Posts for FractalFusion

Posted: 7/9/2011 8:30 PM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

X's run is much more entertaining than Zero's run due to many weapons and no slash delay. I like Slash Beast stage and the final stage the most. Great work, sparky!

Posted: 7/8/2011 1:33 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

Good thing I saw sgrunt's post. But there is another problem. The movie file needs to be fixed further, as the game does not proceed to the credits. It wasn't fixed in the published run.

Posted: 7/7/2011 7:41 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

<TASVideoAgent> New submission! Go and see #3215: x2poet's Arcade The King of Figheters 2002 "KULA+WHIP+ATHENA+VANESSA+LEONA+YURI+MAI+MARY" in 07:25.75 at http://tasvideos.org/3215S.html <TASVideoAgent> New topic by TASVideoAgent (WorkB: #3215: x2poet's Arcade The King of Figheters 2002 "KULA+WHIP+ATHENA+VANESSA+LEONA+YURI+MAI+MARY" in 07:25.75): http://tasvideos.org/forum/t/11474 [a:1] <TASVideoAgent> Post edited by x2poet (WorkB: #3215: x2poet's Arcade The King of Figheters 2002 "KULA+WHIP+ATHENA+VANESSA+LEONA+YURI+MAI+MARY" in 07:25.75): http://tasvideos.org/forum/p/278941#278941 [a:1] <BrandonE> Longest branch of all time <TASVideoAgent> Post edited by DarkKobold (WorkB: #3215: x2poet's Arcade The King of Fighters 2002 in 07:25.75): http://tasvideos.org/forum/p/278941#278941 [a:1]

Posted: 7/5/2011 1:23 AM

Post subject: I am Screenshotman

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

Posted: 7/2/2011 11:13 PM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

I have thought about Hidden Power for a while. HP Grass is only useful for those 4 Pokemon (Piplup, Geodude, Onix, Cranidos) and is pretty much useless afterward. HP Water may or may not be useful. HP Water is probably more fitting on Piplup because of STAB and Torrent, so that may be a possibility. I had also considered catching Abra and using a Hidden Power on it. HP Water, or HP Psychic, or something else? There are a few issues in any case. Although I had not planned it months ago, I might consider looking at the glitched route. This information will be useful anyway.

Posted: 7/2/2011 9:49 PM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

I've started another test run. This time, I cleverly manipulated Chimchar so that its stats are extremely high, and its Hidden Power is Fighting 70! Overworld RNG seed: 380011b7 (ex. 2029-01-01 00:01:54, framecount 0x119a). So the start time is 2029-01-01 00:00:38. Cycle after initial seed: 92. This number is a guesstimate since the number of calls varies (it is controlled by random looking NPCs in the area), but it was around 82 optimally. I had to do a bit of waiting to get to 92. Chimchar's Nature: 4 (Atk+ Spdef-). Sorry for referring to natures as numbers but that's the only way I think of them. Chimchar's IVs: HP IV - 30 Atk IV - 31 Def IV - 26 Spd IV - 30 Spatk IV - 30 Spdef IV - 30 Chimchar's Hidden Power: It is Fighting 70. Go calculate it. Oh yeah, I obviously used a simulator to find it. Because I am not a bot. I chose Fighting 70 because, not only does it work early on, but later when Chimchar evolves, it will become Monferno which is Fire/Fighting, so it will gain STAB on it. Anyway, against the first trainer with a Starly, it is possible to do a crit and a non-crit with Scratch. Better than two crits. Against the second trainer with two Bidoof, the first one can be OHKOed with a crit Scratch. I have not tested completely but it might require a max crit. Which I fortunately got in the test run. The second one goes down with crit Ember or crit Scratch. Against the third trainer with Shinx, it goes down with crit Ember. Against both Abra in the trainer school, they go down with crit Ember or crit Scratch. This is to get Hidden Power. Note that due to high speed IV, Chimchar outspeeds all of them. I get Quick Claw from one of the houses. Against the rival, Starly goes down with crit Ember or crit Hidden Power. Piplup on the other hand, survived a crit Hidden Power at 99% damage variation. Perhaps (and I didn't test), it might just be possible to OHKO it with a max crit. Against Roark, crit Hidden Power OHKOs all of them. This is the first battle which Quick Claw mattered. I discovered that Quick Claw appears to be determined much earlier than I thought, near the beginning of the previous turn! To get Quick Claw to work against, say, Onix, you need to manipulate at the point where you select an attack against Geodude. That's still all I know about Quick Claw. Other stuff: - I chose the boy for now. Assuming the change to girl has the nothing different other than boy -> girl and Dawn -> Lucas, using the boy is faster. - Fast text speed appears to be 1 frame faster at the beginning of some dialogs than scrolling mid speed text fast. And that's it for now. The Poketch entertainment potential is huge but I'll put that aside for now since it's only a test run.

