Super Metroid 14% (aka. glitchless/legit Low%) TAS v1.1 by Saturn

Recorded on Snes9x 1.43-improvement9 (but works on future versions of the Snes9x 1.43 series just fine as well)

Sync Settings

  • WIP 1 Timing: ON
  • Left+Right/Up+Down: OFF
  • Volume Envelope Height Reading: OFF
  • Fake Mute desync workaround: ON
  • Sync samples with sound CPU: ON

Attributes of this run

  • 14% mode (beats the game with the least amount of items possible in legal conditions)
  • aims for fastest possible time, both, real- and ingame time wise
  • takes damage to save time
  • very high luck manipulation (especially to get optimal refills)
  • abuses glitches (but not the major out of room ones the 6% TAS does)
  • Genre: Platformer
  • Total Frames: 153428
Addendum by Bisqwit: "legal conditions" as defined here: Forum/Posts/184660 and Forum/Posts/184775 according to the author.
Suggested screenshot:

About the Game

Super Metroid is a classic platformer with many advanced techniques and a very fine, subpixel-based move engine. The story is about a bounty hunter called Samus who has the goal to explore a foreign planet called Zebes to find and destroy Mother Brain, the evil leader of the galaxy who is breeding a mighty species called Metroids to abuse them for her own benefits. During the journey you will find alot of items such as new ammo, energy tanks, special upgrades that speed you up or allow to pass certain areas you would normally not be able to, and making your character strong enough to be ready for the upcoming showdown against Mother Brain.
Despite being 14 years old by now, it is considered as one of the best games ever made, and is still played by many people on a regular basis. Due to the many different move techniques, a high variety through the new upgrades, and alot of route possibilities, this game is a very common target for many types of speedruns and playthroughs.

Moviemaking & Comments

As promised in the Super Metroid thread, here is my first version of a 14% TAS. I mainly recorded it between July 19, 2007 and January 1 2008 (v1.0), with a small improvement added during the big Metroid escape in July 2008 (v1.1). It's a huge improvement to the previous run, beating it by gigantic 30517 frames (or ~8,5 minutes) of realtime, and achieving a record-time completion of 0:27 minutes on the game clock (or 27:59 to be more exactly), which is definitely the limit for a glitchless 14% run. On top of that I managed to avoid the pause screen / menu entrance completely except to execute the unavoidable Gravity Jump, getting a optimal realtime as well while making the run more entertaining to watch due to less interruptions.
During the making of the run I managed to develop some really cool techs never done before, most noticeably the "WS-Lake horizontal bomb jump" without collecting the Missile pack, and also many new strategies at bosses (especially on Ridley and Draygon) to still beat them very quickly, even with the very limited conditions in ammo and equipment. Finally the run uses a fully optimized route to avoid backtracking as much as possible.
All in all, the quality of the run is very high, and it's only improvable by at most 15 seconds in the early parts of the run (the late ones are pretty much flawless with a few tiny exceptions worth of single frames only). Ammo and energy management is excellent over the entire run, and in fact, the refill drops are so good in this one that I doubt they could be reproduced without sacrificing realtime by entering the menu to manipulate them, which would in any case result in possible slowdowns to compensate that.
Seeing that the 14% category got kind of obsoleted by the 6% NBMB one (although they can't really be compared to each other), I don't expect this to be published, despite the huge and ground breaking improvement. My intention with this submission is to just contribute a unique run that sets a eternal ingame completion record to this site for the many people who want to see it, especially for those who have problems downloading large video files due to slow connections. It's definitely not less entertaining than the 6% NBMB TAS overall, so maybe it's worth to at least include this submission to the description of it without necessary publishing, like it was done a couple of times before. There is also a concept demo section for runs that aim for special goals like this one, so it would eventually fit there as well.

