I'm going to test our new method of encoding 30 Hz motion for youtube on this video. Encoding HD.

I'll be the bad guy and give this a no. Most branches for games on this site are at least conceptually significantly different from all other branches. The difference between real-time and in-game times is such a ridiculously arbitrary distinction that I can't support reinstating the "in-game time" branch. Between the main game and hacks, we already have 5 branches for Super Metroid. It really seems to me like we're at the point where any additional runs need a sufficiently unique gimmick in order to be worth publishing. Unfortunately, I tried to see how this run screamed "I aim for in-game time" and couldn't. It really, really felt like a simple fastest time run, albeit perhaps one that didn't avoid items as often (something I probably wouldn't have even noticed had it not been the fact that I was looking for it). Despite all that, I do have to admit that this is a pretty solid run (although I'm starting to find vanilla SM runs a little stale at this point). I can see the case for grandfathering the branch back into existence, but I personally think it would be better if it stays gone.

My thoughts exactly. Just because the speedrunning community as a whole aims for in-game time due to the convenience of an in-game timer doesn't mean we need a separate branch for a TAS. There are quite a few games with in-game timers and I doubt any in-game oriented run for one of those games would be accepted. Super Metroid shouldn't be an exception. Plus, with the glitched run being 10 in-game minutes faster, this run basically serves the purpose of "glitchless in-game any%", which makes it even more arbitrary. The quality may be as high as the author's ego, but I'm still voting No due to a completely arbitrary category I was glad to see go when Taco and Kriole's run obsoleted it.

There are quite a few games with in-game timers and I doubt any in-game oriented run for one of those games would be accepted.

http://tasvideos.org/Movies-C1012Y.html

There are quite a few games with in-game timers and I doubt any in-game oriented run for one of those games would be accepted.

http://tasvideos.org/Movies-C1012Y.html

None of those have parallel real-time oriented runs. Would a real-time oriented run for any of those games be published alongside the in-game time?

1) It's a great run 1.5) I've not watched it 2) "more branches" "arbitrary goal" is weak, in the (rather large) super metroid community a run of this goal probably has a significant audience 3) we get to put that significant audiences category-since-the-game-came-out category on the front page 4) people come looking for it 5) ad money 6) beer money from google ar·bi·trar·y: Based on random choice or personal whim, rather than any reason or system. The choice for goal of this run is based on a well established system. I'm tired of people complaining about games that are overrepresented. If the game is fun enough to watch/has a large enough audience, then it can handle many categories. Super metroid is fun to watch for many people, and has a large audience. Don't just say "there's too many categories so this can't be published." Why is the currently published run deserve the limited spot more? Why do you think super metroid can't handle any more spots on the list? Why do you think the spots should be limited at all? The goal is not arbitrary, so if that's your angle then pick a different word.

I agree with jimsfriend. There's no reason to limit the number of categories any game can have, assuming all the categories are not arbitrary, as he explained.

There was an in-game time cathegory before, that was obsoleted by Taco and Kriole's movie, but if this one is faster by almost 1 minute (51 seconds) in-game time the in-game time cathegory could be brought back.

I can attest to the importance of in-game time as the traditional method of weighing and judging Super Metroid runs. This game has a rich history of speedrunning and TASing, and I see no reason to reject this video which achieves the best ever in-game time (without game-breaking glitches, of course). That said, I think a parallel real-time category is worthwhile, as, after all, we are talking about a TAS, here. Gigabytes of storage are cheap, these days, so I think having our realtime TAS alongside an in-game-time nod to one of the most popular speedrunning games ever is worth considering.

BTW, for those of you who want to watch this run as it would be timed by the ingame timer, use this lua script

a=0 

while true do 

   if memory.readbyte(0x7e198e) == 128 then 
   gui.drawbox(0,0,255,31,"#000000") 
end 

b=memory.readbyte(0x7E1842) 
if a~=b then 
snes9x.speedmode("normal") 
else snes9x.speedmode("turbo") 
end 
a=b 

   snes9x.frameadvance() 
end a=0 

while true do 

   if memory.readbyte(0x7e198e) == 128 then 
   gui.drawbox(0,0,255,31,"#000000") 
end 

b=memory.readbyte(0x7E1842) 
if a~=b then 
snes9x.speedmode("normal") 
else snes9x.speedmode("turbo") 
end 
a=b 

   snes9x.frameadvance() 
end 

This script automatically fast forwards during any part of the run where the ingame timer is not running. I didn't make this script, BTW. I don't remember who did, though.

