Announcement RBY Tiering/Viability Rankings Timeline (Discussion)

[phoopes]

Hi friends! First of all, I want to make it clear that this is less of an announcement and more of a discussion thread.

As it stands, OU does a yearly "official" VR update, and this year's is slated to take place after SPL is over. I think it's a great idea to have a yearly "official" viability rankings update, and I pretty much want to copy this idea for the lower tiers. However, the question is when exactly we want to do this... when trying to create a timeline I keep running into some potential issues. So that's where y'all come in! I'd like to hear the community's input on this, and hopefully we can come to a kind of consensus as a group as to what to do. Some things to consider...

• Do you think that we should do once-a-year "official" Viability Rankings for all lower tiers?
• What method should be used for lower tier Viability Rankings? (probably "vapicuno method," but thought I'd pose the question anyway)
• Should there even be a timeline? Or should we do Viability Rankings updates for lower tiers "when it feels right?"
• If we do go with a set in stone once-a-year timeline for lower tiers, what should that timeline look like?
• How does said timeline account for tier drops?
• How does said timeline account for any quickbans/suspect tests that may occur?

---

I'm now going to answer my own questions with my thoughts lol

I definitely am a fan of once-a-year "official" Viability Rankings for lower tiers. People post their own personal VRs all the time, which I think is a good snapshot of minor meta shifts during the year. But an overarching community Viability Ranking for the coming year I think is an easier resource to look at for new players and will definitely help define when tier shifts should happen.

I think "vapicuno method" is probably the way to go. A sub-question for this is do we have people are willing to do it/understand the statistics? I think the answer is yes. Another sub-question: how do we decide whose vote counts? This one I'm not 100% sure of, for 2021 UU rankings it was kind of a "you know it when you see it" kind of thing which worked but I would prefer a more set in stone system.

I think there should be an official timeline. Doing something "when it feels right" to me seems like a recipe for having disorganized, out of date Viability Rankings for lower tiers. Especially when it comes to tier drops/quickbans/suspect tests, which I'll touch on more in a later paragraph.

I'm going to answer the next questions kind of all at once haha. First of all, here's my proposed timeline:

1. 20XX RBY OU Viability Rankings happen immediately following SPL 20XX (set in stone).
2. A testing period of X months happens for UU to account for any tier drops that may have happened that year.
3. Quickbans for UU (if needed)
4. Suspect tests for UU (if needed)
5. 20XX RBY UU Viability Rankings happen
6. A testing period of X months happens for NU to account for any tier drops that may have happened that year.
7. Quickbans for NU (if needed)
8. Suspect tests for NU (if needed)
9. 20XX RBY NU Viability Rankings happen

So that answers how my proposed timeline would look. The biggest question is how many months would that testing period be... that I am unsure of. I think quickbans would probably happen within the first few weeks. From there, the meta needs a little bit of time to get settled before we figure out if anything needs to be suspect tested. If not, great! Move on to the Viability Rankings thread. If so, said suspect happens before the official Viability Rankings drop, that way the Viability Rankings aren't out of date almost immediately. Putting a time constraint on this of X months keeps the process moving along, I just don't know what that X should be. So yeah, it's like a "trickle-down" VR in a sense. In a world where tier drops are likely I wanted to propose something that accounted for those tier drops while still allowing for yearly viability rankings. I guess if we want it to all be within a year X should be 6 at max. Maybe that's more than enough time? Maybe it's too little? Idk, you tell me.

I think the "testing period" thing might be a bit of overkill considering the "tier by technicality" thing we have right now. Like people haven't tested UU a ton with Victreebel and Lapras but it seems like there's a general consensus that Lap would be a quickban and Vic is probably fine. However, I wanted to include a testing period in my proposed timeline anyway because the meta with Vic and Lap is definitely unexplored, even though we've been giving a whole year in theory to test it out.

---

So, what do you think? Please let us know in this thread/discuss on the Discord! And hopefully we can figure something out by the time SPL is over. Thanks for reading, and thanks in advance for your responses!

