Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
I'll let it start with a Yes vote. Although I can't read Japanese, I still understood how this game works (partially at least). And the strategy is looking quite insane.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
This movie is incomplete and sloppily played. It even fails in completing all tables -- if you walk through all tables for another time, the ball changes its shape again.
I would prefer to see this run to be published, shouldn't be such a big problem with the rules of the Vault. (And I guess Inzult still has the input file around.)
Just put this movie back to the "rejected" pile.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
CoolKirby wrote:
Vaultable, Yes vote.
If you want it in the Vault, you should have voted no, that's the idea of the Vault... Yes -> Moon (or maybe star), No -> Vault.
I just followed the wishes of the movie creator.
What happens when the score overflows?
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
And how do we handle such cases as the aforementioned Battletoads?
1920M star
2100M faster but a lot less entertaining -> vault + no obsoletion / partial obsoletion ?
Or even worse, Chrono Trigger? Some movie in the middle got a star, the latest movie is also a lot more boring.
I dislike the idea of explicitly publishing movies as "boring crap", but I do like the idea of building a TAS speedrun archive.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
feos wrote:
I'm strictly against separating in speedrun/superplay fashion. Because that means the GOAL would lead to a tier, which is weird.
Well, that's not the plan, but that would make it possible to improve the16 stars run if the same restrictions are applied (and IF "16 stars" actually is a good goal at all, as far as I know this has never ever been an accepted goal). There would still be one more restriction: The improved movie must be entertaining enough, because those movies are neither any% nor 100%. This restriction wouldn't apply for "pure speedrun".
It's the same case with Battletoads (add "don't skip the game"), Donkey Kong (add "don't skip that much"), Chrono Trigger (add "non-glitched"), Po-ke-m-on (we would have only one Pokemon movie if speed is the only matter). Well, each single game would need to be discussed if there is a good reason for some obscure "new" goal to be invented.
The problem with "Vault" is: If we decide to use Vault / Regular, this implies that Vault movies are boring to watch. "Vault" would become "not worth watching". If we just have a "speedrun" branch, this kind of assumption is invalid, but it would require certain movies to be published as "speedrun" and "superplay" at the same time.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
Once upon a time, TASvideos was all about entertaining movies. Now with the Vault tier, it's all about world records. Causual viewers will consider almost all of the movies boring.
Having more publications will make it harder to find the entertaining ones. And as game choice now doesn't matter any more, there will be a lot of runs for boring games. And now even dropping all entertainment requirements, there will be even more boring TASes for them.
Well, that would be the possible "worst case" at least.
OK, let's build a TAS speedrun archive. Let's do the same as SDA already does for non-assisted runs. But let's keep this archive strictly separated from the rest of the site.
If the idea behind the tiers is to build such an archive, it should be enough to have a "speedrun" section, and a "superplay" section on the site. We can still keep stars and moons, their meaning wouldn't be changed. If a "speedrun+superplay" movie is beaten by a less entertaining "pure speedrun", we can modify the goals of the superplay movie afterwards. This will most likely create somewhat esoteric goals ("use MIPS the rabbit to pass through the door", "don't warp to the end of the game"), but this shouldn't be such a big problem.
Reassigning goals for those runs would also allow for future improvements with the same goals and restrictions.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
Yes, I think so. Killing off some enemies should be quite easy, but lag optimization might be more difficult. I would go for killing enemies to reduce lag, because this obviously decreases workload for the console. If BizHawk emulates the actual console accurate enough, some more advanced lag manipulation might be possible. Well, if there actually is anything which can be manipulated in this way. I just don't want to start emulator-dependent optimizations (too bad this already happened for SNES Super Mario World).
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
This is one nice little game. Japanese only, but you don't need any reading skills to play this one.
Very hard to play, even harder if you want to create a good TAS of it. The fastest travelling method in this game is bumping straight into enemies, and hope that they let you bounce in the right direction.
I made a testrun and an encode of it for the complete game. I think it is improvable by a minute or two. And I think it proves this is a good game choice.
At least one level allows you to manipulate luck. Luck manipulation is quite trivial to do, luck based events can be manipulated even one frame before they happen. Enemy behaviour is extremely easy: They walk the same diirection as you, they shoot at you. Or they explode when you are close enough to them while being on the ground.
