1. osu! forums
  2. Beatmaps
  3. Beatmap Management
  4. Archive
posted
https://gist.github.com/Ephemeralis/1f2ab78a2e6af952d468f934c67b9169

This is a draft/WIP of potential changes to the veto system that came up as a result of discussion in #modding. As this only applies to BN, it's posted here for your perusal and feedback.
posted
talked about it in osu!dev server and I agree with it
posted
2c (warning contains some ramble)

Reading this still seems like fixing a symptom rather than an issue

imo vetos could disappear completely if code of conduct would encompass something like vetos, because they're a streamlined talking to one another process that isn't used as such.

essentially what vetos are trying to achieve is a block (that is giving someone the authority to block) of a map unless it is improved upon from its current design.

By the general idea they put the path of least resistence on working out a compromise with the veto person who doesnt agree with the map by addressing their concerns and working out a compromise.

the only part about handling a veto as the recipient that is ridiculously blown out of proportion is disagreeing with the veto reasons completely or disagreeing with the key points it makes (so about 98% of all maps that met somethign drama-esque)

idk if streamlining the process of "talking to each other" is the best course of action after all - just not being able to nominate maps with ongoing debate on them pertaining to the maps concept and design with regards to rankability could also just be covered

i have no clue which of these alternatives is more feasible, but the suggested course of action is an improvement from the current state of matters
posted
That basically means that BNs who decide to veto a map or who dismiss a veto need to make sure that they can react and answer within 48 hours, because otherwise the decision will be finalised. May lead to unwanted situations, especially when the BN who dismisses a veto doesn't make it in time even though his point makes way more sense than the point of the BN who vetoed.
posted
I agree with the entirety of oko's post.
posted
this is mostly decent, it at least makes unvetos require the same amount of effort as vetos, which is the biggest issue with the current implementation. i still dont like all the needless hoops all the party members have to go through before the set can actually move on, but it's at least better than what we have rn.

wording is a bit weird on this tho:

thing wrote:

QAT reviews the case and decides to rule in BN1's favor. The veto is now upheld and the mapper must consider the feedback given by BN1 if they wish for their map to proceed to ranking.
this really sounds like the map just cant get ranked if the qat agrees with the vetoing bn. like even if another bn comes to try to unveto it, the result will just be the same. i don't see how this takes community input into account at all. the main issue of a single bn forming a massive wall to oppose a map is still present, they just need a random qat to side with them. i dont think that's the intention of this tho, am i just misreading?

tho tbh i dont rly see what the point is of the qat moderation anyway, or even the whole back-and-forth being added in. most vetoing bn's aren't interested in discussing, they just don't want the map ranked, so they'll just force the moderation every time. i think it's better to just let bn's vet bn's. let the unvetoing bn unveto, if the issues are actual issues then other bn's can step in to re-veto
posted
Proposal is better than the current system, so i'm for it.

I don't think we should stop here though, and should seek to go more in the direction of the ideas oko put forth in the future, but as long as we don't consider this to be the "be-all end all fix to the veto system" and realize that there is more to improve, then we can start with this proposal.
posted
I don't like the idea that a QAT has to side with a BN.

There's a 3rd negative outcome to a qat siding with a BN. The reason the map was vetoed is broken and corrupt (not always but for this example) and the qat agrees with it, thus forcing the unhelpful change and leaving out the discussion part of the ranking process.

Generally speaking with this change 2 members of the qat / bn with the same mindset can come together and screw over a set in their same, broken way which is a possibility I feel should be addressed.


Reading back on this, I kind of just said what UC said with different wording, but to add onto the discussion; I like the change to unvetoes in the sense it can be a brief post rather than a full-on essay debunking the vetoers points.

However, I don't like the time constraint, as not every BN can work within that time frame effectively. What time frame would be more effective actually? I have no idea. It might be better to remove the time constraint altogether.