 

[pac]

Another sub-question: how do we decide whose vote counts? This one I'm not 100% sure of, for 2021 UU rankings it was kind of a "you know it when you see it" kind of thing which worked but I would prefer a more set in stone system.

I find this question decently intriguing, so I'm going to focus on it. My proposal is to have the UU qualifications be run in a similar manner to the Dragonite Suspect Test where it was based on tour performance. These qualifications would be:

• Qualification in the Dragonite Suspect Test
• Top 4 of RoA Olympics 6
• Top 2 of RBY UU Spotlight Tour #2
• Top 2 of RBY UU Invitational LCQ
• Won at least 2 games in RBYPL II
• Won at least 2 games in UUSD II

Then for future VR reiterations, any voters from the previous VR submission season would auto-qualify while new tour results are added.

Take this same idea as well and apply it to OU and NU (NU may be trickier due to a severe lack of RBY NU tournaments, but OU could definitely make this work).

Thoughts?

 

[Plague von Karma]

Ahhh, feels good to see this picked up. Lutra may have something to share here.

I would like to point out that a timeline has C&C benefits as well. If a Pokemon is added or removed from the VR, this could dictate how we add Pokemon to reservation lists, thus making it all more predictable and rooted in the community goings-on. When adding a format to the dex, I would add every Pokemon from the current VR, and let people work from there. However, this is subject to the VR's accuracy, something that Ubers has been in contention with. I'll get to that later, though.

Do you think that we should do once-a-year "official" Viability Rankings for all lower tiers?

Absolutely, this would be a massive help and solve a lot of historic issues with RBY tiering. It would also promote some gradual shaking up of lower-tiers to preserve interest. I see the "active tiering" of RBY as one of its biggest draws as an old gen, having yearly VRs akin to monthly usage stats would perfectly translate things to RBY. People in this community absolutely love to tier things and discuss rulings, so it works out.

I also see the ability to "qualify" for VR submissions through yearly tournament results as a thing for players to shoot for, adding another reward to participating in the new circuit. Perhaps you could use the circuit points for OU, so if you get a certain amount, you're in. For Lower-Tiers, you could use the same qualification I outlined for the Dragonite Suspect Test, sans suspect tournaments. Or just use that for everything, I dunno. Anyway, point is, integrating this with the tour circuit is very good. If we want to go around and link this in with C&C too, maybe use this as a way to determine who qualifies for QC positions, giving the tour players easier routes to getting badged.

What method should be used for lower tier Viability Rankings? (probably "vapicuno method," but thought I'd pose the question anyway)

Should there even be a timeline? Or should we do Viability Rankings updates for lower tiers "when it feels right?"

So far we've been shooting for the vapicuno method and it works just fine. It's better than anything else we've got, but I do think having a specific couple of people manage the data would be good, given the conflicts that happened as a result of the data interpretation for the UU VR.

I think a timeline is a great thing, as it allows one to see yearly progression of the metagame. I'd say suspend a VR for a year if the metagame's activity dwindles significantly. UU and NU benefit from the progression moreso than any other to me, given they are still developing. I guess Ubers is too.

I will say that Lutra's method was a bit iffy; the result of the voting has (allegedly) made VRs not very representive of the metagame, those being Stadium and Ubers. This was because they weren't well-advertised and didn't have great qualification requirements. Mikon had something to say to me regarding this before during my tenure as C&C Mod, I think she'll explain it better than I have, especially in regards to what makes the VRs inaccurate. I'd trust her with my life on this as a fellow Ubers enjoyer.

If we do go with a set in stone once-a-year timeline for lower tiers, what should that timeline look like?

After all one or multiple big tournament(s) for that tier on the calender, preferably near the end of the year. RBY OU is after SPL, so close to the end of the year should work perfectly. Me and Lutra wanted to do something like this before we stepped down, but it was difficult to figure out. I remember that we wanted Stadium and Tradebacks OU to be incorporated into it, including a Global Championship for the latter, but this is all uncertain now from what I can tell. Of course, it would be nice to see these as part of the discussion, given they're given similar prestige.

How does said timeline account for tier drops?

