forum

Let's make QAT's great again. A proposal.

posted
Total Posts
22
Topic Starter
Monstrata
I want to highlight some fundamental issues that this will hopefully address. Summary/TLDR at the bottom.

Membership to QAT is vague.
  1. This causes people to believe nepotism to be a selection method for QAT. QAT's simply are selecting people who they are friends with, or who because they have known for a long time, are more willing to accept into the team.
  2. Causes some QAH and BN members to feel unappreciated for the work they believe they have done.
  3. Sentiments of "I deserve to be QAT" can be quashed with a clearer selection criteria. Now you can better explain (and maybe even quantify) why someone didn't get into QAT in comparison to someone else besides "they were just better".
  4. Sentiments of "I worked so hard yet some random person who's only been BN for 3 months got chosen instead" can be better explained.
  5. If there are straightforward goals to work towards after BN, people will be more likely to improve themselves instead of stagnate.
QAT do not ensure quality content.
  1. Good examples of this are the recent Hailie maps, but more broadly speaking, just general "low quality" or "controversial" maps that have made it to rank.
  2. Current QAT are too willing to equate "low quality" or "bad" with simply "different".
  3. QAT believe that stricter quality control will create a slippery slope that stagnates mapping progress and creativity, or results in a lot of "controversial" maps being graved.
  4. Conflation of arguments: "no one agrees with any quality standard" and "people want quality standards enforced, but only their quality standards" which will hopefully be resolved with the voting system.
BN's are too indifferent to mapping quality that they are uninvolved in.
  1. Veto'ing beatmaps is very rare as many BN's would rather not get involved in other people's affairs at the price of nominating other maps on their to-do list.
  2. Sentiments of "who cares, just don't play it" result in the absence of quality standards.
Restructuring the QAT


Let me go through some structural stuff that is mostly uninteresting for most of you guys. This is mainly for the higher ups who will be considering these changes.

QAT Role:
  1. The Role of the QAT will be to assure quality. As such, QAT's have the power to issue disqualifications and are able to make the executive decisions that determine the changes a map must make to reach ranked status in the case of disputes that do not favor the mapper.
  2. The current QAT role is far too wide-reaching, and is better served transitioning them to GMT or support roles. If a QAT is currently being kept for reasons outside of quality assurance, we should find a better title or group for them. (Naxess for example, we could move to osu!support team or Development).
  3. Issues regarding Ranking Criteria changes etc... are part of quality assurance and will remain an added responsibility of the QAT's to resolve or amend.
  4. Issues regarding veto's from BN's will be resolved by QAT's however, changes will be made in this area.
  5. Issues regarding the quality of BN modding, and dismissing of BN's will continue to be a responsibility of the QAT as this is indirectly related to mapping quality control.
  6. BN selection will continue to be a responsibility of the QAT as this is related to mapping quality control.
  7. Basically, the QAT name will now reflect what they are supposed to be doing: Quality Assurance on beatmaps.
QAT Membership selection:

Membership selection should become more definable. How does a BN become a QAT? What are the criteria?
My propositions:
  1. BN's who are performing in the Top 25% of BN's in terms of activity, disqualification rate, and involvement in quality control will automatically be on the list for potential selection. This ensures that BN"s appointed to QAT are performing well as BN, and are also involved in quality control. BN's are free not to engage themselves in quality control, but then they should not be considered for a position like QAT which is all about quality control.
  2. From this list, BN's can then elect to be nominated to QAT if they want. (Well, isn't this just circlejerking and nepotism?) To a certain extent, yes. But I think the voting for QAT leader was different to that of the old BAT voting system. The reason is, you are voting for people who are already in the BNG, not voting for people to become BAT's.
  3. Voting will be anonymous.
  4. Why voting? So that those elected into the QAT will also be BN's who were popular among the BN group and will be more likely to represent the concerns that BN's are facing. This is important to mend disconnects between the QAT and the BN.
  5. This process should occur every 3 months, or when positions become available.
  6. Only X number of BN's will be chosen. If the current QAT believes that three new members are needed, then then the three most voted BN's will be promoted. This number must be at least 1 to ensure new blood is always entering the QAT.
QAT tenure:

