I started from 8-1 on 383 (non-FPG "perfect" pace with -3.62) and played through until I got to 8-4, but every time I died I would use the "start from beginning of level on same FR" (Select + B ) - I probably used that maybe 10 times since I "started" the rom from 8-1 on FR 383, including multiple times in 8-4. So yes, both times it happened, I had used the Select+B feature for World 8-4.

I don't know if it's worth noting that I was playing as Luigi - since my understanding was that the ONLY difference in SMB1 was the sprite.

LOVE the rom - and the new version. I would whisper you but I can't find how to contact you directly, so here goes on this thread - My understanding was that the only way to die from the last hammerbro in 8-4 before bowser was to jump from the first frame. However, on the practice rom v1.6 I died twice in the same way to his hammer. If this was possible and I just never have experienced it before, then disregard, but might point to a discrepancy between practice and "actual" rom. Thanks again for the amazing work.

Thought this was good to share with the rest of the SMB1 community.

Xeroxfiend came up with the great idea of having weekly tournaments, which so far has been 2 - the first one was between xeroxfiend, esmo88, and virusereturns. The next week was skipped, and this week (Dec 30, 2017) was a (very chaotic) 6 person tournament between xeroxfiend, darpey, roylgamer, virusereturns, jtknott05, and gtace99

I made a compilation video afterwards. I had planned to make the whole thing in the style of the "practice" video, but it ended up taking more time than I had anticipated, so I made the fancy practice video, then just put all the feeds together for the rest of it.

For the practice video, xeroxfiend's audio is used. For the whole tournament video, xeroxfiend's audio is used for races 1, 3, 4, 5 and darpey's audio is used for race 2.

The first race was agreed to be a "practice" race.

PRACTICE RACE 1st Place - xeroxfiend (5:30) 2nd Place - virusereturns (5:38) 3rd Place - darpey (5:40) 4th Place - gtace99 (5:55) 5th Place - jtknott (6:39) 6th Place - roylgamer - DNF

The entire tournament, using the runners' non-cropped feeds can be viewed here -->

Xeroxfiend's mic cut off during the countdown of the second race, so xeroxfiend and virusereturns stopped (as per xerox's instructions, which were also cutting out a little bit), so it only included 4 of the runners:

2nd Race 1st Place - roylgamer (5:02) 2nd Place - darpey (5:43) 3rd Place - jtknott05 (6:20) 4th Place - gtace99 (8:14)

The next 3 races were the three "official" races

The results are as follows:

Race #1 1st Place - jtknott05 (5:29) 2nd Place - virusereturns (6:38) 3rd Place - darpey (7:34) 4th Place - gtace99 (7:56) 5th Place - xeroxfiend (8:16) 6th Place - roylgamer (DNF)

Race #2 1st Place - xeroxfiend (5:41) 2nd Place - darpey (5:44) 3rd Place - jtknott05 (6:29) 4th Place - roylgamer (7:01) 5th Place - virusereturns (8:51) 6th Place - gtace99 (DNF)

Race #3 1st Place - roylgamer (5:04) 2nd Place - virusereturns (5:26) 3rd Place - darpey (6:07) 4th Place - xeroxfiend (6:21) 5th Place - gtace99 (7:20) 6th Place - jtknott05 (7:47)

Tournament Results (Cumulative time for all three "official" races) 1st Place - darpey (19:25) 2nd Place - jtknott05 (19:45) 3rd Place - xeroxfiend (20:18) 4th Place - virusereturns (20:55) 5th Place - roylgamer (12:05 + DNF) 6th Place - gtace99 (15:06 + DNF)

Best 3 Races (note: this was not the agreed upon metric, and it is further unfair to xeroxfiend and virusereturns, since they only competed in 4 of the possible 5 races, but I thought it would still be worth including, since not all runners completed the "official" 3 races) 1st Place tied - darpey (17:07) 1st Place tied - roylgamer (17:07) 3rd Place - xeroxfiend (17:32) 4th Place - virusereturns (17:42) 5th Place - jtknott05 (18:18) 6th Place - gtace99 (21:11)

If you care to follow any of the runners on Twitch, all 6 runners' Twitch pages can be found below: https://www.twitch.tv/xeroxfiend https://www.twitch.tv/darpey https://www.twitch.tv/virusereturns https://www.twitch.tv/roylgamer https://www.twitch.tv/gtace99 https://www.twitch.tv/jtknott05

You can view the full footage from each player individually here

Xeroxfiend - twitch.tv/videos/213941854 Darpey - twitch.tv/videos/213894989 RoyLGamer - twitch.tv/videos/213868418 GTAce99 - twitch.tv/videos/213901871 ViruseReturns - twitch.tv/videos/213900198 jtknott05 - twitch.tv/videos/213904263

Basically I just assume that it's either dumb Koopa or low... The one you can jump on top of. To me the "high" one looks too much like the low one to tell the difference, so I just die when I encounter it. At least anecdotally though (fortunately) the worst Koopa (high) seems to happen less often than the other two

It's frame rule (as well as level speed within the frame rule) dependent. There are 3 things they can do. Dumb, low, and high. In all three cases you start your jump at the same point and then either hold jump to bop the Koopa or go under. The gif thread has all three combos and how to "defeat" https://www.speedrun.com/smb1/thread/uz8f3

I thought I would post for the benefit of anyone else who was interested in this - I found this web site with a tutorial on how to do exactly what I wanted (make a NES controller split into PC and NES signals, to show on PC while playing on console) --- Thanks to Omnigamer for pointing it out to me.

