Super Mario World "low%" done: small86 begins

2006-04-07 03:59:39 am

Like A Fox

Quote:

So Jimsfriend, what's the status? Are you able to record yuor runs now, and if so, what kind of progress are you making? Are you still doing daily runthroughs of your path? How's the successfulness of the two main tricks?

Sorry about all the questions, but I'd love to know what's going on...

Well, I've been very busy with school stuff recently so I couldn't work on this much. I wasn't home much last week, and when I was the TV wasn't so I couldn't play. Now I have everything back again and there shouldn't be too much school conflict for the next week or two.

Yes I can record runs, but I haven't had a playthrough that was worth keeping. I am still thinking too much about the tape being in, so I die. When I'm not doing that I'm thinking about who knows what... usually things that might help Fabian in some other level so I die. It is hard for me to keep my concentration for a whole 15 minutes as bad as that sounds.

Yesterday I played a bunch and got all the kinks of not playing for a while out, so I can beat the levels again. Today I spent 3 hours trying to make an improvement for DP3 which should be able to put the level completion time at 262. For those of you who don't know, that is the time the current TAS had for that level, and he had a cape and afaik he had knowledge of Mario's oscillating speed. I know it works, I just don't know for sure the exact time completion would turn out to be.

Unfortunately if I ever get it to work I won't be putting it in my run. Too hard. Too much luck involved. Things move faster than my mind can process so I basically have to blindly press the buttons and hope for a good result. After 34 deaths and about 70 attempts that didn't end in death (start+select) I decided to stop. I came too close to breaking my controller. So basically I wasted half an afternoon unlearning DP3 and relearning it a new way, but never completing it the new way. Tomorrow I will probably try some more, maybe I will get lucky and have several successes. Next Monday I will back full swing into recording run attempts, most likely with my old 258 strategy for DP3. I hate not being perfect.

[quote]So Jimsfriend, what's the status? Are you able to record yuor runs now, and if so, what kind of progress are you making?  Are you still doing daily runthroughs of your path?  How's the successfulness of the two main tricks?

Sorry about all the questions, but I'd love to know what's going on...[/quote]
Well, I've been very busy with school stuff recently so I couldn't work on this much. I wasn't home much last week, and when I was the TV wasn't so I couldn't play. Now I have everything back again and there shouldn't be too much school conflict for the next week or two.

Yes I can record runs, but I haven't had a playthrough that was worth keeping. I am still thinking too much about the tape being in, so I die. When I'm not doing that I'm thinking about who knows what... usually things that might help Fabian in some other level so I die. It is hard for me to keep my concentration for a whole 15 minutes as bad as that sounds.

Yesterday I played a bunch and got all the kinks of not playing for a while out, so I can beat the levels again. Today I spent 3 hours trying to make an improvement for DP3 which should be able to put the level completion time at 262. For those of you who don't know, that is the time the current TAS had for that level, and he had a cape and afaik he had knowledge of Mario's oscillating speed. I know it works, I just don't know for sure the exact time completion would turn out to be.

Unfortunately if I ever get it to work I won't be putting it in my run. Too hard. Too much luck involved. Things move faster than my mind can process so I basically have to blindly press the buttons and hope for a good result. After 34 deaths and about 70 attempts that didn't end in death (start+select) I decided to stop. I came too close to breaking my controller. So basically I wasted half an afternoon unlearning DP3 and relearning it a new way, but never completing it the new way. Tomorrow I will probably try some more, maybe I will get lucky and have several successes. Next Monday I will back full swing into recording run attempts, most likely with my old 258 strategy for DP3. I hate not being perfect.

Fabian

2006-04-07 10:06:04 am

I love YaBB 1G - SP1!

It's good to see you're working on this. Sooner or later you'll get it.

I just TASed DP3 for fun as small mario. I got a time of 261, with only 13 frames margin from getting 260. It looks extremely difficult to pull it off on a console run though. If anyone's interested in watching it for fun I could link you to the emulator movie.

