SM RTA Any% Savestates Run in 27min

Reeve

Edit history:

Reeve: 2012-06-21 09:08:34 pm

Reeve: 2012-06-21 09:46:02 am

Reeve: 2012-06-21 09:41:04 am

Reeve: 2012-06-21 09:39:00 am

Reeve: 2012-06-21 09:38:28 am

Reeve: 2012-06-21 09:36:47 am

Reeve: 2012-06-21 09:32:53 am

2012-06-21 09:32:14 am

As some people know, some time ago i decided to make a realtime any% run which would avoid picking up Spazer. I would do the run with savestates and realtime skills, but optimizing the run as much as i could, so even though i pulled off a trick, if i felt it could be done faster, i would try again. Before doing the run i decided to check hotarubi's 31min run to see what would be the differences in average considering door transitions, item collections, in-game time difference in the route and pause screens. Below i listed what i found in hotarubi's run that could be avoided:

Pause to disable ice beam (Red Brinstar)
Pause to enable Spazer and Ice beam (WS)
Pause to disable ice beam (WS)
pause to disable the spazer for the Power Wave Combo on Botwoon (Maridia)
Pause to disable the Gravity Suit (Maridia)
Pause to enable the gravity suit, Spazer and Ice Beam (Maridia)
Pause to disable Ice Beam (Maridia)
Pause to disable the Ice Beam (Leaving Lower Norfair)
Pause to enable the Ice beam to freeze the Metroids and Kill Mother Brain (Crateria)
Pause to disable Varia and Gravity Suit to take more damage from the Big Hopper before the Big Metroid (Torian)
Pause screen to enable the Varia and the Gravity Suit in the Refill Station (Torian)

Two door transitions (Spazer)
Two Screen transitions (Maridia Super Missile)
Two door transitions to get run up for a super short charge to break the pillars in LN
two door transitions (Ridley's Energy Tank)
Two door transitions (Refill Station after Big Metroid)

Item Box collection (Spazer)
Item Box Collection (Maridia Super Missile)
Item Box collection (Ridley's Energy Tank)
Refill Station Box (After Big Metroid)

And below what would be saved in-game time with the route change:

Spazer Route = 0600 frames (in-game)
Maridia Super Route = 0510 frames (in-game)
E.T. After Ridley = 0130 frames (in-game)
Torian Refill Station = 0270 frames (in-game)

Of course there could be some minor differences in the numbers, but it's very close.

A personal comment about hotarubi's run is that for someone with his skills and that decided, as he said, go for realtime, he picked up the wrong route. The reason i say that is he decided not to use Murder Beam on Mother Brain to avoid pause screens and lag, but if he was going to do that anyway, he should've picked the non-spazer route to save some in-game time and plus realtime to compensate the timme lost by not using Murder Beam, so in the end he wouldn't be using Murder Beam but would still be skipping around 10/11 seconds in-game time and plus two door transitions and one item collection box. Of course it was back when he was probably the only one who was playing the game in such a level, so he had to think about everything on his own, and today we have more speedrunners skilled enough to think about route possibilities and stuffs like that.

Anyway, I did the run skipping the missile pack near the charge beam which costs around 7 seconds to collect (reason is you can collect two missile drops faster than 7 seconds) and skipped the missile pack in the WS lake to avoid another 7 seconds realtime, but for that i couldn't do a single shinespark to Wrecked ship and this would cost some in-game seconds, but much less than 7 seconds realtime. Also, the reason why i skipped those two missile packs is that you only need 3 missile packs and 2 super missiles for the first phase of Mother Brain and although it's a too hard strategy to go for, the run was going to be kinda of a test run. also, in this run i did i collected the E.T. after Ridley and then proceed to the end and finished the run. I even uploaded it to youtube and some people probably watched it. Then a friend suggested skipping the energy tank after Ridley and i found it was a great idea, so i deleted the run from youtube and redid the run, but this time i decided to collect the missile pack near charge beam but still skipping the missile pack in the WS lake, though i should have collected it too, but given it's just a test run to help people, i don't see much of a problem. Anyway, I don't remember the date when i finished the run, but i decided to keep it as a secret so if i decided to do a run i would go for that route, given that if i could master it i could go both for in-game and realtime, given that the difference is just a few seconds.

After some time i started watching some of Garrison's attemps and noticed that he was going for realtime, but would only pick up the non-spazer route if he did a big mistake in the beginning of the run, but eventually he noticed the route was not that much of an issue and started using it. So now i decided to post the smv and some calculations i did in case it can help anyone.

Below is what the differences would be between a run using the same route i used in the smv (but collecting the WS missile pack) and hotarubi's 31min run (youtube).

11 Pause Screen31mins = 2200 frames -
08 Door transitions = 1815 frames --
02 Screen Transitions = 0360 frames ---
03 Item Box Collection = 1200 frames ----
01 Refill Station Box = 0050 frames -----> 5625 frames -> 93 seconds and 45 frames
Spazer Route = 0600 frames (in-game)
Maridia Super Route = 0510 frames (in-game)
E.T. After Ridley = 0130 frames (in-game)
Torian Refill Station = 0270 frames (in-game)
-------------------------------------------------------
Total Realtime saved = 7135 / 60 = 118 seconds and 55 frames

Note: If my calculations are right and if you want to get a run that has around the same in-game time as hotarubi's run but using the same route as in the smv (avoiding pause screens and door transitions) you would have to get a run that is around 70 seconds faster than hotarubi's run. Also, to skip the two door transitions and the refill station message box in Torian you have to pull off a Big Metroid skip, so if you can't you'll have to refill in the remote station, so you can add around 12/13 seconds to your calculations, so instead of 70 seconds you would have to be 82/83 seconds faster.

