the emulation itself is a nightmare to track since it will be psxjin vs bizhawk 2.10...
Im trying to reference those on a spreadsheet but its difficult, let alone the +- for gameplay or not.
It's better to have actual gameplay improvements that work in both versions than to redo the entire movie for a skipped cutscene and potentially lose more time on something unskippable, especially on a completely different emulator that'd make it even harder to definitively compare.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
That one is quite different as it also defeats Counter Hunters for Zero's parts to achieve best ending. It is not exactly Low%, as Zero's parts count towards the percentage of game completion. On top, it has a unique route designed to get Zero's parts faster.
Regular Low% category would disregard that, and will clear the game without collecting a single upgrade or any of Zero's parts. Route will be different compared to the movie you mentioned.
I do intend to obsolete first the us version with minor gameplay improvements.
Thing is despite the difference in non-gameplay parts (that I didnt calculated yet), the total dialogue for shaft, before and after fighting it, ammount approx to 1min40sec... so thats quite consequent.
On a different matter, I see the old pacifist "low level" was relabel just "pacifist" but the new one is still label "luck mode" and does not obsolete it, what gives ?
That one is quite different as it also defeats Counter Hunters for Zero's parts to achieve best ending. It is not exactly Low%, as Zero's parts count towards the percentage of game completion. On top, it has a unique route designed to get Zero's parts faster.
Regular Low% category would disregard that, and will clear the game without collecting a single upgrade or any of Zero's parts. Route will be different compared to the movie you mentioned.
Low% is a standard goal now, so I'm pretty sure both "no upgrades" movies are fine, as long as they're decently optimized.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Let's go with completing all the discs. But what does that actually entail here? I would like to tas from beginning to end, from the start of the first disc all the way to the end of the third disc. So as far as a stopping point when switching discs should it be the very next frame after I press fire to defeat the last enemy in the first disc? Should it be the very frame after the last save point? Or should it be after the credits? And yes, each of these discs has its own credits as though they are individual games.
1) Are each of these discs individual completable games?
2) If not and if these should all be considered one long contiguous game, then at what point should I switch discs from 1 to 2, and from 2 to 3?
What's the minimal required action in the previous disk that would properly continue in the next one? Is there any in-game prompt to swap a disk?
I think it makes sense to play them individually as separate goals, and then one continuous play as yet another goal since the game lets you carry over your stats.
The minimal required action to get to the next one is...nothing. There are no requirements needed to start up the next disc. However if you don't load the previous game's data then you lose out on powerups, but the saving and loading takes long and the powerups are typically not worth the time tradeoff. There is no in-game prompt to swap to the next disc.
Sorry saw your replay late.
However if you don't load the previous game's data then you lose out on powerups, but the saving and loading takes long and the powerups are typically not worth the time tradeoff.
Then maybe 3-in-1 is not worth it for standard? There, savegame only makes sense if it makes things quicker.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
However if you don't load the previous game's data then you lose out on powerups, but the saving and loading takes long and the powerups are typically not worth the time tradeoff.
Then maybe 3-in-1 is not worth it for standard? There, savegame only makes sense if it makes things quicker.
The Movie Rules article says that ROM hacks need to be in a "well-known database" and then lists some examples. I imagine this extends to homebrew titles if I'm not mistaken. The game I'm TASing is currently only released on GitHub. Does this count as a "well-known database"?
Edit: The game is also listed as a "beta". Would it still be acceptable to submit this TAS?
Homebrews are judged like normal games, there's no database for them like there are for hacks.
Prerelease versions are fine if there's no release ones, and they may be obsoleted by the movie of the latter.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
I'm interested in getting a grand prix TAS published for the game F-Zero GX.
I wanted to get verification on two speed-entertainment trade-offs that will probably need to be noted.
Nearly every GX TAS in existence imposes these two restrictions because they change the gameplay so drastically.
1. No Hyper speed side attacks (HSSA): by alternating the accelerator and side attack button every frame you can spam side attacks faster than the usual cap of one every 8 frames.
A run that uses this technique will typically spam it to gain height and pseudo-spacefly to the end of the lap.
Here is an example of HSSA in action: https://www.youtube.com/watch?v=PRC4hYYRaz4
2. No major checkpoint/lap skips/spaceflying shortcuts (there is no official name for this tactic so it can go by a few different names. My good friend midiman proposed the term "Ultra Shortcuts" as a catch-all term for these types of exploits.
The game normally allows you to skip a certain number of checkpoints and you can even progress your checkpoint status while airborne.
However these flying shortcuts are different because they exploit the checkpoint system by flying through the air and precisely passing through specific checkpoint planes. This allows you to finish laps without traversing much of the track at all.
Midiman is easily the most knowledgeable person on this topic and has documented all the different types of spaceflying shortcuts. He also helped me write several scripts including this section that provides a quick overview on "Ultra Shortcuts"
"-Ultra Shortcuts
Ultra shortcuts are glitches that exploit the checkpoint system to allow massive lap skips without having to drive through the rest of the track. If you're currently somewhere before the 9th checkpoint and you could somehow teleport behind the starting line and drive across it, your lap will count. No other checkpoints are required. However, if you try to cross the starting line backwards, the game will subtract from its internal lap counter and won’t count your lap when you cross it again forwards, for obvious reasons. And going around the outside of it won't work because the starting line is actually a plane that extends indefinitely in all directions, so there’s no way to go backwards without crossing the plane itself… or is there?