Also, in my new 96 exit TAS the completion time for this level is 263. VIPer7 was very close to getting it too, I'm 9 frames faster than him in that level, which was enough to get the better time.

jimsfriend

2006-04-07 09:21:08 pm

Like A Fox

Just want to make sure you use this strategy to get the time, or maybe it was another. Either way I just wanted to know what the best possible time would be, I will stick with the 258 strategy because I can do that every time, and the jumps required for the route in the picture are too difficult to be pulled off consistently on console. For me anyway.

Note* run on the moving platforms when it benefits you as long as possible. And Fabian, how many frames are in a game time unit? Is one unit actually 1 second?

Fabian

Edit history:

Fabian: 2006-04-07 11:53:59 pm

2006-04-07 11:52:38 pm

I love YaBB 1G - SP1!

I ran on moving platforms whenever it saved time. My strategy differed slightly from yours though. I did not stop to pick up a shell at the beginning of the level, instead I slowed down slightly before the tricky part with the ON/OFF switches. What I did was to make a long jump from the moving platform and hit the koopa right before the first ON/OFF switch. I then turned back for a few frames so I picked up the empty shell. This was just a side effect though, since the point of the waiting was to make the platform from which I jump to the second ON/OFF switch move up slightly. This meant I could jump to the second ON/OFF switch directly, without a need for a shell, so I threw that away soon after.

This required very precise control of the B-button though, something like hold B 10 frames, then release it for a few frames and then hold it again (holding it longer would mean bumping in to the first ON/OFF switch) without much room for error. Basically not doable on console.

I didn't use the hopping glitch or anything like that btw, since the point was to see how fast it could be done "regularly"

About the time units. 60 frames per second. 41 (or possibly 42, can't remember) frames per time unit. So a time unit is basically 0.7 seconds

I ran on moving platforms whenever it saved time. My strategy differed slightly from yours though. I did not stop to pick up a shell at the beginning of the level, instead I slowed down slightly before the tricky part with the ON/OFF switches. What I did was to make a long jump from the moving platform and hit the koopa right before the first ON/OFF switch. I then turned back for a few frames so I picked up the empty shell. This was just a side effect though, since the point of the waiting was to make the platform from which I jump to the second ON/OFF switch move up slightly. This meant I could jump to the second ON/OFF switch directly, without a need for a shell, so I threw that away soon after.

This required very precise control of the B-button though, something like hold B 10 frames, then release it for a few frames and then hold it again (holding it longer would mean bumping in to the first ON/OFF switch) without much room for error. Basically not doable on console.

I didn't use the hopping glitch or anything like that btw, since the point was to see how fast it could be done "regularly"

About the time units. 60 frames per second. 41 (or possibly 42, can't remember) frames per time unit. So a time unit is basically 0.7 seconds

jimsfriend

Edit history:

jimsfriend: 2006-04-08 03:54:56 am

2006-04-08 12:01:17 am

Like A Fox

Yeah after much frustration last night, and then spending 2 hours undoing habits learned last night today, I definitely won't be trying that. I believe your strategy would be slightly faster, but only because p speed can't be immediately regained when picking up the blue shell; you have to fall first. I guess it all depends on how much delay was required to allow your super long jump. But this is unimportant because I will simply use the 258. Much better.

EDIT: The hop glitch can be done "regularly". I just won't be attempting it that late in the run. On console it is only reasonable to hope for one or two hops, and the only place I try it is YI3. I know if it works when I get 14 stars for cutting the tape instead of 15.

thegreginator

2006-04-08 08:17:40 pm

Hey jimsfriend I've been following this topic and let me first just say awesome progress and I'm really excited to see this finished run.

From your pictures, I've noticed that DP3 could be improved by a little bit. In a few cases, you land on a rotating gray platform while its moving backwards, which slows you down. An example of this is on the set of 3 gray platforms right after pipe A (the 6th one total). If you landed on the top gray platform instead of the lower one, you could be rid of the slow down of landing on the backward-moving bottom platform and actually gain speed if you stay running on the top gray one. To do this, simply jump off that 2nd to last flying koopa and land on the yellow pipe, and then when you jump off the yellow pipe you'll be on the top, forward-moving platform. The other times you land on a backward moving gray platform (besides the 6th set) are on the 3rd set of 3, the 5th set of 3, and the 7th set of 3 (when a platform is moving counterclockwise you want to land on the bottom). If its not possible to land on the bottom of these it would still be faster than the current method if you land on the little block that the platforms stem from. I dont know how possible this is in these cases, and its not really that big of an improvement, but its worth a look (it made the difference between 258 and 259 for me on YI3 on console).