I used memory watch to calculate the door transitions, pause screens and item box colletions, and although the item box colletion always have the same number of realtime frames, the door transitions or pause screens can change due to lag (for doors) and what the player will enable or disable in the pause screens, also there could be some differences between the emulator and the console, so the numbers could be higher than that for the realtime that's beeing saved.

Thread title:

As some people know, some time ago i decided to make a realtime any% run which would avoid picking up Spazer. I would do the run with savestates and realtime skills, but optimizing the run as much as i could, so even though i pulled off a trick, if i felt it could be done faster, i would try again. Before doing the run i decided to check hotarubi's 31min run to see what would be the differences in average considering door transitions, item collections, in-game time difference in the route and pause screens. Below i listed what i found in hotarubi's run that could be avoided:

Pause to disable ice beam (Red Brinstar)
Pause to enable Spazer and Ice beam (WS)
Pause to disable ice beam (WS)
pause to disable the spazer for the Power Wave Combo on Botwoon (Maridia)
Pause to disable the Gravity Suit (Maridia)
Pause to enable the gravity suit, Spazer and Ice Beam (Maridia)
Pause to disable Ice Beam (Maridia)
Pause to disable the Ice Beam (Leaving Lower Norfair)
Pause to enable the Ice beam to freeze the Metroids and Kill Mother Brain (Crateria)
Pause to disable Varia and Gravity Suit to take more damage from the Big Hopper before the Big Metroid (Torian)
Pause screen to enable the Varia and the Gravity Suit in the Refill Station (Torian)

Two door transitions (Spazer)
Two Screen transitions (Maridia Super Missile)
Two door transitions to get run up for a super short charge to break the pillars in LN
two door transitions (Ridley's Energy Tank)
Two door transitions (Refill Station after Big Metroid)

Item Box collection (Spazer)
Item Box Collection (Maridia Super Missile)
Item Box collection (Ridley's Energy Tank)
Refill Station Box (After Big Metroid)

And below what would be saved in-game time with the route change:

Spazer Route		 = 0600 frames (in-game)
   Maridia Super Route	 = 0510 frames (in-game)
   E.T. After Ridley	 = 0130 frames (in-game)
   Torian Refill Station = 0270 frames (in-game)

Of course there could be some minor differences in the numbers, but it's very close.

A personal comment about hotarubi's run is that for someone with his skills and that decided, as he said, go for realtime, he picked up the wrong route. The reason i say that is he decided not to use Murder Beam on Mother Brain to avoid pause screens and lag, but if he was going to do that anyway, he should've picked the non-spazer route to save some in-game time and plus realtime to compensate the timme lost by not using Murder Beam, so in the end he wouldn't be using Murder Beam but would still be skipping around 10/11 seconds in-game time and plus two door transitions and one item collection box. Of course it was back when he was probably the only one who was playing the game in such a level, so he had to think about everything on his own, and today we have more speedrunners skilled enough to think about route possibilities and stuffs like that.

Anyway, I did the run skipping the missile pack near the charge beam which costs around 7 seconds to collect (reason is you can collect two missile drops faster than 7 seconds) and skipped the missile pack in the WS lake to avoid another 7 seconds realtime, but for that i couldn't do a single shinespark to Wrecked ship and this would cost some in-game seconds, but much less than 7 seconds realtime. Also, the reason why i skipped those two missile packs is that you only need 3 missile packs and 2 super missiles for the first phase of Mother Brain and although it's a too hard strategy to go for, the run was going to be kinda of a test run. also, in this run i did i collected the E.T. after Ridley and then proceed to the end and finished the run. I even uploaded it to youtube and some people probably watched it. Then a friend suggested skipping the energy tank after Ridley and i found it was a great idea, so i deleted the run from youtube and redid the run, but this time i decided to collect the missile pack near charge beam but still skipping the missile pack in the WS lake, though i should have collected it too, but given it's just a test run to help people, i don't see much of a problem. Anyway, I don't remember the date when i finished the run, but i decided to keep it as a secret so if i decided to do a run i would go for that route, given that if i could master it i could go both for in-game and realtime, given that the difference is just a few seconds.

After some time i started watching some of Garrison's attemps and noticed that he was going for realtime, but would only pick up the non-spazer route if he did a big mistake in the beginning of the run, but eventually he noticed the route was not that much of an issue and started using it. So now i decided to post the smv and some calculations i did in case it can help anyone.

Below is what the differences would be between a run using the same route i used in the smv (but collecting the WS missile pack) and hotarubi's 31min run (youtube).

11 Pause Screen31mins	= 2200 frames -
08 Door transitions	= 1815 frames --
02 Screen Transitions	= 0360 frames ---
03 Item Box Collection	= 1200 frames ----
01 Refill Station Box	= 0050 frames -----> 5625 frames -> 93 seconds and 45 frames
   Spazer Route		 = 0600 frames (in-game)
   Maridia Super Route	 = 0510 frames (in-game)
   E.T. After Ridley	 = 0130 frames (in-game)
   Torian Refill Station = 0270 frames (in-game)
-------------------------------------------------------
   Total Realtime saved	= 7135 / 60 = 118 seconds and 55 frames

Note: If my calculations are right and if you want to get a run that has around the same in-game time as hotarubi's run but using the same route as in the smv (avoiding pause screens and door transitions) you would have to get a run that is around 70 seconds faster than hotarubi's run. Also, to skip the two door transitions and the refill station message box in Torian you have to pull off a Big Metroid skip, so if you can't you'll have to refill in the remote station, so you can add around 12/13 seconds to your calculations, so instead of 70 seconds you would have to be 82/83 seconds faster.

