I'm doing a smw 100% console speedrun. I was curious if there were any wips of smw 100% TASes that I could possibly view? I'm sure the majority of tricks used would not be possible on console, but I don't want to miss anything.
Is there a map route change from the current 100% smw TAS? The map route seems pretty optimal, but there seem to be quite a lot of options to choose.
Also, have there been any major level shortcuts in any of the non-any% TAS levels (with glitches and such) that I should know about?
Any extra smw info would be appreciated. I would like to know the fastest methods people have come up with since the old 100% smw TAS.
Super Mario Bros. console speedrunner
- Andrew Gardikis
There's this, which was posted here almost exactly 2 years ago:
100% WIP up to the Big Boo fight
Keep in mind this is between 2 and 3½ years old. As far as I know, ism has been working on a new, good version since then, but I haven't seen it.
Notice the different overworld route in the star worlds from the currently published run.
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
His other wips also can be downloaded in his blog. He restarted his run after the wip was released, and he's optimizing Gnarly. I'll maybe post his map route plan later.
Retired because of that deletion event.
Projects (WIP RIP): VIP3 all-exits "almost capeless yoshiless", VIP2 all-exits, TSRP2 "normal run"
Oh, ISM's WIP 100% is the greatest ever! How I love this run!
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Joined: 6/25/2006
Posts: 138
Location: Fort Collins, CO
bahamete wrote:
Very impressive run, the improvements are great :)
I was thinking while watching it, in VD1, could you split the yoshi coin to get a key/wings/balloon block, and then obtain a balloon to use the left+right trick to fly through the wall with the pipe? I would think it would be faster than going through the pipe.
I made a demo of VD1 a few years ago where I duplicated the block up through the ceiling to make a hole then key jumped across and eventually under a wall without going through the pipe. I cannot seem to find the file though (must've been on my old computer)
Thanks a bunch for all of the positive feedback everybody!
Very impressive run, the improvements are great :)
I was thinking while watching it, in VD1, could you split the yoshi coin to get a key/wings/balloon block, and then obtain a balloon to use the left+right trick to fly through the wall with the pipe? I would think it would be faster than going through the pipe.
I made a demo of VD1 a few years ago where I duplicated the block up through the ceiling to make a hole then key jumped across and eventually under a wall without going through the pipe. I cannot seem to find the file though (must've been on my old computer)
Thanks a bunch for all of the positive feedback everybody!
That must of been quite difficult to optimize? I'm guessing it worked out slower. And please, carry on the run :)
Thanks for the info' on fadeout lag and score.
In Super Demo World, with better score control, the first level can be improved by a frame. Level 2 should be improvable as well. For an any% run, should score always be controlled to minimise lag? I'd guess an obvious yes, but I'd like to be certain.
Edit:
I redid the first few levels, taking score into account. Up to the start of level 4, I managed to improve jimsfriend's WIP by 6 frames (if his WIP were to finish level 3).
smv
If nobody else is already working on this, I may consider giving the run a serious attempt.
An any% run should be as short as possible (duh), which includes minimizing frames lost to fadeout lag (and regular lag). Ideally, a 100% run should also aim for this, but, in the past anyway, it has often been kind of ignored because it's a tough thing to optimize over 96t or 120 exits. Plus it was boring and stuff :)
Fadeout lag probably will not matter in the slightest to the average Joe, however. I can't say I agree that TASers should spend time on such stuff.
But as always, it's an opinion, and it's the one of the TASer that counts. There are no rules for this stuff, is there? So long as it beats the old run by a reasonable amount of time, with out without saved frames from fadeout lag, it's OK, so far as I know?
Generally-speaking TASers compare frames to find improvements, and in a game like SMW lag frames and fadeout frames will definitely need to be accounted for if you want to beat the previous TASer. I suspect probably the only time that most people here wouldn't have a problem with the TASer ignoring lag frames is when the TASer is brand-new and they're running a game that hasn't been submitted before. Once you have some experience under your belt it's more or less expected now that you start using memory watch, pay attention to the lag counter, etc.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Just posting this because it beats the 96-exit speed run by 30 in game seconds. I'm using a yoshi and I don't have a cape, but I think if someone does a new run, this might be useful.
http://www.youtube.com/watch?v=2hZFM_9fqsM
Don't comment on the video, comment on the forum.
I also found a new speed trick while doing this if its helpful to anyone.
Since I needed to maintain sprinting speed, I couldn't do the hopping glitch. Instead, while in air, I let go 2 frames before hitting subpixel 49, held it like that for 5 frames, then I started moving again to keep my running speed for 5 frames, then repeat.
What I'm trying to say is I let go of > for 5 frames, pressed > for 5 frames, and repeat.
This changes the running pattern to 49(x6), 48, 47, 48, 47, repeat. I might be slightly off with that. If you let go for 6 frames, mario's speed will fall to 46 for one frame, and it'll take longer to get it back to 49.
I cannot find the microstorage because I haven't had to use in it months, but if someone could post a link that would be great.
Joined: 10/28/2007
Posts: 1360
Location: The dark horror in the back of your mind
Merged post from Tool-assisted laboratory.
bullethead2202, please post game-specific info to the thread for that game.
Also, Microstorage is at http://dehacked.2y.net/microstorage.php
I also found a new speed trick while doing this if its helpful to anyone.
Since I needed to maintain sprinting speed, I couldn't do the hopping glitch. Instead, while in air, I let go 2 frames before hitting subpixel 49, held it like that for 5 frames, then I started moving again to keep my running speed for 5 frames, then repeat.
What I'm trying to say is I let go of > for 5 frames, pressed > for 5 frames, and repeat.
This changes the running pattern to 49(x6), 48, 47, 48, 47, repeat. I might be slightly off with that. If you let go for 6 frames, mario's speed will fall to 46 for one frame, and it'll take longer to get it back to 49.
I cannot find the microstorage because I haven't had to use in it months, but if someone could post a link that would be great.
This concept has been well analysed and used for a long time now. Releasing > for 6 frames does not make Mario's speed fall to 46 when > is pressed again, but releasing it for 7 frames does. The optimal input is to keep releasing > for 6 frames on 49 and then holding > for 5 frames. Its common name is '6/5'.
Joined: 12/28/2007
Posts: 235
Location: Japan, Sapporo
I forgot to note pipe entrance information in smw any% run's text, and I'd share it here.
When you enter a pipe from the left, you can stick yourself into the pipe before entering it. If you can get yourself stuck in the pipe by n pixels, you'll get map transition 2n frames faster than usual. Then by what amount of pixels can you stick yourself into a pipe (or wall, etc)? Saying from the conclusion, you can get 3 pixels with speed 51 or 49, and 2 pixels with speed 37 or 33 (in water level, with item). But you must be landing on the ground 1 frame before you enter the pipe, so you end up losing your speed and the pixels by which you can stick yourself into the pipe before entering it are:
3 pixels with speed 51
2 pixels with speed 49 or 37
1 pixel with speed 31-33 (in water level, with item)
Here are samples with speed 51:
http://dehacked.2y.net/microstorage.php/info/1994256128/smw-pipe1.smvhttp://dehacked.2y.net/microstorage.php/info/286646643/smw-pipe2.smv
The trick is used in the former but not in the latter. In pipe1, I sacrificed 1 frame to stick mario into the pipe by 3 pixels (instead of 2 pixels), and got 1 frame as a result.
This technique should be used in YI4 of any% run, but a corner-boost at the first block made us end up being stuck into the pipe by three pixels so we didn't have to do the trick. I think this is why I dropped this infomation in the text of the current any% run. :p
Phallosvogel:
bahamete wrote:
