GlitchMan's Mega Man 4 in 34:03.38

Main Priorities

  • Aims for fastest time
  • Uses no passwords
  • Abuses programming errors in the game
  • Takes damage (which I did a lot of by the way) to save time
  • Manipulates luck

Emulator Used

I originally used FCEU 0.98.28 at the time I recorded this but then converted it to fm2 format on FCEUX 2.1.2-interim so that I could look at the lag counter and see how much lag each stage had. Luckily for me the conversion worked.

The History of My Speed Run/My History of Mega Man TASing

Being my first Mega Man TAS submission, I decided that I wanted to go over how it all started. Being such a big fan of Mega Man games I came across Mega Man TASing after looking up gameplay movies of it on youtube.com back in mid-2007 which I thought was interesting to watch seeing how fast a game of Mega Man can be beaten. I have been watching those speedruns of Mega Mans 1-6 on vision.ameba.jp (it is so convenient because it has TASes of numerous Mega Man games on there). I TASed NES Mega Man games ever since January 2008 (starting with Mega Man 6) and have just been getting better ever since. I began to TAS this one in the summer of 2008 after seeing the published run the year before which inspired me to do my own run of this game and this run was the result of it. This TAS took me three months to complete and I finished this in November of 2008. I wasn't sure if I should submit it or not but then I decided "You know what? It's been a while since there has been a new Mega Man 4 run. Sure it's not perfect but I'll go ahead and submit it anyway to give them a new and improved Mega Man 4 TAS."

New Glitches/Tricks I Used

I made more use of the chargeable Mega Buster during the stages which helped me to save time in my movie. Another thing I took advantage of is the one frame movement after the boss's gate closes and I noticed that the '06 movie didn't take advantage of this. There is a glitch where after climbing a ladder that is right next to a wall I slide, go the opposite direction for one frame, then start the slide again which saves 1 frame each time and there are many opportunities for this throughout the game.

The Speed Run

I saved exactly 31 seconds off of the June 2006 TAS due to better sliding opportunities and lag reduction techniques along with some other time savers that will soon be explained.

Pharaoh Man

The first two rooms of this stage: I use the Mega Buster in this part because it kills the enemies faster for reduced lag. I also minimized the lag by luck-manipulating the flying enemies into dropping only one bullet on the screen at a time.
From the checkpoint on, I used better sliding techniques to reduce a lot of lag in this stage and killed enemies at the right time so that lag doesn't build up where it can easily be avoided.
I improved the uncharged shot in the boss fight. I missed a glitch that could have saved 70-80 frames which was jumping in the center of the room right when the ending music played.

Bright Man

For this stage I used the same strategy as in Pharaoh Man's stage combined with some new strategies. I didn't switch to the balloon right at the start of the second room because I needed to kill enemies to reduce lag.
For the bossfight I used the Pharaoh Shot glitch where Mega Man charges up a Pharaoh Shot and slides then throws an uncharged Pharaoh Shot as soon as the slide stops. This causes him to have two Pharaoh Shots above his head at the same time (he can have up to three). This saved a lot of time on this boss because there is no waiting time for charging up.

Ring Man

In the ladder climbing section of this stage I used a strategy where I jump off the ladder, throw a balloon, and jump on it to save some climbing time. I kill the enemies after the second miniboss so that I can get in more sliding and reduced lag. I also needed to collect small powerups for my weapon or else I won't have any to use against Ring Man. The Pharaoh Shot glitch is used on Ring Man but I couldn't find a way to position him so that time isn't wasted on the energy balls disappearing off the screen after he is killed.

Dust Man

I discovered a trick where the shield enemies can be killed at the front so I used Pharaoh Shot against them because this type of enemy only requires to shots of it to be killed. I take damage off Dust Man to get him to stop using his vacuum which makes him invincible.

Skull Man

Instead of switching to the Ring Boomerang which I thought was slower, I kept the balloon because not only does it save almost 80 frames but I can save tons of climbing time with the Balloons. I also pulled off a zipping glitch with the Balloon near the end of the first room. In the last room with the big red sky and lava pits below (that's how red it is) I took damage off the last Skeleton Joe (that's actually what it's called; look it up on megaman.wikia.com; too off subject, sorry) at the end but I'm not sure if this saved time or not.
For the boss fight, my main strategy was to luck-manipulate Skull Man into keeping his shield on for the shortest possible amount of time. He responds to B button and left/right D-pad movement. I also needed to get the Dust Crusher particles off the screen as soon as possible so that I could fire another one.