If you know some time-saving secrets about this game let me know.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
You can't get a run published if you rush those runs like that. Good runs require a lot of time. A few weeks or months is quite normal. Some take years.
Monopoly is a really bad game for a beginner. The game itself is quite uninteresting, but heavy luck manipulation counterfeits that. Manipulating luck like in the published runs is a really difficult task. Choose some more interesting game please, there are plenty of it. Fast platformers are a good choice to start with.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
You're right, a single core Pentium M 1.5 GHz is not a good system...
Turning rewind off increased my speed (still slower than realtime), then SNES got 20 fps. Moving the mouse cursor out of the game window finally increased it to 30 fps. So the perfomance penalty is about 30% for the mouse cursor, and 50% for the rewind feature.
30 fps, that's good enough. Ironically, SNES performance seems to be better than GB performance on my old laptop.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
The most recent SVN seems to fix the D3D-error I got. Good thing, one bug fixed.
But for some reason, I have a horrible performance with BizHawk. All systems I tested run with 20 fps or less. Average is 10 fps (GDI) to 12 fps (Direct3D). lsnes gets 30 fps. I'll need a new computer before TASing with BizHawk is fun enough.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
JWinslow23 wrote:
Why do I have to cancel all of my movies?
Well, an acceptable movie takes a lot of time to make... The "fast" and experienced movie makers need a few days up to a few weeks for a short TAS. Each of my movies took several months of time actively working on them (well, and long breaks in between.) Some movies even take years of work, for example the upcoming Super Mario 64 - 120 stars TAS will have taken 5 to 6 years to be completed...
Don't expect fast success. Try to beat yourself, and try again, and try again, and try again. If you find some, watch already existing TASes. Most of the time, you'll need to do some research on the game (RAM watch and the like), which can take long. Different strategies need to be tested. If you try hard enough, you will eventually beat the current world record. And finally, when you can't think of any further improvements, your movie is ready.
---
And about a 7 CPU run: this might be especially interesting if you give yourself totally horrible starting conditions, like: 10.000$ for every CPU, 1$ for you + 3 turns in jail... just my opinion. You would win despite facing a totally hopeless situation. Such ideas should be discussed beforehand.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
Well, we already have published movies for this game: [520] NES Monopoly by Acmlm in 00:30.03 and [717] NES Monopoly "4 CPUs" by FractalFusion in 01:07.67. The first one wins against one CPU, the second one wins against 4 CPUs.
There is one really interesting thing in the second movie: The input ends very early. Luck is manipulated for 8 turns in advance! If you get this movie improved for 14 turns in advance (to allow beating 7 CPUs after the end of input), that would for sure obsolete the 4-player run -- if you don't need to wait too long for the right result... But that will be VERY hard to do. Some disassembly was done beforehand, and you need to understand that first.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
Yes, for the sake of video game history. A site which deals with thousands of video games must have a video of the worst video game, even more if it is that important.
Well, I have seen some opinion that this game is actually average compared to its competitors -- if you have read the manual very carefully. But who wants to read the instructions first? No one, you most likely want to start right away... but this strategy is a miserable failure for this game, it just doesn't work.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
I used to play this game for hours during my childhood. I got the 10 million after a few hours of continuously playing.
The game has a lot more to offer than just 10 million points. The ball can become a triangle, and later on it becomes a cube. I think it's more fun to activate all extra lives, play all tables once, and get those helpers which kick your ball back to the center of the table.
For a special challenge, you could show how you survive with both flippers destroyed (there are rockets in two levels which can damage the flippers), or even advance to the next table afterwards. I don't know if that is even possible though.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
Well, now I also think that the current run is improvable.
25 lines to clear. 12 lines filled with garbage. 7 line clear animations required.
We need to manipulate a "good" garbage pattern. We require garbageblocks mod 4 == 2, and as much garbageblocks as possible. manipulating 4 additional blocks should take no more than 30 to 60 frames, manipulating 8 blocks shouldn't take more than 60 to 120 frames (exact numbers need to be tested). Also, as much garbage as possible should be at the bottom (less pieces need to be dropped down deep.)