Posted: 7/2/2011 2:01 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

Vykan12 wrote:

I feel like the game's trolling me, which is very strange.

This. I've never seen a run that does everything it can to make its viewers want to stop watching it. It's neither entertaining, nor educational, nor technically impressive, nor funny (except for the first stage, and the very end when you realize what you've just gone through).

Posted: 6/30/2011 7:06 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

FuzZerd wrote:

it took you ~2 weeks to TAS, it took me ~1month to watch. voting yes.

I think it took DarkKobold longer than 2 weeks. Especially if you count all the preparation time. RPGs are hard to plan out.

OmnipotentEntity wrote:

Since Mister Epic gave up, I'll give this encode a go.

Last time I heard, dada was also working on the encode.

Posted: 6/29/2011 5:13 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

Thanks for the input. I have replaced the screenshot with the Steven's dying Skarmory one. This one was already one of my two preferred screenshots for publication. As much as everyone likes the LOL screenshots, I would like to stand by my opinion that LOL screenshots should not be used. From http://tasvideos.org/PublisherGuidelines.html :

Pick a screenshot that represents the movie or the game well.

Most likely, an action screenshot (one that shows user action, or the product thereof) is most suitable for this purpose. However, for some reason, people seem intrigued by these LOL screenshots as to want them for publication screenshot. I have thus compromised somewhat, by allowing the LOL screenshot for the first few days of publication, then replacing it with the one I like. I did it with the Breath of Fire III publication as well (apparently, more people like naked women than I thought). The only few screenshots that I liked the most were the Magneton flinch, Wattson's dying Magneton, and Steven's dying Skarmory. There were others of course, but they don't have the NVE message, although you could make the case that critical hit message is good because that's what the run is about. The Magneton flinch spoils too much in my opinion (although if you think about it long enough, you might wonder "how could this possibly happen?"). Wattson's Magneton only gets its appeal from the fact that it is Wattson's, and that it has Shock Wave, neither of which are immediately apparent. That leaves the Skarmory one. For those who liked the "I'm going to give you a little taste of pain! Resign yourself to it!" screenshot, I'll place it at the top of the submission text.

Posted: 6/28/2011 10:44 PM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

OK, here's some more stuff I've found so far: - The PID of the starter Pokemon is generated by the first two RNG calls, and the IVs by the next two RNG calls.

aaaa....      PID: bbbbaaaa
bbbb....

cccc....      IVs: cccc -> (binary) .eeeeeff fffggggg    hpiv: ggggg, atkiv: fffff, defiv: eeeee
dddd....           dddd -> (binary) .hhhhhii iiijjjjj    spdiv: jjjjj, spatkiv: iiiii, spdefiv: hhhhh

- The PID of wild Pokemon appears to be generated on a frame with the fourth and third last values for PID and second last and last values for IVs, in the same manner as above. Not much else is known. - For the Pokemon catch tutorial, Dawn/Lucas Pokemon has PID determined on a frame by first two calls and IVs by last two calls. The wild Pokemon has PID determined on the next frame by first two calls and IVs by last two calls as well. Though it doesn't look like I can get a knockout this time. - In battle, the Quick Claw appears to use the same mechanic as in Ruby/Sapphire; that is, it is determined before the round of attacks. However, what determines it and how is still unknown. - The game keeps track of two RNG sequences, both using the RNG formula above. -- The first one is the "overworld RNG", which is initially seeded right after the "red Gyarados screen" with the clock and appears not to be modified under any other circumstances. It determines everything related to natures and IVs. -- The second one is the "battle RNG". Before every battle, the overworld RNG is placed aside and the RNG is seeded with the clock (as in the post above). Once the battle is finished, the overworld RNG is placed back as the RNG. As an addendum, I do see reports in many places regarding mechanics and probability, including Bulbapedia, that contradict some stuff that I've found in Ruby/Sapphire and in Diamond. I do not know if it refers to the latest generation (possible), or is done on a different ROM, or is just plain anecdotal evidence that has been passed along the years. That is why sites that document exact information must have their data cross-checked frequently. At least it isn't as bad as the Pokerus "information" situation.