People seem to like this movie a lot. It's better than the existing run in some ways. It's not better than the existing run along all metrics. Separate publication seems to be the clear choice. "I'm tired of Super Metroid runs" is not a compelling argument. Edit: And now, having watched the run, I am voting Yes. There's a lot of parts of the game in this run that I haven't seen for quite some time. I do wonder why get the Varia suit, though.

People seem to like this movie a lot. It's better than the existing run in some ways. It's not better than the existing run along all metrics. Separate publication seems to be the clear choice. "I'm tired of Super Metroid runs" is not a compelling argument. Edit: And now, having watched the run, I am voting Yes. There's a lot of parts of the game in this run that I haven't seen for quite some time. I do wonder why get the Varia suit, though.

Varia Suit halves the damage of Mother Brain's hyperbeam attack (the only thing it does that the Gravity Suit does not, as far as I know). Therefore, you need 6 E-tanks to survive against Mother Brain if you don't pick it up.

There are quite a few games with in-game timers and I doubt any in-game oriented run for one of those games would be accepted.

http://tasvideos.org/Movies-C1012Y.html

technically i aimed for in-game time in my fusion 100% tas, but this hardly matters. anyway i voted yes. i think all metroid games should always aim for in-game time.

Saturn sure likes submitting runs that will inevitably lead to flame wars and heated debate.

Saturn sure likes submitting runs that will inevitably lead to flame wars and heated debate.

This is very true. He even refrained from submitting his Redesign run, which would have almost certainly been published, highly praised and uncontroversial.

I don't like in-game as a goal for this game. It has historical value and would clearly be of interest to Super Metroid speedrunners, so it should be 'published' on YouTube or other places clearly visible to Super Metroid speedrunners. As a TAS we already have a very good metric for how fast a TAS is; the framecount. As aiming for lowest framecount does not result in ugly pauses (like a Sonic TAS would), I think it's a much more pleasing goal. Also, lower framecount means less door transitions, which can only be a good thing. Voted no.

This run looks very optimized, with good planning and execution, including a spiffy new Baby Metroid Escape. Well done, Saturn! As for the question of publication, I think it should be linked from the any% page, but not published in a separate category. Routes By now, the routes are as follows: to red brinstar PBs -> phantoon -> high jump + E-Tank -> SB -> Wave -> Ridley -> Ice -> Kraid -> Maridia -> Draygon -> Plasma -> Tourian to red brinstar PBs -> phantoon -> Kraid + E-Tank -> SB -> Wave -> Ridley -> Ice -> Maridia -> Draygon -> Tourian The ingame-run collects two additional items and does Kraid later. (The realtime run has to do Kraid earlier for the E-Tank, since the E-Tank near the high jump is obviously skipped.) This results in two large detours, a slightly quicker Kraid kill and a faster fight against Mother Brain. Everything else looks similar, with minor alterations in room strategies, partially due to differences in items (HJ/Varia), partially due to different needs for manipulating door transitions and lag, and partially due to stylistic choices. It's hard to tell which one looks "better": the high jump allows for a nicer exit after ridley, but makes for a clumsy jump in the pre-wave room, so it all evens out. Note that this was different when we had two any% runs published. Back when we decided to keep those runs separate, the routes had major differences. The arguments that warranted separation back then may not hold any more. Tricks Does this run have any unique tricks or situations the other run cannot show off? I could only find one instance where Saturn's run truly surprised me: the improved Baby Metroid-Escape. However, that's a trick that could be applied to all unglitched runs and thus wouldn't warrant a separate category. Since the runs only feature minor differences and don't have their unique tricks to show off, I don't think publishing both is beneficial for the site. ingame vs. realtime Of course one might argue that we should feature the ingame-run on the frontpage, and have it obsolete the any% run. It's difficult to tell which one is more optimized since the runs diverge after a few door transitions, and only a handful of SM TASers could spot inaccuracies at this level of optimization in either run. This run adds ~85 seconds of boring stuff like door transitions, menus, item-jingles, elevators and delayed frames in order to lower the ingame-timer by ~51 seconds. While removing playing time is always a noble goal, doing so by adding a disproportionate amount of non-playing-time still doesn't sound like a good idea to me. In addition to that, IMHO it looks more impressive to run through the game without having to grab the big guns. Most of the savings come from the Mother Brain fight. But I'd prefer to see two minutes of well-timed wave PB combos to one minute of detours for plasma and one minute of senseless plasma spamming. So altogether, I remain convinced that the realtime category serves this game better. The short form of this post is a "Meh" vote.

I actually agree with Tub. This run has a number of functional improvements over the published any%—a surprising number, I may add—but they could be applied to the published run to lower its ingame time just as well. Maybe not enough to lower it to 0:22, but close to it at the very least. That being said, Saturn, I'm glad you still aren't done with this game. Your attention to it is valuable regardless of your submissions being published or rejected. This submission, in particular, has quite an amount of tricks that could be used in further iterations of runs in other categories.

Voted no. Certainly not different enough than the current published run.

Honestly, I think it's quite unjust that some people are voting "no". There has ALWAYS been a "in-game time" category for Super Metroid in addition to the "real time" category. It's not some wonky or arbitrary new category; it's existed since Cpadolf's original run in March of 2008. (Which was extremely well-received, by the way, and didn't have tons of people voting "no" because it was a new category at the time) Now, it just so happens that the run by Taco and Kriole obsoleted both categories at the time. Now that Saturn has obsoleted one of those categories, his run should be published, just like Cpadolf's "in-game time" runs before. It's only fair. It's an incredible run and an easy yes vote.

There has ALWAYS been a "in-game time" category for Super Metroid in addition to the "real time" category. It's not some wonky or arbitrary new category; it's existed since Cpadolf's original run in March of 2008.

The old adage goes: Past mistakes don't justify repeating them. (This doesn't mean the in-game-timing run shouldn't be published. It just means that "we have done so in the past" is not a good argument for doing so.)

The old adage goes: Past mistakes don't justify repeating them. (This doesn't mean the in-game-timing run shouldn't be published. It just means that "we have done so in the past" is not a good argument for doing so.)

In theory. However, when is the last time that an improvement over an early run of a crappy game has gotten rejected? By your reasoning, said improvement could be rejected if it's retroactively decided that the game stinks, right? In this case, we're talking about an improvement over a GREAT run of an amazing game. I don't see the issue. Easy yes vote.

In theory. However, when is the last time that an improvement over an early run of a crappy game has gotten rejected? By your reasoning, said improvement could be rejected if it's retroactively decided that the game stinks, right?

This is mainly a problem because of the informal "no unpublications" policy of the site. The question is, AFAIK, open and not formally decided.

Does Super Metroid get to be the exception where both a real-time and in-game time run can co-exist? If everyone were slobbering over a game like... say, Star Ocean: The Second Story, which also has an in-game timer used to judge console speedruns, would the community accept two separate runs if one has a time-wasting trick to lessen the in-game time? I seriously doubt it, because SO2 isn't Super Metroid. The arguments for publication strike me as pure favouritism. I have yet to see a single good argument for publication. Even Cpadolf's posting of the list of in-game time runs lacked any power, since none of those runs have a corresponding real-time run. Super Metroid is the only game that has had both. I just don't see why it's acceptable for any one game to have near-identical runs side-by-side while the rest have none. It should be either one or the other, not both, unless it's suddenly okay to submit... say, something like this: A Sonic TAS where the author intentionally wastes time in certain stages to reduce the score countdown after each act, resulting in a 20-30 second lower real-time at the cost of in-game time. Would such a run be published alongside the currently published in-game time runs? Possibly, though we've rejected strategies that offer faster real times at the expense of finishing the game slower "in-game". I admit that it's not a perfect parallel. Should this run be published, which I am still against, I'd much rather it obsolete the real-time run instead of existing alongside it. I just don't see why it should be published on the grounds that it satisfies a small group's wishes while the majority of people who watch speedruns have absolutely no idea what constitutes the separation.

Btw, for the record: This run is very good and I like it a lot. I'm just holding my vote until I can decide whether the in-game timer goal is sensible or not.