I used memory watch to calculate the door transitions, pause screens and item box colletions, and although the item box colletion always have the same number of realtime frames, the door transitions or pause screens can change due to lag (for doors) and what the player will enable or disable in the pause screens, also there could be some differences between the emulator and the console, so the numbers could be higher than that for the realtime that's beeing saved.

Reeve

2012-06-21 09:36:17 am

Damn, forgot to attach the smv. Here it goes:

attachment:

Super Metroid any% RTA.smv

Quietus

2012-06-21 11:42:23 am

Super Secret Area - Dead Ahead!

Hotarubi's run should be beatable anyway, even using the exact same route. His run is nearly six years old, and the standard of general play has risen since then.

Reeve

Edit history:

Reeve: 2012-06-23 12:39:17 am

2012-06-21 08:58:47 pm

It's possible to beat hotarubi's run for sure, but i'd say it's still hard to find someone with the same level of skills than him. I'd say the difference today is that more tricks were found and that's what would make it possible to beat horatubi's run by in-game time. Anyway, the idea of a run like the one i posted would be that, it makes possible to beat hotarubi and hotarubi's run by around 2 minutes, and not only that, but doing a run skipping spazer would be much better considering it would be skipping 11 pause screens, 6 door transitions and 3 item collection boxes.

Btw, i edited the first post with some numbers i realized now.

Gari

2012-06-22 03:57:02 am

Very helpful post, Reeve. I'll go ahead and share some of the basic stuff that I've learned from both the Japanese community and my own attempts some of the differences between gametime and realtime using this route. Hotarubi's run was timed as 55:39 from the moment he resets his SFC to the moment the BGM ends after the See You Next Mission screen. We've used a different timing on SpeedRunsLive for our RTAs and races, which start from Start Game (from an 0:00 file) to the moment you lose control on top of the Gunship. To convert this over, we had to time how long he spent resetting the game, how long the intro was, and how long it takes from losing control the end of the BGM. There were issues with the encoding process (some of this can be explained in this thread http://forum.speeddemosarchive.com/post/important_analog_capture_frame_rates_and_mysterious_gained_seconds.html), so unfortunately the numbers were kind of vague. Timing these things myself on my own SNES netted me a time 6:14, which brings hotarubi's run to 49:25. Unfortunately, I hadn't accounted for changing the control scheme, so the number gets even more vague. At the absolute lowest, however, hotarubi's run ends up clocking in at 49:22 with an ingame time of :31.

With that said, I've been using this route more or less (sans the difference in missiles) and beat hotarubi's time with a 49:08 with a gametime of :33. I'll be the first to admit that I'm not of the caliber that hotarubi was at his prime, but I do have a passion for running the game with realtime in mind. That said, my run is highly improvable by at least a minute, even with a slowest possible phantoon pattern. I've found a few things that were pretty funky with my gametimes since switching over. For example, I've gotten a :33 using hotarubi's route (sans the crab supers) and ended with a realtime of 51:10. Using the no spazer route, however, I've gotten a :34 with a 50:05 realtime and iirc, this was before I was skipping ridley's etank.

I was first informed that skipping spazer might be potentially faster by the japanese community (http://www40.atwiki.jp/niconamarta/pages/343.html) that they've been using it a for a while for RTAs only. If I remember correctly, their best times for RTAs are 49:43 and 49:52 using the route, both with mid :33s. I also did my own savestate run, though probably with a lot less care than yours, ended with a 46:52 realtime and a gametime of 00:30:59'85. Using just basic mental math, I've been assuming that :31 starts somewhere around 48:05.

I dunno, I just enjoy running the game with realtime in mind more than I enjoy gametime, so I'm glad to see some of the figures. Thanks for the post, Reeve.

Very helpful post, Reeve. I'll go ahead and share some of the basic stuff that I've learned from both the Japanese community and my own attempts some of the differences between gametime and realtime using this route. Hotarubi's run was timed as 55:39 from the moment he resets his SFC to the moment the BGM ends after the See You Next Mission screen. We've used a different timing on SpeedRunsLive for our RTAs and races, which start from Start Game (from an 0:00 file) to the moment you lose control on top of the Gunship. To convert this over, we had to time how long he spent resetting the game, how long the intro was, and how long it takes from losing control the end of the BGM. There were issues with the encoding process (some of this can be explained in this thread http://forum.speeddemosarchive.com/post/important_analog_capture_frame_rates_and_mysterious_gained_seconds.html), so unfortunately the numbers were kind of vague. Timing these things myself on my own SNES netted me a time 6:14, which brings hotarubi's run to 49:25. Unfortunately, I hadn't accounted for changing the control scheme, so the number gets even more vague. At the absolute lowest, however, hotarubi's run ends up clocking in at 49:22 with an ingame time of :31.

With that said, I've been using this route more or less (sans the difference in missiles) and beat hotarubi's time with a 49:08 with a gametime of :33. I'll be the first to admit that I'm not of the caliber that hotarubi was at his prime, but I do have a passion for running the game with realtime in mind. That said, my run is highly improvable by at least a minute, even with a slowest possible phantoon pattern. I've found a few things that were pretty funky with my gametimes since switching over. For example, I've gotten a :33 using hotarubi's route (sans the crab supers) and ended with a realtime of 51:10. Using the no spazer route, however, I've gotten a :34 with a 50:05 realtime and iirc, this was before I was skipping ridley's etank.

I was first informed that skipping spazer might be potentially faster by the japanese community (http://www40.atwiki.jp/niconamarta/pages/343.html) that they've been using it a for a while for RTAs only. If I remember correctly, their best times for RTAs are 49:43 and 49:52 using the route, both with mid :33s. I also did my own savestate run, though probably with a lot less care than yours, ended with a 46:52 realtime and a gametime of 00:30:59'85. Using just basic mental math, I've been assuming that :31 starts somewhere around 48:05.