Ice VS Speed

In a 14% run, you must either take the Ice Beam, or Speed Booster. Both items, as different as they are, have the same purpose: They are required to get past the pre-Botwoon room and through the Zebetites in Tourian. I collected the Ice Beam in this run because my tests have confirmed that it is at least 75 ingame seconds (and even more realtime sec) faster in the end, despite of the slower running speed and the missing shinesparks. The reason is that Ice Beam does 50% more damage than the normal beam you have to use when picking up Speed Booster, and therefore speeds up boss battles with much HP alot. In fact, the Mother Brain fight alone already compensates the entire losses of the lacking Speed Booster, with additional big time saver at Ridley and the Metroids (which you would have to slowly PB otherwise).

Thanks

  • Terimakasih -- for his previous 0:35 run. He used a nearly same route, which helped alot in the decision to pick Ice over Speed Booster, giving me a pretty accurate estimation before even starting this run that the former will be a faster choice especially due to early Wrecked Ship.
  • Cpadolf -- for his new technique to escape the big Metroid even faster. It was the key to get the 0:27 time in this run.
  • Hero of the Day -- for his any% v1 in 0:25 run. It was of great help to me when comparing the Speed Booster gains and allowed me to estimate the rough completion time of this run already back at Ridley.
  • Moozooh -- for the Torizo Skip, and for motivating me to start this run parallelly to his own 14% one that would use the Speed Booster route. I actually hoped he would finish his one so that we would have TASes for both routes and a better comparison between them, but unfortunately it didn't want to happen.
And everybody else who has contributed something useful for SM-TASing.

Notes

For those interested, this run is also available on YouTube.
Enjoy!

mmbossman: I’m rejecting this run for the following reasons:
Primary reason - Quoting the Judge Guidelines: ‘’Keep the number of different branches per a game minimal.’’ This run is an attempt to resurrect a movie branch that is now obsolete, as there is already a true low% Super Metroid run published, and I see no need to resurrect a dead category. Considering that the previous 14% run was left unimproved for 4 years, while multiple other Super Metroid runs were improved (and added) several times, it shows that there is a lack of interest in the Super Metroid TAS community in this category, which can likely be correlated to a lower interest in this category from the TASvideos community in general. When compared to the three currently published "non-glitched" runs, this movie provides very little additional entertainment value for viewers who do not closely follow Super Metroid techniques and tricks.
Secondary reason - The goals of this run, along with the execution of those goals, are nebulous. The author has stated an obvious preference towards ‘’maximally optimized subpixel position’’ techniques in the latter half of his run, while rationalizing intentionally missed techniques in the first half of the run by saying they ‘’add up to the entertainment level of the run’’. I am not against entertainment tradeoffs at the expense of speed, however those exceptions should be made clear in the submission text, not explained away afterward (and especially not in the passive aggressive manner shown by the author). Additionally, very little attempt is made in the submission text to define what makes a run ‘’legit’’ in the authors mind, and the issue is further clouded by admitting to abusing other glitches. This lack of clear goals is the second reason this submission is rejected.

Saturn:
Dear mmbossman, the goals of this run are as clear as they can be: Completing the game with the minimal amount of items without skipping bosses and use major glitches. The explanation of "legal conditions" wasn't stated in the submission text because I expected it to be a obvious thing to anybody who knows at least a bit about Super Metroid (since those who don't, wouldn't even care to watch this run). Besides, it was explained in the discussion tread here, or here. You apparently still fail to realize that the 6% run you labeled as "true low%" skips 90% of the game and all bosses, so it can't be put into the classic low% category this run represents, that exists for almost any game.
Also, the clear majority of people have stated that they enjoyed the run and think the category is good. I'm not sure if ignoring all this people and just go by your own biased opinion is a good thing, especially for the TAS community as a whole, which because of that will miss a very unique and entertaining quality run that would only enrich this site due to the many people who would definitely be interested to see it. Too bad.

mmbossman: For my rebuttal to Saturn's complaints, see the following links: Link 1 Link 2 Link 3 Link 4

Saturn:
Since we're at it, mmbossman forgot Link 5, for a clear and detailed explanation to his "multiple goal problem". Didn't seem to arrive yet, but hey, it's not too late.