  1. A core group of QAT's will have to be maintained. QAT's hold tenure after 6 months, which is long enough to be considered osu!alumni. After 6 months, they become a permanent part of the QAT.
  2. QAT's who are within their first 6 months of QAT-ship and are no longer performing in the top 30% of BN's will be demoted back to BN. This is to ensure that BN's who are successful in becoming QAT will continue to uphold their quality standards and activity level. Because it's expected that with additional responsibilities, QAT members will not be as active in nominating maps, the criteria is 30% instead of 25%. Their involvement in quality control will stay the same though of course, if not become higher.
Approach to "controversial maps" and "vetoes":

The meat of the issue.
  1. QAT's will begin enforcing a higher standard of quality than what is currently the standard. How much will they increase their quality standard? We will have to balance that out.
  2. Higher standards will be maintained in an effort to draw more opinions. The most important thing with regards to quality standards, is getting a more accurate picture of how a map's quality compares when viewed by more of the community. The goal of BN's is to represent the community and push forward content the community will enjoy.
  3. Therefore, QAT's will no longer be solely responsible for disqualifying maps or absolving veto's.
  4. Instead, we will now pool from a group of randomly selected BN's and QAT's to vote on the map's acceptability. Let's call this the approval score. We will use this score to determine the level of quality control we want to enforce. The vote will comprise of one simple questions: "Should the map, in its current state, be permitted for rank? Yes/No" . Ideally we will have 5 BN's and 3 QAT's in every vote. BN votes are worth 1 point, and QAT votes will be worth 2. A map will need 7/11 approval score or all 3 QAT's approval to get ranked. I think we can change this number (8, to 6, to 5) to control the extent to which quality standards are enforced. For example, if the current system seems too restricting, I think QAT can make the executive decision to reduce this number to 6. Basically, this number will dictate how acceptable a map has to be given a random pool of BN's and QAT's to be able to meet the subjective quality standards.
  5. If the map does not meet the necessary number of approvals or doesn't have unanimous approval from the QATs, it will now be up to those 3 QAT's who were responsible for voting, to now decide what steps the mapper must take in order for the map to reach ranked status. If these steps result in the destruction of a mapper's core concept or style, then the mapper has three options: Grave the map, get the map Loved, or ask for a re-vote. Re-vote will be up to to discretion of the QAT and should only be considered if destruction of the mapper's core style is inevitable, the mapper has good reason not to implement the QAT's solutions, and the mapper has good reason to believe that the current vote was skewed or very close (6 when the minimum was 7).
  6. If the vote number is ever changed (lets say it was 7 and was reduced to 5) All veto'ed and disqualified maps that were previous rejected but now would have passed, are no longer veto'ed. This means if my map received a 5/11 approval score back when the score was 7, my map got vetoed. But if the number was dropped to 5 following QAT revisions to their quality control standards, my map is no longer veto'ed and can be pushed for ranked again.
  7. Similarly, if a map has passed the approval score (lets say its 7/11) but the score was raised to 9/11 before the map reached the qualified state, the map is now vetoed. (But this probably will never happen).
BN involvement and how a veto proceeds:

I've talked a bit now about BN's being involved in quality control matters, as well as the approval score. Now lets flesh that out:
  1. BN's can opt out of voting if they do not want to concern themselves with quality control. BN's who do this will not be contacted to cast votes on controversial and vetoed maps. In return, they will not be able to work for the QAT title as they are choosing to not involve themselves in quality-related issues which is the core of the QAT position.
  2. BN's can issue vetoes and attempt to resolve them with the mapper. If the BN and mapper are unable to resolve the issues, a QAT can be brought in to arbitrate. If the QAT agrees with the mapper, then the veto is invalidated and the mapper can continue with their ranking process again (Calling the previous BN who bubbled the map, and their second BN to qualify). However, if the QAT also agrees with the BN, then the mapper is either forced to fix, or call for a vote.
  3. If the QAT sides with the mapper, and a second BN or QAT comes in and disagrees with the QAT, then a vote can be called.
  4. Calling for a vote is an option the mapper always has. But they must respect the results of the vote. Calling for a vote is possible for a BN/QAT as well, but only when a veto has been lifted. The same BN/QAT who veto'ed cannot also call for a vote.
The voting procedure for vetoed controversial maps etc:
  1. 5 random BN's will be selected from a pool of BN's who volunteer to be involved in quality control. BN's who choose not to involve themselves in this are basically saying "others can decide what the quality standard should be, I will simply agree with them or not care".
  2. 3 random QAT's will be selected from a pool of all QAT's.
  3. BN votes will be anonymous, QAT votes will not, because QAT's have to tell the mapper how to proceed if the map is unsuccessful in a vote.
  4. The question will be simple: "Should the map, in its current state, be permitted for rank? Yes or No." BN's are not obligated to check all discussions, or mod the map at all. They only have to look at the map themselves and judge it.
  5. The approval score as I previously mentioned will be 7/11 with BN's all getting 1 point per vote, and QAT's getting 2 points per vote, for a total of 11 possible points.
  6. The approval score can be changed later to better reflect quality control standards, whether increasing to 8 to enforce a higher approval rate, or reduction for more leniency.
BN performance:
  1. As we know, BN performance is tracked by QAT. Adding a score to this, much like the old BN scoring system, would be a good way to track how well BN's are performing.
  2. For modding/nomination activity, a logarithmic graph can be used so any number below (let say) 4 mods per month and 2 nominations is penalized. And above 4, the benefit of modding more and nominating more begins to diminish (So this portion doesn't become overweight).
  3. For disqualification, QAT's already track this but it's basically the frequency in which a map that is nominated by a BN get's dq'ed for foreseeable errors (unrankables).
  4. For involvement with quality control BN's can opt out of this, but in doing so cannot be considered for QAT. BN's who are involved with quality control will get a subjective score from QAT's based on how often they are active and engaged in quality concerns, and other factors like how often their veto is upheld versus dismissed, or how long it takes for them to cast a vote (lets say after a week) when called.
Becoming BN and Staying BN:
  1. The 3-month cycle will be replaced with a 1-month cycle.
  2. BN intake will occur more frequently, and in lesser quantities. Think 3-6 instead of 15-20 per 3 months.
  3. Staying as a BN will be slightly more difficult. I think if we implement a BN-performance system, or scoring system, something like this would be neat:

    Lets give some arbitrary numbers here, so you can see what I'm thinking:

    Modding a map: 50 points decreasing by a logarithmic value after 5
    Nominating a map: 50 points decreasing by a logarithmic value after 3
    Participating in a veto arbitration/vote: 50 points decreasing by a logarithmic value after 3
    Disqualifications: 0-100 points depending on seriousness.
    Poor behavior: 0-5000+ points depending on seriousness.

    Now let's say every month, a BN must achieve at least 500 points.

    Hobbes2 is a BN. He has modded 10 maps, nominated 5 of them, and has received 1 disqualification. He has also participated in two veto votes. His score for the month is 500 for modding, 250 for nominating, -100 since he nominated a map with 2B elements, and 100 for quality control participation. His total is 750. He only needs 500, so after the month is over, 500 is subtracted from his score, leaving him with 250.

    Hobbes2 throws a temper tantrum and posts angry QAT memes on reddit and receives -100 points. His score is now 150. He decides to go on a long vacation and does not go on osu for the rest of the month. The next month, he loses 500 points, and is now at -350 score. This means, the following month, he will be in the negative. If he is unable to get the value back to above 0, he will be removed. Anyone who has a negative value *before* the end of the month when the -500 deduction occurs, is removed.

    I think a system like this that blanket subtracts X amount from BN's every month will allow for better flow in the BNG. Obviously we will have to clean up the numbers a bit to make them more realistic. But a system that constantly deducts some points from BN's will make the team more active and encourage doing at least minimum activity requirements in BN (Something small of course, not like 15 mods and 10 nominations a month).

