Best Times

2007-07-02 01:27:31 pm

Yes, this looks very possible to get the last 16 Missiles + 1 Super in. However you will have to reduce your speed towards him to the minimum in order to succeed from this position.

Regarding SM TAS, yes, I hate it to do multiple runs with same category of the same game. I prefer it much more to rather take all the time needed and only concentrate on one attempt of the run to finish it as good as possible.

The only thing that will always be improvable in SM TASes is luck manipulation, because you don't have full control of it, and thus always have to hope for good refill drops especially after Bosses. If you don't depend on luck it's very well possible to get a part of the run (a single or multiple rooms without refill problems for example) perfect. It's of course very hard to do, but not impossible.

And of course I will keep up the TAS work. I'm glad you like it. It will just take a while to finish them, since I don't have much freetime left for TASing in this days and the near future.

Yes, this looks very possible to get the last 16 Missiles + 1 Super in. However you will have to reduce your speed towards him to the minimum in order to succeed from this position.

Regarding SM TAS, yes, I hate it to do multiple runs with same category of the same game. I prefer it much more to rather take all the time needed and only concentrate on one attempt of the run to finish it as good as possible.

The only thing that will always be improvable in SM TASes is luck manipulation, because you don't have full control of it, and thus always have to hope for good refill drops especially after Bosses. If you don't depend on luck it's very well possible to get a part of the run (a single or multiple rooms without refill problems for example) perfect. It's of course very hard to do, but not impossible.

And of course I will keep up the TAS work. I'm glad you like it. It will just take a while to finish them, since I don't have much freetime left for TASing in this days and the near future.

2007-07-02 02:16:57 pm

Almost happy

Well ofcourse it is theoretically possible to do a perfect run but I can bet a finger that it will never be done (although the better the run the harder it is do see if it's improvable or not) also let's say the perfect run was done, who would have the right to say it's perfect because no one can actually know if it is. Nuf ranting.

To hear that it was possible to do the 1 round from there only makes it hurt more that the movie desynced. I'm really taken aback by this and have no lust to try again at least until tomorrow.

2007-07-02 02:56:25 pm

I am a strong believer that a perfect run is technically impossible. There have been many TASes which were thought to be perfect, yet improvements have been made. Keep in mind that most of the runs that got improved were very very simplistic compared to Super Metroid physics, which means the odds of perfecting Super Metroid is much less likely. Don't get me wrong though, I do think that a run can be within 2-3 seconds of theoretical perfection. Only problem is that once a "perfect" run is complete, someone will come around and introduce a new trick or timesaver and thus the "perfect" run would not be perfect anymore.

Cpadolf, I would suggest that you not spend too much time on phantoon. If you need to do a 2 round kill, it would not really cost much time and you get the added benefit of having more energy/missile drops during the fight. Spending too much time on a single part of the run can really suck the fun out of it. For example I have spent the past few days trying to get missile drops in the purple diagonal cavern area without slowing down much, but I cannot do it without losing over 30 frames. I have spent so much time on this one room, that my enjoyment of the game has slowly decreased. If you really want to you can post your movie and I could take a crack at phantoon for you, but if you would prefer the work to be only yours that's cool too.

I am a strong believer that a perfect run is technically impossible. There have been many TASes which were thought to be perfect, yet improvements have been made. Keep in mind that most of the runs that got improved were very very simplistic compared to Super Metroid physics, which means the odds of perfecting Super Metroid is much less likely. Don't get me wrong though, I do think that a run can be within 2-3 seconds of theoretical perfection. Only problem is that once a "perfect" run is complete, someone will come around and introduce a new trick or timesaver and thus the "perfect" run would not be perfect anymore.

Cpadolf, I would suggest that you not spend too much time on phantoon. If you need to do a 2 round kill, it would not really cost much time and you get the added benefit of having more energy/missile drops during the fight. Spending too much time on a single part of the run can really suck the fun out of it. For example I have spent the past few days trying to get missile drops in the purple diagonal cavern area without slowing down much, but I cannot do it without losing over 30 frames. I have spent so much time on this one room, that my enjoyment of the game has slowly decreased. If you really want to you can post your movie and I could take a crack at phantoon for you, but if you would prefer the work to be only yours that's cool too.

P.JMan

2007-07-02 03:13:36 pm

I like Big Butts and I can not lie

A run that is 1 frame long is indisputably perfect

2007-07-02 03:14:51 pm

Almost happy

Then first I have 2 curiosity questions for you Hero, though of course you don't need to answer if you don't want to.