As briefly mentioned in part 2 of this video series, First-Corner type shortcuts are just one of the types of shortcuts that make this possible. At the first corner are non-parallel checkpoint planes that, when extended indefinitely, eventually intersect with the starting line’s plane. As long as the first corner has fewer than 8 checkpoints, flying around this area will allow you to cross the starting line backwards without decrementing the game’s internal lap counter. Courses with 90 degree first corners are the easiest to spacefly on because the point where the planes intersect occurs right next to the starting line. This is just one of many ways a player can exploit the checkpoint system."
Here is an example of a spaceflying run. It clears one of the longest tracks in the game in 14 seconds: https://www.youtube.com/watch?v=Y-JdQU_7yY8&list=PLgBXlc-aPuyixmvCOiHb2blK_acGibqYB&index=14
-
So long story short restricting these techniques still allow for insanely fast times, but they would traverse more of the track and showcase much more of the gameplay.
Please let me know if these restrictions would be allowed on a published run because that is my ultimate goal. I can also provide more information if needed.
Thanks,
Sam
1. No Hyper speed side attacks (HSSA): by alternating the accelerator and side attack button every frame you can spam side attacks faster than the usual cap of one every 8 frames.
A run that uses this technique will typically spam it to gain height and pseudo-spacefly to the end of the lap.
Here is an example of HSSA in action: https://www.youtube.com/watch?v=PRC4hYYRaz4
Is there a video of how it'd look without HSSA? Not necessarily that same track, but ideally 1 with HSSA and 1 without for the same track.
vsamxzero wrote:
As briefly mentioned in part 2 of this video series
What series?
vsamxzero wrote:
So long story short restricting these techniques still allow for insanely fast times, but they would traverse more of the track and showcase much more of the gameplay.
Please let me know if these restrictions would be allowed on a published run because that is my ultimate goal. I can also provide more information if needed.
How much longer would a movie avoiding them both be compared to absolute fastest with them, approximately?
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Is there a video of how it'd look without HSSA? Not necessarily that same track, but ideally 1 with HSSA and 1 without for the same track.
Funnily enough I made a TAS for the same track that the HSSA TAS was made.
Fire Field Undulation with Fat Shark Using HSSA (Maxspeed) - 22"082 [TAS] by midiman
https://www.youtube.com/watch?v=PRC4hYYRaz4
Fire Field Undulation with Fat Shark No HSSA (Maxspeed) - 31"175 [TAS] by me
https://www.youtube.com/watch?v=Zk0wiFxmn9Q&list=PLv6WH9UuWwlDanusfhglYBxJw-8RJyfyH&index=9
That first video may be the only TAS to use HSSA because no one seems interested in using it I guess. It saves a significant amount of time but that also varies depending on the track.
I'd like to revisit and improve that TAS some day but I can't imagine it going below 22 seconds.
feos wrote:
What series?
I've made a series of long form videos over the years that cover the history and discovery of advanced mechanics in GX. Midiman was a huge help in making them and provided all the information about spaceflying.
GROUND-BREAKING Discoveries That Changed F-ZERO GX FOREVER - [Part 2]
https://youtu.be/iTjh99dE87Y?si=EVv9avYu7FhCOyy4&t=1197
The part of the video referenced in that quote is at 19:57. It covers the discovery of Ultra Shortcuts and explains one of the first to be discovered called a "First Corner Shortcut".
feos wrote:
How much longer would a movie avoiding them both be compared to absolute fastest with them, approximately?
It is highly dependent on the track but spaceflying/HSSA will typically save a significant amount of time. The strategies they use have been optimized where they only have to fly the shortest distance possible to advance the checkpoint status and clear laps. Even RTA spaceflying runs will typically be faster than a non-flying TAS on most tracks since they don't have to actually traverse the track. The spaceflying video on Big Blue Ordeal I sent in the previous message is probably the most poignant example because you can complete the track by circling the very first corner on the track. A non-flying TAS couldn't possibly beat that time.
F-ZERO GX: MCTR (Flying) 12"925 [TAS]
https://www.youtube.com/watch?v=oGBgjO8VSL4
This might be the only spaceflying TAS ever made because all others choose not to do Ultra Shortcuts. A non-flying TAS hasn't been made for that track but it would probably be closer to 20 seconds.
This isn't always the case though. Some tracks can be optimized so much that they can beat the spaceflying times.
Sonic Oval is a short track that is famously the only case of the snaking WR (and the shaking TAS by extension) being slightly faster than the spaceflying WR.
F-ZERO GX: Snaking ~ Mute City - Sonic Oval - 0'24"790
https://www.youtube.com/watch?v=GVqUNXp64io
F-ZERO GX: Mute City - Sonic Oval - 0'11"245 【TAS】
https://www.youtube.com/watch?v=afE4xzKfSOs
This port town TAS is probably the peek of GX optimization while adhering to the two restrictions. It skips large sections of track but still manages to adhere to the checkpoint limit by strategically
returning to touch parts of the track to avoid exploding.
Looking at the description shows that CGN noted a third restriction for this run too. "No SpaceShaking" which is essentially shaking in the air to gain even more speed/height. I'm not sure if that should be a third restriction that should be noted. At this point even I am confused about how to handle all of this.
F-ZERO GX: Port Town - Aero Dive - 0'21"363 【TAS】
https://www.youtube.com/watch?v=SMKhCh_pmhk
That's not a trade-off then, but an entirely new goal, because gameplay is vastly different. Sounds like a great goal for the Alternative class, I just can't come up with a nice label for it that wouldn't be too long and still specify the 2 avoided tricks.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
On a different matter, I see the old pacifist "low level" was relabel just "pacifist" but the new one is still label "pacifist, luck mode" and does not obsolete it, what gives ?
Sorry to ask again but its been a month, is theres something blocking obsoleting the old tas with the new one? luck mode only improve the time as the movie not using it was sent into playground for that very reason.