    Ideally this sort of system can be used to measure BN performance, but who knows how much tinkering there will need to be. I'm fine with formulating more detailed ideas if this is feasible though.
  4. Basically, membership to BN should become slightly more difficult. However, because we are making it easier to get kicked, we are compensating by making entrance to BN much more frequent. I think with this change we can see BN's take on more risks, such as knowing how many "points" they have and can afford to potentially lose nominating a very controversial map. Quantifying this allows BN's more knowledge of their standing.
  5. This system will hopefully also allow more modders to become BN at least to "try" and see if they can keep afloat.
  6. We may want to cap the number of points a BN can receive a month, and have successive DQ's every month cost more and more points so BN's with say 1500 points can't just yolo qualify 15 maps and not care if they get dq'ed cuz "ha I got all these points, i'm safe no matter what".
  7. An alternative idea is having this system in place only for probationary BN's and having the probation term spanning 3 months. This would automate the BN probation review process somewhat, though QAT's obviously still want to review BN work regardless.
Summary


Introduction of a trial/voting system for controversial and vetoed/disqualified mapsets.
  1. A jury comprised of 5 random BN's and 3 QAT's will vote on maps. BN's have 1 vote, QAT's have 2 votes. A map needs 7/11 votes in favor to pass.
  2. By using a voting method, we can now quantify the quality of a map through its approval rating given a random selection of BN's and QAT's. Why is this good? Because mapping quality is determined by BN's and QAT's. A map's quality should be acceptable to a certain extent when surveyed. We can always change this number later to reflect how lenient we want to be.
Introduction of a more transparent selection criteria for QAT, as well as QAT tenureship to remain as QAT (6 months).
  1. Only BN's who perform in the top 25% of BN's can be nominated for QAT. This is to ensure a base level of proficiency.
  2. BN's will vote for these nominated BN's every 3 months, and the top X are promoted to QAT.
  3. QAT's are expected to perform well after promotion, and must keep their work up for at least 6 months. If they fall below the top 30% of BN's the will be demoted.
Quantifying BN work some-what for the purposes of those interested in QAT promotion.
  1. Mainly introducing a scoring system to what QAT's already track, so we can record BN performance.
  2. Not well fleshed out and probably requires major tinkering by someone who can do math.
BN Selection and Removal will become more frequent and more rigorous.
  1. BN's will be selected every month instead of every 3. In smaller batches of course.
  2. Some scoring system can be in place, whether just for probationary BN's, or for the group as a whole. A flat score is deducted every month from the BN's total score, and they must mod/nominate/etc to receive points and stay above a score of 0. The amount deducted will be equivalent to a typical and expected BN activity (4 mods / 2 nominations or something).
  3. Ideally we will see more BN's potentially fail to pass probation or simply get removed for a negative score (meaning their proficiency or activity or attitude was just not good enough to stay BN).
I hope this can help to resolve some deep-rooted problems I've been seeing in the community. Give me comments and criticisms too I guess, I'll try and answer what I can, but I obviously don't have the answer to everything. Fundamentally though, I want to see the QAT and BNG become a bit more diplomatic in terms of how they handle controversial maps, instead of saying "lets agree to disagree" and overlook that maybe a significant majority of the community may be objecting a view as opposed to just one BN disagreeing with another BN.

Some things I want to expand on: BN selection (how a regular modder becomes a BN) and BN membership (what a BN has to do to remain a BN) but I think those are less important to the core issue of restructuring the QAT, so we can discuss those later.

wow, i ended up rambling a bit, okay, maybe i'll clean this up later, but i think the fundamental ideas are stuff i've been thinking about for a while now.
hi-mei
I kinda agree on most of the stuff, tho I think we should basically redesign what QAT/BN is at this point. Gonna message you pm.
Topic Starter
Monstrata
Oh, I just realized that by using a trial system, we can also resolve the issue of "QAT's being unwilling to make the final decision on a map and kill mapper spirits". By issuing a vote, the verdict becomes shared among at least 8 other people and the mapper basically has to admit their map is not up to many people's quality standards, not just one QAT or one BN. Also, since the QAT's job will be more focused on helping the mapper get their map back to a status that is rankable, QAT's don't have to take as much of the blame and can also help with resolutions.
Loctav
some core ideas of yours are also contained in the initial proposal I brought up to the QAT internally, who are in fact still working on what I offered them to bring along a suggestion of change. You basically came to comparable methods like elections, maintaining a core team for management and having the QAT proactively APPROVE beatmaps instead of disapproving them. I hope that Mao will soon finish his write-up on this and post it publicly so maybe your and my proposal complement each other in a way that they can fix each other's potential flaws and shortcomings.
Kibbleru
Great job with this.