Just ordered the Arduino and plan to give it a shot. Thanks to all who commented.

https://github.com/jaburns/NintendoSpy/blob/master/docs/guide-evilash25.md

Yeah I saw that was down - For now, I have another copy here --> goo.gl/z9b1RP -- Not sure why the first was pulled, if it was pulled.

I'd be happy to take this link down if told to, but for now - that's a link to it.

@Timmiluvs - you clearly know your stuff about this -

Any reason why a setup like this wouldn't work? Since NES controller --> usb adapters already exist.

1. http://a.co/ddKlJQ0 - NES to usb adapter cable ($15 from Amazon)
2. Genuine NES controller

open the wiring to the controller and the adapter cable, splice them together

I don't know where in the CABLE itself the conversion happens - I would assume it would be NES until the usb piece... or would it be a "usb" cable in the cable? no idea. Even worst case though, you could operate on the NES end and connect the wires together as they would if they were male/female connected, right?

or is the Arduino a necessary component in your mind?

Does anyone know how people like Darbian () who play on a console (not emulator) display their controller's input? Do they splice the wires and have one go to USB? Anyone know the answer to this question?

Would apply to any console that doesn't have easy interception (e.g. wireless controllers)

Anecdotally - it appears to be the case that slowing down in a level does affect enemy behavior later in the level.

The 8-2 over/under koopa is the same on each "framerule" you're on, but you just kinda gotta feel for it if you don't know what rule you're on.

Your questions:

1. At least to me, it seems that IF you go the same speed, make no kills, and get no items, then all enemy behaviors will be the same. If you slow down, kill enemies, or get items, then the behaviors will deviate.

2. No, the "framerule" only applies to the beginning of a new level. At least in my experience, if you slow down, even by 1 frame, it will eventually affect what the baddies later on do.

3. If you slow down (in my experience), it does affect enemy behavior.

Most levels "slowing down" isn't really an issue if you don't explicitly mess up, except 8-2, because of the high jump, and the turtles immediately after.

There are 3 different patterns that the "parakoopas" can be on - which will act the same way every time you (1) enter a level on a certain framerule -AND- (2) go the same pace, make the same kills, and get the same items

The three different types of koopas and how to react to them is listed on the Gif tutorial post HERE --> https://www.speedrun.com/smb1/thread/uz8f3 --- but here it is again:

In all three cases, you START your jump on the exact same visual cue, and adapt based on what you see the koopa doing.

End stairs koopa low ~~~ For low koopa, do a maximum jump and bounce on the koopa

End stairs koopa high ~~~ The toughest one. For high koopa, let off jump and go under the koopa as shown

End stairs koopa stupid ~~~ For stupid koopa, let off jump and jump on the block shown

http://a.co/28CpHVT - "I-o DATA USB connection video capture GV-USB2"

roughly $35 (United States Dollars) on Amazon

It's what I use, and I think it's what Darbian uses, and probably a good chunk of other runners that use console. All of the instructions AND the install CD are in Chinese, so it's slightly tricky to install (but not that bad), but once it installs it works great.

It changes the RCA cables (red white yellow) into a USB signal that OBS can detect under "video capture"

You will obviously still want to PLAY the game on a CRT (or whatever you have been playing on), since there is a sizeable delay to change the signal, but it works great for streaming the video.

Added a new version of the 1-1 FPG with frame counter... I think the old one might have been missing frames somehow. Also selectively slowed the speed on the technical portions. It's at 25% speed, then 3% speed for the frames that "matter", and added a detailed description of the trick. Let me know if anything is inaccurate.

Thanks to Aldyn/Lul_ecks_dee for coming up with the original video, and thanks to Sockfolder for coming up with the setup for RTA - as well as YouTube user "Gunnery Sergeant Hartman" for his video explaining the setup - youtube.com/watch?v=ucBAwdS3YB4

You must land on the 3rd block from the top on the first pixel while not holding "B" - on the first frame with mario in the "running" animation (as opposed to the jump/falling animation), you hold only right (not "A" or "B"), then for exactly the next 2 frames, hold "B", then do a full jump on the 3rd frame (as shown). Hold all inputs until the jump is completed. By the time you reach the flagpole block, you will need to release "A" and "B", and hold "left" for exactly 3 frames (as shown), then jump on the 4th frame. If you do the first "jump" half of the trick correctly, you will clip into the block assuming you do a full jump, even if you mess up the "second" part. If you do not clip into the block, it means you messed up the jump.