Follow the old tiering decision made by Hipmonlee and this yearly proposal should work as intended, providing predictable tier drops that allow for quick, easy suspect tests if people test the metagame as intended. However, I will say that the testing of Victreebel and Lapras in UU has so far not been too fruitful, since people see the latter as a nuisance in the metagame already. Perhaps this could be looked into further. We wanted it to be 2 VR revisions, therefore making drops take 2 years; however, revisions could be done earlier to increase the speed of drops, making it malleable.

I do agree that the test period is a bit overkill, and I think a few people grossly overreacted to the idea of Lapras and Victreebel dropping and forced a compromise that was not warranted. Speaking as the person who worked behind the scenes, I profoundly regret giving those people as much leverage as they had, because the result feels like putting your head in the sand and avoiding tiering in favour of some weird zen world. Perhaps making the initial drop zone VR count is the best option: they're put in the drop zone, if they're there the next year, they drop. This is predictable enough for anyone with more than 2 brain cells fighting each other for survival to understand.

One idea I had a long time ago was to use the RoA Spotlight's Staff Picks as a way to support the test periods of "by technicality" Pokemon, and those who top the ladder and spotlight tournaments could qualify for the eventual suspect test, which, in my opinion, could be a mandatory measure when a drop occurs. Maybe this could be explored further.

How does said timeline account for any quickbans/suspect tests that may occur?

Suspect test creation should be made as linear and easy as possible, the bureaucratic approach with Dragonite frankly sucked and I really didn't like it. I think a proper test should take no more than a week or two to get started once public opinion has been gauged, and thereafter, the suspect live tours (which imo were a very good thing) should be done in the next week. Therefore, a suspect test of a Pokemon should take 2 weeks at minimum, and preferably no more than 3. If something like Tradebacks were to get looked at for a tier (since you can technically remove the clause in any tier, not sure why that would happen though) then you may want something longer.

This quickness is helpful, as it accounts for multiple drops, see how Victreebel and Lapras are going right now. Should they drop, you could get them tested within a month if you do them concurrently, which I would personally prefer for a more qualitative approach. One Pokemon should not influence the perception of the other in the test, in my opinion.

If anyone has any questions, lemme know.

 

[Volk]

I'm going to take a quick stab at this. I might come back to this later and refine some points, but I feel like typing some thoughts right now so I will.

So right now we are following two different systems for VR updates and how they affect lower tiers. For our new tiers, namely NU and the nascent PU, we simply update the VR and the changes take effect immediately. When Venusaur rose to UU, NU lost it immediately and when Raichu and Poliwrath fell from UU, NU took them immediately. To me, this makes perfect sense and we should continue to follow it. These newer tiers don't really deserve any level of stability (especially considering UU hadn't even really settled at the time) so them seeing frequent and fast changes is a good thing.

For UU, however, we are operating a bit differently. Right now we send Pokemon that drop from OU to "OU by Technicality" and maintain the status quo until they begin to appear consistently below the OU Standard. They then finally drop to UU, though this hasn't actually happened yet under the modern system (I think this is a slight oversimplification, but don't focus on this). The logic of this system is to protect the more firmly-established UU tier from sudden major changes. It's somewhat preservationist. This system has a few issues that I think we need to address. The biggest one is definitely how we handle bans and suspect tests. Theoretically, we are supposed to be testing Lapras and Victreebel in UU right now in preparation for their possible drop. This comes with some difficulties. For starters, nobody really wants to do it. Playing practice matches with these two Pokemon seems pretty pointless, as it won't help you win matches and it isn't even really fun because the "meta" is so under-developed. Speaking for myself at least, I want to play UU, not UU + Lapras and Victreebel. Aside from this, it's tricky to examine two Pokemon at the same time. Ideally, we should be testing a meta with Victreebel, with Lapras, and with both to get the best idea of the health of each Pokemon. Again, nobody wants to do this. Furthermore, laddering in RBY UU isn't always easy, so how we are going to conduct a suspect test on a dropped Pokemon remains a little inconclusive. I don't have a solution to this problem right now, but I want to acknowledge that it is a problem.