I dunno, I just enjoy running the game with realtime in mind more than I enjoy gametime, so I'm glad to see some of the figures. Thanks for the post, Reeve.

Gari

2012-06-22 03:59:41 am

Oh yeah, also, if you guys want to check out the Japanese runners, they stream on twitch sometimes and they're worth checking out.

http://www.twitch.tv/hypooooo34
http://www.twitch.tv/fusiana
http://www.twitch.tv/emodoran
http://www.twitch.tv/hakuten
http://www.twitch.tv/danetch

Dessyreqt

2012-06-22 06:48:16 am

Why couldn't you one-shot over to Wrecked Ship? I find it's almost easier to one-shot it if you accidentally jump over the missile.

Also, some notes on your run:

- Your SRL time comes out to about 43:09, which is about 14 seconds faster than my 43:23 that gets a 26:52.70 gametime, but it's 39.30 ingame seconds slower. Such is the power of skipping spazer, as it not only cuts out two door transitions and an item box, but you can also do the run with only 2 pauses. That said, pause glitching the energy refill while going to Maridia is slower in realtime than not pause glitching it, so I'm not sure why you used it. I'm guessing a force of habit?

- Your Kraid quick kill is improvable by a second or two, using the method discovered by Taco and Kriole back in 2008. This isn't a huge gain, of course, and a lot of racers use the old way, but there are at least a couple that use Taco and Kriole's method.

- Similarly your Draygon could have been completed in two rounds using Saturn's PB-Wave method from 2006. This is the method I personally use most, though I don't spark the second round, instead opting to use ammo. Since actually using the bluesuit in the large mochtroid room is a bit of a wash, I'm not terribly surprised you didn't opt to use it there.

- The pillars thing. You do end up going faster into the pillars room, about 0.8125 pixels per frame faster. But your speed jump loses ~40 ingame frames and ~80 realtime frames to my shinespark. A few more realtime frames if you account for the fact that SNES9x 1.43 doesn't emulate lag properly (it's currently banned in SRL races as it gives ~15 second advantage during an any% race.)

- My only other thought is minor, the ball boost leaving lower Norfair is actually faster than going "as fast as possible." Any reason you didn't do it in this run, but you did in your 28:30?

Anyway, very nice run, it serves as a pretty good blueprint for a true RTA route.

Why couldn't you one-shot over to Wrecked Ship? I find it's almost easier to one-shot it if you accidentally jump over the missile.

Also, some notes on your run:

- Your SRL time comes out to about 43:09, which is about 14 seconds faster than my 43:23 that gets a 26:52.70 gametime, but it's 39.30 ingame seconds slower. Such is the power of skipping spazer, as it not only cuts out two door transitions and an item box, but you can also do the run with only 2 pauses. That said, pause glitching the energy refill while going to Maridia is slower in realtime than not pause glitching it, so I'm not sure why you used it. I'm guessing a force of habit?

- Your Kraid quick kill is improvable by a second or two, using the method discovered by Taco and Kriole back in 2008. This isn't a huge gain, of course, and a lot of racers use the old way, but there are at least a couple that use Taco and Kriole's method.

- Similarly your Draygon could have been completed in two rounds using Saturn's PB-Wave method from 2006. This is the method I personally use most, though I don't spark the second round, instead opting to use ammo. Since actually using the bluesuit in the large mochtroid room is a bit of a wash, I'm not terribly surprised you didn't opt to use it there.

- The pillars thing. You do end up going faster into the pillars room, about 0.8125 pixels per frame faster. But your speed jump loses ~40 ingame frames and ~80 realtime frames to my shinespark. A few more realtime frames if you account for the fact that SNES9x 1.43 doesn't emulate lag properly (it's currently banned in SRL races as it gives ~15 second advantage during an any% race.)

- My only other thought is minor, the ball boost leaving lower Norfair is actually faster than going "as fast as possible." Any reason you didn't do it in this run, but you did in your 28:30?

Anyway, very nice run, it serves as a pretty good blueprint for a true RTA route.

Reeve

Edit history:

Reeve: 2012-06-23 12:50:40 am

Reeve: 2012-06-22 07:53:25 am

Reeve: 2012-06-22 07:50:59 am

Reeve: 2012-06-22 07:46:11 am

2012-06-22 07:45:30 am

Garrison:

Thanks for sharing your infos, also, as Dessyreqt said snes9x isn't accurate mainly in terms of lag frames, so the final results of my tests could be wrong by some seconds. I imagine that door transition lag could be different between the emulator and the console, if this is the case, you could be saving a bit more than what i calculated and if this is the case in order to equal hotarubi's time you would have to shave even more from the run. Thanks for the links too, i was looking for some super metroid runners.

Dessyreqt:

What i thought at first was more about consistency, i mean, even if you can do that, you can end up not getting a shinespark in the right height and hit the door or the wall or even shinespark too low and get the missile pack. so the best i thought was to shinespark from the previous room in a way that would be fast and also safe. But as i said it's safer to get the missile anyway so you can just try and shinespark straight through the door.

Yeah, the plus of skipping spazer is you almost don't need to enter the pause screen along with skipping two door transitions and one item collection box and also saving around 10 seconds in-game time. Regarding the pause in the refill station, yeah, it was more habbit, i only realized i should test it after the run was done. If i recall correctly you loose 1 second in-game time but saves 1 second realtime? But yeah, if i was going to redo the run or something like that i would not pause there.

Regarding kraid, yes. After i did the run i tested that trick found by Kriole and Taco and adapted for realtime by forensics and realized it was easier than i first thought. So it's another thing i would change in the run. But as you said it wouldn't save that much anyway, but it's something i would go for in a run.

