(Link to video)
Submission Text Full Submission Page
In an attempt to make a more entertaining run in the "minimum buttons pressed" category, I ran SMB3 with this goal. I figured this was a better choice than SMB because it has autoscrollers and a more distinct path from the main run. After several attempts and improvements, I managed to complete the game in 235 presses.

Game Objectives

  • Emulator used: FCEUX 2.1.5
  • Attempts to complete the game with the least amount of button presses possible
  • Avoids wasting time
  • Aims to maximize score / kills without jeopardizing the above goals
  • Takes damage to save presses
  • Luck manipulation

The Run

Title: 1 press

I press start at the title screen, but only once the intro finishes playing. Otherwise, I'd have to press start an additional time to skip the sequence.

Title -> 1-1: 3 presses (4 total)

I have to press right, up, and then A to enter 1-1. Pressing up on the map, or any other directional button for that matter, while another is pressed does nothing.

1-1: 9 presses (13 total)

I held up throughout this level just to show that it doesn't affect gameplay. This would be useful if I had to go up in the map again after finishing the level, but that never happens in the warped run. This level was fairly trivial as I was able to boost off of a Paratroopa to the clouds, saving many jumps.

1-1 -> 1-2: 1 press (14 total)

I have to release B when exiting a level to the map as otherwise, the item menu would come up. I can keep holding right, though, to keep moving in that direction. Therefore, I just have to press A on the map and repress B in 1-2.

1-2: 9 presses (23 total)

I get the mushroom here as I'm able to run into a pipe afterwards. This seemingly wastes 1 or 2 presses, but that's better than my best strategy for getting it in 1-1.

1-2 -> 1-3: 1 press (24 total)

Same as 1-1 -> 1-2.

1-3: 12 presses (36 total)

I found an interesting trick when trying to collect the item in this stage. It seems that when you get a mushroom as opposed to a leaf, you can shift Mario's gravity enough to boost off of both of the Goomba's to the next section. Sadly, I need to get the leaf in this stage instead. I found this glitch while trying this.

1-3 -> Sub-fortress: 4 presses (40 total)

Left, down, left, A.

Sub-fortress: 15 presses (55 total)

Other than having to reduce my acceleration at the beginning to avoid getting hit by a light, this is fairly self-explanatory.

Sub-fortress -> Tank: 6 presses (61 total)

I keep holding B from the sub-fortress to prompt the item menu to come up and then select the whistle with A. When I arrive at the warp zone, I have to press B and A again. From there, I continue to hold B as it won't have any affect and I can use it in the tank. I also have continued holding right from the sub-fortress so that I can move to World 8 and select it with A. From there, I go down and right to the tank.

Tank: 21 presses (82 total)

Many of these autoscrollers employ the same strategy: default to running right, try to boost off of cannonballs when possible, stopping temporarily, and using the scrolling wall to get a slight speed increase. I tried to defeat as many enemies as I could within these confines for entertainment purposes. While fighting the Boomerang Bro, I have to press left to avoid the boomerang he throws. This is OK as I have go to the left side of the screen for the item to spawn on the right side of the screen, anyway.

Tank -> Ship: 0 presses (82 total)

I continue holding right from the tank to enter the ship level.

Ship: 17 presses (99 total)

This is probably the hardest auto-scroller to complete. Same strategy as the tank, but with some nice twists. See the subtitles for more specific notes. I only need 3 presses to defeat Boom Boom as if I run into the wall, he will move towards me.

Ship -> Airship: 7 presses (106 total)

I have to go up and press A to enter the pipe. Once inside, I go right, slide down the hill, and then jump into the pipe. Although I have to release the right button and press down to slide, this doesn't waste any presses because I need down to enter the pipe, and I don't need to go right on the map once I exit. From there, I hold left at the right time to cross the panels without the hands grabbing me and go up to the airship.

Airship: 26 presses (132 total)

Same as ship. This time, Boom Boom can fly, so I go left to jump on him again before he starts to fly. This doesn't waste any presses because I need to go left once I re-enter the map anyway.

