[deleted as part of purging my old post history]

I personally believe that there's no real need to change the current guideline, it says in a quite clear way already what the mapper can do and that the circumstances must be special enough to allow him to do something different from the usual.

In the end, it always ends to be a case-by-case situation when we're talking about this, so these minor changes suggested wouldn't really affect or change anything imo.

The only thing that I would agree to adjust to the wording is that only if the changes that conflicts with this guideline actually makes sense and are actually approved, then only in that case the mapper would be allowed to avoid this guideline and do those changes for the benefit of their beatmap, but still emphasizing the fact that this isn't something that can be done in every beatmap, but only in rare cases where the mapper explains their reasoning really well and actually makes sense.

i think the difference between these will be irrelevant to 99% of people reading the rc lol

proposal 2 seems ok, it is closer to how people respond to mods. if people are ok with this (or have no opinion, as the lack of responses seems to show), it could go through imo

I can't tell the difference between the original and the proposal. It seems just like rewording.

If the breaking of a guideline doesn't have a positive effect on the mapset, I see no reason for it to be broken, as these guidelines are in the RC for a reason, because they lead to questionable scenarios or something either unfitting to the song or map for the most part.

Whether or not it negatively affects the map doesn't seem relevant to me, because in 99% of cases, guidelines should be treated as rules, and this change makes it seem like breaking guidelines is fine if there's no negative changes, which brings a question of why stop there? If breaking a guideline is fine if it doesn't negatively affect the map, why should breaking rules not be fine if they also don't negatively affect the map (and in some cases, can even help quality)?

The problem with this proposal is that it seems as if it views guidelines as just unnecessary restrictions, rather than something similar to rules that are set for a reason.

I also feel like this proposal wouldn't really change anything because people don't take the RC word by word (at least in this case). If anything, guidelines need to be taken more seriously and they should basically be seen as rules like MBomb said. For example, I've seen lots of Normal diffs as the lowest diff of a set lately which break one or even several guidelines without any explanation from the mapper and the maps were just ranked like that without anyone bothering to check them. However, I'm not sure if the definition of a guideline can even be reworded in a way that would make guidelines more important.

[deleted as part of purging my old post history]

added proposal 2 in https://github.com/ppy/osu-wiki/pull/2156

Changing the idea of guidelines from "If following it doesn't make it worse, follow it" into "If what you have is fine, you don't need to change" would make them way laxer than they currently are, which is the opposite of what the RC rework one or two years ago was aiming to do, hence how the wording came to be like this in the first place.

In the past, guidelines were essentially just a way of saying "this can be ignored in pretty much any scenario, but is good learning material for newer mappers", and even with more serious guidelines being added, them being called "guidelines" was seen by people as being unimportant due to this (many of the current mappers and modders have kept this mindset since then). The RC rework essentially got rid of all the unimportant guidelines that didn't need to be followed and kept the ones people should actually follow in most scenarios. Because of this, changing the definition made sense, as it wouldn't obstruct creativity more than rules already do.

The definition being like it is, is mostly a way of making sure nominators can convince mappers to change bad cases in an easy way, which probably wouldn't be possible under the proposed definition. So it doesn't really matter how people do this argumentation in practice, as long as it can be enforced and done in the intended way when needed. Arguing for why keeping the current is fine is usually much easier than arguing for why making a change is bad, which is probably why so many naturally go that way. This is comparable to how people tend to want to be proven wrong rather than to have to prove themselves right, since it places the burden on the other person.

For this reason we're removing the proposed change from the pull request.

If you have an example of a guideline that, under the current definition, prevents you from doing something that is widely considered good, do bring this to us (preferably proposing a new wording in a thread here and getting feedback on that) so we can change it.