Yeah, Draygon can be killed a bit faster, but again something i went for in this run was Consistency, and to make all PW combo to hit Draygon is too hard and the idea of Shinesparking Draygon to death in an any% is more about saving ammo for later but i've been thinking about another strategy that would allow me to use the supers i have on him. The problem here is stopping in that bugs after botwoon is not a good strategy imo, you loose too much in-game time (realtime too) and you can't count on Botwoon to drop enough supers and you already need two for the door in the mocktroids room and another one for the eyed door entering Draygons's lair. so i think the best strategy is to shinespark Draygon to death, saving the supers for Ridley or later. Regarding the shinespark back i don't think it's worth in a run. Iirc it saves around 2 seconds in-game time, but considering you managed to shinespark all the way before Draygon, you'll be already low in health for a second shinespark back and if you intend to get Draygon's drops you'll waste so much time that the shinespark back will end up beeing a slower strategy. Plus if you don't collect Draygon's drops it can be a problem for Lower Norfair and the shinespark back will probably drain all your health, so you'll have to stop to refill before Ridley and that will be slower again. So the blue suit would only be worth (if it is at all) in a 100% run.

After we talked in twitch tv while garisson was speedrunning i tested it again and realized i forgot to measure the realtime losses by lag, so yeah, but again we come to consistency and health management. The shinespark before Draygon costs a lot and the others on him costs a good chunk of health too, so you already have low health after Draygon. Stopping to refill wouldn't be a good choice and you'll loose some health in Lower Norfair already, first from the damage boost in the head in the wall before the LN elevator and then, if you have luck, in the double kago room, then you loose some more diving in the lava after the Space Pirates and then you have to Fight Ridley and end the fight with enough health to survive leaving Lower Norfair (iirc Garrison died a lot in the red ki-hunters in the way out of LN). Also, there's the problem with consistency charging the shinespark without falling down, which would cost some time and make the shinespark option almost equally fast as the jump. I saw Garrison pressing down frames later than he was supposed to and not getting the shinespark but instead he would see Samus falling with a blue suit. I really think he would be more consistent with a jump there than charging the shinespark. If he fails to charge the shinespark he will probably loose the run given the time to destroy all pillars and the lag generated by the power bombs and the lava rising, not counting he could get hit and loose health too.

I don't know if i tested it right or not, but i remember it beeing just some frames faster so i think i forgot about it due to the fact it almost doesn't save time.

Also, thanks for watching the run and i apreciate your comments.

Garrison:

Thanks for sharing your infos, also, as Dessyreqt said snes9x isn't accurate mainly in terms of lag frames, so the final results of my tests could be wrong by some seconds. I imagine that door transition lag could be different between the emulator and the console, if this is the case, you could be saving a bit more than what i calculated and if this is the case in order to equal hotarubi's time you would have to shave even more from the run. Thanks for the links too, i was looking for some super metroid runners.

Dessyreqt:

What i thought at first was more about consistency, i mean, even if you can do that, you can end up not getting a shinespark in the right height and hit the door or the wall or even shinespark too low and get the missile pack. so the best i thought was to shinespark from the previous room in a way that would be fast and also safe. But as i said it's safer to get the missile anyway so you can just try and shinespark straight through the door.

Yeah, the plus of skipping spazer is you almost don't need to enter the pause screen along with skipping two door transitions and one item collection box and also saving around 10 seconds in-game time. Regarding the pause in the refill station, yeah, it was more habbit, i only realized i should test it after the run was done. If i recall correctly you loose 1 second in-game time but saves 1 second realtime? But yeah, if i was going to redo the run or something like that i would not pause there.

Regarding kraid, yes. After i did the run i tested that trick found by Kriole and Taco and adapted for realtime by forensics and realized it was easier than i first thought. So it's another thing i would change in the run. But as you said it wouldn't save that much anyway, but it's something i would go for in a run.

Yeah, Draygon can be killed a bit faster, but again something i went for in this run was Consistency, and to make all PW combo to hit Draygon is too hard and the idea of Shinesparking Draygon to death in an any% is more about saving ammo for later but i've been thinking about another strategy that would allow me to use the supers i have on him. The problem here is stopping in that bugs after botwoon is not a good strategy imo, you loose too much in-game time (realtime too) and you can't count on Botwoon to drop enough supers and you already need two for the door in the mocktroids room and another one for the eyed door entering Draygons's lair. so i think the best strategy is to shinespark Draygon to death, saving the supers for Ridley or later. Regarding the shinespark back i don't think it's worth in a run. Iirc it saves around 2 seconds in-game time, but considering you managed to shinespark all the way before Draygon, you'll be already low in health for a second shinespark back and if you intend to get Draygon's drops you'll waste so much time that the shinespark back will end up beeing a slower strategy. Plus if you don't collect Draygon's drops it can be a problem for Lower Norfair and the shinespark back will probably drain all your health, so you'll have to stop to refill before Ridley and that will be slower again. So the blue suit would only be worth (if it is at all) in a 100% run.

After we talked in twitch tv while garisson was speedrunning i tested it again and realized i forgot to measure the realtime losses by lag, so yeah, but again we come to consistency and health management. The shinespark before Draygon costs a lot and the others on him costs a good chunk of health too, so you already have low health after Draygon. Stopping to refill wouldn't be a good choice and you'll loose some health in Lower Norfair already, first from the damage boost in the head in the wall before the LN elevator and then, if you have luck, in the double kago room, then you loose some more diving in the lava after the Space Pirates and then you have to Fight Ridley and end the fight with enough health to survive leaving Lower Norfair (iirc Garrison died a lot in the red ki-hunters in the way out of LN). Also, there's the problem with consistency charging the shinespark without falling down, which would cost some time and make the shinespark option almost equally fast as the jump. I saw Garrison pressing down frames later than he was supposed to and not getting the shinespark but instead he would see Samus falling with a blue suit. I really think he would be more consistent with a jump there than charging the shinespark. If he fails to charge the shinespark he will probably loose the run given the time to destroy all pillars and the lag generated by the power bombs and the lava rising, not counting he could get hit and loose health too.