Bisqwit starts concentrating on a new spell.
Bisqwit holds up a diamond and whispers the magic words "herää".
Pieces of #2136: Saturn's SNES Super Metroid "glitchless low%" in 42:37.13 surface from various positions in the ground
and come together forming a living submission!
Eww! The pieces stink like grue excrement ― hardly surprising,
considering the circumstances in which they were deposited.
Bisqwit starts concentrating on a new spell.
Bisqwit empties a bag of fairy dust at #2136: Saturn's SNES Super Metroid "glitchless low%" in 42:37.13 and sings "wunderbaum, wunderbaum".
The foul smell of #2136: Saturn's SNES Super Metroid "glitchless low%" in 42:37.13 is neutralized!
Submission 2136 has been resurrected from the dead.

Bisqwit: Decision: Submission postponed until a time we can have arbitrary goal movies without making the site unmanageable.

adelikat: Decision: rejected until a time we can have arbitrary goal movies without making the site unmanageable

Nach: Since the above has now come to fruition, accepting.
feos: Processing.......


1 2
12 13
TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 15649
Location: 127.0.0.1
This topic is for the purpose of discussing #2136: Saturn's SNES Super Metroid "glitchless low%" in 42:37.13
Joined: 9/30/2007
Posts: 103
Oh my god, the arrogance. Setting that aside though, very entertaining run. Some notable parts were the bomb-jumping-over-the-sea, the Draygon fight and gate-trick to get the ice beam. The slow, slow Ridley and Mother Brain fights are, sadly, unavoidable, but well, there's a handy turbo-button for that. A quality run, no doubt. 9 dot something in entertainment, probably the same in tech.
Joined: 5/2/2006
Posts: 1020
Location: Boulder, CO
I really don't care for the long boss fights. Well made it looked like, but not so hot for entertainment. Also, as you pointed out, publishing a "kinda low %" is a little silly, but that's not my decision.
Has never colored a dinosaur.
Joined: 8/7/2006
Posts: 344
I like how you basically reviewed your own TAS in your comments. Anyway, I agree with Neophos on most parts, and was very entertained other than having to watch you spend so long plinking away at Mother Brain. Unavoidable but bleh. 9 for tech, 8.5 for entertainment.
Skilled player (1445)
Joined: 7/15/2007
Posts: 1468
Location: Sweden
Well it's good to see that you finally released the smv for this run, I didn't excpet you would but I really wanted you to (uncomparable quality versus youtube, and the possibility to watch in frameadvance and try out improvements). Anyway, this is a smashing run. The first part up to phantoon is pretty dull since it is sloppy compared to what we have now, but from there on out (prehaps with a few exceptions) it's a great ride with a lot of new cool tricks due to the restrictions you have. The optimization in the latter part of the run is also very very good, though I still refuse to think that there are not at least a few seconds of improvement, making the run in whole improvable by about 20-25 seconds without new discoveries. Entertainment was handled good in the waiting parts, and ridley was not nearly as dull as I thought that fight would be so overall it is a fantastic run. However I don't think it should be published (and it won't either), since the category is obsolete. But it should definitely be mentioned in the NBMB run text, like my first any% run was in the realtime any% runs text. 9 for entertainment and 8.5 for technical (~7.6-7.7 for the first part of the run, about up to kraid, and ~9-9.1 for the rest). Also yay for flamers voting <1 for technical...
Agare Bagare Kopparslagare
Player (206)
Joined: 2/18/2005
Posts: 1451
Cpadolf wrote:
The optimization in the latter part of the run is also very very good, though I still refuse to think that there are not at least a few seconds of improvement, making the run in whole improvable by about 20-25 seconds without new discoveries.
I don't blame you for truly believing that the later part from Ridley on (at which I started to optimize the run much more after realizing how close it will get with the completion time) is improvable by 10 seconds from optimization only just like that, but you are always free to try proving me wrong, having the smv to compare to your improvements now. :-) And yeah, props to the vote abuser who aren't able to judge the run objectively. Keep it up.
