Video of strat attempts:

After getting a 10:18, I've been trying to implement some advanced strats. The three things I address in this post (Ghost House, Dark Room, and cloud glitch) are three points in the 11 exit run where things often go wrong for me, but I don't know why. It's probably going to take a lot of effort to reply to this post completely, so any help is appreciated.

Ghost House: I've been trying to implement the strat where Mario jumps through the boo rings in the first and second room, but the position of the boo ring in the second room seems to be inconsistent. It feel like it has something to with how fast or slow I do the first room, but in the video above, I included what looked to me like nearly identical attempts but had different results in the second room. Any help or information on this?

Dark Room of Bowser's Castle: When I try to keep P-Speed in the dark room, I have to spin-jump on the second mecha-koopa in the section with the short ceilings. Sometimes I kill the koop and other times the koop kills me, as I show in the video above. I've practiced the jump hundreds of times and have tried jumping at many different points, but nothing seems to consistently work. Am I missing something?

Cloud glitch: I'll write out what I understand about getting the cloud to see if my struggle to obtain the cloud is just due to some misunderstanding of what I've read. After reading the SMW Cloud google doc, I understand that you have to get either a B4 or B5 address by jumping on the second to last koopa. If I get B5, I discharge the shell after I spawn Yoshi, and if I get B4, I discharge the shell before spawning Yoshi (I don't go for fast Yoshi). From there, I place the green shell at the top of the hill and let it get hit by the fire ball. I hold L, R, and X (and right) after jumping off and sticking Yoshi's tongue out to collect the coin. After collecting the coin, I should have the cloud, right? Is there some randomness or variability when you go for the cloud? In the video above, I included two attempts with what I believed to be B5 (one that worked and one that didn't), and two similar attempts with B4. Maybe I was wrong about the B4 and B5 in these attempts, but I've practiced enough to the point where I feel like I know what memory addresses I've gotten just by looking. Anyone have any tips to make this more consistent? One day I got the cloud 5 out of the first 9 attempts, and yesterday, I got it 2/27 attempts.

In all these cases, I find it frustrating to press all the right buttons at the right time (although maybe I'm not and that's my problem) and still have the trick not work. I probably used more words than was necessary in this post, so thanks in advance to anyone who bothers answering.

I’ve not researched it extensively by any means, so I may make myself look foolish, but I’m wondering what time saves are left in 11 Exit, and whether or not it would it be worth it to put a lot of time into the category.

I ask this mainly because I’m trying to decide what game to speedrun. I’d like it to be a 2D Mario game, which I realize are all quite optimized. In SMW, I’m currently working on consistency in getting the cloud and would consider putting significant time into RTA’s, but am not sure if I would ever get close to the record. What I’m mainly worried about would be investing my time only to be working towards a record that could be improved by a couple seconds or less.

Since I’m much more knowledgable about SMB1, I’ll use that as an example. There are already elite runners (darbian, Kosmicd12, andrewg1990, and somewes) at the top of that game, just as there are in SMW. They’ve put thousands of hours in and their records are extremely optimized. darbian himself has said that once any% reaches 4:55, it is essentially dead. This is one of the main reasons I’ve stayed away from running SMB1 myself, especially any%. It would be like running a marathon only to find a insurmountable wall just before the finish line. Sure, you ran 26.19 miles, but there’s something significant about the last .01.

Maybe one would say I have the wrong attitude about speed running if I’m only looking for records and good times. Maybe one would say I’m being unrealistic. I do genuinely enjoy running SMW, but I also see the value in running a game that even a beginner has the ability to develop if he invests enough time into it.

I know I got slightly off topic, but any advice on decision making would be appreciated, as well as any specific remarks on 11 exit.