I don't know if i tested it right or not, but i remember it beeing just some frames faster so i think i forgot about it due to the fact it almost doesn't save time.

Also, thanks for watching the run and i apreciate your comments.

Reeve

2012-06-23 01:04:46 am

Btw, Garrison, iirc you got a 49:08 in your attempts but you got a 00:33. I'm curious to see which realtime number will hit 00:32, so it will be possible to know for sure around how much time the non-spazer route saves and how much you would have to shave to get a 31min run. If your calculations are right and 48:05 realtime would be a 31:59, a number around 49:05 would be a 32:59. Good luck with your attemps and i look for the moment you'll get a 00:32.

Gari

2012-06-23 01:57:24 am

Thank you Reeve, I think a :32 with this route will be a lot more satisfying than my :32 using gametime strategies. Your posts have been very helpful :)

MSDS

2012-06-24 06:34:17 pm

Thanks for posting this Reeve. As I'm sure Garrison will agree, realtime runs and gametime runs should really be two separate categories for any% at this point. The strategies for each are so vastly different that they truly change the composition and appearance of the run.

That being said, I personally think the gametime of a run that's optimized for realtime is mostly irrelevant. I realize it's going to be difficult for anyone who's been involved in the speedrunning scene for this game for the past 10+ years to accept that the gametime is irrelevant in any situation. But the runs are so becoming so divergent that they just shouldn't even be compared to each other, much like the way PAL and NTSC runs should not be compared.

Keep in mind my opinion is a bit biased. As a speedrunner of this game, I myself am currently more interested in pushing the boundaries of realtime for any%. The main reason is that there's more healthy competition in a realtime category because you have access to the seconds of the final time. As it stands now, my personal best realtime is 49:24. And while I'm not quite on Garrison's level, knowing that I'm only ~16 seconds away from his WR is encouragement for us both to improve our times. In the gametime category, our runs could be as much as 59 seconds apart, but they would still be considered TWR (tied world records). If there was a way for gametime seconds to be accessed on console that would be ideal, but we all know that's probably never gonna be possible.

Again, my ultimate goal here is to get people into the mindset of separating the two categories. I certainly don't want to take anything away from gametime runs, because I think the skill set involved is entirely different and has merit in it's own right. If you watch hotarubi's and/or behemoth's :32, for instance, they make alot of pauses in the first 20 minutes up to powerbombs that allow them to make more precision jumps/movements that are not done in a real time run. These tricks are not easy to pull off by any stretch and deserve their place in the game.

So I think the two runs can co-exist and not be compared to each other. Super Metroid is IMO the greatest game of all time, not only for casual play, but ESPECIALLY for speedrunning. The Deer Force development team (R&D1) spent countless hours creating a masterpiece that is so intricate and diverse, so it could be played and enjoyed on so many levels of gameplay. Not only does it make sense to have multiple categories for speedrunning to match that diversity, I think it's our responsibility as gamers to fulfill that duty.

Anyway, that's just my humble opinion on the matter, thanks for reading :)

Thanks for posting this Reeve.  As I'm sure Garrison will agree, realtime runs and gametime runs should really be two separate categories for any% at this point.  The strategies for each are so vastly different that they truly change the composition and appearance of the run.

That being said, I personally think the gametime of a run that's optimized for realtime is mostly irrelevant.  I realize it's going to be difficult for anyone who's been involved in the speedrunning scene for this game for the past 10+ years to accept that the gametime is irrelevant in any situation.  But the runs are so becoming so divergent that they just shouldn't even be compared to each other, much like the way PAL and NTSC runs should not be compared.

Keep in mind my opinion is a bit biased. As a speedrunner of this game, I myself am currently more interested in pushing the boundaries of realtime for any%.  The main reason is that there's more healthy competition in a realtime category because you have access to the seconds of the final time.  As it stands now, my personal best realtime is 49:24.  And while I'm not quite on Garrison's level, knowing that I'm only ~16 seconds away from his WR is encouragement for us both to improve our times.  In the gametime category, our runs could be as much as 59 seconds apart, but they would still be considered TWR (tied world records).  If there was a way for gametime seconds to be accessed on console that would be ideal, but we all know that's probably never gonna be possible.

Again, my ultimate goal here is to get people into the mindset of separating the two categories.  I certainly don't want to take anything away from gametime runs, because I think the skill set involved is entirely different and has merit in it's own right.  If you watch hotarubi's and/or behemoth's :32, for instance, they make alot of pauses in the first 20 minutes up to powerbombs that allow them to make more precision jumps/movements that are not done in a real time run.  These tricks are not easy to pull off by any stretch and deserve their place in the game.

So I think the two runs can co-exist and not be compared to each other.  Super Metroid is IMO the greatest game of all time, not only for casual play, but ESPECIALLY for speedrunning.  The Deer Force development team (R&D1) spent countless hours creating a masterpiece that is so intricate and diverse, so it could be played and enjoyed on so many levels of gameplay. Not only does it make sense to have multiple categories for speedrunning to match that diversity, I think it's our responsibility as gamers to fulfill that duty.

Anyway, that's just my humble opinion on the matter, thanks for reading :)

Reeve

2012-06-24 11:06:06 pm

Hello MSDS and thanks for posting your opinion.