See my perfect 100% movie-walkthroughs of the best RPG games on http://www.freewebs.com/saturnsmovies/index.htm Current TAS project (with new videos): Super Metroid Redesign, any% speedrun
Joined: 4/30/2006
Posts: 480
Location: the secret cow level
The lack of crazy speedbooster tricks hurts the entertainment value, but holy Draygon fight.
Joined: 11/11/2006
Posts: 1235
Location: United Kingdom
Saturn wrote:
And yeah, props to the vote abuser who aren't able to judge the run objectively. Keep it up.
I think you'll find everyone is entitled to their own opinion. Though the fact they chose not to explain themselves here pretty much nullifies the vote..
<adelikat> I am annoyed at my irc statements ending up in forums & sigs
Joined: 5/2/2006
Posts: 1020
Location: Boulder, CO
Is this http://www.youtube.com/watch?v=3LLwcB5-WcE method faster then what was used in the video? It looks like it, but I have been wrong before. The video says you found this trick, but it really really looks like the video is faster then what was used in the run.
Has never colored a dinosaur.
Mitjitsu
He/Him
Banned User
Joined: 4/24/2006
Posts: 2997
Saturn wrote:
And yeah, props to the vote abuser who aren't able to judge the run objectively. Keep it up.
As one little bird told once told me on IRC. 'If it's not perfect, then it's not optimized' Since I haven't watched the run, given how incredibly boring the previous run was compared to the other valid categories. I won't be rating this, but I'm sure you did a good job overall.
Player (89)
Joined: 11/14/2005
Posts: 1058
Location: United States
Twelvepack wrote:
Is this http://www.youtube.com/watch?v=3LLwcB5-WcE method faster then what was used in the video? It looks like it, but I have been wrong before. The video says you found this trick, but it really really looks like the video is faster then what was used in the run.
Yes it is about 1 second faster than what is done in the run. Saturn chose not to use this trick because he did not want to spoil it for his any% run. There are quite a few other time savers he intentionally left out of this run for the same reason. This is why he says the run is improvable by about 10 seconds. As for the run itself, I liked it. It does have a very high level of optimization, and shows off some nifty tricks that would only be needed in low% item conditions. As for publication, I am not too sure. Published or not, great job Saturn.
They're off to find the hero of the day...
Experienced player (829)
Joined: 11/18/2006
Posts: 2426
Location: Back where I belong
AKA wrote:
Saturn wrote:
And yeah, props to the vote abuser who aren't able to judge the run objectively. Keep it up.
As one little bird told once told me on IRC. 'If it's not perfect, then it's not optimized'
And the corollary to that: It's perfect except for all the tricks that weren't included.
Living Well Is The Best Revenge My Personal Page
Joined: 5/2/2006
Posts: 1020
Location: Boulder, CO
If its not even updated to tricks that have been known by the author for 2 years, its technical quality is not very high. Leaving out stuff like that intentionally is either lazy or obnoxious. A good run should be neither. Also-- Your submission text says that you aim for both in game time as well as real time, but in which order? In order to avoid another ugly, annoying discussion about this in the future, you should say which has been minimized.
Has never colored a dinosaur.
Skilled player (1445)
Joined: 7/15/2007
Posts: 1468
Location: Sweden
Twelvepack wrote:
Your submission text says that you aim for both in game time as well as real time, but in which order? In order to avoid another ugly, annoying discussion about this in the future, you should say which has been minimized.
Ingame time is the main priority, the one thing that was not abused for ingametime as far as I know is using multiple pausescreens for whenever that can save time.
Agare Bagare Kopparslagare
Joined: 7/16/2006
Posts: 635