1: Did you manage to improve JXQ's ceres escape?
2: Besides of torizo skip, how much have you improved on your last run?

And ok I give up I'll let you do phantoon (Though of course don't feel rushed do it whenever you like) and I think the main reason I give up is because if I have to do it myself I don't know how long it will take until im upp for it again, spending ~half of the rerecords in the entire run so for on phantoon solely to get him to a point were he was defeat able just to see the movie desync before the fight even started just took the juice out of me.

I'm of course deeply grateful of you to do this for me and will naturaly let everyone know you did that fight and not me.

EDIT:

Quote from P.JMan:

A run that is 1 frame long is indisputably perfect

ZOMG your right I never thought of it that way Rolling Eyes

attachment:

CpadolfTASv2.smv

2007-07-02 03:39:55 pm

Cpadolf, next time you should at least playback your run more often while making it, to spot such desync errors in time.

And I never said that a perfect run can be done. I mean that some rooms that don't require luck manipulation can be passed as fast as possible and thus perfectly, which doesn't mean that it applies to the entire run. Aside of that I agree with hero that even with all the luck manipulation problems it's possible to get a whole run done that is only improveable by 100-200 frames in total, and since this is a matter of 1-3 sec, it wouldn't affect the in-game timer in any case. As said before, having more than 30 sec (or ~2000 frames) to catch up from a already near perfect any% run estimation in order to get 0:22, is just too much to be possible, at least unless new major shortcuts get discovered, which I highly doubt will be.

Quote from hero of the day:

Only problem is that once a "perfect" run is complete, someone will come around and introduce a new trick or timesaver and thus the "perfect" run would not be perfect anymore.

Won't happen, since all possible tricks/shortcuts already got discovered (unless you mean some out of area stuff, which makes it a non-legit run).

Cpadolf, next time you should at least playback your run more often while making it, to spot such desync errors in time.

And I never said that a perfect run can be done. I mean that some rooms that don't require luck manipulation can be passed as fast as possible and thus perfectly, which doesn't mean that it applies to the entire run. Aside of that I agree with hero that even with all the luck manipulation problems it's possible to get a whole run done that is only improveable by 100-200 frames in total, and since this is a matter of 1-3 sec, it wouldn't affect the in-game timer in any case. As said before, having more than 30 sec (or ~2000 frames) to catch up from a already near perfect any% run estimation in order to get 0:22, is just too much to be possible, at least unless new major shortcuts get discovered, which I highly doubt will be.

[quote="hero of the day"]Only problem is that once a "perfect" run is complete, someone will come around and introduce a new trick or timesaver and thus the "perfect" run would not be perfect anymore. [/quote]
Won't happen, since all possible tricks/shortcuts already got discovered (unless you mean some out of area stuff, which makes it a non-legit run).

072

2007-07-02 03:48:03 pm

coral to complement blue

All possible tricks and shortcuts were discovered 3 years ago, too.

2007-07-02 03:58:47 pm

I am always willing to answer any questions you have, Cpadolf.

1. I am actually 2 time units slower than JXQ's escape time. I got a 48"99 and he got a 49"01. While the escape time clock is slower, I actually tied his frame count. The discrepancy is from a change in emulator versions, which slightly altered the way in which the escape timer counts. I am using the new snes9x 1.51.

Not comparing to JXQ, My escape is improvable by 4 frames. I intentionally delayed the game for 4 frames in order to achieve better steam locations in a couple of the rooms. If I was aiming in-game time, I would have delayed the Samus mission statement in order to get perfect steam luck. If I were to delay the mission statement however, it would lead to a longer delay in real time. So I sacrificed 4 in-game frames to save over 30 real-time frames. For this reason, my escape time is not the best possible time. Saturn currently holds the best escape clock time, but once he switches to the new Snes9x 1.51 it will not be reproducible.

2. I haven't been actively comparing my current run to my old run, but I would say that I am about 1 second ahead (not including the Torizo). The 3 main reasons are that the moozooh jump, the catnap jump, and the old MB room continuous walljump each saved a decent amount of frames.

Saturn:
By timesavers I meant things like the moozooh jump, catnap jump, the hero red brinstar shot. These are things that were overlooked until recently. There may be a few more of these that no one knows about yet.

I am always willing to answer any questions you have, Cpadolf.

1. I am actually 2 time units slower than JXQ's escape time. I got a 48"99 and he got a 49"01. While the escape time clock is slower, I actually tied his frame count. The discrepancy is from a change in emulator versions, which slightly altered the way in which the escape timer counts.  I am using the new snes9x 1.51.