Keep up the good work!

Hey jimsfriend I've been following this topic and let me first just say awesome progress and I'm really excited to see this finished run.

From your pictures, I've noticed that DP3 could be improved by a little bit.  In a few cases, you land on a rotating gray platform while its moving backwards, which slows you down.  An example of this is on the set of 3 gray platforms right after pipe A (the 6th one total).  If you landed on the top gray platform instead of the lower one, you could be rid of the slow down of landing on the backward-moving bottom platform and actually gain speed if you stay running on the top gray one.  To do this, simply jump off that 2nd to last flying koopa and land on the yellow pipe, and then when you jump off the yellow pipe you'll be on the top, forward-moving platform.  The other times you land on a backward moving gray platform (besides the 6th set) are on the 3rd set of 3, the 5th set of 3, and the 7th set of 3 (when a platform is moving counterclockwise you want to land on the bottom).  If its not possible to land on the bottom of these it would still be faster than the current method if you land on the little block that the platforms stem from.  I dont know how possible this is in these cases, and its not really that big of an improvement, but its worth a look (it made the difference between 258 and 259 for me on YI3 on console).

Keep up the good work!

jimsfriend

2006-04-08 08:47:10 pm

Like A Fox

The difference between 258 and 259 on this level is the difference between a super easy level and a challenging level. Thanks for the advice, but I would rather just lose one time unit and keep it simple.

I can tell you that whenever I land on a platform that would make me go slower because it is moving backwards I get off of it as fast as I can. And I prefer to keep my jumps short because they are easier to control that way, so I will be landing on more of those moving platforms than is absolutely necessary.

As for landing on the gray piece in the middle... it's a lot easier to land on a platform of length 3 than to land on a spot of length 1. In levels where super hard tricks are not required for completion (such as DP3) I like to avoid doing things that are too hard so I have a better chance of beating the levels where I can control the difficulty.

Psonar

2006-04-14 02:15:04 pm

Long live SF64!

Good call Jimsfriend....I agree that a darned good completion should be done before an almost perfect, just because you may never finish a run if you didn't make some simplifying decisions like that.

I was wondering, if you have time, if you could say what has been killing your runs, or what you run practices are. Do you always go all the way through, do you restart at each major mistake, or is it the big trick levels that hold you up?

Of the tricks that tend to kill the run what are the chances of you getting the trick after a long series of attempts. 1 in 10? 1 in 100? Good luck man!

I loved seeing the map of the whole level with your path...I think it would be neat if only for this forum to have a collection of those for each level, but I know that's a lot of extra work and I don't want you to do it if you don't have time...plus it'd probably just stir up a bunch of people saying, "why not go that way instead?!"

jimsfriend

2006-04-14 05:21:09 pm

Like A Fox

Considering the difference between "darned good" and "almost perfect" will probably be no more than 10 seconds for the entire run (barring any major finds, and TAS) I will never try to get better than 258 in that level for an actual run.

If my run gets off Yoshi's Island (I make silly mistakes at the beginning often, and while still on YI I will reset and start over) then I will let any non-major mistakes fly. If I mess up the DP1 trick (1/5 success about) I reset because there is no way to quickly try again. The next level that ruins runs (death) is DP4 when I miss a very narrow window for any of three jumps, but my success rate for the level is still about 2/3. Then VD1 I have a success rate of about 1/2, such a pain to do; if I screw up I have one alternative method I will try and if that succeeds I lose about 5 seconds overall, if it fails I reset because too much time is lost (forced death). Then SW4, the bane of any run that arrives, success rate of 1/10 on a lucky day, usually 1/20 (this is when practicing). The difference between practicing and doing it on an actual run is that in practice I can play the same level 100 times in a row and my muscles learn from previous mistakes, in a run I get one try after thinking about other levels for 15 minutes. I practice Bowser a whole lot so if I ever get there I won't die. Any death will be a reset because too much time is lost for such a short run.