Airship -> 8-1: 8 presses (140 total)

I go left and enter the pipe. This time, I can't slide down the hill because I need to continue right on the map. I continue holding B and select the star; this will be explained later. From there, I continue moving right, down, then select 8-1.

8-1: 14 presses (154 total)

Instead of taking damage from the Piranha Plants, I'm able to go unscathed thanks to the star. This makes me use 1 extra press (To select the star). In addition, I duck underneath a cannon instead of taking damage twice to simply walk underneath it, using an extra 1 press (I continue holding down from the map, so I merely have to release right). This is necessary because I want to remain as Raccoon Mario, for reasons which will be revealed shortly. After doing some breathtaking Bullet Bill boosts, I jump over the last Pirahna Plant and finish the level with a mushroom.

8-1 -> 8-2: 3 presses (157 total)

Left, down, A.

8-2: 5 presses (162 total)

I use the fairly slow but helpful shortcut underneath the quicksand to save many jumps. I'm able to enter the pipe there without pressing down again as I hold it from the map. Also, I don't start running until I get through the pipes as otherwise, I would run on top of the pipe.

8-2 -> Sub-fortress: 2 presses (164 total)

Continue holding right, up, A.

Sub-fortress: 28 presses (192 total)

I could get a leaf here, but it would require that I waste 2 presses instead of the 1 I used in 8-1 to keep it (If I didn't duck, then I wouldn't be able to get a leaf here anyway because I'd be small Mario). Why I actually need the Raccoon suit will be explained later. To enter the first door, I duck, jump into a block, boost off of a Dry Bones, move under the other block, and enter the door. I could also jump off of the Dry Bones to break the second block and jump again to the door instead of ducking, but that would be slower. The rest of the level is fairly straightforward except that I have to release right to avoid the light in the second to last room.

Sub-fortress -> Tank: 6 presses (198 total)

Right, up, left, pipe (Don't slide), right to the tank.

Tank: 16 presses (214 total)

Same as the airship. This time, however, I need to continue moving right when I re-enter the map, so I have to let Boom Boom fly around for a while in order to defeat him in the minimum 3 presses.

Tank -> Bowser's Castle: 1 press (215 total)

I continue moving right and select the castle.

Bowser's Castle: 20 presses (235 total)

I get to the door in 2 jumps. I would need to use 3 if I selected the star and wanted to kill the Thwomp with it. Instead, I fly to and walk through it. This saves an additional 3 presses as it takes 1 press to fly as opposed to 4 to wall-jump. Although I could have also simply boosted off of the Dry Bones to the Thwomp, I would still need to be Raccoon Mario as I need to be Big Mario to glitch through the wall. I fly instead simply for speed purposes. At the final falling platform, I'm able to run off of it to the next platform instead of jumping. This only works some of the times I've tried it, and I think it has something to do with the collision system or sub-pixels, but I haven't studied it thoroughly. From there, I enter Bowser's room. I stop running at the best frame to manipulate Bowser. Interestingly enough, you can won't take damage from Bowser if you stay still. This allows me to wait around until his final jump, at which point I jump to the door and enter it when it appears.

Lua Script

See this section of my first submission for details on the Lua script used to display the input in the encodes.

Known Improvements

Immediately after submitting, Tompa found two improvements in the sub-fortress using wall glitching.

Possible Improvements

The only thing I can think of on the presses side is that there might be a way to get a Bullet Bill between the two pipes at the end of 8-1, which would allow me to boost over the final Piranha Plant instead of jumping. I don't think this is likely, though. I certainly think my route is good at this point. During my first attempt, I used the wall-jump at the end, not realizing how many presses this would end up wasting. I have since thought it through very hard, and would be quite surprised if I ended up doing the power-up handling wrong.
On the speed side, I'm sure there are plenty of things that could be done faster, especially at the beginning. Two clear mistakes I've notices are that I jumped on the pipe instead of the hill at the end of 1-1 and that I started flying too late in the sub-fortress (I've tested this out and I can save at least 4 frames). Still, this might not factor out to be that much at the end, especially when you factor in the luck with Bowser (Which I might have done suboptimally as well).
In any case, I couldn't merely edit these fixes in, so I decided against re-running the entire game just for them. My primary goal is to find the baseline for presses. Once I find that (It's very likely a viewer will find an improvement or two), then I can focus on. Perhaps a SMB3 guru might team up with me to accomplish this feat just like HappyLee will be doing with me shortly for SMB.