Not comparing to JXQ, My escape is improvable by 4 frames. I intentionally delayed the game for 4 frames in order to achieve better steam locations in a couple of the rooms. If I was aiming in-game time, I would have delayed the Samus mission statement in order to get perfect steam luck. If I were to delay the mission statement however, it would lead to a longer delay in real time. So I sacrificed 4 in-game frames to save over 30 real-time frames. For this reason, my escape time is not the best possible time. Saturn currently holds the best escape clock time, but once he switches to the new Snes9x 1.51 it will not be reproducible.

2. I haven't been actively comparing my current run to my old run, but I would say that I am about 1 second ahead (not including the Torizo). The 3 main reasons are that the moozooh jump, the catnap jump, and the old MB room continuous walljump each saved a decent amount of frames.

Saturn:
By timesavers I meant things like the moozooh jump, catnap jump, the hero red brinstar shot. These are things that were overlooked until recently. There may be a few more of these that no one knows about yet.

2007-07-02 03:59:07 pm

Just watched your run, Cpadolf. Looks very very good, really. If you continue in this quality, you can indeed get 0:24! Good luck on that. :)

Quote from 072:

All possible tricks and shortcuts were discovered 3 years ago, too.

Last known shortcut was the Torizo skip, discovered a bit more than a month ago. But thats most likely the last one now.

EDIT:
Possible, hero, but unlikely. I for one learned alot from my overlooks of such moves, and test much more of them now.

2007-07-02 04:01:15 pm

Almost happy

I did check the run quite often IMO, I never played more then 1 ingame minute at a time which is not much for me since I don't spend hours of perfecting. And I did a check before phantoon with no desync so only the battle itself got lost.

Quote from 072:

All possible tricks and shortcuts were discovered 3 years ago, too.

072 has a good point here, we can't know if everything is discovered yet, although I find it highly unlikely that something big will be discovered. The game is to thoroughly searched and as Moozooh said after he made the torizo skip possible, "this may well be the last big thing for this game, now that everything else is done by Kejardon, Saturn and all the others." And I do belive him but would be more then happy if I was to be proven wrong.

EDIT: The main reason for the big increase in quality is that im now actively having current room strategies played on a second emulator (To this point mostly from heros run) to keep from using old tricks.

Oh and what is the hero red brinstar shot?

I did check the run quite often IMO, I never played more then 1 ingame minute at a time which is not much for me since I don't spend hours of perfecting. And I did a check before phantoon with no desync so only the battle itself got lost.

[quote="072"]All possible tricks and shortcuts were discovered 3 years ago, too.[/quote]

072 has a good point here, we can't know if everything is discovered yet, although I find it highly unlikely that something big will be discovered. The game is to thoroughly searched and as Moozooh said after he made the torizo skip possible, "this may well be the last big thing for this game, now that everything else is done by Kejardon, Saturn and all the others." And I do belive him but would be more then happy if I was to be proven wrong.

EDIT: The main reason for the big increase in quality is that im now actively having current room strategies played on a second emulator (To this point mostly from heros run) to keep from using old tricks.

Oh and what is the hero red brinstar shot?

072

2007-07-02 04:08:44 pm

coral to complement blue

Yeah, the game in question is Super Metroid, you know, that game that's been pushed to its "limits" about a million times.

CpadolfTASv2-phantoon1-round.smv

2007-07-03 02:47:23 am

The hero Red Brinstar shot is the one you do in the very first room with the long vertical shaft with the rippers, right after Green Brinstar.
You shoot the top block to walljump up through it, which avoids the usually required continuous midair morph bomb jumping. It's the most major of the minor shortcuts so far, saving a good second alone.

I don't know when hero will do the 1-round for you, so in case you want to continue very soon again, you can try this smv. Didn't took me much effort to do, since I just did it systematically:

attachment:

JaggerG

2007-07-03 02:53:50 am

Shifty Leader

That M2k2 guy

Is that trick doable in real time?

Also, I'm guessing it's not the expected trick in SMI due to the green hatch on top. I'm not sure I could have pulled it off.

2007-07-03 05:15:24 am

Almost happy

Thank you VERY much Saturn I can now conutiniue :P
The only bad part, only 1 SuperMissile drop :o

JaggerG: Probably not, it requires very fast walljumping. Plus in SMI you have to keep the shot onscreen for a longer part which probably is impossible even in frameadvance (Though I don't know)

2007-07-03 10:03:31 am

Sorry, I didn't have much time to work on the phantoon 1 round yesterday. Nicely done Saturn. I also wanted to mention that the quality of your work is getting very high Cpadolf. At this rate I foresee a very possible 24, keep it up :P

2007-07-03 01:40:01 pm

Almost happy

Finally done with WS, and I'm still a good 15 seconds ahead of you hero. and so begins the route change grin new

I know that this route is ~10 seconds faster then terris, but how much faster is it then your ingame time route, more, less...?