As such, I don't really have a great solution for handling drops to UU (and eventually drops to NU, once it stabilizes). My only idea as of now is to kind of do it heuristically. I would generally default to quick bans when dealing with new Pokemon. This preserves popular tiers and reduces the chances of chain drops (which aren't really a bad thing, but it simplifies things a bit). I wouldn't ban everything immediately, but if a Pokemon seems unhealthy, just axe it immediately. For example, I don't want to see Lapras anywhere near UU. I don't want to test it, or suspect it, or anything like that. I want it to go from OU to UUBL, spending exactly zero days in UU.

Also, for what it is worth, I don't really want Victreebel in UU either. I'm more open to allowing it to drop and get tested, but I really don't think it'll be a good addition for the meta for a couple of reasons I don't really feel like getting into right now. For stuff like this, it honestly might be worth setting up a council or taking a vote on new Pokemon to determine if we even want to deal with them. Like if we can just cut to the chase with these two, especially Lapras, that would be great. I also want to note that all of these issues will inevitably hit NU as well. I hope to have the first true NU VR posted soon and I would suspect we would have a pretty stable NU within a year or two thereafter. If some random Pokemon in UU begins to struggle and lose favor, I assume the NU playerbase would want some kind of holding period to protect their meta and prepare for changes.

As far as general VR thoughts go, here:

• I think VRs should be staggered semi-annually. OU generally lines up with the end of SPL, meaning VR Season is April and the final version is posted around May. I think UU should go six months after (October or so, though December/January looks nice too with the Rotational Ladder being up). Then NU will go six months after that. Basically, I think OU and NU should see revisions in the Spring/Summer and UU and PU should see revisions in the Autumn/Winter (pardon my Northern Hemisphere bias). Ubers can go in either because I don't predict any drops coming from there anytime soon. I think this is a nice and easy to follow system that also helps changes in lower tiers (NU to PU) occur seamlessly.
• There is no reason not to use the Vapicuno Method for all of these VRs. The only real opposition to this is if the data set is too small, in which case I think a simple aggregation (average/median) is fine. I don't see this being a challenge until PU or even lower.
• As far as qualifications go, I'm not sure how hard and fast they need to be. We kind of know who the good players are and what kind of VR we should take seriously, so I don't really want to exclude anyone just because they may have been a little busy or just had one or two bad showings. It also doesn't jive super well with team tournaments. Like Thor helped the Viridian Victreebels win RBYPL though he ultimately didn't play any matches. I feel like his support in that tournament shouldn't go unnoticed and he should be able to submit a VR if he feels so inclined (I don't know if he has other qualifications, just imagine a hypothetical where this is his only qualification).
• I want a proper NU VR up by the summer and to start PU immediately after (though it may be considered Pre-Alpha at that point).
• UU is probably due for a new VR. I don't know if we want to do that now or wait until the next occurrence of its scheduled update.

I'll probably go back and edit some other stuff in later. This is basically just a stream of consciousness, so bear with me. Thanks for reading and look out for edits (I'll write a note in bold for ones that are non-trivial).

 

[Plague von Karma]

Not to completely divert from the thread, but since the topic of how we handle suspect tests came up, I'll give some thoughts on how I wanted it to go. This isn't a direct reply to anything Volk said, but adding to the points he's laid out.

The biggest one is definitely how we handle bans and suspect tests. Theoretically, we are supposed to be testing Lapras and Victreebel in UU right now in preparation for their possible drop. This comes with some difficulties. For starters, nobody really wants to do it. Playing practice matches with these two Pokemon seems pretty pointless, as it won't help you win matches and it isn't even really fun because the "meta" is so under-developed. Speaking for myself at least, I want to play UU, not UU + Lapras and Victreebel. Aside from this, it's tricky to examine two Pokemon at the same time. Ideally, we should be testing a meta with Victreebel, with Lapras, and with both to get the best idea of the health of each Pokemon. Again, nobody wants to do this.

