Now to the Demo
tier Category.
The problem we all had with it is that we couldn't define any clear criteria for it, there were lots of ideas of which particular runs
seem to fit into Demo, but none of them was solid enough.
Here's how it would work in the proposed system:
If a run can't fit into any other Category due to breaking their rules, but nonetheless has positive feedback, it should go into Demo.
Records Category demands being any% or 100%, beating existing records (or matching them), not using ROM modifications, standing out from an unassisted play, etc. Moons wants all of these too, but also demands high entertainment value.
And there are some runs that couldn't fit into these Categories, but were well received.
This submission predated the Tier system, so the question was still "Should this be published?" But people still posted quite a lot about how much they were entertained with the run.
The rejection reason was that even though cheat codes aren't completely banned, one should not use them to make his life easier, because it doesn't look that impressive anymore. And the problem with that is, they
are still impressive to people! It's just they are not an overwhelming majority.
Now see what we did about these people not in majority. We agreed to accept speed records that are
not entertaining to the majority, but still have value to
some people. That's how Vault was added. Even if the verbal reasons for it are different ("yes we want to be a records site too, bla bla"), the very reason why adding a whole new tier was even considered is because of the people who wanted not-so-entertaining records to ever be published. Again, they are not in majority, but that way we annoy the viewer the least, and include more runs entertaining only to the few.
This run received clear Moon feedback, most of the posters enjoyed it. It was rejected because it didn't stand out from the normal play.
Then there was a
6-page thread about how and why do we want some way to have such runs published, and we couldn't come up with something certain. Some suggested adding board games to Vault, others wanted to add a special tier for board games (inclusively).
What is clear: people do want board game runs published. However, not all of them:
Scrabble was received poorly (48% (11/10/12)), people wanted to see a playaround instead.
Then, there's a whole bunch of Gruefood Delight runs that were well made, but couldn't make it for different reasons.
Not much to say except that its post feedback matches the votes. Judgment isn't quite clear to me, but it's pre-tier.
Fantastic research&development piece, great feedback, simply rejected because it's real-time slower that its emulator-aimed analog.
Regardless of decent votes, not a single post about it being entertaining (or boring). An example of a run to be pondered, and then rejected for no support.
Thought there are people disliking it too, the overall feedback is okay. Rejected for the quality of the hack, but people didn't have much of a problem with it, except for those voted No. Pre-tier, so wasn't considered for Moons, and there's KMW3 published, so there's probably some content overlap, but I'd let the crowd decide if they are similar enough to the point when one iteration of the hack series obsoletes another, or maybe we could allow both.
Quite borderline feedback in my opinion, novel goal, nice script, so this might be an interesting run to ponder once again, with the new Category in mind.
ABSURD goal, but good feedback. Rejected for "too much branches" and the fact that this goal is also doable in real time (lol, RBO is doable too, its TAS is still great).
And so on...
Runs canceled by the author are out of the question, because they could be vaulted.
Runs that desync or are suboptimal can't be considered either (see how I changed my mind about
Star Control 2, as it can still be redone to sync and get accepted, same with Perfect Dark).
Runs that can't even be emulated aren't eligible either, as there's nothing to verify (and it can simply be redone on an actual emulator now).
And no "
uncompleted game" runs, simply because nothing prevents us from completing them.
Finally, there's Technical / Showcase tier thread and wiki page.
Not posting any of its suggestions though, because using the above approach, it can also be decided, which of the runs mentioned there could make it into Demo.
This is the best TAS I've see in a long while. Its skips are so amazing and obvious that I'd even consider it for NewcomerRec. If not these Puppy Love levels. Damn, if there was only one, it's still be a star, but they had the cheek to put in there THREE of them!
Now what gives me some hope is that it must be very improvable by knowing the exact RNG code, and using it to make him throw the 