Dive Man

I don't switch to the Flash Stopper in this stage and I figured that would take too much time. I also took a different route at the beginning of the stage while killing enemies which saved some time. I think that taking damage was made up for with the huge amount of sliding frames gained throughout the stage. Boss fight was basically the same as the old run.

Drill Man

I switch to the Balloons to save a lot of climbing time for the first two rooms then I switch to Flash Stopper. On Drill Man, I take damage while sliding during the explosions to reduce lag.

Toad Man

Not really any optimization here other than less lag and better physical movement. Once again, I take damage to reduce lag from the explosion (this time I'm using the weapon so I have control over it).

Dr. Cossack Stage 1

This is where the charging up came in handy (the first room) and because of it I get more frames of sliding speed as opposed to walking speed. The room with the spiked walls sliding up and down the ladders was really difficult because there was so many of them to not take damage from and lag to avoid but I got through it just fine. Not much I could do about the giant robot in the next room. The boss fight was the same only this time I jump during the explosion to reduce lag. Jumping and not moving reduces lag in NES Mega Man games.

Dr. Cossack Stage 2

I make better use of the Balloon item because of the trick I used back in Ring Man's stage where I jump off the ladder to use the Balloon. This came in handy because I was able to climb up the ladders using Balloons.

Dr. Cossack Stage 3

I reduced lag in this auto-scroll stage by having Mega Man jump over the edge of the screen to the point where his sprite disappears. I also didn't switch to Ring Boomerang just to fill up two bars of it. The first time watching the other run when it did that I wasn't sure what happened. The Pharaoh Shot trick once again saves time in this boss fight but only for the first Cockroach Twin (the second one was at the bottom on spikes and there was no way I was going down there!).

Dr. Cossack Stage 4

I found a faster way to maneuver Mega Man through the walls while he is in his sliding mode. Pixelation has to be adjusted then I can alternate left and right every frame which is twice as fast as the previous run. In my battle against Cossack I used the uncharged long-distance shot technique which saved a little bit of time.

Dr. Wily Stage 1

I don't take damage off the Met in the third room like the old run did. At the end of the underwater portion I switch to the Balloons as opposed to the Rush Jet which has the Rush animation you have to wait an additional 30-40 frames for. For the boss fight I tried so many ways to get rid of that long waiting sequence where you have to wait for the Giant Met to pop back up but I couldn't find a way. I even tried using the Pharaoh Shot glitch but the weapon kept bouncing off its helmet unless it was up in the air.

Dr. Wily Stage 2

Same as the old run but with improved movement and cutting back on the lag.

Dr. Wily Stage 3

This time I stick with my regular weapon instead of switching to Drill Bomb because I wanted Pharaoh Man to be killed first so that I don't have to waste a weapon switch by going to another weapon then coming back to it. This helps because I don't waste any weapon HP. I took this route for boss re-fights: Pharaoh-Toad-Dive-Skull-Drill-Bright-Ring-Dust. During Dust Man's fight I made use of a Ring Boomerang twice which saves me from having to get a small power-up for this weapon; I do the same for the first form in the Wily boss fight.

Dr. Wily Stage 4

Nothing to mention about this stage except that I charge up on some of the caterpillar enemies so that I can get more sliding time in. The Pharaoh Shot for the boss in this stage is used differently in this run than the 2006 run. I defeat Dr. Wily right in the center of the screen which reduces time even after the movie! Not sure if that matters or not but I just thought I would add that... :)

Possible Improvements

  • Using 4 uncharged shots against Pharaoh Man instead of just 1 could save time because it takes 92 frames for each charged shot and this means I need no more than 30 frames to make an uncharged shot to charge back up.
  • Instead of going to Bright Man after Pharaoh Man, I should consider going to Drill Man instead so that I get the Rush Jet. Using it in Skull Man's level would save about 4 seconds off of this run. Also, using charged Pharaoh Shot on Drill Man does the same amount of damage as his actual weakness and because I have the Pharaoh Shot glitch it won't lose me that much time.
  • Maybe I should have switched straight to Dust Crusher instead of the Mega Buster for the boss fight in Wily Stage 1 (weapon switches cost nearly 80 frames).
  • It could probably be faster to switch straight from the Mega Buster to Rush Jet in Wily Stage 2 (another weapon switch saved).
  • Place the bosses closer to the center when killing them in order for the energy balls to disappear off the screen faster.
  • Jump in the center after defeating a boss when the ending music plays. This trick can save 30-40 frames in six stages and 70-80 frames in Pharaoh Man's stage.
  • Use Pharaoh Shot on the Pharaoh Man boss re-fight in Wily 3.
  • Find more ways to minimize lag and discover more glitches and newer techniques.
  • Find a way to get past long "waiting points" such as the second hippo in Ring Man's stage, the boss in Cossack 2, and the Giant Met in Wily 1.