Posted: 6/25/2011 11:21 PM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

This is one weird gladiator battle and it has absolutely nothing to do with hockey. Don't know how this compares with a TAS of Mutant League Football, though. Anyway, I wonder if it would be better playing as the worst team against the best, instead of the other way around.

Posted: 6/24/2011 7:40 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

Back here, for some more RNG analyzing. - As you know, the same formula is for advancing the RNG: value <- value*[0x41C64E6D] + [0x6073] - The battle system hasn't changed much. This time, the RNG cycles once every two frames, rather than twice a frame. This makes expected delays twice as long. If the RNG cycles more than once on a frame, it is determining something. A series of four calls (each two frames apart) determines the following:

1:
........
...x....         x=0 critical hit
2:
........
...y....         y damage variation 0=max
3:
........
4:
........
aaaa....         aaaa accuracy

This is done right after the move is selected. - Something new: the game appears to set initial seeds very often throughout the run, mostly right before battle. Any time the game sets an initial seed, it is as follows: xxyy00zz+dddddddd xx=month*day+minute+second mod 256 yy=hour zz=last two digits of year dddddddd=framecount=value at address 21c4834 And that's it! Pretty simple. Edit: It was on smogon.com too: http://www.smogon.com/forums/showthread.php?t=52180 - I initially had trouble trying to find the starter's Pokemon data in RAM, as the developers encrypted the values well. Of course, I had the idea of dumping RAM after the starter data was determined, then going back, poking another value for the RNG, and then dumping in the same spot again. Then I compare the two files with a difference reporter. Very simple; I wish I had thought of that sooner. Anyway, the PID for the starter Pokemon is the first two RNG values thrown after selecting the starter. PID is: yyyyxxxx, first value (top half) xxxx, second value (top half) yyyy That's all for now, but I plan to find some more stuff.

Posted: 6/24/2011 12:55 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

Interesting. I tried to avoid that one as it is a bit of a spoiler, but now that I look at it again, it is awesome. Viewers who know something about Pokemon Ruby/Sapphire will be baffled by this screenshot for ages.

P.JBoy wrote:

Is there any reason you didn't set the sound to stereo as you exited the lab after getting Mudkip?

I didn't set it to stereo because I had a hard time figuring out why I should bother. Meh, the music sounds the same.

You asked for it.

:)

Chamale wrote:

You mention that May's Treecko has a Def- nature, which would presumably make the battles slightly faster as a male protagonist.

Yes. Because of Def- nature, it saves a critical hit. That's about 1.5 seconds. Aside from that, and from what GoddessMaria said, I really wanted to try out the girl. At least to do something different than what was done before.

Posted: 6/23/2011 12:18 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

Good job, sparky! Looks like X will be very entertaining. Something interesting I noticed lately: A TASer called kannagi0611 (a.k.a. 神代凪) uploaded a Rockman X4 + Rockman X5 dual run WIP. http://www.youtube.com/watch?v=CM8e-RCizDg But he appears to be busy with Rosenkreuzstilette TAS, so don't expect anything soon.

Posted: 6/22/2011 4:58 AM

Post subject: WARNING: INCOMING SCREENSHOTS

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

You can zoom out until the screenshots are two per row, if you like.

All these were taken from the Youtube upload (using Avidemux). Some of them are sized differently than the others. By the way, I think the screenshot used for publication should be in JPG form. But it should be taken from the emulator itself, or from a lossless dump. Dada, what's the status of your encode?

Posted: 6/22/2011 12:30 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

trytoguess wrote:

why did Game Saisoku decided to use that Pokémon?

I know it was already answered, but, just to point out, "Game Saisoku" (or "Saisoku Game", whatever), is the blog whose listing contains cstrakm's TAS. Maybe it wasn't clear from the submission text. If you want to be technical, the blog's full name is "ゲーム最速クリア動画のまとめ [ge-mu saisoku kuria douga no matome]", or "collection of movies of video game fastest completion". It does not specifically say TAS or not TAS, though the subtitle mentions that it is mainly TAS movies.

petrie911 wrote:

Is it faster to use the female character over the male one?

