Submission Text Full Submission Page
TAS by DyLaX.
I make "Game Over" in Level 3 Glitch.
  • fceux v2.1.6
  • usa
  • warp
  • level 3 glitch
  • one death by player 2
  • Fastest possible?
  • Battletoads is a game in which we overcome the obstacles, in order to save kidnapped princess by the Dark Queen.
  • The hardest part was calling the "Game Over", but with the right parameters, it worked.
  • TAS has been corrected for the 1 and 3 levels by DyLaX.
Movie speeds almost completing the game by ~10 minutes. The last movie is this movie: http://tasvideos.org/1920M.html

feos' comment

input after suicide creates tons of broken objects that get read from $0B in RAM and then the pointer jumps over $B addresses each time (object descriptor line length), thinking he reads from ROM. giving him needed values is pretty much possible. especially creating objects by direct input (that is stored in $15 and $16). but some objects need special attribute values, that aren't read from descriptor lines, but loaded other ways. finding the ways to load the needed values, and figuring out the very values and addresses - will take time until this submission is obsoleted

Nach: I've taken a few days to mull over the situation here, and was mindful of everyone's comments in the forum, as well as on IRC.
The previous run found a shortcut in level 3 which essentially leads to a buffer overflow. Instead of abusing that glitch to its fullest, it simply took the data initially written there and skipped a single level. This run now fully abuses that glitch by writing new data and using it to skip 14 levels straight to the final boss fight room.
Now entertainment wise, I vastly prefer the previous run, and even find that run star worthy. However since the same initial glitch is used, even if not fully abused, the precedence in this situation is to have the faster run obsolete the other, despite a drop in entertainment.
In general, based on what we now know about Battletoads, I find that this game deserves 3 main branches.
  1. Glitched (this branch) - fastest completion, even if not very entertaining, skips as much as possible.
  2. 1p warpless - so you can see how quick the game can be beaten without being able to fully beat up the enemies, and make use of team attacks, and without skipping anything.
  3. 2p warpless - which in theory is the fastest way to beat the game without massive glitches, and significant skips, showing off the full game.
Of course another branch for an absolutely crazy playaround would also be acceptable.
I've found this run to live up to our standards, and the feedback overall was very good. Based on the precedence, I am accepting this run to obsolete the previous.

Nach: Since the previous movie is terrific, and many have asked for it, to give it the recognition it deserves, I have starred it, as short as that may last. This run should not be published with a star. A moon would be fine.

Guga: Processing...

PJ
He/Him
Experienced Forum User
Joined: 2/1/2011
Posts: 178
Location: Western NY
hegyak wrote:
That was weird. You warp to level 3, die, then go to the boss and get a game over, then win? Sure. Why not.
When you're dealing with Battletoads, that is actually a pretty tame result. The emulator didn't even break!
Experienced Forum User, Player (12)
Joined: 6/17/2006
Posts: 471
Now THAT is a speedrun! Yes vote! I believe this is another case where a new branch should be created because of memory corruption. See [1945] SNES Super Mario World "game end glitch" by Masterjun in 02:36.40 for reference. By the way, is there a way to kill Zitz on the way without losing time? It would make the "final battle" shorter after the end of input. @Radiant: The jump to stage 3 is achieved through a normal warp.
Ambassador, Experienced Forum User, Experienced player (585)
Joined: 7/17/2004
Posts: 985
Location: The FLOATING CASTLE
Awesome! Would it be possible to speed up the unseen fight by dying a couple more times in turbo tunnel?
Experienced Forum User, Skilled player (1638)
Joined: 11/15/2004
Posts: 2200
Location: Killjoy
Arcade TASer of 2010PCE TASer of 2010Arcade TAS of 2010Lucky TAS of 2009Lucky TAS of 2007
SmashManiac wrote:
Now THAT is a speedrun! Yes vote! I believe this is another case where a new branch should be created because of memory corruption. See [1945] SNES Super Mario World "game end glitch" by Masterjun in 02:36.40 for reference. By the way, is there a way to kill Zitz on the way without losing time? It would make the "final battle" shorter after the end of input. @Radiant: The jump to stage 3 is achieved through a normal warp.
This is more like the SML2 setup - the warpless run does the same glitch, to only skip a single level. This one does the same glitch, but more effectively, skipping the entire game. This is unfortunate, because the warped run is pretty awesome.
Sage advice from a friend of Jim: So put your tinfoil hat back in the closet, open your eyes to the truth, and realize that the government is in fact causing austismal cancer with it's 9/11 fluoride vaccinations of your water supply.
Experienced Forum User
Joined: 5/27/2012
Posts: 70
Location: Wisconsin
WOW i knew this game had its fair (or Unfair) glitches but wow yes vote but only if it stays as a glitched run i kinda do like seeing this game get completed with a tas but not to a point where its warp, warp glitch, end glitch, i'm just saying here thats all Good job at exploiting the new end of level glitch
AnS
Emulator Coder, Experienced Forum User, Experienced player (697)
Joined: 2/23/2006
Posts: 682
This one-glitch demonstration should not obsolete existing warped run, since the entertainment offered here fades away after initial surprise (it's not like you'd want to watch this movie twice, unlike the previous Battletoads submission). Sure, technically it's the same glitch, but the result is different enough, that's all that should matter.
Experienced Forum User, Experienced player, Site Admin (974)
Joined: 4/17/2010
Posts: 10599
Location: RU
NES TAS of 2011
BimmyLee wrote:
discuss it in the workbench imo
Okay! First of all, I have to say more deeply, what happens here and in my and MESHUGGAH's run. Dying on a jet with race checkpoint not activated breaks the object pointer. It starts reading from RAM, jumping over 11 addresses each 4 frames, thinking he creates objects as when the level normally starts. Then it stops jumping. When it is at $16, it reads from $16 the temp input value. We created level end in previous run by pressing all buttons but A. That's all, the level ended. I see no memory corruption here, just lucky bug. But you can go deeper. You can stick the pointer to addresses you want, spawning objects, whose IDs were read from some irrelevant place, you can add them attributes that affect the game logics (some cause huge glitches, some don't). Insane debugging is involved here, and this is the essence of memory corruption. Here, this memory corruption was used to reach the authour's GOAL - break to the game end. It's interesting, that the game has other ways to reach that glitchy goal. It can break the object counter (slot pointer written to X or Y registers), reading irrelevant values as objects and attributes. This can write to some addresses, that may appear critical and cause huge glitches, but we didn't have time to investigate it. Here it is in Russian. Note, that this has nothing to do with the jet suicide. It needs only heavy lag. We also have Reset button. It is able to screw the objects in some way too, and it can definitely end the game, as was seen here. Above we have 3 ways to reach the GOAL - glitched game end. All of them, as long as they succeed to do so, and obsolete each other by speed, must be considered the Glitched branch. They may even be combined to make the thing even faster.
Now about the possible decision around this submission. People say, that if the same glitch is used in Default and Glitched branches, they must obsolete one another. And the glitch this way becomes exclusive for 1 branch, and abandoned for others. Okay! We obsolete the Default branch by Glitched here, abandoning the jet glitch. Then I edit out that glitch from my Warps (Default branch) run and it gets restored in it's rights as NEW Warps (slower than OLD). Then we suddenly succeed to use the reset glitch in stage 1 and end the game right there. The jet glitch becomes UNABANDONED again, I resubmit the OLD Warps run and it obsoletes the NEW Warps (that didn't use the jet glitch), because the OLD one was faster. Let's recall Rockman. It uses the bank switching glitch to end some levels. It didn't become the Glitched run, because it still BEATS the game. But what if it used the bank switching in the first level and break to the ending? Would it obsolete the Default branch? Why do we need that cross-goal obsoletion at all? To reduce the amount of branches? From the above, we will still get both Default and Glitched, just abandonong/unabandoning some arbitrary METHOD in one or another of them several times with no productive point.
Conclusion. METHODS can obsolete each other, GOALS can only if they are comparable. 2P Warpless can obsolete 1P Warpless, because it beats the game just the same logical way, but better in entertainmebt/speed. 2P Warps obsoleted 1P Warps because it was the fastest method to BEAT the very game. And if we use deadly glitches to skip the game to the ending - it is not PLAYING the game, it's BREAKING the game totaly. We must compare not by arbitrary tricks, but by the Superplay type involved, am I wrong? Memory corruption is pretty outstanding thing and must not be mixed with playing the game logically.
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. If TASing is meta-play, TASVideos Movie Rules are meta-meta-play!
Experienced Forum User, Player (12)
Joined: 6/17/2006
Posts: 471
Hmm... it indeed uses the same glitch as DarkKobold mentioned, and many publications have been obsoleted because of this. On the other end, it is also true that feos and MESHUGGAH did not corrupt the memory in their run, at least not in the sense of manipulating it. Quoting from the SMW discussion:
adelikat wrote:
I'm fine with this being a new category as long as we can have a clear line as to what constitutes this category vs the any%. The difference here is ram corruption from what the 'experts' have told me. This is probably going to be clear cut enough, though I suspect there will be at least one TAS that challenges this potential gray line.
I believe we have our challenger...
Experienced Forum User
Joined: 6/8/2005
Posts: 236
Location: Madison, Wisconsin
Whaaaaa. That is an impressive way to break the game! But I do agree, the run's real value is in the shock value, and after watching it once loses it. Still definitely worth publishing if only to have people see it once and get the same surprise!
Editor, Experienced Forum User, Publisher, Skilled player (1228)
Joined: 10/12/2011
Posts: 6014
Location: The land down under.