Many of the core ideas overlap with the one Loctav proposed, once that comes out, we should honestly merge that into the same thread or something for easier discussion.

But in general there are some good ideas here that we have not thought of.
Topic Starter
Monstrata

Loctav wrote:

some core ideas of yours are also contained in the initial proposal I brought up to the QAT internally, who are in fact still working on what I offered them to bring along a suggestion of change. You basically came to comparable methods like elections, maintaining a core team for management and having the QAT proactively APPROVE beatmaps instead of disapproving them. I hope that Mao will soon finish his write-up on this and post it publicly so maybe your and my proposal complement each other in a way that they can fix each other's potential flaws and shortcomings.
I've heard bits and pieces about your proposal. Maybe it's something similar to what ideas we've discussed in the past. the BN performance section was borrowed from the BN scoring system you set up in the past too, which was randomly dismantled (peppy said so? idk...?). Well, lets see how this goes.
Net0
The problem of any system that takes into account "BN performance" as a factor that takes quantity of nominations/amount of DQs will always privilege BNs who push generic maps more often compared to BNs that pushes different/edgy maps foward. In the long term the result of this idea was seen already and people complained a lot. You may say "oh but it only accounts for unrankable issues" but that's NEVER the case in the end for DQs since "on-going discussion" still a valid reason for DQing maps even now
Topic Starter
Monstrata

Net0 wrote:

The problem of any system that takes into account "BN performance" as a factor that takes quantity of nominations/amount of DQs will always privilege BNs who push generic maps more often compared to BNs that pushes different/edgy maps foward. In the long term the result of this idea was seen already and people complained a lot. You may say "oh but it only accounts for unrankable issues" but that's NEVER the case in the end for DQs since "on-going discussion" still a valid reason for DQing maps even now
On-going discussions, and subjective issues, have not been a factor considered in penalizing BN's for a while now. Obvious any map can get dq'ed for subjective issues. BN's clearly cannot foresee that happening, and obviously won't be penalized for failing to consider what one random person out of 13 million users might have issue with.

That's why only unrankable issues and foreseeable issues are taken into account.
abraker
Monstrata, do you think the quantity of BNs and QATs will allow this proposed voting system to be effective in non std gamemodes?
Topic Starter
Monstrata

abraker wrote:

Monstrata, do you think the quantity of BNs and QATs will allow this proposed voting system to be effective in non std gamemodes?
Definitely an issue with other modes. Something like 3 BN's and 1 QAT might be okay though? I'm not too worried about other modes though, because vetoes and quality concerns typically only occur in standard. The other thing too is that their community is smaller so the number of BN's and QAT's don't need to be that high. The important thing is having a system in place that allows for issues to be resolved via some sort of majority voting instead of just on the word of one or two people.
abraker

Monstrata wrote:

abraker wrote:

Monstrata, do you think the quantity of BNs and QATs will allow this proposed voting system to be effective in non std gamemodes?
Definitely an issue with other modes. Something like 3 BN's and 1 QAT might be okay though? I'm not too worried about other modes though, because vetoes and quality concerns typically only occur in standard. The other thing too is that their community is smaller so the number of BN's and QAT's don't need to be that high. The important thing is having a system in place that allows for issues to be resolved via some sort of majority voting instead of just on the word of one or two people.
I am worried about a small minority having control of what type of maps get ranked. It's somewhat an issue even now with modding queues since there is not enough of them for diversity. With your suggestion, I'd prefer something that ensures that will not happen on the voting side of things.
timemon
Very interesting proposal, but I do have a few questions.

1. During the voting procedure can the 8 persons discuss with each other, or the 5 persons (excluding QAT) have to judge purely based on their own perception of quality. As in the voting result is hidden until everyone's votes are cast and you cannot discuss directly or indirectly.
2. Can you change your vote in case of a change of mind?

Thank you for reading.
clayton
I'm excited to hear the details of Loctav and Mao's proposal because it seems as though yours and theirs will work very well when combined

I will save my comments for when Mao's write-up is available, but overall I think these changes would be very beneficial. nice work
Topic Starter
Monstrata

timemon wrote:

Very interesting proposal, but I do have a few questions.

