Submission Text Full Submission Page
Super Mario Advance 2 is the GBA port of Super Mario World. The gameplay and levels are almost the same, with some minor differences. This run beats the previous one by 500 frames.

Game objectives

  • Aims for fastest time
  • Takes damage to save time
  • Abuses programming errors

Differences between the SNES and GBA versions

  • Items dropped in the GBA version lose x speed
  • Jumping off an item makes Mario lose his x speed
  • Major Yoshi glitches are not possible
  • There is only one button for running and flying

Level by level comments

Yi4: 15

Optimized the shell jump.

Iggy: 33

Minor optimizations and a new method of jumping off the p-switch saved some time.

Dp1: 35

Used a faster method of getting the feather.

Ds1: 72

Got the red shell faster, avoided a fish, and improved the end.

Dsh: 224

Clipped through the stairs and flew in the second room.

Sw1: 306

Used ISM’s smashola-like glitch by wallcatching the cave wall.

Sw2: 333

Keeping flying status when getting the key saves 27 frames.

Sw3: 346

Optimized the ending by swimming down with the key.

Sw4: 377

Started flying earlier, optimized the end, and avoided having to slow down to get the shell.

Fd: 403

It is faster to take damage in room 2 than to keep the cape (pointed out by Mister).

Bowser: 500

Using the new Bowser strategy and hitting the mecha-koopas lower saves 97 frames.

Suggested screenshots

17912, 21943

Addresses

  • 03003FA9 - x speed
  • 03003FAD - y speed
  • 03003FD0 - x position
  • 03003FD4 - y position
Special Thanks: ISM and Mister, for discovering and pointing out some tricks in this run.

feos: Many places were mentioned where this run loses time and doesn't use the tricks that are known for years. Rejecting due to sub-optimality. As for the region, lag will likely count when choosing the version, otherwise only solid gameplay will matter.


TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 15564
Location: 127.0.0.1
Editor
Joined: 11/3/2013
Posts: 506
Is this becoming a frame war?
Spikestuff
They/Them
Editor, Publisher, Expert player (2633)
Joined: 10/12/2011
Posts: 6437
Location: The land down under.
thatguy wrote:
Is this becoming a frame war?
Same runner, about 4 seconds saved. NOPE. Still questioning the Bowser strat and why the SNES way wasn't used
WebNations/Sabih wrote:
+fsvgm777 never censoring anything.
Disables Comments and Ratings for the YouTube account. Something better for yourself and also others.
Joined: 6/4/2009
Posts: 893
Spikestuff wrote:
thatguy wrote:
Is this becoming a frame war?
Same runner, about 4 seconds saved. NOPE.
maybe he is schizophrenic ?
Post subject: WHAT IS A TAS?
Spikestuff
They/Them
Editor, Publisher, Expert player (2633)
Joined: 10/12/2011
Posts: 6437
Location: The land down under.
Link to video Got told in IRC due to 30sec. lag in Jap but I digress like always I mentioned this and I'm mentioning it again. Giving it a no. That is all. Because it's still improvable even if Jap is slower U/EU can be improved more.
WebNations/Sabih wrote:
+fsvgm777 never censoring anything.
Disables Comments and Ratings for the YouTube account. Something better for yourself and also others.
Experienced player (588)
Joined: 2/5/2011
Posts: 1417
Location: France
No too; I did found a frame improvement in overworld also.
Current: Rayman 3 maybe? idk xD Paused: N64 Rayman 2 (with Funnyhair) GBA SMA 4 : E Reader (With TehSeven) TASVideos is like a quicksand, you get in, but you cannot quit the sand
Editor, Skilled player (1535)
Joined: 7/9/2010
Posts: 1319
Lag or text doesn't count for slower/faster version to use, only different glitches count.
Favorite animal: STOCK Gt(ROSA)26Sortm1.1(rtTA,EGFP)Nagy Grm7Tg(SMN2)89Ahmb Smn1tm1Msd Tg(SMN2*delta7)4299Ahmb Tg(tetO-SMN2,-luc)#aAhmb/J YouTube Twitch
Former player
Joined: 6/30/2010
Posts: 1107
Location: Zurich, Switzerland
TASeditor wrote:
Lag or text doesn't count for slower/faster version to use, only different glitches count.
The rules don't mention lag specifically, besides text it just says: "differences in title screen, cutscenes, and menus (unless menus are the game's main control interface). " I guess that this does include lag, but if the TAS can improved by gameplay anyway, why not doing it on the version that lags less?
Current project: Gex 3 any% Paused: Gex 64 any% There are no N64 emulators. Just SM64 emulators with hacky support for all the other games.
Spikestuff
They/Them
Editor, Publisher, Expert player (2633)
Joined: 10/12/2011
Posts: 6437
Location: The land down under.
andypanther wrote:
TASeditor wrote:
Lag or text doesn't count for slower/faster version to use, only different glitches count.
The rules don't mention lag specifically, besides text it just says: "differences in title screen, cutscenes, and menus (unless menus are the game's main control interface). " I guess that this does include lag, but if the TAS can improved by gameplay anyway, why not doing it on the version that lags less?
Crash Bandicoot is an example that we can use. Pirohiko uses Japanese to avoid death abusing to get Box counts (Killing self still generates box counts). but Marcusboy95 (NOT MUKKI) uses the American version and has to death abuse in order to save time to skip box counts. The math apparently is NTSC-U is the faster one due to it completely skipping the box count scene and going to map (This is calculation of each of the fastest deaths and every time there is a box count scene compared to box count scenes).
WebNations/Sabih wrote:
+fsvgm777 never censoring anything.
Disables Comments and Ratings for the YouTube account. Something better for yourself and also others.
Player (173)
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
I'm not going to interfere in competition, but I have to mention obvious (possible) improvements which have been well-established in the SNES runs, before it enters a rather meaningless frame war. If anyone is working on an improvement, then I'm very sorry to make this post; the following ideas are all well-known and provide nothing new. Also, because I watched your run via the nico encode, I can't check the exact input, and the following could contain some errors. Thank you for your understanding.
  1. C1. My point might not have understood. The method in the submitted run looks better than before, but spinjumping with small mario is still useful in this situation so as to control your height. That's demonstrated in this movie. You'll possibly go further before catching the fence.
  2. C1, Iggy's room. I'm wondering if it is useful to keep pressing the jump key at the very beginning of the Iggy fight. It makes your falling speed mild so you lose less speed on the ground. I'm not sure if you've already checked it. It's mentioned just in case.
  3. DS1. Are you sure if there's no loss when entering the first pipe? I mean, there could be loss of a frame if you are just keeping 1/1. Details are explained in this movie. Roughly, you have only to stop 1/1 just before the first pipe entry.