Special Thanks

Thanks to Tompa and bobmario511 for helping me with the glitches in Bowser's Castle. In addition, I'd like to thank Tompa for telling me about the glitch that allows me to not take damage from Bowser. Finally, thank you to all of the people I thanked in my previous submissions for helping me in my quest to establish this category.

DarkKobold: I guess I should handle all of the submissions in this new category.

DarkKobold: Rejecting. See the decision from here, as it applies identically: #3556: Brandon's NES Super Mario Bros. "minimum button presses" in 05:48.04

TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 14908
Location: 127.0.0.1
Experienced player (759)
Joined: 6/17/2008
Posts: 146
Watched the run without commentary subtitles and eyecancerous input display. It was great. 8-1 and the Bowser fight were especially awesome.
Joined: 4/6/2012
Posts: 44
Location: Lawn, PA
First of all, let me say that this run is indeed more interesting than the SMB1 minimum-presses run. I'd also like to say I've rethought what I think of such runs after watching this, including the usage of the DDR Lua, because it's necessary for those viewing this category in order to understand how the presses are being played out. With these things being said, I did watch this the entire way through (I did the SMB1, also, but stopped really caring halfway in; this actually kept my attention throughout), and although the Lua script was no 13-footer, it was also more entertaining to watch than it was in your SMB1 submission. This is probably the most entertaining minimum-button-presses run I've seen yet; however, the fact remains that this is, indeed, such a run. I can't deny that some people might want to watch this, and I wouldn't want to deny them the opportunity in this case, but a minimum presses run can only be so entertaining per game. This'd be a new category for SMB3 on this site. Unlike SMB1, we don't have a zillion categories for this game already, so I'd be alright with its addition. Having said what I have, I cannot vote No on this submission. However, this did not leave me so greatly impressed that I would beg for its publication, as so many submissions around here do. Therefore, I'm forced to vote Meh for this submission. As that one neutral guy in Futurama said: "I have no strong feelings one way or the other."
1/60 of a second is important; every frame matters.
Expert player (2454)
Joined: 12/23/2007
Posts: 822
Having watched the run, I'd say this game truly is a better choice than SMB. This movie is different than the current SMB3 speedrun, and has its clever parts. I think Brandon has shown us that MP TASes can be interesting in its way, nicely done! But because most of the run still seems a little tardy for a TAS to me, so meh vote.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Brandon
He/Him
Editor, Player (190)
Joined: 11/21/2010
Posts: 913
Location: Tennessee
Much to my horror surprise, Tompa found two improvements to World 8's sub-fortress that would result in 2 presses saved. I would request that you all will judge this run regardless of the improvements. What I need to know is whether or not this run is even publishable. Based on how the community reacts, I am more than willing to resubmit or replace the submission file before publication if accepted. In fact, I'm very open to a partnership with Tompa, as I'm sure he can help me find more presses and could definitely help me with the time factor.
All the best, Brandon Evans
Spikestuff
They/Them
Editor, Publisher, Expert player (2312)
Joined: 10/12/2011
Posts: 6342
Location: The land down under.
I watched with closed captions... I read quite a-bit great run YES vote very interesting
WebNations/Sabih wrote:
+fsvgm777 never censoring anything.
Disables Comments and Ratings for the YouTube account. Something better for yourself and also others.
YoungJ1997lol
He/Him
Player (53)
Joined: 7/4/2011
Posts: 550
Location: U.S.A.
voted meh. not entertaining, but probably worth it.
So yea, how's it going? Currently TASing: Nothing
mklip2001
He/Him
Editor
Joined: 6/23/2009
Posts: 2224
Location: Georgia, USA
This run does make a marginally better choice than Super Mario Bros 1 for a minimum-presses run. I especially found 8-1 and the airship autoscroller the most interesting. The autoscrollers turned more into a guessing game for me, trying to predict when you'd jump next before hugging the walls. I definitely think this is publishable. Especially with this run, luck manipulation, or very precise reactions, play a key role in timing your bounces to save jumps. That provides a good demonstration of superhuman play.