PSX TASer of 2016
I'm going to look for parts of my brain now, thanks a lot. YES VOTE
WebNations/Sabih wrote:
+fsvgm777 never censoring anything.
Disables Comments and Ratings for the YouTube account. These colours are pretty neato, and also these.
Experienced Forum User, Player (230)
Joined: 7/30/2006
Posts: 207
Location: Alefgard, USA
TheAxeMan wrote:
Awesome! Would it be possible to speed up the unseen fight by dying a couple more times in turbo tunnel?
I would like to ask that question as well. It seemed like you had to force the game over by killing him a couple of times. It seems he could die in level 3 on the way to the hovercraft? racers, and then just die once in the "final" fight.
Active player, Experienced Forum User (316)
Joined: 9/14/2011
Posts: 349
Sega TAS of 2014
Commit suicide, get game over screen, telepors to nowhere, fight with something invisible, and win the game instantly. Make sense to me. I give you Yes vote.
Experienced Forum User, Skilled player (1638)
Joined: 11/15/2004
Posts: 2200
Location: Killjoy
Arcade TASer of 2010PCE TASer of 2010Arcade TAS of 2010Lucky TAS of 2009Lucky TAS of 2007
feos wrote:
Okay! We obsolete the Default branch by Glitched here, abandoning the jet glitch. Then I edit out that glitch from my Warps (Default branch) run and it gets restored in it's rights as NEW Warps (slower than OLD). Then we suddenly succeed to use the reset glitch in stage 1 and end the game right there. The jet glitch becomes UNABANDONED again, I resubmit the OLD Warps run and it obsoletes the NEW Warps (that didn't use the jet glitch), because the OLD one was faster.
This logic fails, because we now we know that the jet glitch can be used to end the game, so putting it back in the warped run would be foolish - you'd be using a known glitch sub-optimally. SMW is a perfect example - if using the same glitch could be used to warp 3 levels ahead instead of to the credits, it still wouldn't be allowed for the any% run. So, that argument fails. And having a slower run obsolete a faster one is not a new thing here. There is precedent for that. Having two runs use the same glitch, where one uses the glitch in a more optimal manner, does not have a precedent here.
Sage advice from a friend of Jim: So put your tinfoil hat back in the closet, open your eyes to the truth, and realize that the government is in fact causing austismal cancer with it's 9/11 fluoride vaccinations of your water supply.
Experienced Forum User, Experienced player, Site Admin (974)
Joined: 4/17/2010
Posts: 10599
Location: RU
NES TAS of 2011
Is it abandoned for Warpless run as well then? And I still need a clear definition on what is considered the "SAME GLITCH". EDIT: Another situation. I record a Warpless run without jet glitch. But I use object counter glitch to generate myself a stick (there's NO proof right now, that object counter glitch may end the game or someting). Then I succeed to use that counter glitch to end the game for Glitched branch. Will it obsolete Warpless as well because ot the glitch suboptimality? How can I know (doing the first run) if the glitch can or can not end the game, that info would be pretty sudden, and may never appear at all. EDIT 2: To clarify about the "editing out" thing again. The game has 4 RNG addresses, 1 of them is primary. They have insanely complex calculation of the resulting values, and they tick when there is free processor time. You can take away some of that free time by random input and moves, but IN NO WAY you can manipulate the exact values you need directly. This means that editing the game in the 3rd level will favour me with resyncing the whole freaking game over. And together with no gameplay finds, standing behind that editing work, I insist that this cross-goal obsoletion would have NO PRODUCTIVE SENSE. It can encourage no one to do the better and improve the movie gameplay-wise. It just demands dumb editing, and has much long-term effects of ambiguity and weirdness.
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. If TASing is meta-play, TASVideos Movie Rules are meta-meta-play!
BigBoct
He/Him
Editor, Experienced Forum User
Joined: 8/9/2007
Posts: 1681
Location: Tiffin/Republic, OH
I don't see that we necessarily have to obsolete the published warped run because it uses the same glitch this one does. I would argue thusly: Publish this as a new category, and stipulate that any new 2-player run must either be warpless, or use the glitch to its' fullest effect as this run did. EDIT: Derp-a-derp, I forgot that we already HAD a 2p warpless. Color me confused, then.
Previous Name: boct1584
Mitjitsu
He/Him
Banned User, Experienced player (503)
Joined: 4/24/2006
Posts: 2997
TAS of 2007N64 TAS of 2007
WTF just happened?
Post subject: Explanations and Informations.