1. During the voting procedure can the 8 persons discuss with each other, or the 5 persons (excluding QAT) have to judge purely based on their own perception of quality. As in the voting result is hidden until everyone's votes are cast and you cannot discuss directly or indirectly.
2. Can you change your vote in case of a change of mind?

Thank you for reading.
Personally, I think it's fine for those 5 and 3 people to discuss among themselves if they want. After all, you'll probably know who they are based on who is checking your map and stuff anyways. For BN's, voting should be anonymous so no one is pressured to give their opinion, and it would rather be a "collective" opinion, so the mapper can't blame or single out individuals, nor will the blame be on a single individual.

As for changing your vote, I think it's okay if you have good reason to, or if you overlooked something critical. No one's perfect. However, what I don't want to see is mappers trying to single out BN's who voted "no" and try to convince them to change their vote to "yes". I can see that happening with QAT's but personally, I think most QAT's (and future QAT's) will be people who really have a consistent opinion on something and are not that easily swayed.
Topic Starter
Monstrata
I added a short blurb about potential changes to the BN system too. If we decide to go with a score system, we should base it on fundamental activities that define good and bad BN's. Actively modding, nominating, and (optionally) participating in veto arbitrations will result in points. Lack of activity, or avoidable DQ's, and poor behavior will result in point deductions.

Basically, a system that is more rigorous will improve BN activity and BN quality. However, with a score system, BN's will also know their standings which I think is super important because it allows BN's to see whether they are in danger of getting kicked and should be more careful, or if they are actually fine, and can take some risks and not just nominate normal, easy, "safe" maps. A system that is more rigorous should be paired with a selection system that is more lenient.

Easier to get in, easier to get kicked, easier to get back in.
Myxo
overall this seems like a good proposal. i'm not so sure about the random selection of bns to vote on a map, cause it can basically depend on rng if your map gets through or not. the current mapping community as well as the teams of bn and to a certain extent also qat involve a lot of people who are very biased towards or against certain types of maps and don't really differentiate between quality within a category (with categories i mean things like wub maps, something from the uc/hailie/etc circle, maps with an older style, ...). so it will really highly depend on luck if you get a bn to vote on your map that is willing to accept that style of mapping in general or not. sure, it's to some extent balanced out by the fact that there are 5 bns and taking stochastics into account the chance will be pretty low to get only bns with a similar mindset to vote on a single map, but i think i'd still prefer if there were some kinda way to prevent cases like this.

with how far the mapping community has been split and standards of people have diverged heavily, what about splitting the bn into identifiable groups with differing mindsets (like people mainly interested into wub maps, people more into older style maps, etc, ofc people could join multiple groups if necessary). then, for vote shenanigans bns from specific types of groups could be chosen. for example, if there's an issue with mapping fundamentals (this is how i'd describe recent controversies) it would be nice to pick bns from different groups, but if it's something more specific and less impactful, it's enough to get mainly bns who are into the type of map in the first place. if bns are going to be forced into checking certain maps due to their random selection, it's better to pick them for maps they are generally interested in most of the time, simply to not kill motivation (we've been through getting assigned to maps in the old qat system, it sucks hard after a while).
Pata-Mon
As we known, now new BNs are determined by QATs. If "BN's will vote for these nominated BN's every 3 months, and the top X are promoted to QAT. ", then it means, BNs decide who will be the new QATs, and QAT decide who will be the new BNs... If somebody always vote to his friends, and his friends do the same, after several cycles, I'm afraid of there may be a circlejerk or something.
I think this situation should be avoided.

QATs determine new BNs √
BNs determine new QATs √
QATs determine new BNs while BNs determine new QATs ×?