Even if you can't get a lower displayed time, the game keeps track of game time down to seconds/frames, right? I'd like to see this fully optimized, to know what the "actual" lowest game time is.
Player (206)
Joined: 2/18/2005
Posts: 1451
Titus Kwok wrote:
The lack of crazy speedbooster tricks hurts the entertainment value, but holy Draygon fight.
Yeah, but on the other hand it makes the run very unique especially in the later parts, since there is no other run form that allows this conditions while still having a solid goal, hence the many unique tricks it allows to present. It's kind of comparable to the "small only" Super Mario TASes, just with much more restrictions, making it a pretty special run.
Twelvepack wrote:
Leaving out stuff like that intentionally is either lazy or obnoxious. A good run should be neither.
I defintely wasn't lazy with this run. Just look how much planning and luck manipulation work has been done in every part of the run. There are only 3 notable tricks I left out of this run (new OTES climb, Gravity Spike Tech and the mentioned Norfair Walljump) which together would save around 2 seconds only, and would barely make a difference in the end overall. In fact, I think that the old techs look even cooler anyway, so it only adds up to the entertainment value of the run. As for the reason, hero summed it up pretty well.
Twelvepack wrote:
Your submission text says that you aim for both in game time as well as real time, but in which order? In order to avoid another ugly, annoying discussion about this in the future, you should say which has been minimized.
Well, since I skip every single menu entrance, even if it slows down the ingame time slightly (examples seen at the Tourian recharge station, or various opportunities where switching off Ice Beam would allow for slightly faster refills), I would say realtime. Luckily I managed to get the lowest possible game clock completion time nevertheless, so it basicly satisfies both goals in my eyes. But yeah, primary goal was realtime in that case.
See my perfect 100% movie-walkthroughs of the best RPG games on http://www.freewebs.com/saturnsmovies/index.htm Current TAS project (with new videos): Super Metroid Redesign, any% speedrun
Skilled player (1445)
Joined: 7/15/2007
Posts: 1468
Location: Sweden
Saturn wrote:
Well, since I skip every single menu entrance, even if it slows down the ingame time slightly (examples seen at the Tourian recharge station, or various opportunities where switching off Ice Beam would allow for slightly faster refills), I would say realtime. Luckily I managed to get the lowest possible game clock completion time nevertheless, so it basicly satisfies both goals in my eyes. But yeah, primary goal was realtime in that case.
Well optimizing doors for realtime makes quite a big difference (as shown in my 100% run), so that's the reason I saw ingame time as the main priority here.
Agare Bagare Kopparslagare
Player (206)
Joined: 2/18/2005
Posts: 1451
Yeah, thats the only exception. Optimizing doors for realtime looks pretty sloppy and in most cases loses much more ingame time than it saves realtime, so the exchange wasn't worth it for me. Besides nobody will notice the few frames of longer door scrolling, but everybody will notice the sloppiness when entering them, which I wanted to avoid.
See my perfect 100% movie-walkthroughs of the best RPG games on http://www.freewebs.com/saturnsmovies/index.htm Current TAS project (with new videos): Super Metroid Redesign, any% speedrun
Editor, Experienced player (735)
Joined: 6/13/2006
Posts: 3300
Location: Massachussetts, USA
Saturn wrote:
I defintely wasn't lazy with this run. Just look how much planning and luck manipulation work has been done in every part of the run. There are only 3 notable tricks I left out of this run (new OTES climb, Gravity Spike Tech and the mentioned Norfair Walljump) which together would save around 2 seconds only, and would barely make a difference in the end overall. In fact, I think that the old techs look even cooler anyway, so it only adds up to the entertainment value of the run. As for the reason, hero summed it up pretty well.
