"Unlike most 2D Mario games, Super Mario Bros. Special didn't have any warps, and this was faithfully kept in this romhack. This means that every level has to be completed." A quote from the
movie page of the newly published TAS of this hack.
But that's not true. In fact, not every level has to be completed, because you can freely choose a starting world in the title screen.
The goal of my TAS is clear: fastest completion. The real any% run of this hack should start at World 8.
This is not cheating, because it's simply allowed by the hack creator (Frantik). He stated it in "Readme.txt": "There are two special features accessible from the main title screen: Press B to select the starting world..."
After saving the princess, you'll get the very same ending text & credits. I see no reason not to use it in an any% TAS.
Other than that, it's just an ordinary SMB hack TAS.
Frankly, it's not my favourite hack, and far from my favourite TAS. But to me, if this hack is allowed to be published on this site, it deserves a real any% run.
Suggested screenshot (frame #9175)
Samsara: Well, this was quite a ride.
First, I must clarify up front that the decision made for this run is mostly independent of HappyLee's temporary ban, in that the ban only affects one possible outcome, and only temporarily. While it is true that the run would have been cancelled or rejected had his ban been indefinite, the defining factor behind this decision is based entirely off of the nature of the TAS itself. On TASVideos, we occasionally receive submissions that require us to re-evaluate our
MovieRules. This is in no way a bad thing, this is something that we want as staff. For myself especially as Senior Judge, I want nothing more than to ensure our rules are clear, readable, understandable, and malleable. If the community decides that there needs to be a change, there will be a change.
This has been happening quite often lately, and I'm proud of the work we have all done as a community to make things easier and more reasonable, from the TAS authors who make runs that challenge the rules to the community members who discuss them and come to a consensus.
This was, in a way, one of those submissions. It found a weakness in the rules, and that weakness was corrected. However, unlike other submissions that lead to edits of the rules, this one found
a weakness of omission and not a weakness of complication. The usage of this form of level selection as a time saving technique has never been allowed on TASVideos, and in my opinion it's unlikely that it ever will be. I don't believe any speedrunning community, RTA included, would ever count skipping 87.5% of a game through a level selection system as a legitimate strategy for an any% speedrun. Keep in mind, however, that this is an opinion and not a firm statement. There is a possibility the overall community's thoughts on this could change in the future, and we will change accordingly to fit the desires of the community.
That being said, there are ways in which this run could have been treated differently, and I'd like to go over three of the more notable ones.
In cases where we accept individual level TASes, we have a
clause in the rules that handles how to treat these runs, namely that they will always be obsoleted by
full game runs of
their respective games, due to what we call "full content overlap". That is, the entirety of an individual level run will be contained in a full game run, so there is no need to have both published alongside each other. Were someone to create and submit a full run of Biker Mice From Mars, it would obsolete the "final round" TAS linked above. For this run, though,
the full game run already exists and has been published. so even if we determined this run to be a valid individual level run, we still cannot accept it. Of course, since this is an SMB1 ROM hack, it is following SMB1's presentation of a long series of connected stages, and as such any usage of the title screen level select is explicitly skipping required content. This is confirmed further by the published run's statement that there are no warp zones, meaning all worlds and levels are normally required to be played. In short, we have to rule out this run being acceptable as an individual level.
Another possibility is that the input itself could be salvageable outside of being published. If this is a strict improvement to the published run's World 8, perhaps the two runs could be spliced together as a compromise solution. This presents a (thankfully) much simpler to describe set of problems, namely that this run and the published run cannot be adequately compared. Any improvement that this run contains would need to be recontextualized in the full game run, as conditions between the two could be completely different due to the level select. Notably, the full game run is fully powered up through World 8 while this run remains small Mario, which leads to this run actually being slower overall as it needs to wait for a piranha plant in 8-4 that the full game run can just kill, so there's no real compromise solution here either.
The final possibility is to accept the run as a new branch altogether, putting it in Alternative or Playground. Alternative is clearly ruled out by the community reaction, with the run receiving a wealth of No votes. PG, on the other hand, was quite literally made to support runs like this. There is a major problem with this option, though: The run was submitted as an any% run of the hack. As such, without any changes, it does not qualify for Playground. Placing it there would require us to change the run and treat it as something else. Not only is this going against HappyLee's original intention for the run, but given his temporary ban, he is unable to make any statements regarding how the run is presented. We should not, under any circumstances, make any changes to an author's work without their explicit consent, so as of right now we will continue to treat this run as HappyLee submitted it, which means it cannot currently be placed in Playground. However, it is still an option for the future, once HappyLee returns and is able to publicly consent to it.
I believe that's everything I would like to address. If anyone has any questions about how this submission was handled from a judgement and rules perspective, feel free to ask me directly. If need be, I will continue to update this judgement with further notes and clarifications to ensure that my thoughts and actions are understood and not misinterpreted.