All I have for now, later xp
posted
Following recent trends neither the mapper nor the vetoing bn can usually reach a consensus on things, mostly either agreeing to disagree or one side giving up. Assuming this proposal goes through, and people still fundamentally disagree with each other, this would end up taking a lot of time and mostly leaving qats in charge of which ways vetoes go, which I doubt is the intended result, as this would lead to vetoes being even more of a pain to deal with for everyone involved except the unvetoing bn.

Currently when a map is vetoed, another bn can check over the map and offer their explanation to why they don't think the concerns are problems, and then have that veto overruled. This means the mapset is both checked completely by both nominating parties and one bn can stop one bn and no more or less.

There are essentially 3 main issues with the current system:
1. The unvetoer needs to check the whole set, whereas the vetoer can simply check one difficulty, which creates an imbalance in effort believed to be unfair.
2. It takes additional effort for the unvetoer to respond to the veto in detail, compared to a non-vetoed map, making vetoed maps less attractive to nominate.
3. Vetoing usually doesn't lead anywhere, often doesn't improve quality significantly and is tedious for everyone involved.

The proposed solution solves the above problems in the following way:

1. The unvetoer can address the concerns that the vetoer brought up without having to check the rest of the mapset, allowing the initial bn to renominate. However, the tables would turn and put most effort on the vetoer, creating the same problem but the other way around.

2. The effort required to respond to a veto is split into two, initial briefly addressing the overall concern of the veto and the second most likely simply being a "I disagree" kind of response, following current trends. The unvetoer won't be able to nominate the map and must wait for a qat to mediate the situation, possibly not even going in the unvetoers favour, even if the unvetoer checked the whole mapset and wants to nominate. This breaks the principle of one bn being able to stop one bn and no more or less, as one won't be able to do anything about what the other did without the qat agreeing to that.

3. Rather than solving this point, the proposal actually makes this problem more of an issue, since it forces the nominators to continue discussing even if a conclusion is already reached, wasting more time for everyone, including a random qat. Only thing this is easier for is the bn who tries to invalidate the vetoing bn's points, meaning mappers might be even more unwilling to discuss things knowing they can get their map renominated if they ask some bn to unveto and it passes.

tl;dr: The proposal makes it easier for mappers to ignore vetoes, should qat lean towards unvetoing (for better or worse), whereas if qat lean towards keeping vetoes, the unvetoing bns won't be able to do anything about the vetoes, even if they checked the whole mapset as they can now.

In my opinion, the current system, even if it still has issues, seems to be better than this one as far as I can tell, but I wouldn't be against trying it to see how it turns out.
posted
The proposal makes it seem like vetos are way too easy to do. If you're vetoing for something subjective you need to spend a good amount of time explaining the issue, why it's an issue, and possible solutions. If you don't provide enough weight to your side of the argument, people will just complain and ask why the veto was needed or just disregard you in general.

That said, if the requirements for overturning a veto is reduced then I would expect the need to write paragraphs about issues to go down with it. It would be more encouraging to actually place veto on sets that need it, most of the time I just can't be bothered to spend the time needed to make such little progress.
posted

Sinnoh wrote:

The proposal makes it seem like vetos are way too easy to do. If you're vetoing for something subjective you need to spend a good amount of time explaining the issue, why it's an issue, and possible solutions. If you don't provide enough weight to your side of the argument, people will just complain and ask why the veto was needed or just disregard you in general.
I agree with what Sinnoh said here. A bigger problem I think should be addressed with vetoing is that both the current and proposed systems contain a voluntary response bias from whoever is placing or uplifting a veto. That basically means that whoever goes through the effort to place/lift a veto has a stronger than usual (be it good or bad) opinion on the map They usually care enough to nitpick everything through the map and keep it from getting ranked while facing backlash from the mapper. This also leads to problems because in both the current and proposed systems, BN1 and BN2 will basically be arguing with almost opposite opinions. This might be the case because vetos aren't seen as a usual part of the ranking process. They're for "special cases."