Experienced Forum User
Joined: 9/4/2012
Posts: 9
Please do not forget the main thread! We are doing everything, to every person on earth could understand what happened here (step by step, as you need). Most important in this movie "is not" a assignment to the category, is only what really happened in the game code between the 3493 and 3494 frame. If we have to explain the assembler code, we'll do it. At the moment, we know that: * "Created player" in 12th slot, cannot be manipulated (modified) too much (in the sense that "it" cannot be killed, for example - is best, when "it" is not touched by something or for glitch). "The parameters must be stable". * In addition to what Feos wrote, we know that the "Introduction" (is what's going on before you press the start button), has an impact on the "Game End" event (It can reduce the value and the game starts normally). Now, we are working on a faster end of the game (not just using the level 3 glitch).
Post subject: Re: Explanations and Informations.
Experienced Forum User, Skilled player (1589)
Joined: 9/17/2009
Posts: 4883
Location: ̶C̶a̶n̶a̶d̶a̶ "Kanatah"
GBA TASer of 2010
TheZlomuS wrote:
Now, we are working on a faster end of the game (not just using the level 3 glitch).
Nice to see this game get broken further.
Post subject: Re: Explanations and Informations.
Experienced Forum User
Joined: 9/4/2012
Posts: 9
jlun2 wrote:
Nice to see this game get broken further.
For me, this is not nice, I've seen the... such much...
Experienced Forum User, Experienced player, Site Admin (974)
Joined: 4/17/2010
Posts: 10599
Location: RU
NES TAS of 2011
Am I right that FASTEST NON-GLITCHED branch was just terminated forever?
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. If TASing is meta-play, TASVideos Movie Rules are meta-meta-play!
Emulator Coder, Experienced Forum User
Joined: 3/9/2004
Posts: 4588
Location: In his lab studying psychology to find new ways to torture TASers and forumers
feos wrote:
Am I right that FASTEST NON-GLITCHED branch was just terminated forever?
Not necessarily. If you can find a way to make a fastest non glitch branch which doesn't seem redundant with a fastest non warped branch, then we'll add it. I'm not sure you'll be able to do so though.
Warning: Opinions expressed by Nach or others in this post do not necessarily reflect the views, opinions, or position of Nach himself on the matter(s) being discussed therein.
Post subject: Movie published
TASVideoAgent
They/Them
Experienced Forum User, Moderator
Joined: 8/3/2004
Posts: 12381
Location: 127.0.0.1
This movie has been published. The posts before this message apply to the submission, and posts after this message apply to the published movie. ---- [2100] NES Battletoads "game end glitch" by TheZlomuS & DyLaX in 01:02.68
Experienced Forum User, Experienced player, Site Admin (974)
Joined: 4/17/2010
Posts: 10599
Location: RU
NES TAS of 2011
This is all legit, but I think I need to repeat my question. If I use object counter glitch to generate myself a stick to kill enemies faster and publish the full run, what will you guys judge it like, if the counter glitch then also appears helpful in a glitched branch? It's also weird that you're unable to define the term GLITCH you now use to set the branch, instead of GOAL. You are not? Then point me into the actual RULES (not submission texts) that talk clearly about the new GLITCH-not-GOAL politics.
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. If TASing is meta-play, TASVideos Movie Rules are meta-meta-play!
Emulator Coder, Experienced Forum User
Joined: 3/9/2004
Posts: 4588
Location: In his lab studying psychology to find new ways to torture TASers and forumers
You can use any glitch in a "non-glitched" run that can't be abused to force a very quick game finale.
Warning: Opinions expressed by Nach or others in this post do not necessarily reflect the views, opinions, or position of Nach himself on the matter(s) being discussed therein.
Experienced Forum User, Experienced player, Site Admin (974)
Joined: 4/17/2010
Posts: 10599
Location: RU
NES TAS of 2011
Nach wrote:
that can't be abused
I don't read the future. Anything may appear to be able to force a very quick game finale. And we NEED some sane rule about that, because this way GOALS don't matter anymore. And GOALS are about gameplay and will to show off a certain aspect of superplay. You guys seem to want to mess up these clear definitions. Player has no way to know how his gameplay decisions will treat him in the future. It is also strange how you prefer "negative" precedents to "positive" ones. You generally don't fear to lose the total level of awesomeness in the runs, but sometimes you do. The "no zips" decision for some reason wasn't meant as a precedent. Precedent for what? For saying "All future runs for that category shall be done without trick A, and will obsolete the runs using tricks A and B, that are still published for the viewers' pleasure". A 100 minute GLITCHED run getting obsoleted by a 1 minute GLITCHED run is perfect. A SRAM abusing run being obsoleted by a GLITCHED run is still fine since it was a side branch anyway. But players DO CARE about obsoleting a gameplay-based run with a non-gameplay based run. PS: I do not care of "my run" obsoleted. The only 2 things I care about (you know it very well) are creativity and my ability to put it into a new superb run (be it warpless or whatever!) That's why I keep questioning - to make sure my work won't be eaten in vain.
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. If TASing is meta-play, TASVideos Movie Rules are meta-meta-play!