Frame counter will obviously be different depending on your pace - as long as you hit the first pixel on the block, it doesn't matter if you do a full or partial jump. In order to "save" the framerule, you will need to be 4 frames from perfect on the coin pipe section, which means that some "380"s will not be fast enough to save the rule.

25% with slowdowns

50% no slowdown

The difference between "individual level" and "real time" is bigger than you'd think, because it's multiplicative - I'm going to pull some #s out of the air for this example, don't burn me on exactness pretty please, but these are probably my average #s

1-1 "380" ~ 35% of the time 1-2 ~ 35% of the time 4-1 ~ 85% of the time (because I kill Lakatu, and even though he ends my runs sometimes, Lakatu just has to die #worthit) 4-2 "374" or "375" - 40% of the time 8-1 Good or Bad judges - 40% of the time 8-2 "slow" - 40% of the time 8-3 fast (372 with flaggrab) - 60% of the time 8-4 WITH the wall jump and no deaths or bad timesinks - 35% of the time

Now - multiply those #s (none of which are THAT difficult on their own by the way) 0.35 x 0.35 x 0.85 x 0.4 x 0.4 x 0.4 x 0.6 x 0.35 and you get... 0.0014 or... or 0.14% or 1.4 in 1000 - which means I should expect to successfully make my "sub-5" goal with these percentages about every 714 runs or so - but there's a HUGE amount of fluctuation and luck in that obviously, so it could take me 100 runs, or it could take me 2500 runs.

That, of course, also varies with skill level. BUT you could even create your own likelihood #s by running each level 20 times, and getting a percentage for each level, then multiply it out like that. Just make sure you don't cherry-pick the #s you use.

For sure - (first off, take what I say with a grain of salt, because I haven't been running SMB1 that long) when I feel like I'm not enjoying the game anymore (for example after a multi-hour session) I take a break. It's all about enjoyment of the game. If you're not enjoying it anymore, stop, at least for a little bit. Focus on having fun, as cheesy as that sounds. 80-90% of this game in the low to sub-5 is playing 1-1 a bajillion times. If you can Buddha your mind into relaxing and enjoying the "grind" of it, then you're winning, so to speak. If it starts to feel like homework, you really do need to reevaluate why you're running it.

I also coach cross country. You have kids that love to run and kids that hate it. There are kids that love improving their times, and kids that don't care - and kids that are the 3rd fastest on varsity and think they're "not good" kids at the back of the team that are having the time of their life, and vise-versa.

^ "time" is all relative. Even the fastest kid on the cross country team is flat out SLOW compared to an olympic athlete. But even the kids on the slower end of the team are faster than most of the kids in the high school. The same goes for SMB (or really anything in life for that matter) - Be proud of a time that's in the mid-5's. Currently, there are 540 runs submitted - so the "average" run is at 270th place, and it's a 5:29. And those are people that worked hard enough and cared enough about their times to submit it to the WORLD RECORD leaderboards. That's pretty dang good imo.

If I were you I'd totally go for a sub-6, and be pumped when you get it. I was super happy to get my first completed run (5:35). Kosmic (former world record holder) is a GOD at this game, it's amazing to see him play - but if the ONLY thing that fueled his mind was getting the WR back, I doubt he'd still be playing. I'm a relatively "bad" runner, or an "amazing" runner depending on what perspective you're taking.

I'm sounding preachy now - TL/DR is (1) Proactively focus on relaxing your mind and enjoying yourself, and when you're not enjoying yourself, take a break (2) Like Roopert said, I would start with smaller goals, like sub-6... or shoot, even 5:30. Either way, it's going to be much less frustrating than going for 5 or sub-5 to start.

Good luck figuring it all out =)

Fwiw I think xeroxfiend was planning on getting together an any% tournament at some point soon as well. Would be fun to have some events for smb1 :)

That above video is really good.

Yeah, it is very similar to any% just without flagpole glitch (0.35 seconds saved per FPG), bullet bill glitch (0.7 seconds saved), wrong warps in 4-2 and 8-4 (a few seconds saved) - so just any% but a few seconds slower.

Darpey used one - I bought an actual NES a couple weeks ago and have been using it exclusively since. I made a post about this - the AVS appears to run slightly slower, (over a 5 minute run, I figured I was slow by ~2 seconds) which shouldn't be a problem, at least in my mind. Should be effectively the same as running it on VC.

added two methods of the faster "1 frame" wall jump for 8-4 and fixed both FPG gifs to be less clunky and made them uniform size with the rest of the gifs

also if anyone can make a video with inputs for the remaining FPG, the 1-2 wall clip, or the 4-2 wall clip, I can put it up here with runner credit.

Added this (and the "traditional" 1 frame) to the gif post.