-----------
Maybe keep old qats, and add a class between bn and qat, they can dq map, they are voted by BNs. And old qat carry on determining new BNs.
Loctav
as I said before, this proposal shares a lot of similarities with my proposal that I worked on with the QAT, so please give it a read https://osu.ppy.sh/community/forums/posts/6910042
Topic Starter
Monstrata

Loctav wrote:

as I said before, this proposal shares a lot of similarities with my proposal that I worked on with the QAT, so please give it a read https://osu.ppy.sh/community/forums/posts/6910042


1. This is the tier system all over again, but I'd criticize it not just because "the past systems have all resulted in failures" but why a tier system like this won't work to begin with. The "Core Team" which is replacing the current QAT, is still going to have unreliable and uncertain criterias. One of the changes we're looking for is to get a more clear criteria for acceptance into the QAT. Rebranding QAT to something else does not change the issue. Your proposal does a good job of defining the "New QAT" (Basically BN) but the issue is that under your new system, only the titles have really changed, but the core issues remain. Cleverly though you've hit the goal of "making entry to QAT more visible" but really, at this point people are looking at getting into the Core Team which is the real QAT.

My proposal is different because it doesn't make this distinction. I don't believe there needs to be a distinction other than by tenure. Otherwise, the value of QAT is completely replaced.

2. Another big issue is making the ranking system a queue. This is no different from Tier 1 and Tier 2 BN's which failed. Replace BN's with Probationary BN's and QAT's with BN's, and you get exactly the same system that was implemented before. But what's changed? Yep, now there is no QAT. This system effectively rebrands the Tier 2 BN's as QAT's like i've said before. Except now, the Tier 2 BN's are only seeking out maps that are nominated (instead of whatever they want). And if they do try and nominate a map, a BN cannot qualify. Sure, you can say that they can just get a BN to nominate it first, but people also like to mod of their own accord. This is basically saying that modding of your own accord is not efficient because the system is now a queue.

This system also (as was shown in the past) results in low content, followed by Tier 2 BN's and QAT's yolo nominating stuff without thoroughly checking, and yes, this is true.

3. This proposal doesn't address one of the main issues which is quality control. It seems you are just assuming that the new QAT's will have some form of quality control that might be different from what it is now. It also doesn't address the issue of vetoes due to subjective issues.

Anyways, this proposal seems not really well fleshed out yet since honestly, it just seems like a rehash of previous systems, but with changed names. A name change won't solve anything. It doesn't matter if QAT becomes easier to get into if everyone knows QAT's are just Tier 2 BN's now. It's like you own a fast food shop, and rebrand all your cashiers to the title "Manager". Well great, everyone's a manager, so everyone has a special title, but it doesn't mean anything because they still all have to listen to the "Core Manager". Their job hasn't changed, only their title has.
Topic Starter
Monstrata

Pata-Mon wrote:

As we known, now new BNs are determined by QATs. If "BN's will vote for these nominated BN's every 3 months, and the top X are promoted to QAT. ", then it means, BNs decide who will be the new QATs, and QAT decide who will be the new BNs... If somebody always vote to his friends, and his friends do the same, after several cycles, I'm afraid of there may be a circlejerk or something.
I think this situation should be avoided.

QATs determine new BNs √
BNs determine new QATs √
QATs determine new BNs while BNs determine new QATs ×?

-----------
Maybe keep old qats, and add a class between bn and qat, they can dq map, they are voted by BNs. And old qat carry on determining new BNs.
Voting for friends will always happen. Even current QAT is comprised of that to some degree. I'm sure other people with conspiracy theories can make better arguments. I won't get into that. But you have to face the fact that in any small system like this, there will always be circlejerks. It's how you can contain them, or have circlejerks not become a dominant factor in determining the course a map takes. For example, my proposal of having random BN's judge a map is there so that even if one circlejerk really believes the map should not be ranked, they cannot shut the map down.

Also, as for promotion to BN, I feel we can flesh out that system more, but we could honestly keep the current system since it doesn't seem that biased to me honestly.