I already have the map for most of the levels drawn up, but I won't post them here because it would be a major spoiler. They are pretty much optimized for console play, and if you really want to see them I could pm them to you, just pm me.

Considering the difference between "darned good" and "almost perfect" will probably be no more than 10 seconds for the entire run (barring any major finds, and TAS) I will never try to get better than 258 in that level for an actual run.

If my run gets off Yoshi's Island (I make silly mistakes at the beginning often, and while still on YI I will reset and start over) then I will let any non-major mistakes fly. If I mess up the DP1 trick (1/5 success about) I reset because there is no way to quickly try again. The next level that ruins runs (death) is DP4 when I miss a very narrow window for any of three jumps, but my success rate for the level is still about 2/3. Then VD1 I have a success rate of about 1/2, such a pain to do; if I screw up I have one alternative method I will try and if that succeeds I lose about 5 seconds overall, if it fails I reset because too much time is lost (forced death). Then SW4, the bane of any run that arrives, success rate of 1/10 on a lucky day, usually 1/20 (this is when practicing). The difference between practicing and doing it on an actual run is that in practice I can play the same level 100 times in a row and my muscles learn from previous mistakes, in a run I get one try after thinking about other levels for 15 minutes. I practice Bowser a whole lot so if I ever get there I won't die. Any death will be a reset because too much time is lost for such a short run.
 
I already have the map for most of the levels drawn up, but I won't post them here because it would be a major spoiler. They are pretty much optimized for console play, and if you [i]really[/i] want to see them I could pm them to you, just pm me.

Psonar

2006-04-14 05:54:01 pm

Long live SF64!

Sounds cool...
This thing's going to be monumental once you're done with it. Discounting other chances of death/retry as negligible, I multiplied out what you said and got 1/300 as the chance of a successful run. Of course you aren't a random number generator, and I think you'll defy all odds whenever you actually do it, because mood, clarity of mind and memory of exactly what you need to do all go into it, so on some random good day you could just bust out with it...
I play a lot of Star Fox 64, and there's levels where I've done all the necessary things to get a top score, but I just can't get them to all coincide in the same run. I know what it's like.
Oh and good thinking about not posting spoilers. I await the shock and awe of the real deal!

trihex

2006-04-23 06:54:16 am

Yoshi's eggs are at my mercy!

jimsfriend, you are inspiring with your determination. I cannot wait until we get to see a finished product! Cheesy

Not to ... sound ... stupid, but you speedrunning the prequel, and me doing the sequel kind of makes us speedrunning brothers!

...

Er, it sounded better in my head, anyway, remember I'm still here anxiousily awaiting!

ReTrOgAmEr

2006-04-23 07:38:40 pm

Take This! Ultra Frisbee-Disk Attack!

That's good news to hear jimsfriend, I definitely can't wait (as most other people can't either) to see this awesome run and wait for more news from you each time you have progress. I hope the recording you eventually do will be the best possible time that you could do. Best of luck.

jimsfriend

2006-04-29 01:06:59 am

Like A Fox

Quote:

jimsfriend, you are inspiring with your determination

Cool

It's been 2 weeks since I said anything so here:
I haven't had success. I don't really make many recording attempts every day any more, only about an hour a day although I will also go through the levels individually and play them a couple of times to make sure I get every level every day. This next week and a half I probably won't be able to play much at all, I have to study. After that school is (basically) over so it will be Mario only. Hopefully I finish before I graduate, but at the rate things are going I won't get my hopes up. I'll just keep trying forever, but don't worry, I have already passed the insanity threshold so that can't break me. Sorry that all sounds pessimistic, I get that way when I'm exhausted.