I'll post my personal opinion as well: It's no secret that i'm more into in-game time for any cathegory (any%, 100%, low%, RBO, etc...) and the reason is the in-game time is where you see the player's skills. You can't see it in pause screens or door transitions or message boxes. The idea of doing a realtime any% run is i realized that it's actually the same route but skipping a part of the game which is spazer. IIrc, despite Murder Beam, there's nothing you can't do without Spazer. Most people agree that Murder Beam is not something pretty to see, so as hotarubi did, you can use the spazer route and choose not to use Murder Beam in MB. Now if you consider both routes not using Murder Beam, the Non-Spazer route is actually faster in-game time too. So skipping Spazer is not only a realtime route, It's just a coincidence that you can achieve the best time for both in-game and realtime. Another thing about the route i took in the smv is you skip the Energy Tank after Ridley, which saves around 3 seconds in-game time and coincidently skips two door transitions and one item. Also, skipping the Big Metroid with more than 3 energy tanks will allow you to skip the Recharge Station which again will save around 5 seconds in-game time and coincidently skip two door transitions and one recharge box message. So, you can see that all these strategies are not only realtime strats but in-game time too. Of course this route will need a higher level of precision from the player, but it's nothing that a player in the level of hotarubi would have problems with. So, honestly, for any% runs, i don't think there's the need for two cathegories.

Hello MSDS and thanks for posting your opinion.

I'll post my personal opinion as well: It's no secret that i'm more into in-game time for any cathegory (any%, 100%, low%, RBO, etc...) and the reason is the in-game time is where you see the player's skills. You can't see it in pause screens or door transitions or message boxes. The idea of doing a realtime any% run is i realized that it's actually the same route but skipping a part of the game which is spazer. IIrc, despite Murder Beam, there's nothing you can't do without Spazer. Most people agree that Murder Beam is not something pretty to see, so as hotarubi did, you can use the spazer route and choose not to use Murder Beam in MB. Now if you consider both routes not using Murder Beam, the Non-Spazer route is actually faster in-game time too. So skipping Spazer is not only a realtime route, It's just a coincidence that you can achieve the best time for both in-game and realtime. Another thing about the route i took in the smv is you skip the Energy Tank after Ridley, which saves around 3 seconds in-game time and coincidently skips two door transitions and one item. Also, skipping the Big Metroid with more than 3 energy tanks will allow you to skip the Recharge Station which again will save around 5 seconds in-game time and coincidently skip two door transitions and one recharge box message. So, you can see that all these strategies are not only realtime strats but in-game time too. Of course this route will need a higher level of precision from the player, but it's nothing that a player in the level of hotarubi would have problems with. So, honestly, for any% runs, i don't think there's the need for two cathegories.

Smokey

2012-08-05 08:21:33 pm

*Claps* impressive work, Reeve. Did I just witness the current room-by-room strat? I really liked WS escape method.

Also, I'm really liking the idea of the No Spazer route taking precedence over the traditional route. One thing though that doesn't fully convince me is the fact to ignore that ET after Ridley in an unassisted run. I'm probably too tired to review it now to see if it is the optimal choice, but the overall route is very well done. Probably old news, but this is the first time I saw the route in action. Good work.

Reeve

2012-08-06 10:16:28 am

Wow, it's really great to see you around smokey and it's great to know you still have interest in the game, if not for speedrunning, for watching runs and maybe playing around.

Yeah, the strategies i used in the run, but there's a few timesavers i should have used there which can save some more seconds, not much really, but it can help a bit.

Regarding the Energy Tank skip, i asure you it's something you can go for. The biggest problem would be the red ki-hunters in the lava rising room when leaving LN, but considering we use some strategies before Ridley that wastes less health and considering you have luck with his drops for health, you will be ok. Garrison has been using this route for some time and he managed to get a 32min in-game time run. The difference between the route Garrison uses and the one in the smv is that he doesn't use the refill station before Maridia, but he has to stop to refill in the bug pipes after Botwoon. And then he uses the Refill Station after the Big Metroid as scaping him is tricky.

Wow, it's really great to see you around smokey and it's great to know you still have interest in the game, if not for speedrunning, for watching runs and maybe playing around.

Yeah, the strategies i used in the run, but there's a few timesavers i should have used there which can save some more seconds, not much really, but it can help a bit.

Regarding the Energy Tank skip, i asure you it's something you can go for. The biggest problem would be the red ki-hunters in the lava rising room when leaving LN, but considering we use some strategies before Ridley that wastes less health and considering you have luck with his drops for health, you will be ok. Garrison has been using this route for some time and he managed to get a 32min in-game time run. The difference between the route Garrison uses and the one in the smv is that he doesn't use the refill station before Maridia, but he has to stop to refill in the bug pipes after Botwoon. And then he uses the Refill Station after the Big Metroid as scaping him is tricky.

Gari

2012-08-07 03:59:30 am

Actually, the strategies have changed a little bit in maridia for our realtime runs. I wanted to find a less ammo demanding strategy for draygon and accidently found a variation of the two round w/ bluesuit that's both consistent enough and easy enough to do in a single segment run. Due to swapping out the 8 or so super missiles I used to farm before botwoon for an extra powerbomb, I'm able to skip the bug refill in favor of refilling after the wrecked ship. I haven't timed just how much is potentially saved with this, but I'd estimate somewhere around 15-20 seconds of realtime if things go optimally. Of course, with two more room transitions and an extra text box, it's also more rewarding gametime. I don't have an up to date video highlighted of the new strategy nor do I have an smv, but if you substitute the super missile I used in this video with a second pb combo, you can probably get the gist of it: http://www.twitch.tv/garrison_tt/b/325711902 (sorry for the chatter, it was during a race).