Also about voting in new QAT's, remember there are only limited positions every 3 months, and only the best BN's will be eligible. So you have to also be a good scoring BN. Having a "circlejerk" won't mean anything if you just aren't good at modding or lacking in some area. A popular vote also means even if there is a circlejerk, that isn't necessarily bad, because it means that BN (now QAT) represents a good chunk of the BNG too.
Topic Starter
Monstrata

Myxomatosis wrote:

overall this seems like a good proposal. i'm not so sure about the random selection of bns to vote on a map, cause it can basically depend on rng if your map gets through or not. the current mapping community as well as the teams of bn and to a certain extent also qat involve a lot of people who are very biased towards or against certain types of maps and don't really differentiate between quality within a category (with categories i mean things like wub maps, something from the uc/hailie/etc circle, maps with an older style, ...). so it will really highly depend on luck if you get a bn to vote on your map that is willing to accept that style of mapping in general or not. sure, it's to some extent balanced out by the fact that there are 5 bns and taking stochastics into account the chance will be pretty low to get only bns with a similar mindset to vote on a single map, but i think i'd still prefer if there were some kinda way to prevent cases like this.

with how far the mapping community has been split and standards of people have diverged heavily, what about splitting the bn into identifiable groups with differing mindsets (like people mainly interested into wub maps, people more into older style maps, etc, ofc people could join multiple groups if necessary). then, for vote shenanigans bns from specific types of groups could be chosen. for example, if there's an issue with mapping fundamentals (this is how i'd describe recent controversies) it would be nice to pick bns from different groups, but if it's something more specific and less impactful, it's enough to get mainly bns who are into the type of map in the first place. if bns are going to be forced into checking certain maps due to their random selection, it's better to pick them for maps they are generally interested in most of the time, simply to not kill motivation (we've been through getting assigned to maps in the old qat system, it sucks hard after a while).
I mean, your criticism is true for basically any country with a judicial system that involves random juries lol. There's always a chance the 5 people picked out of 30 million for Canada are all extreme radical leftists or something lol. The possibility of this occurring should not completely deter this from happening. As I said, a way to prevent this is for the mapper to try and request a revote. And this should only have to happen if the QAT's have proposed changes that fundamentally destroy the concept of a map. But remember, ranking is a privilege not a right. I really feel we need more quality control anyways so if the mapper isn't willing to make changes, maybe they shouldn't get their map ranked because the community has deemed it not fit for ranking.

Remember, this is a vote for acceptability, not for goodness. Personally I find a lot of maps acceptable, but I sure wouldn't consider them good maps.

I understand this proposal might seem threatening, especially for people who are trying to pioneer radical styles and whatnot. But I think if you want to push these radical maps, you shouldn't be able to flout quality control completely. And recently it seems QAT's are too indecisive with drawing the line because they don't want to be blamed. I'm hoping that this system will absolve that blame issue.
Naxess
This seems interesting, but how do you ensure that bns won't just go all out quantity over quality in order to weigh the score system in their favour? I realize "quality" depends on who you ask, but since this is something the overhaul is meant to place more focus on I'd think it should at least be accounted for in scoring. It's from the score that you reach the QAT in this system, after all, so unless we want to rename it to "Quantity Assurance Team", the quality factor should probably be taken into account in some way.

In the past we've had ideas like increasing score for spotlights, but that would hardly be consistent or frequent enough to be reliable. Disqualifications and vetoes do kind of work in a sense, but they're all negative reinforcements and only happen in a select few cases and as such don't help differentiate good from good enough. The QAH system, which looks to be the equivalent of the "Quality Control" system (may need to be elaborated otherwise), helps to some degree, but isn't very consistent since there's a large amount of people with varying standards involved and is easily circlejerked or rushed without consequence, hence why the reward for involvement here is somewhat limited. As such, with the current way we track these things, the "score value" is really only feasible in determining who is struggling, not who is performing well, and that doesn't seem to change with this proposal, unless perhaps I've missed something?

Another thing, who do you propose will manage and develop the scoring system? For what it's worth we've attempted to work with the devs to make a scoring system in the past on multiple occasions but in the end received no support for this. The current "scoring system" isn't actually a scoring system at all, and turning it into the one as described here could take months to set up properly, especially when it comes to how it is interacted with and maintained.

Forcing the QAT to look through all members manually every month is a huge workload with the amount of data this requires, and will definitely lead to certain members needing to focus on management rather than checking or dqing maps, in order to keep everything running properly, as we have currently. This leads back to your issue of the QAT being too wide-reaching, and explains why this is currently the case, and probably always will be unless some other group is willing to support the QAT in managing things, who can be delegated this task. All the other proposals seem to incorporate a solution to this in some way, but it seems like it's missing here.
Please sign in to reply.

New reply