This topic is for the purpose of discussing #768: Phil's NES Super Mario Bros. 2 "warps" in 07:53.63

So you used exactly the same route than sleepz did, tried your very best to get some frames here and there and in the end it's about half a second faster? I know it's faster but as with your earlier SMB1 submission which was one(?) frame faster than pom's, somehow I don't feel comfortable with obsoleting sleepz run with this one... but as usual, it's a matter of opinions.

If you're unhappy then do it yourself. Btw, it was 3 frames faster and new pom's version is 1 frame faster than me so... And if you have suggestions for better route then tell it and stop complaining.

If it's faster, post it.

It may be a small improvement, but it's still impressive. There is a limit for every game on how fast you can (theoretically) complete it, and it looks like we're getting closer to that limit with this one, so even if it's just a small improvement in absolute terms, it still deserves to be posted. Of course, if you disagree and don't think that it's an impressive improvement (I'm looking at you, Fihlvein), feel free to go ahead and submit your own, much better version. :) But if you find you can't make it substantially faster, then maybe you should ask yourself whether Phil's new run is really that unimpressive after all.

Naturally I'm interested in what's changed. I'm not going to be able to spot a half-second difference on my own... Oh wait, yes I can. (Fires up dual run encoder)

I'm not saying that it's not impressive, not at all. Phil's TASs are usually really entertaining and I'm sure this one is as well. All I mean is that I don't find really really good motive for doing such a run, unless there are clear slowdowns in published version. Aiming for perfection is a very good reason I know, but do you really think there will be no more new tactics, glitches and such in the future of SMB2? Of course there's nothing wrong with trying to perfect current tactics to the maximum and it's great to see that. I will probably vote yes because it's a fastest existing run. All I'm saying is that sleepz did an awesome work and in a way I feel bad when such work is being obsoleted by something which isn't possibly any more entertaining, only noticed in the end as some frames improved. Of course, it might be that you were more entertaining overally, that I will find out after I watch your submission.

Of course, if you disagree and don't think that it's an impressive improvement (I'm looking at you, Fihlvein), feel free to go ahead and submit your own, much better version.

Sorry I'm not doing TASs myself, but does that mean I can't critisize them? Does every music critic make music by themselves? Of course not. You don't need to make music to judge music as you don't need to make TASs to judge TASs.

The most obvious improvement seems to be in the second area 4-1. Obvious of course is a relative term when looking at a half-second improvement.

I agree with Filvhein, in terms. Getting an obsoletion by improving 30 frames is kind of backstabbing the previous player. But there are games and games. This is a mario game we're talking about, and its a reference to other players, so the faster the better. Making a mario movie means getting into a hard competition, so, such improvements are predictable as big candidates for obsoletion. Based on that i'm gonna vote.... after i watch it hehe.

I vote yes. The only improvement I could see was in 4-1, where phil avoids slowing down to jump over the fly-guy-on-a-cannon by throwing an enemie and double jumping over it.

I agree with Filvhein, in terms. Getting an obsoletion by improving 30 frames is kind of backstabbing the previous player. But there are games and games. This is a mario game we're talking about, and its a reference to other players, so the faster the better. Making a mario movie means getting into a hard competition, so, such improvements are predictable as big candidates for obsoletion.

I think Foda described it pretty well. I too agree with Fihlvein about it being a small improvement, but on the other hand it is like Foda said, a Mario game. Also, didn't Phil had the first record on this? Or maybe the first "improved" record over Bisqwits older? I can't remember really. Edit: My 900th post apparently.

To be honest, I really couldn't see a difference other than how you handled the part before the tall stairs part of 4-1 like De said. Also, you messed up the end of 4-1 by not doing the traditional enemy throwing trick where they get stuck in the floor. Aparently, you just threw him and said screw it and went on. I'm very tempted to "Meh" this becuase of of end of 4-1 and the fact I can barely see anything different, but becuase it is faster, and we're aparently reaching the speed plateau with this game, I voted yes.

I can't edit my previous post for some reason.. Well, I can only remember one case where an improvement haven't been published and that's the 1 frame over Pom's SMB1 run. So I sort of think this is settled already ;)

Well, maybe I should write when, why and how those improvements happen but I prefer avoiding this work. You know, Sleepz's version was almost perfect so I am lucky I was able to save 36 frames. When I started doing this run, it was to improved 4-1, where most people had noticed the improvement. Even if it would be 1 frame faster I would have submitted it. And if someone beat me by 1 frame I will vote yes. This game does have a lot of competition. If it would be SMRPG and only 36 frames faster, evidently it deserves to be refused but SMB2 is a short game.

Clear yes from me. True, less than a second isn't an amazing difference, but faster is better. It helps that in this game there is a clear end point, and Phil has clearly not lost any entertainment value in the process. There's no reason this shouldn't obsolete Sleepz version.

To be honest, I really couldn't see a difference other than how you handled the part before the tall stairs part of 4-1 like De said. Also, you messed up the end of 4-1 by not doing the traditional enemy throwing trick where they get stuck in the floor. Aparently, you just threw him and said screw it and went on.

Is it because he was carrying a Flurrie instead of a Shyguy? I don't know. I think it's still cool that the Flurrie went through him, though.

Interesting that half a second isn't considered a lot in SMB2, which is only 7:54 long, but it would be greeted with such fanfare if it was SMB1, where everyone is desperate to see it listed as 4:59. If only sleepz's run had been 8:00, those 36 frames would have broken a psychological barrier and created a new milestone.