Special Thanks/Closing Remarks

I want to thank everyone on this board for providing many cool glitches and movies of glitches of their own, finding my 34-minute "WIP" on microstorage ^_^", and your support and encouragement in helping me make this TAS. This is my first official Mega Man submission and it has been an amazing experience TASing this game!
Suggested frame shot: frame 122603 which is the last frame of the movie! :D
Edit: Okay, maybe that wasn't the exact screen shot I was talking about but the other one in Bright Man's stage looks better.
I think I have pretty much covered everything that needs to be said about this run.
Enjoy the speedrun!!!

Nach: I enjoyed watching this run. However, some of it looked a bit sloppier than it should have been, and I think this run can be improved quite a bit. Possibly even more time can be shaved off of this run than this run did to the existing one. I look forward to your next submission for this game, which I understand is already in the works. I am rejecting this run as per Judges' Guidelines: "However, if the submitted movie is clearly improvable as well, it (usually) should be rejected just as a non-published submission might be".

Nach: Here's the video to watch directly: (Flowplayer module removed)

Nach: Being an enjoyable run to a popular game, this was a great run to force us to reexamine our rules. The decision on this is now pending our judges coming up with a new set of rules. Hopefully the new rules will allow this run to be published.

Nach: We have a new clause in our rules. This run is now accepted as an excellent improvement to an already published run.