The current solution requires 10 line clear animations, which takes a known amount of frames (something around 30 frames, you know it better than me).
Lets assume that we know the solution for every seed with the lowest amount of line clears. So we need to check at worst all earlier seeds and the next 30*(10-7)*2 = 180 seeds if they allow solutions with less than 10 line clears. (Well, if the current seed allows a solution with less than 10 line clears, it will almost definitely be improvable.) I would guess that only one out of four of the solutions satisfies garbageblocks mod 4 == 2, so that would only be ~45 seeds to check.
On average, we need to clear 2 lines of garbage, and 2 extra lines. If we can manipulate a garbage pattern which allows 3 or even 4 garbage lines to be cleared at once, that would be nice. This would increase chances for a solution with 7 line clears.
Clearing more than 2 lines of garbage at once is most likely impossible. We can afford to clear only 3 lines exactly 3 times. We can also afford to combine 1 line of garbage with 2 or 3 extra lines.
If we have can choose where we do the extra lines, then they should be done as early as possible.
Well, and one thing which makes all of this more complicated If the blocks don't need to be dropped that deep, we require less time until the next block starts. It might be well worth the time to do the extra lines very early and do an extra line clear later on. Doing line clears at the top requires less time than doing them at the bottom. If N is the number of frames it takes to get a block down one line, we may save ~2,5*N frames for clearing a line one position more at the top.
Let's build a formula for exactly calculating the time.
One more idea At the end, manipulate garbage and build the bottom in a way which allows input to be ended before the last pieces are dropped.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
At about 045 you're running straight into a wall, and after glitching through it, you run straight into the pipe and stop your forward movement for a moment.
That's quite an interesting game, go on with it.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
That's a nice little script, and it's even quite intuitive in this version. Well, it would have been, if there were instructions on how to use it... But finally I got it
So here is how this one works if you want to try it for a two-player game
("the lights" are those circles directly below the input display)
- savestate, and play around a bit with player one. press "z" (-> those red dots on the bottom turn green), then "l" (-> lights at the bottom turn red, and the second controller on the bottom gets activated)
- loadstate, and now use the same buttons to play around with player 2. when you're satisfied with the results for player two, press "z" (-> lights turn green), then "l" (-> lights stay green, first controller gets activated), then "z" (-> lights turn red). savestate.
- play around as player one. press "z" (lights turn green), then "l" (-> 2nd player is activated, lights stay green), then "z" (lights turn red). loadstate
- play around as player two. press "z" (lights turn green), then "l" (-> 1st player is activated, lights stay green), then "z" (lights turn red). Now you should know how it works...
which controller is active, and which buttons are to be held can be seen clearly at the bottom.
there is also a "sticky" input mode which can be activated and deactivated by pressing "x", which is essentially the same as "typing input". for example, pressing x,d,frameadvance,frameadvance one after another will hold down "right" for the first frame, and release it for the second one. pressing x,d,d,frameadvance will result in not having pressed "right" at all.
when the lights are red for a specific controller, previous input for these buttons is ignored. when the lights are green, then either the previous input is used, or -- if this player is active at that time -- the new input is XORed with the previous input.
I think this script could also support "mixing" red and green lights. This would turn out to be useful if you want to try different frames for a specific button, but want to keep the rest of your input untouched.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
Feature request Lua function to parse those lsnes memory watch expressions like C16zC0x7e0071zw*C16zC0x7e0075zB/+
This should basically be a one-liner to implement, because lsnes can already parse these statements.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
Just to let you know I could get about two thirds of this movie to sync in lsnes with minimal modifications. 40 frames slower due to even more waiting time between the levels. I'll upload it either when I could get it to sync completely or when someone requests the current result.
Experienced Forum User, Published Author, Player
(136)
Joined: 9/18/2007
Posts: 389
Somewhat late for a reply, but why not...
That's sort of strange game... Almost everyone likes to hurt TinTin for no real reason.
As for the desync problem I could solve my sync problems for Road Runner by turning "Sync Sound with CPU" ON and turning "Flexible Sound Samples Mix" OFF. This will result in absolutely terrible sound while playing, but everything will be fine when there comes an encode of it.
As Snes9x is now deprecated, you might want to switch to lsnes or BizHawk.