"All in all, the quality of the run is very high, and it's only improvable by at most 15 seconds in the early parts of the run (the late ones are pretty much flawless with a few tiny exceptions worth of single frames only). Ammo and energy management is excellent over the entire run, and in fact, the refill drops are so good in this one that I doubt they could be reproduced without sacrificing realtime by entering the menu to manipulate them, which would in any case result in possible slowdowns to compensate that." Make up your mind, either the run is of high quality and improvable by only 2 seconds from 3 tricks, or it's improvable by 15 seconds like you say in your submission text.
Homepage ☣ Retired
Player (206)
Joined: 2/18/2005
Posts: 1451
The "entire" run is improvable by ~15 sec with 95% comming from the early part of the run. 2 of this seconds would be saved with the mentioned 3 tricks, the rest from just better optimization and a few techniques I wasn't aware about back then (improved Torizo Skip, down-reverse fall and hero-knockbacks in green/red brinstar are the major ones). Is there anything not clear enough yet? And keep in mind that 15 sec is the peak. In fact, I'd estimate even less than that. Don't ask for exact numbers though. It's impossible to know for sure.
See my perfect 100% movie-walkthroughs of the best RPG games on http://www.freewebs.com/saturnsmovies/index.htm Current TAS project (with new videos): Super Metroid Redesign, any% speedrun
Tub
Joined: 6/25/2005
Posts: 1377
It's definitely not less entertaining than the 6% NBMB TAS overall
well, I'd give the 6% run a solid 9.5 on entertainment, so this'll have to get 10.5 because Saturn said so.
All in all, the quality of the run is very high
voting 11 on tech quality, because Saturn said so. Now that that's cleared up, on to actually watching the run.. A movie should be as good as can be at the point of creation. While I enjoyed the new tricks that were squeezed out of the long-time neglected low% category, intentionally skipping tricks for selfish purposes looks like a case of "artificial restrictions" as defined in the guidelines. So - category issues aside - I'd vote gruefood.
m00
Joined: 7/1/2008
Posts: 272
tech quality i have no problems with. crazy amounts of luck manipulation indeed. i gave it an 8. entertainment kinds suffered in some places, mostly because of the overexposure to SM. but there were some nice things. i really liked how quickly you refilled your ammo before Ridley. i'll give entertainment a 7. this run should be able to coexist with the 6% run because that run uses gamebreaking glitches, where this one doesn't.
Joined: 5/2/2006
Posts: 1020
Location: Boulder, CO
Saturn wrote:
Twelvepack wrote:
Leaving out stuff like that intentionally is either lazy or obnoxious. A good run should be neither.
I defintely wasn't lazy with this run.
Then I guess that narrows it down.
Has never colored a dinosaur.
Joined: 4/3/2005
Posts: 575
Location: Spain
Satoryu wrote:
this run should be able to coexist with the 6% run because that run uses gamebreaking glitches, where this one doesn't.
That argument would work if there weren't other published runs of this game that don't use gamebreaking glitches, either, while also being faster, better at showing all features in the game, and/or cooler.
No.
Player (206)
Joined: 2/18/2005
Posts: 1451
Tub wrote:
It's definitely not less entertaining than the 6% NBMB TAS overall
well, I'd give the 6% run a solid 9.5 on entertainment, so this'll have to get 10.5 because Saturn said so.
I didn't consider you thought that almost 4 minutes of slowly X-Ray climbing up 11 screens while having a still picture without any movements at all (other than the climb animation) is worth a 9,5 entertainment rating for you. Sorry for the misunderstanding. As for technical, no need to exaggerate. It is, other than at the very beginning and the few skipped tricks, very high, and especially in the late run barely improvable (otherwise Cpadolf would already post his claimed 600 frame improvement smv from the later part already, which didn't happen as you see).
See my perfect 100% movie-walkthroughs of the best RPG games on http://www.freewebs.com/saturnsmovies/index.htm Current TAS project (with new videos): Super Metroid Redesign, any% speedrun
1 2
12 13