Another problem I have with the proposed system is the timeframes set for each BN to respond to eachother, as well as the forced QAT intervention after 2 responses from each BN. This is almost like saying all mapping debates are similar and should take the same amount of time to be resolved, when this is not the case. If the new veto system could feel more like a conversation about mapping rather than a forced process, I feel like it would be a lot more productive.

This is definitely a start, but in my opinion I think we can go in a different direction. There have been a lot of maps recently that have caused a lot of uproar and still passed through the ranking system without any vetos to show that concern.
posted
any continuation whatsoever on this?
posted
please something anything
posted
@Sinnoh, they are because even with dumb reasoning they still require a huge-ass amount of effort from the recipient to make it abundantly clear that the veto reasoning is bogus

----

Adding on my previous line of thought is:
What i think is actually broken is that people go in to veto a map with the mindset of not wanting to talk about this further than what they say in their issues - in such cases the veto should be void to begin, in my opinion because they don't wish to provide more context or help the mapper find the compromise.

I mean you're free to debate and reach the conclusion to disagree and have someone else decide, but going in with just that intention from the get go seems counterproductive - that's what the process is commonly used for at least: go in, dump your reasoning as to why the thing shouldnt be ranked, never bother with it again and let the mapper struggle his way out of it.

I dont think the current proposal addresses this at all, but just makes it easier for the mapper to get out of this deadlock which is a fair enough start but i doubt this is going to have much of an effect to the current situation.

I think the process of vetoing a map is meaningful - but everything is so gray.
If a qualified group of people think the map is bullshit and can explain why sufficiently, dismissing their opinion as "that's your opinion, haha!" seems pretty wrong, so ranking it anyways would be wrong. Giving the decision to veto to individuals might be what's wrong here then?

I have no idea where i'm going with this, but agree with Naxess' points if we want to improve the situation we may need to start rethinking the principle of making ranked status a vote of BNs
posted
48 hrs seems a bit too short. GG no re if you have anything remotely interesting to do irl
posted
This is how the new proposal be interpreted:



I think the 'two cycle' system is a bit clumsy as the stances of the veto side and the obejct-veto side will not be changed, especially when they consider their given evidences are valid. No matter how many cycles, the stance is still unchanged. What both sides need are actually pulling out facts and objectively consider the changes before or after the veto's suggestion. Which one can contribute to the map itself better, that should be adopted.

QAT interception should be served as a intermediate person, just like how we (BNs) solve problems when we run into obstacles which could not be simply judged by us. (like controversial issues, weird snappings, timings, etc. we have a lot of these examples already yeah?) I am aware of the last purple highlighted sentence: a random QAT will be picked ...... At least QATs are divided into groups before they are randomly picked, which considers the modes, specialties (metadata/ timing/ SB), and general parts. For example in Taiko, we have barline omission and that often triggered a lot of controversies. Not every BN, or even QAT, who are not familiarized with Taiko, could handle it well and properly, especially the thread of the mapset is in a heat discussion. We prefer a calm and chill, and specialized QAT who can provide an objective solution to solve the conflict, and actually this is what we have been doing so far.

Let's give a real example that happened half-month ago: https://osu.ppy.sh/beatmapsets/713086/d ... ll#/150924

This is a veto (raised by me) to timing settings in Taiko, involves a lot of ms shifting and barline omission.

I pointed out the problems solidly, but the mapper disagree with it. Then I preferred him to get a second person to judge the issues I mentioned. He got a few BNs and the result was tie. We cannot seek help from one of the Taiko QAT since he nominated, so we want to wait for another one's response, but it's too time-consuming since the latter is so busy.

The mapper then tried to talk to me individually to address my concern. It worked after I explained a bit, but the mapper wanted extra ensure so he found a Timing Specialist QAT who said both ways (my suggestion, and the mapper's original settings) worked. The case was solved finally.


^^This is exactly what I suggest in the top 2 paragraphs. 2 cents here.
posted
You can find how the new process works here from now on: https://osu.ppy.sh/help/wiki/People/Beatmap_Nominators/Beatmap_Veto

Thanks to everyone who participated in the discussions!
Please sign in to reply.