  4. DSH. In fact, we can slide through the stairs as in the SNES run. That's performed in the same movie in a highly optimized form. Just copying inputs from that run will not work. You must try hard to pull it off, or learn more about qualification. It is possible, being verified by ISM. Cf. Post #300412.
  5. DSH. In the second room, you can turn around at grabbing the p-switch, just as done in the recent runs, e.g. 1944M.
  6. Bowser. Did you really make sure whether being caped would be faster or not? It's not so easy to compare two ways without examining.
  7. Though I didn't mention any other standard optimization tricks such as cornerboost and cornerclip, there should be a lot of points at which they are useful to save time. It's rather within a frame war area in my opinion.
Wow, what a huge list... There are a lot to work on. Good luck, and thanks again. Oh I was about to forget to vote. No vote from me.
Retired because of that deletion event. Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Post subject: Re: WHAT IS A TAS?
Active player (264)
Joined: 4/15/2010
Posts: 197
Location: England
No due to improvements some of which have been known for many years. I comment only to mention as to why using USA version is a reason to reject when J or EU version is faster. In SNES SMW, in the E version, Mario runs/flies faster and Door #5 in BACK DOOR can be skipped without waiting for any spikes at all. I have a bad encode of the trick here: SMW - Door 5 break (I originally intended as April Fool's joke). Plus, this version has much less lag and might be faster for the TAS by several seconds. The J version on the other hand is faster due to text on save dialog and switch palaces so either of these versions would be better. Our SNES SMW runs aren't rejected for version difference so I don't know why it should apply here. Maybe I just misunderstand, I don't know.
Retired smw-96, smw any%
Player (122)
Joined: 1/22/2012
Posts: 45
bahamete wrote:
Post.
Well, if I redo this run, should I do it on the Japanese version?
Mister wrote:
Post.
Oh, I didn't know about some of that stuff. Thanks! By the way, do you know the addresses for x speed and position in the Japanese version? I'm testing the C1 idea right now.. I also feel like ds1 and sw1 are un-optimized. Also, for the ds1 improvement, does the same apply for getting the blue Yoshi in sw2? I tried to optimize it by stopping 1/1.
Spikestuff
They/Them
Editor, Publisher, Expert player (2633)
Joined: 10/12/2011
Posts: 6437
Location: The land down under.
gbreeze wrote:
bahamete wrote:
Post.
Well, if I redo this run, should I do it on the Japanese version?
Spikestuff wrote:
Got told in IRC due to 30sec. lag in Jap but I digress like always
WebNations/Sabih wrote:
+fsvgm777 never censoring anything.
Disables Comments and Ratings for the YouTube account. Something better for yourself and also others.
Site Admin, Skilled player (1251)
Joined: 4/17/2010
Posts: 11475
Location: Lake Char­gogg­a­gogg­man­chaugg­a­gogg­chau­bun­a­gung­a­maugg
gbreeze, what are your plans?
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.
Player (122)
Joined: 1/22/2012
Posts: 45
feos wrote:
gbreeze, what are your plans?
I plan on improving it in the future, but I don't have the addresses for the Japanese version. It'll also take a while for me to finish it, so don't expect a new run anytime soon from me. Eventually I want to submit a better run, but not nearly soon enough to cancel this run. Edit: Just found the x speed address. The run still won't be done in quite a long time though.
Active player (264)
Joined: 4/15/2010
Posts: 197
Location: England
gbreeze wrote:
bahamete wrote:
Post.
Well, if I redo this run, should I do it on the Japanese version?
If that question was directed at me, maybe you should actually read my post. I don't think this movie should be rejected based on version differences, only because of known improvements. By the way, maziari is working on improvement to this run, he has saved 13 frames up to dp1 (I hope he won't mind me saying) - maybe you could contact him and work together rather than 2 separate runs being created. I look forward to an optimal one. :)
Retired smw-96, smw any%
Site Admin, Skilled player (1251)
Joined: 4/17/2010
Posts: 11475
Location: Lake Char­gogg­a­gogg­man­chaugg­a­gogg­chau­bun­a­gung­a­maugg
If it gets TASing with public WIPs and posts in a thread, the chance of not losing time at all would become way higher. We won't be seeing various people coming and telling how much time was lost.
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.
Player (122)
Joined: 1/22/2012
Posts: 45
bahamete wrote:
gbreeze wrote:
bahamete wrote:
Post.
Well, if I redo this run, should I do it on the Japanese version?
If that question was directed at me, maybe you should actually read my post. I don't think this movie should be rejected based on version differences, only because of known improvements. By the way, maziari is working on improvement to this run, he has saved 13 frames up to dp1 (I hope he won't mind me saying) - maybe you could contact him and work together rather than 2 separate runs being created. I look forward to an optimal one. :)