Skilled player (1099)
Joined: 8/26/2006
Posts: 1139
Location: United Kingdom
This decision is disgraceful, unfortunately due to the state of the rules Nach had every right to decide this as he did. Reform of the site's rules is needed to avoid such things happening in future. Most of what is manifestly ridiculous here has already been said, so I won't dwell on it, however one thing here really gets me. Nach's decision has been vocally objected to by three of the site's judges (two of which have not commented on the run either way), however as things stand a judge can single-handedly make a call that the others disagree with provided that they make it first. I've noticed a convention here that judges don't (rather than 'can't') overturn others decisions and as a result this will probably stand. Would it be at all possible to introduce a quorum rule that will require at least three judges to make a decision on controversial submission?
Player (121)
Joined: 2/11/2007
Posts: 1522
Pilling on a "why wasn't this published" vote. Also, Mukki's idea is great.
I make a comic with no image files and you should read it. While there is a lower class, I am in it, and while there is a criminal element I am of it, and while there is a soul in prison, I am not free. -Eugene Debs
Skilled player (1652)
Joined: 11/15/2004
Posts: 2202
Location: Killjoy
There is not yet a majority of judges - currently, there are six judges - Nach, adelikat, Truncated, mmbossman, Baxter, and cpadolf. So far it is 3 in favor, (adelikat, mmbossman, and Baxter), 1 against, Nach, and 2 unreported (cpadolf, Truncated). 1 more judge, and you'd have a majority in favor of publication.
Sage advice from a friend of Jim: So put your tinfoil hat back in the closet, open your eyes to the truth, and realize that the government is in fact causing austismal cancer with it's 9/11 fluoride vaccinations of your water supply.
Emulator Coder
Joined: 3/9/2004
Posts: 4588
Location: In his lab studying psychology to find new ways to torture TASers and forumers
I'm not against.
Warning: Opinions expressed by Nach or others in this post do not necessarily reflect the views, opinions, or position of Nach himself on the matter(s) being discussed therein.
Joined: 4/3/2005
Posts: 575
Location: Spain
I think one of the reasons of that rule was to prevent publishers to spent several hours encoding the movie when it was known a better one was in the way, and maybe prevent people to watch an unoptimized movie that could detract them from watching the optimized one in a few weeks. I myself was going to watch this run, but knowing that it can be made 30 seconds faster makes me wait patiently until it's done. Judges might show the same reticiency. However, even though I can understand the above reasons, if there are people that want to encode the movie and people that want to watch it anyways, I don't see any bad in allowing it, giving than this run: a) doesn't show sloppy play, and b) it's better than the previous run in all fronts.
No.
Experienced player (961)
Joined: 12/3/2008
Posts: 940
Location: Castle Keep
I watched encode quite soon, i liked it, but somehow i forgot to vote, or tought it wouldnt be necessary, anyway it seem to be back on bench, so ill vote yes this time to push toward publication. Imo regardless whatever know improvements, 30 seconds is (largely) enought. Now if you look how frame wars work, like the smb one, when theres 2 runs in bench and one is faster than the other, generaly the guy just cancel... and if not then its obvious to choose the one to reject. That rule preventing an improvement is somewhat making the site sloppy, if it happen theres so many frame wars requiring the staff to hold down publication pace in order to unsure highest quality content, then just delay judging, to offer more time to contenders for submit improvements. As such the delaying status seem well apropriate and enought to handle such situations, rules are maybe not necessary (leaving also more flexibilty on the staff side) However, this is pretty clear nobody is claiming here to submit anything in close futur, as such rejecting seem wrong for everybody
Joined: 6/4/2009
Posts: 893
GlitchMan wrote:
That's it!!! I'm definitely going all out on finding Mega Man glitches!! As Mega Man would say, "I'm going to defy the laws of physics!"
that's the spirit, man, this run should never been rejected in the first place we should only judge movies on what they have and not on "what they don't have" especialy for a megaman run based on speed "improvement of the current published movie" should have been far enough. if we begin to reject for "improvements that are not there" we should also reject all the other runs "because they don't use improvements that will be found soon...." if a run don't use a known improvement, then someone else should resubmit another run obseliting the run. improvement's the game, frame war's the name happy (lee) new year btw :)
Joined: 7/2/2007
Posts: 3960
I think (and please correct me if I'm wrong) that there's two main reasons the "known improvements" rule is in place. The first is for framewars, where a run is on the workbench, someone says "Why didn't you do X?", and the runner (or someone else) says "Good point, I'm going to give that a shot, update in a day or two." In that situation, we don't publish to prevent a bunch of redundant publications from getting made, which would clutter up the newsfeeds and generally muddy up game histories (not to mention waste a bunch of time on the part of the judges and publishers). The second reason is to prevent things like that Ocarina of Time submission we had awhile back. Thanks to new glitches, it was faster than the currently-published OoT run, but it had a significantly lower quality-of-play throughout and thus looked sloppy and poorly-played (I'm just going by what people said here; I didn't watch it myself and am not generally qualified to judge the quality of OoT runs). Neither of these two rules seem to apply here: first, while there are known submissions (EDIT: I meant to say "improvements" here), we won't get a new run using them for several months at least, and possibly longer. Second, people who know the game don't seem to consider the gameplay to be sloppy (and speaking as an educated amateur, I certainly didn't notice anything that looked off).
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
adelikat
He/Him
Emulator Coder, Site Developer, Site Owner, Expert player (3576)
Joined: 11/3/2004
Posts: 4754
Location: Tennessee
Derakon wrote:
Second, people who know the game don't seem to consider the gameplay to be sloppy (and speaking as an educated amateur, I certainly didn't notice anything that looked off).
More importantly, the gameplay in this movie is not only "not sloppy", it is clearly superior to the published movie. Also the oot example is exactly why we need a rule of this nature (though for the record, I consider the gameplay of the published movie about as equally as bad as that submission, but that is for another debate). There was a boy and his blob submission a long time ago that better illustrated this concept. We had a 5ish minute optimized TAS. A shortcut was found that shaved several minutes off the route. A submission using this but was obviously sloppy (where the published run was not) was made. It was rejected because, while faster, was much less impressive looking. Regardless of the rules, we do need an "out" in the guidelines for this (rare) situation.