Used to be a frequent submissions commenter. My new computer has had some issues running emulators, so I've been here more sporadically. Still haven't gotten around to actually TASing yet... I was going to improve Kid Dracula for GB. It seems I was beaten to it, though, with a recent awesome run by Hetfield90 and StarvinStruthers. (http://tasvideos.org/2928M.html.) Thanks to goofydylan8 for running Gargoyle's Quest 2 because I mentioned the game! (http://tasvideos.org/2001M.html) Thanks to feos and MESHUGGAH for taking up runs of Duck Tales 2 because of my old signature! Thanks also to Samsara for finishing a Treasure Master run. From the submission comments:
Shoutouts and thanks to mklip2001 for arguably being the nicest and most supportive person on the forums.
Joined: 4/24/2012
Posts: 129
Location: Atlanta, GA
This definitely feels like a better fit than SMB1 for a "minimum buttons" category, no doubt about it. I think this does appeal to certain kind of user; this is the sort of thing I mess around with sometimes, and it's certainly impressive, regardless. It's clear that there was a lot of thought put into it.
Joined: 2/16/2005
Posts: 462
Can't most of the levels be done without the B button? I guess the B button has to save the use of just 1 jump to be worth it but I'm not sure if it always does even that.
This signature is much better than its previous version.
RachelB
She/Her
Player (127)
Joined: 12/3/2011
Posts: 1579
This is wayyyyy better than smb. I'm still not sure it's worth publishing though. I may have been able to justify a yes vote if i knew it was optimal, but 2 improvements have already been found, and i'm sure there will be more. I am going to vote meh.
Joined: 7/2/2007
Posts: 3960
I enjoyed it. It's a different take on the concept of a TAS; more leisurely, less "must be doing something every frame or else the viewer will get bored!" I don't think this kind of approach will be valid for all games, and I could believe that eventually we'd get tired of it (as we did with the multi-run concept), but I think this is worth publishing (with the known improvements of course).
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Joined: 5/2/2009
Posts: 656
Way more interesting than the SMB run.It's clever, entertaining and interesting. The Bowser fight is simply brilliant. Yes vote.
My first language is not English, so please excuse myself if I write something wrong. I'll do my best do write as cleary as I can, so cope with me here =) (ノಥ益ಥ)ノ
Experienced player (576)
Joined: 2/23/2008
Posts: 266
Location: CA, USA
I also thought this was more interesting than the SMB run and showed off several nice tricks. Yes vote, and I'd like to see what else Tompa finds if anything.
Brandon
He/Him
Editor, Player (190)
Joined: 11/21/2010
Posts: 913
Location: Tennessee
asteron wrote:
Can't most of the levels be done without the B button? I guess the B button has to save the use of just 1 jump to be worth it but I'm not sure if it always does even that.
If you find such a level, please point it out. I'm pretty sure that B saves at least 1 press in every level.
rog wrote:
This is wayyyyy better than smb. I'm still not sure it's worth publishing though. I may have been able to justify a yes vote if i knew it was optimal, but 2 improvements have already been found, and i'm sure there will be more. I am going to vote meh.
Argh, no, Tompa looked through it all and said that's all he could found. Those were totally obscure glitchy optimizations that were near impossible to guess. I already said I would fix these problems before it actually gets published...
All the best, Brandon Evans
RachelB
She/Her
Player (127)
Joined: 12/3/2011
Posts: 1579
Brandon wrote:
rog wrote:
This is wayyyyy better than smb. I'm still not sure it's worth publishing though. I may have been able to justify a yes vote if i knew it was optimal, but 2 improvements have already been found, and i'm sure there will be more. I am going to vote meh.
Argh, no, Tompa looked through it all and said that's all he could found. Those were totally obscure glitchy optimizations that were near impossible to guess. I already said I would fix these problems before it actually gets published...