EDIT; Your Realtime route I meant of course, didn't matter much since you understod me anyway.

2007-07-03 02:32:53 pm

It should be about 55 seconds faster than my real time oriented run on the in-game time. Only because I skip the plasma/space jump. I designed my route with the intention of minimizing door transitions, as a result I saved a ton of real time but end up with a slightly slower in-game time. I estimate that when I finish my run I will get around a 24:30, while Saturn says the in-game route will get about a 23:35.

moozooh

2007-07-03 08:58:11 pm

Quote from JaggerG:

Is that trick doable in real time?

Most likely not, unless you're some kind of an Elite Beat Agents champion. %)

Cpadolf, just watched your WIP (haven't watched v1, though), nice to see you've taken a couple pages from my book as well. :)
Some advice you may need to achieve a better time:
1) If you haven't accelerated to Samus's full running speed, it makes more sense to release run button than to stop completely in case you need to slow down. That costed you some time in OMBR on the way back, and then a couple rooms later when entering the diagonal room in Crateria.
2) Don't forget that spinjump has more momentum than aiming-down jump, so don't aim down unless/until you need the reduction of vertical hitbox -- otherwise you'll just fly slower for no good reason.

Other than that, all the situations were handled pretty good so far, this is pretty enjoyable to watch despite the inaccuracies.

2007-07-04 06:13:50 am

Almost happy

Thanks moozooh, I will consider this through the rest of the run. Unfortunately I think the I have done at least the second mistake 1 more time since the WIP. Probably the first one to but I don't think it will matter very much, Although if i got a 25:01 that would suck pretty hard :/

EDIT: Just thought of something, when Saturn finally do release his 00:23 any% run, will it replace heros run since it has lower ingame time or will it not replace heros run?

2007-07-04 02:14:04 pm

Quote from Cpadolf:

Just thought of something, when Saturn finally do release his 00:23 any% run, will it replace heros run since it has lower ingame time or will it not replace heros run?

No, it definitely shouldn't replace heros run, since:

1. It's faster in realtime (which is the prefered goal of TASvideos in general)
2. A totally different run category (aims for realtime)

If my run should be published, it should get a different category (in-game oriented, legit any%).

I already talked with Bisqwit about this problem we will have to face once the run is done, and so far there is no clear decision what will happen with it. I personally wouldn't have a problem if my run won't be published at TASvideos due to a less accepted goal, as I mainly do it for all the Super Metroid fans, who generally see the in-game timer as the more accurate one to express speed and quality of a run. But of course it would be nice to have it on TASvideos as well.

2007-07-04 03:07:58 pm

Almost happy

I've come up with something that might work though, in the description of heros run (or vice versa) there could be a link and a description like " Also see Saturns Any% run which uses a different route and aims for ingame time"

That way both runs would be up and the site would not be bloated with too many SM runs.

2007-07-04 04:43:44 pm

This is a very difficult subject to address. My preference would be to allow both runs to be published in two distinct categories. The routes are massively different, and each one shows very unique strategies. Hopefully bisqwit will implement a new page layout that allows for as many runs as possible for each game without cluttering the page up. Though it may be a long time before we finally see an any% in-game oriented run submitted.

As for my run, I have made some good progress. I am on my way to the wrecked ship (I just got on the crateria elevator). I am exactly 24 seconds ahead of my old run (in-game frames). I managed to find so many little timesavers that it added up nicely over the course of the run. Once I use the CWJ on crateria lake and 1 round phantoon, I will be well over 33 seconds ahead of my old run.

2007-07-04 06:08:25 pm

Yeah, a new page layout would solve the problem. Otherwise, like Cpadolf suggested, mentioning it in the description of the other run would also be a good option to consider. But yeah, it will be a while until I finish my version of it, so there is enough time left to solve the problem.

Hero:
WOW, this is a very huge improvement already!
When you said that you could get 24:30 with your route, I first doubted it, since the Torizo skip saves ~20 sec, and we all thought your 1st run (which is 25:30) is improvable by not more than 15 sec, making a 24:55 from that calculations. But if you continue to save so many seconds that eary already, then a lower time than 24:55 is indeed very well possible. Keep it up man!