As for skipping the ridley tank, I've been practicing low% for a while now and I feel pretty comfortable escaping lower norfair with much less health than I used to. My survival rate went from about 50/50 to 90/10.

Actually, the strategies have changed a little bit in maridia for our realtime runs. I wanted to find a less ammo demanding strategy for draygon and accidently found a variation of the two round w/ bluesuit that's both consistent enough and easy enough to do in a single segment run. Due to swapping out the 8 or so super missiles I used to farm before botwoon for an extra powerbomb, I'm able to skip the bug refill in favor of refilling after the wrecked ship. I haven't timed just how much is potentially saved with this, but I'd estimate somewhere around 15-20 seconds of realtime if things go optimally. Of course, with two more room transitions and an extra text box, it's also more rewarding gametime. I don't have an up to date video highlighted of the new strategy nor do I have an smv, but if you substitute the super missile I used in this video with a second pb combo, you can probably get the gist of it: http://www.twitch.tv/garrison_tt/b/325711902 (sorry for the chatter, it was during a race).

As for skipping the ridley tank, I've been practicing low% for a while now and I feel pretty comfortable escaping lower norfair with much less health than I used to. My survival rate went from about 50/50 to 90/10.

Smokey

2012-08-08 02:17:40 pm

Thanks mate, it's also great to see you and all of you mastering the game, especially mastering my favorite run category.

Yeah, I guess it really depends on how comfortable you are with your health-management after Ridley. And yeah, the problem is the animation of the red Ki-hunters at the time you encounter them. They usually have a nice habit to block the entrances, which forces you to risk taking damage while attempting the shots. Is there any workaround for this behavior?

Gari

2012-08-08 08:52:05 pm

The Ki Hunters, as far as I know anyways, cannot be manipulated in a fashion that would guarantee the problematic Ki Hunter to be in a good position. There's a strategy that I'm sort of dubious on in regards to its age in kraid's hideout on the way back that involves jumping through the door as opposed to running through in order to manipulate those particular Ki Hunters not to swoop down into the hole. We've tried to incorporate this same principle to the Lower Norfair escape to no avail. Maybe it could be done for those who aren't as comfortable hitboxing them with little health.

As for the shots themselves, they're pretty simple to time going up vertically, just as they are horizontal, they just require different timing and adjustments. The biggest problem lies when the top two gang up together forcing you to essentially hitbox both of them in succession, typically in the same jump. These first Ki Hunters are the real problem for me, as they're more difficult to avoid taking damage from than the Ki Hunters that usually deal the killing blow. For the most part, I feel comfortable with the escape as long as I have 200 energy leaving Ridley's Room. The way I figure it is that I give myself enough wiggleroom to run through the acid in the refill room, take a hit from a desgeega if I fail to hitbox them, take a hit from the first set of Ki Hunters and leave myself with just enough energy in case things go bad with the second set. Typically, only one of the three dangers actually hits me, but if I get sloppy, I want to be able to at least survive it.

Another thing I've been doing in terms of health and missile management is that if I have less than 200 energy leaving Ridley's Room, I kill one of the kung fu pirates to not only bolster my health a little bit more, but to also get whatever super missiles drop to ensure that I either have a faster tourian, or negate getting screwed over by metroid drop rng. The 25/10/5 ammo balance works well for speed, but I'm sure everyone knows that if you kill the last of the metroids with less than 7 super missiles, you're going to lose a lot of time due to having to open at least one door with missiles and then refill. It kind of sucks in terms of gametime when this happens, but in terms of RTA timing, it's a crippling blow that can add anywhere from 30 seconds to your run, to a full minute. So while I may waste time killing a pirate in lower norfair, I guess the idea is that I have more to work with in tourian in case my luck goes sour.

The Ki Hunters, as far as I know anyways, cannot be manipulated in a fashion that would guarantee the problematic Ki Hunter to be in a good position. There's a strategy that I'm sort of dubious on in regards to its age in kraid's hideout on the way back that involves jumping through the door as opposed to running through in order to manipulate those particular Ki Hunters not to swoop down into the hole. We've tried to incorporate this same principle to the Lower Norfair escape to no avail. Maybe it could be done for those who aren't as comfortable hitboxing them with little health.

As for the shots themselves, they're pretty simple to time going up vertically, just as they are horizontal, they just require different timing and adjustments. The biggest problem lies when the top two gang up together forcing you to essentially hitbox both of them in succession, typically in the same jump. These first Ki Hunters are the real problem for me, as they're more difficult to avoid taking damage from than the Ki Hunters that usually deal the killing blow. For the most part, I feel comfortable with the escape as long as I have 200 energy leaving Ridley's Room. The way I figure it is that I give myself enough wiggleroom to run through the acid in the refill room, take a hit from a desgeega if I fail to hitbox them, take a hit from the first set of Ki Hunters and leave myself with just enough energy in case things go bad with the second set. Typically, only one of the three dangers actually hits me, but if I get sloppy, I want to be able to at least survive it.

Another thing I've been doing in terms of health and missile management is that if I have less than 200 energy leaving Ridley's Room, I kill one of the kung fu pirates to not only bolster my health a little bit more, but to also get whatever super missiles drop to ensure that I either have a faster tourian, or negate getting screwed over by metroid drop rng. The 25/10/5 ammo balance works well for speed, but I'm sure everyone knows that if you kill the last of the metroids with less than 7 super missiles, you're going to lose a lot of time due to having to open at least one door with missiles and then refill. It kind of sucks in terms of gametime when this happens, but in terms of RTA timing, it's a crippling blow that can add anywhere from 30 seconds to your run, to a full minute. So while I may waste time killing a pirate in lower norfair, I guess the idea is that I have more to work with in tourian in case my luck goes sour.