Oh, okay! Well, I started a run on the Japanese version, and 59 frames of lag are saved from using that version. Oh alright! I actually found his twitter account; it seems he saved 17 frames (11 on yi3 and yi4). I was able to save over 6 frames on C1, so maybe we could work together. I'm currently trying to figure out how fadeout lag works.. it seems like it works with the frame (and maybe the x position) the goal post is touched..
Buddybenj
He/Him
Joined: 1/12/2013
Posts: 166
Location: USA
gbreeze wrote:
bahamete wrote:
gbreeze wrote:
bahamete wrote:
Post.
Well, if I redo this run, should I do it on the Japanese version?
If that question was directed at me, maybe you should actually read my post. I don't think this movie should be rejected based on version differences, only because of known improvements. By the way, maziari is working on improvement to this run, he has saved 13 frames up to dp1 (I hope he won't mind me saying) - maybe you could contact him and work together rather than 2 separate runs being created. I look forward to an optimal one. :)
Oh, okay! Well, I started a run on the Japanese version, and 59 frames of lag are saved from using that version. Oh alright! I actually found his twitter account; it seems he saved 17 frames (11 on yi3 and yi4). I was able to save over 6 frames on C1, so maybe we could work together. I'm currently trying to figure out how fadeout lag works.. it seems like it works with the frame (and maybe the x position) the goal post is touched..
There is no reason you have to use the J version. You can use U or J (U is preferred however). Read this: http://tasvideos.org/MovieRules.html#UseTheCorrectVersion
Projects: Interested in TASing N64 Mario Golf. GBA Mario Tennis: Power Tour is on hold.
Spikestuff
They/Them
Editor, Publisher, Expert player (2633)
Joined: 10/12/2011
Posts: 6437
Location: The land down under.
gbreeze wrote:
Oh, okay! Well, I started a run on the Japanese version, and 59 frames of lag are saved from using that version.
Spikestuff wrote:
gbreeze wrote:
bahamete wrote:
Post.
Well, if I redo this run, should I do it on the Japanese version?
Spikestuff wrote:
Got told in IRC due to 30sec. lag in Jap but I digress like always
I'm sensing deja vu on what I already wrote
Buddybenj wrote:
There is no reason you have to use the J version. You can use U or J (U is preferred however). Read this: http://tasvideos.org/MovieRules.html#UseTheCorrectVersion
As for you Buddybenj read what I wrote HERE. On the same topic!
WebNations/Sabih wrote:
+fsvgm777 never censoring anything.
Disables Comments and Ratings for the YouTube account. Something better for yourself and also others.
TASVideosGrue
They/Them
Joined: 10/1/2008
Posts: 2785
Location: The dark corners of the TASVideos server
om, nom, nom
Player (122)
Joined: 1/22/2012
Posts: 45
I understand how this run is sub-optimal, but does it not still beat the current run that is published at this site? I may not understand the TASVideos rules completely.
Site Admin, Skilled player (1251)
Joined: 4/17/2010
Posts: 11475
Location: Lake Char­gogg­a­gogg­man­chaugg­a­gogg­chau­bun­a­gung­a­maugg
A movie must use known tricks. Otherwise: imagine someone obsoletes an old movie by a second, and we ignore timesavers it missed and accept it (due to being faster by raw numbers). Then the other week another 1-second improvement is submitted. We accept it too. And so on. One would be able to drawl really improving movies for years. To avoid such spam of small improvements, we require the movie to use all known tricks. Sometimes some of them are still missed, but it's another story when known improvements are considered not worth redoing - it is when a run is overall optimal, and of a very high quality. When it's overall suboptimal (too much known improvements), we won't publish it even it is an improvement. #4041: McBobX's Genesis Knuckles in Sonic the Hedgehog in 14:09.43
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.
Buddybenj
He/Him
Joined: 1/12/2013
Posts: 166
Location: USA
gbreeze wrote:
I understand how this run is sub-optimal, but does it not still beat the current run that is published at this site? I may not understand the TASVideos rules completely.
There is a TAS of this game on TASVideos already? I didn't see it. Could you link me to it?
Projects: Interested in TASing N64 Mario Golf. GBA Mario Tennis: Power Tour is on hold.
Spikestuff
They/Them
Editor, Publisher, Expert player (2633)
Joined: 10/12/2011
Posts: 6437
Location: The land down under.
Buddybenj wrote:
There is a TAS of this game on TASVideos already? I didn't see it. Could you link me to it?
Are you serious: [2229] GBA Super Mario Advance 2: Super Mario World "warps" by Masterjun in 10:07.98 Not that hard to find. Fairly old (2011) and completely sub-optimal but it was accepted never-the-less.
WebNations/Sabih wrote:
+fsvgm777 never censoring anything.
Disables Comments and Ratings for the YouTube account. Something better for yourself and also others.