No clue. But I used the female because everything's better with girls. Anyway, what I know is that May's Treecko (as the male character) has a lower defense than Brendan's Treecko (again, nature is different depending on trainer data). There is some different text whenever the two meet up. The male character starts in the left house instead of the right. Other than that, I don't know.

Oscillator wrote:

Movie desynced for me at Slateport museum, at the time the description says you're doing luck manipulation for a King's rock.. the text is being delayed for like a minute, then the character just keeps walking into walls (which I don't think is part of the luck manipulation). What could be the cause of this? I use VBA v23.5 with realtime clock and flash 128k. Edit: Tried v22; desync occurs in the exact same way.

Don't know for sure; try a different ROM.

Sir VG wrote:

I looked over the submission text over and over...but I've yet to find what kind of debt you owe Abra?

:D I owe Abra nothing, except maybe an apology for leaving it in the PC.

Posted: 6/20/2011 9:07 PM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

Sir VG wrote:

3. Why the boots? Those things are awful and worthless.

First, sparky is doing 100%. Second, even if not, it is required to do the Split Mushroom midboss skip trick without Nova Strike. http://tasvideos.org/forum/viewtopic.php?p=272233#272233

Posted: 6/20/2011 5:30 AM
(Edited: 6/20/2011 5:30 AM)

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

henke37 wrote:

Who is "A-chan"?

A-chan is the speedrunner's character. Also, the Japanese version TAS's character as well. Unfortunately I got them mixed up that one time. Thanks for telling me.

was0x wrote:

and I made fixed Japanese note. http://pastebin.com/HF90zGGJ

Thank you, was0x. I linked to it now.

Brandon wrote:

Also, FractalFusion, I highly recommend the use of captions for your encodes in the future, as mine are very well received, and they'd probably shorten your submission text significantly. It's a lot easier to read a play by play than an accompanying essay.

I was thinking that too, but I like long submission texts. :( I can try to cook up some captions, but I don't know whether I'll end up in a caption addiction or not, something that is very hard to get out of. Oh, and regarding screenshots: All the dialog-only ones are silly and are only there just to get a laugh from how bad they are. In serious, I would prefer screenshots like those used in battles, although there isn't that much to go with since disabling battle animations is what makes it fast. For publication screenshot I'd go with either the dying Magneton or the dying Skarmory one, not only because the opponent is dying, but also because of the "not very effective" as well.

Posted: 6/20/2011 1:25 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

N. Harmonik wrote:

And this is faster by the Japanese TAS of Sapphire by how much?

Unfortunately, due to different text, I was not willing to figure out the exact differences. The TAS of the J version of Sapphire is around 1:27.

Posted: 6/14/2011 5:20 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

Pictian, I do not know if you understand the purpose of this site or not. Please read the welcome page if you haven't already. We aim to complete games with an extremely high level of precision of input, which constitutes a form of art. The movie you have presented appears to be more like a casual play without death. Please also read sgrunt's post above, as the submission breaks some rules.

Posted: 6/13/2011 6:01 PM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

Aqfaq, I think that's enough for now. Let's spread the fun steadily over the year shall we? Something that April Fools submitters cannot have the right mind to do.

Posted: 6/12/2011 6:11 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

Good run, hellagels. I like when Zero runs up the wall as fast as he dashes forward. Also, I never knew that resetting at the beginning allows you to skip the very first two screens. Yes vote.

Posted: 6/10/2011 12:55 AM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

The movie requires Snes9x v1.43, not Snes9x v1.51 or Snes9x v1.52 . Or if you mean that it desyncs, the ROM has to be the correct one. See submission text for details.

Posted: 6/9/2011 10:43 PM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

I think Metal Slug 3 is too long with all the autoscrollers and stuff. If a Metal Slug game gets a star, I'd say it is Metal Slug X. You suggested it earlier too.

Posted: 6/9/2011 10:25 PM

Quote

FractalFusion

Editor, Experienced Forum User, Published Author, Skilled player (1941)

Joined: 6/15/2005
Posts: 3247

I haven't seen the previous run, but since this one is a big improvement, I will vote yes. I had been worried about the run using the warp barrels throughout the first two worlds but the rest of the run doesn't use any, and that's what makes it good. The stage where you use the spider is my favorite part of the run.