I totally agree with Phil, I would vote yes on even a 1 frame improvement. I agree that it would suck to be obsoleted by only a few frames, but that should inspire you to try harder. quote from the guidelines to this page: "Competition breeds perfection"

Sorry I'm not doing TASs myself, but does that mean I can't critisize them? Does every music critic make music by themselves? Of course not. You don't need to make music to judge music as you don't need to make TASs to judge TASs.

I didn't say that. But if you're not doing these things yourself, then how are you able to judge what's impressive and what isn't, anyway? You still can say "I don't like this", of course, but that's just your own opinion. The same goes for everyone else who made similar comments, too, like FODA and Walker Boh. I have a lot of respect for you both, but I honestly don't understand you here. Look at it this way: If there is still room for a big improvement in a game (several seconds, at least, possibly more), then of course, a new movie should try to exploit that, and it shouldn't be published if it doesn't do so despite obviously being able to. However, that's not always the case. It may just as well be that the game does not allow for big improvements anymore; and in that case, 30 frames may actually be quite impressive an improvement. What I really mean, I guess, is that whether something's a big improvement or not does not only depend on the absolute number of frames saved, but also on the potential for saves that still exists in the game. The SMB1 runs are a good example of this - even an improvement by a single frame can be a remarkable feat when it was previously thought that the theoretical limit for the game had already been reached. And FWIW, the site's about perfect movies, anyway, isn't it? It's not about good TAS movies, or even great ones, but *perfect* ones - so if a published movie is obsoleted, then the new movie should definitely be published, unless it's reasonable to assume that an *even* better movie will be submitted soon (in that case, not publishing the first improvement is justified since it'd just be extra work to publish a movie that'd be obsoleted itself again soon, anyway). For that reason, it's also rather stupid (sorry) to argue that it's "unfair" to another player to obsolete their movies. I mean, come on - that's like saying that it's unfair to the previous record holder if an athlete breaks a world record in sports. If you don't like your own submission being obsoleted, then prepare a better one again; and if that's not possible because the newly-published movie *is* at the theoretical limit, then maybe ask yourself why on earth it should not be published.

I think people were proud of Sleepz's accomplishments and feel sorry for the work he put into obsoleting Genisto's movie just to be obsoleted by another movie that is only 36 frames faster.

A lot of work is put into planning, route testing, glitch discovery, etc. If you find a place that can be improved by looking at someone else's run, and copy the whole of it (let me make myself clear that i dont believe this was the case here) and change only at that place, that's unfair. It would be better to point out the mistake to the person who made the movie so he can improve it himself. But if you find major flaws and significant improvements, it's ok to do it yourself. The line between one case and the other is blurred, and everyone has thought about it before, but little has been discussed on this subject.

If Sleepz submitted this run, would we even be having a discussion? Here's how I'd think this topic would have instead played out: "Wow, Sleepz, you did it again!" "I didn't expect this improvement. Nice work!" "And I thought it couldn't have been done any faster. Good job at finding that small improvement." "We're getting closer to perfecting this run; I really liked how you handled 4-1" Now I'm a Sleepz fan myself, but to remain objective, it is important to judge the run and not its author. You have my yes vote, Phil. Edit: Oh, and after Sleepz would have his run published, we'd get these comments: "What codec do you use for this avi? I can't watch this run!" "how do you do that double-jump thing?!"

Ok, I didn’t say anything when Phil redid the LoZ run, since a minute and a half is a significant improvement, and taught me the importance of meticulous accuracy, but Phil, this is a joke. You redid this whole run copying mine move for move just to improve by 36 frames. My improvement over the previous run was 20 whole seconds, contributing several new tricks and ideas. And yours is 36 frames, contributing nothing? Why would you even bother to do something like this if you can’t add anything to the table? As FODA said, this is really a stab in the back. And it’s not the thought of my run being beaten that’s bothering me, if that’s what you’re thinking. When Genisto beat my first run, I was very pleased since he added some neat things like killing Birdo with the key and DJs. His run became and instant favorite of mine. Your run on the other hand adds nothing new and doesn’t even get an extra second. It’s still 7:54. Now to critique the run itself, I must ask Phil, why do you still do those annoying fast left/right, up/down movements? There was already a discussion or two about this, and as I recall, everyone said they hated it. Your run is full of it, while I made an effort to avoid it in mine. Even Bisqwit thinks it’s stupid; he mentions it openly in the guidelines page. It’s like you made the run uglier in exchange for a few measly frames. I thought that a slightly slower run that is more pleasing to the eye (not to mention original) takes precedence over a slightly faster one. But now people are voting for it because all they can see is the “36 frames faster!” Is that all you people care about? Seriously, if this is what TAS making is all about now, I don’t even see the point in making them anymore. Who cares about making cool videos when we can compete over who can scavenge for the most frames instead. For those of you who keep up with human speedruns, this is exactly why I’m keeping my LoZ run from the public’s eye. Someone can just rip-off your hard work, make a carbon copy of your run with a dinky little improvement to obsolete it, and claim it as their own. I don’t mean to come off sounding like a jerk or to degrade your work, but when people do this shit, whether it be in gaming, business, or any other aspect of life, I just lose respect for them.

Well, Genisto was working on newer version and if you would have tell it this new trick(I am talking about that bug in 1-1 before Birdo), I am sure it would be as fast as mine version, if not faster. The other tricks in your video were already been found. For the left+right, I must recall that people were displeased because in Castlevania 3, the screen is shaking which I admit is somewhat a little annoying. Other than that, see videos by other authors and you will notice that I am not alone to do this. I even notice, I don't remember which game, that Bisqwit does this. I think it's his WIP Solomon's key run or maybe that SMB3 demo.