It's hard to look this good. My TAS projects
BigBoct
He/Him
Editor, Former player
Joined: 8/9/2007
Posts: 1692
Location: Tiffin/Republic, OH
I think "improvements are known" is a bad grounds for rejection, unless the movie is also poorly played, which this one clearly is not. Plenty of movies have been published with improvements known. (For example, Chef Stef's 120-exit Super Demo World; he didn't use 1/1 swimming because it wasn't discovered until he was well into the run.) Adding another yes vote to the pile.
Previous Name: boct1584
Joined: 4/19/2008
Posts: 31
Location: Chicago
Coming here to agree with pretty much everyone else on this, and not just this run, but the rule. It's ridiculous. If it's faster and more entertaining, it deserves to be published. The rule should state something like "If improvements were known at the time of the making of the movie and purposefully left out" etc, although I'm not sure of anyone that would do that besides for entertainment value.
Joined: 1/26/2009
Posts: 558
Location: Canada - Québec
Glitchman, did this submission should have more glitch!?? I mean what about editing at least few part by adding some funny trick like this one, etc..? As long as there no desync, there no problem... right?
Joined: 7/2/2007
Posts: 3960
BadPotato: unfortunately, the glitches in that video require picking up Wire in Dive Man's stage, which is a time-costing detour.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Experienced player (634)
Joined: 8/28/2008
Posts: 443
You know, I agree with what all of you all said! Just because a TAS doesn't look faster than its predecessor doesn't mean that it actually isn't faster! Also, this could've been the first published run of 2010!
mklip2001
He/Him
Editor
Joined: 6/23/2009
Posts: 2227
Location: Georgia, USA
As has been pointed out, there is a need for a rule that allows some sort of rejection of shoddy improvements. How about a requirement that says if a run is to be rejected for "not being enough of an improvement", evidence needs to be demonstrated for how the run appears sloppy or suboptimal? For instance, with the OoT runs, the last one was rejected by pointing at Groobo's videos. The idea behind this restriction is to prevent rejections based on theories or radical route changes which have not been tested. (In many cases of sloppy runs, someone on the forum has posted a WIP and said "Look, this WIP beats your run by x seconds", thus demonstrating the sloppiness. This is the kind of thing I had in mind.) Also, if people are worried about immediate improvements to a movie that just got published, why don't we make a minimum time requirement before a movie is judged? I know it's been suggested before to have a policy where a movie should be judged no sooner than a week (maybe two) after its submission, and no later than four weeks (these numbers can be altered). The sole exceptions will be for submissions that clearly break other important rules, like plagiarizing other people's work.
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.
Skilled player (1652)
Joined: 11/15/2004
Posts: 2202
Location: Killjoy
Additionally, that rule may be out-dated, since we now have amazing encoding volunteers, that publishes runs almost immediately after acceptance. TASVideos has never accepted low-quality runs, obsoleting or not, and so perhaps this rule could be completely eliminated? Glitchman's run is clearly not low-quality, simply out-dated.
Sage advice from a friend of Jim: So put your tinfoil hat back in the closet, open your eyes to the truth, and realize that the government is in fact causing austismal cancer with it's 9/11 fluoride vaccinations of your water supply.
Joined: 7/2/2007
Posts: 3960
It seems like the goal is to have a rule that allows us to block runs that can be improved not necessarily due to using new glitches, but simply due to tighter gameplay, whether that be through more precise movements, more consistent use of existing glitches, not making obvious mistakes, whatever. If your run can only be significantly improved by making use of new glitches, then "could be improved" shouldn't be the cause for rejection.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
sgrunt
He/Him
Emulator Coder, Former player
Joined: 10/28/2007
Posts: 1360
Location: The dark horror in the back of your mind
Here are some attempts to articulate current thoughts on this matter: "A run should have at least a comparable, if not higher, level of optimisation to any predecessor runs; having a faster time is not by itself sufficient for acceptance." "If there are known improvements that a) were not known at the time the relevant segment of the run was created and b) would require a significant amount of effort to implement (usually involving re-running subsequent segments and/or the run as a whole), they shall not be considered grounds for rejection by themselves." Does that seem plausible and reasonable?
Experienced player (634)
Joined: 8/28/2008
Posts: 443
Okay. Now I know what you guys meant by "it was sloppy." But when you compare it to the other run but just ignore the fact it was outdated, it is actually a good run.
Former player
Joined: 3/23/2004
Posts: 95
Just saw this got rejected/delayed because of it needing to implement known improvements...all I have to say is this: Why, then, did the first run of New Super Mario Bros. get accepted when many of the flaws that the run had were pointed out at the time of submission?
Former player
Joined: 4/16/2004
Posts: 1286
Location: Finland
Therealssjlink wrote:
Just saw this got rejected/delayed because of it needing to implement known improvements...all I have to say is this: Why, then, did the first run of New Super Mario Bros. get accepted when many of the flaws that the run had were pointed out at the time of submission?
See this post
Experienced player (634)
Joined: 8/28/2008
Posts: 443
BadPotato wrote:
Glitchman, did this submission should have more glitch!??
Sorry for not including "more glitch" to it (which I will)! ;) I told you guys this was before I acquired my new username so I didn't have any cool username to live up to at the time. I WILL add more glitchiness and side-effects to this game in my new and improved run. :D Being the encouraging TASer I am, I will always find more ways to make this faster!!! :D
Post subject: Movie published
TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 15629
Location: 127.0.0.1
This movie has been published. The posts before this message apply to the submission, and posts after this message apply to the published movie. ---- [1444] NES Mega Man 4 by GlitchMan in 34:00.02
mklip2001
He/Him
Editor
Joined: 6/23/2009
Posts: 2227
Location: Georgia, USA
Yay! I'm glad to see the new rule in the guidelines. It looks pretty solidly-worded. Also, congrats again, GlitchMan!
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.
Experienced player (634)
Joined: 8/28/2008
Posts: 443
Thanks everyone!!! YYEEEESSSSS!!!!!!! :D:D:D I finally did it! It got published! I knew my determination would get this movie in!!! P.S. I like the image you selected for it! It rocks! Invisible Mega Man! :D