Oh, so you're 100% sure there aren't any more improvements?
Brandon
He/Him
Editor, Player (190)
Joined: 11/21/2010
Posts: 913
Location: Tennessee
rog wrote:
Brandon wrote:
rog wrote:
This is wayyyyy better than smb. I'm still not sure it's worth publishing though. I may have been able to justify a yes vote if i knew it was optimal, but 2 improvements have already been found, and i'm sure there will be more. I am going to vote meh.
Argh, no, Tompa looked through it all and said that's all he could found. Those were totally obscure glitchy optimizations that were near impossible to guess. I already said I would fix these problems before it actually gets published...
Oh, so you're 100% sure there aren't any more improvements?
No, but you're implying that I didn't test everything meticulously. These were in no way obvious improvements. These improvements being found immediately after submission was not an issue of me overlooking something obvious to Tompa, but him figuring them out at a really bad time (He's known about the run for a few weeks now). I just wanted to defend that. Again, if this run gets accepted, I'm going to re-run it no matter what. I just want to know if this run, in a vacuum, is acceptable. Unless Tompa's improvements are enough to change your vote from a no/meh to a yes without us knowing about the, I'd ask that you vote yes to approve of this kind of run being published, whereas with SMB1, it seemed that the branch was unwanted regardless of the improvements.
All the best, Brandon Evans
RachelB
She/Her
Player (127)
Joined: 12/3/2011
Posts: 1579
I didn't really want to vote yes anyway. Unlike smb though, it's definitely right on the border. I'm ok with this being published, but i don't think it should be. I don't really have any strong feelings on way or the other on this. It's interesting, but not particularly. It's certainly not boring the way the smb run was, but i'm not sure that's enough to be published. Honestly, i probably would have voted meh regardless of the improvements. When i first watched it i was intending to not vote at all, but i was still leaning a little closer to meh than yes. When i saw there were already known improvements, it just made it an easier decision. I can't say for sure i would not have voted yes if not for the known improvements, but i don't think i would have. Again though, this is an interesting concept. It's definitely nice seeing it done with a game that isn't as terribly suited for it as smb. If the next game we see an MP tas of is any more interesting than this, i'll gladly vote yes on it.
Brandon
He/Him
Editor, Player (190)
Joined: 11/21/2010
Posts: 913
Location: Tennessee
rog wrote:
Honestly, i probably would have voted meh regardless of the improvements. When i first watched it i was intending to not vote at all, but i was still leaning a little closer to meh than yes. When i saw there were already known improvements, it just made it an easier decision. I can't say for sure i would not have voted yes if not for the known improvements, but i don't think i would have.
OK, fair enough.
All the best, Brandon Evans
Player (33)
Joined: 3/8/2012
Posts: 398
Location: Windfall Island
I found it interesting, but not extremely entertaining. Still voting yes because it seems very efficient and I enjoy this game.
IronSlayer wrote:
Your counterargument would be like me saying that the Earth is round and then you telling me that I need to show it's flat so I can "prove us all wrong".
Seems legit.
Joined: 7/10/2008
Posts: 56
Watched the run, It was entertaining for me and showed off a few tricks that cannot be shown in a pure speed run. Voting yes for this to be a new category. I am also eager to see what improvements could be made
i see...
Editor, Experienced player (608)
Joined: 11/8/2010
Posts: 4012
Entertaining both to watch normally and to count the presses. Excellent job, Brandon! This will be a great first "minimum buttons pressed" publication (once the improvements are put in). YES vote.
Joined: 10/20/2006
Posts: 1248
I've found this to be pretty boring in the autoscroll levels. Meh.
Guga
He/Him
Joined: 1/17/2012
Posts: 838
Location: Chile
Brandon, you did a good job in this movie, but I think this is not a good branch, no matter what game is chosen. I will not vote.
Skilled player (1707)
Joined: 9/17/2009
Posts: 4952
Location: ̶C̶a̶n̶a̶d̶a̶ "Kanatah"
So...how much games need a "minimum buttons pressed" catagory?