DateHell

Edit history:

DateHell: 2006-04-29 07:46:31 am

2006-04-29 07:45:49 am

po0oq

Despite the many ppl awaiting your vid, I'd like to say/ask/beg you not to rush anything
keep trying until you are 100% satisfied, because if you submit a 10-sec-slowish-but-lets-show-it run, you won't ever work on an improvement Smiley

Psonar

2006-05-05 06:17:10 pm

Long live SF64!

That's the spirit, and no you don't sound pessimistic, you sound insa....no it's like you said you're PAST the insanity threshold.

I'm glad to here that you are a determined type of person. Just be patient with it and keep practicing and trying and it'll all fall together one day, and the more effort you put into it, the greater it becomes because we see the worth and the cost behind the endeavor, and part of the greatness of some speedruns is the sheer quality of willpower some people have to not stop until they've found their idea of the perfect run.

Study hard for school and treat that with your Mario-worthy determination ... good luck!

rhilty05

2006-05-27 08:32:22 am

I just wanted to see how the run was coming and I hope all is well. And on an inspiring note: I BELIEVE IN YOU!!!
~Peace

jimsfriend

Edit history:

jimsfriend: 2006-05-27 06:41:47 pm

2006-05-27 06:40:45 pm

Like A Fox

Summer is coming up. That means I will soon have a whole lot of time... sort of. Right now with the end of school and graduation crap I don't really have much time. Basically right now all I have enough time to do is not get worse, although I lost the ability to complete DP1, it shouldn't take more than 20 hours or so to get back, so that isn't too bad. I'm getting better at my SW4 so it is no longer luck that determines if I make it. I also decided to throw in another trick which saves 6-8 time units in VS1.

Unfortunately, I won't be able to play at all next week because this room will be off limits. This room happens to be where my SNES is. After that it is just a race against time to see if I can finish in under a year from starting. That and the crazy DVD requirement which would immediately disqualify me. I'm not sure how that works, maybe having such a simple animation game would exempt me, haven't read much about it.

Fulgore

2006-05-31 03:01:28 am

I will definately watch this run. The tricks that have been discussed are just awesome and I can't wait to see you pull them off. Good luck!

ReTrOgAmEr

2006-06-03 02:46:42 am

Take This! Ultra Frisbee-Disk Attack!

Hopefully you can get back to playing ASAP to continue and ultimately finish what is appearing to be a very good run in the making. I know about the delays and mishaps with running games.

thegreginator

2006-06-04 02:50:19 am

jimsfriend, you pull off some incredible tricks in this run. When you get it polished, it will be amazing. I eagerly await watching the video! I am considering working on a small mario run myself after I finish my 11-exit run. I'm sure it won't be as good as yours, but it's just for fun and I won't be submitting it. I was wondering if you could post your times (in timeunits left on the timer) for each level so I could compare.

Thanks and keep up the good work!

jimsfriend

2006-06-04 03:17:04 am

Like A Fox

I'll do this from memory. Left will be my best time ever, right will be my usual time. I'll probably explain some levels times more, we'll see. If there is only one time then best and usual are the same.

YI2-360
YI3-259 + a little something special
YI4-275 (I don't even attempt bouncing off the shell to avoid the water, too hard for too little time gained)
C1-240 239
DP1-353 352
DS1-245 244
DGH-282 280
DP3-258
DP4-This one is wierd, it depends on several precise jumps so 260 is the best and usually get it when successful, but sometimes only 259. Anything less is probably because I died.
C2-This one is wierd also. I think the wall scrolly room depends on the state of the game when I enter the room, but I would have to have that confirmed (the wall waits before coming out or something). I do the same thing every time and sometimes I get 327 entering the big red door and sometimes as much as 330. I don't know when the level ends because I start+select before the timer stops when I kill Morton in practice, and in run throughs I don't look at the timer after boss fights.
VD1-371 370 (this one is about 50/50)
VS1-281(I have only gotten that once, ever) 280 (time entering the green pipe. Chuck is random so I wait for him to throw the football. This time is using a trick not discussed anywhere on this site. The method discussed here gets a time of 274. Using my new method I will accept messing up once (it is late in the run) to get a time of 272 entering the pipe). Cutting the tape is anywhere from 265-275, so take that how you want.
SW2-261 Recently I haven't practiced this level much, so my time is 260, but it shouldn't take much to get back. Once I get back to playing 2 hours a day things should start to go smoothly again.
SW3-286 285 if nothing goes horribly wrong
SW4-252 makes me happy, I haven't succesfully completed this level in a run yet but will settle for anything above 230. Of course I would try to make a better run before submitting if that happened.
Bowser Castle-351 usually 347-348. I take doors 2 and 8 (7 is too random). I try to keep the Bowser fight interesting and I practice it enough that it won't kill me to make it interesting.

So there you have it.

I'll do this from memory.  Left will be my best time ever, right will be my usual time. I'll probably explain some levels times more, we'll see. If there is only one time then best and usual are the same.

YI2-360
YI3-259 + a little something special
YI4-275 (I don't even attempt bouncing off the shell to avoid the water, too hard for too little time gained)
C1-240      239
DP1-353    352
DS1-245    244
DGH-282    280
DP3-258
DP4-This one is wierd, it depends on several precise jumps so 260 is the best and usually get it when successful, but sometimes only 259. Anything less is probably because I died.
C2-This one is wierd also. I think the wall scrolly room depends on the state of the game when I enter the room, but I would have to have that confirmed (the wall waits before coming out or something). I do the same thing every time and sometimes I get 327 entering the big red door and sometimes as much as 330. I don't know when the level ends because I start+select before the timer stops when I kill Morton in practice, and in run throughs I don't look at the timer after boss fights.
VD1-371     370 (this one is about 50/50)
VS1-281(I have only gotten that once, ever)     280  (time entering the green pipe. Chuck is random so I wait for him to throw the football. This time is using a trick not discussed anywhere on this site. The method discussed here gets a time of 274. Using my new method I will accept messing up once (it is late in the run) to get a time of 272 entering the pipe). Cutting the tape is anywhere from 265-275, so take that how you want.
SW2-261     Recently I haven't practiced this level much, so my time is 260, but it shouldn't take much to get back. Once I get back to playing 2 hours a day things should start to go smoothly again.
SW3-286   285 if nothing goes horribly wrong
SW4-252 makes me happy, I haven't succesfully completed this level in a run yet but will settle for anything above 230. Of course I would try to make a better run before submitting if that happened.
Bowser Castle-351    usually 347-348. I take doors 2 and 8 (7 is too random). I try to keep the Bowser fight interesting and I practice it enough that it won't kill me to make it interesting.

So there you have it.

thegreginator

2006-06-04 03:31:54 am

Wow those are some incredible times! How do you manage to get a 261 on SW2? On my 11 run I've only been able to get a 260, and I'd like to be able to improve that if you could tell me how (I'm assuming cape vs. small doesn't affect this level).

jimsfriend

2006-06-04 04:05:47 am

Like A Fox

Hehe. I love that last sentence, because that is why you never assume. Small mario is one square tall (hopefully you understand what I mean by "square" I don't know how to better describe it) and cape mario is two squares tall. When you exit the pipe at the beginning the timer is already going and it takes cape mario longer to get out of the pipe because he is taller. Also it is easier to not hit a wall at the end of the level with small mario becaues he is smaller. Also you watched JXQ's tas? Try to grab the key like he did where you let go of yoshi just before hitting the key so you have max velocity the longest amount of time, and all the acceleration from the key is in the leftward direction. Good luck.

ReTrOgAmEr

2006-06-04 05:51:39 pm

Take This! Ultra Frisbee-Disk Attack!

Damn, jimsfriend, those are some impressive numbers. With stats like these, this promises to be a very exciting run to watch.

skamastaG

2006-06-15 04:26:29 am

I was really bored yesterday so I started playing Bowser's castle through each door. I thought about your times and I think taking door 5 is much faster than taking door 8. Even though you have to wait for the spikes in 5, it is still faster than running all the way to door 8. I got to the dark room around 367 I think. I was about a half of the way through door 8 and was already at 367.
Am I right about this or is there something I am forgetting?