So this is essentially what I meant by how there is no incentive for actually participating in a testing period. You could theoretically take the approach with an RoA Spotlight like I mentioned earlier, but it sucks, I don't think the ladder would be active and the tour would work but not be great. Maybe a New Frontiers-style tour could function...

The point of playing the matches is to develop the metagame and see if it's better or worse than before, but as mentioned by Volk here, it only really helps with policy, rather than anything competitive, so it's hard to justify making it work without further incentive. If this were a modern gen tier, it would function, but it doesn't right now, as modern gens work by having active ladders. RoA Room Tours are basically all you get right now, and tour nights no longer exist to support us, which was part of the intention of the system I helped the RBY Council put in place: they were experimental and Rage jumped at the chance to support tiering.

Furthermore, laddering in RBY UU isn't always easy, so how we are going to conduct a suspect test on a dropped Pokemon remains a little inconclusive. I don't have a solution to this problem right now, but I want to acknowledge that it is a problem.

In regards to this, I think the Dragonite methodology is fine. While nobody there has necessarily played a metagame with the Pokemon in it, they probably would in the coming weeks after it's dropped in some form, and they have a stake in the tier by qualifying at all. Therefore, I think that while objectivity may be worse, the results should end up fine. I trust and respect the intelligence of people who qualify for VR revisions, and using that as a metric for suspect votes and such seems fair as well.

As such, I don't really have a great solution for handling drops to UU (and eventually drops to NU, once it stabilizes). My only idea as of now is to kind of do it heuristically. I would generally default to quick bans when dealing with new Pokemon. This preserves popular tiers and reduces the chances of chain drops (which aren't really a bad thing, but it simplifies things a bit). I wouldn't ban everything immediately, but if a Pokemon seems unhealthy, just axe it immediately. For example, I don't want to see Lapras anywhere near UU. I don't want to test it, or suspect it, or anything like that. I want it to go from OU to UUBL, spending exactly zero days in UU.

Also, for what it is worth, I don't really want Victreebel in UU either. I'm more open to allowing it to drop and get tested, but I really don't think it'll be a good addition for the meta for a couple of reasons I don't really feel like getting into right now. For stuff like this, it honestly might be worth setting up a council or taking a vote on new Pokemon to determine if we even want to deal with them. Like if we can just cut to the chase with these two, especially Lapras, that would be great.

Perhaps using UUBL over OU by Technicality and then suspect testing to selectively drop the Pokemon that seem fine is the approach to take. UUBL technically means the Pokemon has dropped and got instabanned by the council, but...I don't know, it seems strange to "catch" the drop and then just leave it for whenever someone goes "let's suspect test this now". It seems vulnerable to not giving Pokemon with a legitimate argument for their inclusion due process for ill-thought reasoning. I do like the inevitability factor of Pokemon dropping because it means the tier can and will get spiced up, and allows for tiering to take place properly and predictably, properly being, well, experimenting with the new drops, seeing what stuff looks like, and getting rid of the problem children. The problem is that people don't like experimenting right now due to lack of incentive, which is where the system falls apart.

It's all a difficult conundrum, as the general approach to RBY tiering is that due to how stable much of the generation is, rises and drops should be rare. Maybe this notion is wrong, though...

 

[Lutra]

• I think it would be good to aim for an annual rankings for every tier. 2 years seems too long to wait for a new survey.
• I’d use my own method if I was the one doing the viability rankings.
• I think there should be a schedule. I think it can work similar to a tournament one.
• I think the schedule should follow the order of tiers (Ubers, OU, UU etc.) and be spaced throughout the year.
• Very difficult to find a good solution; but I think allow a month for a tour following the potential higher tier VR change and then do the new VR after that.
• I think you are just going to have to wait for the next VR to reflect the post-suspect-test metagame.

I will say that Lutra's method was a bit iffy; the result of the voting has (allegedly) made VRs not very representive of the metagame, those being Stadium and Ubers. This was because they weren't well-advertised and didn't have great qualification requirements. Mikon had something to say to me regarding this before during my tenure as C&C Mod, I think she'll explain it better than I have, especially in regards to what makes the VRs inaccurate. I'd trust her with my life on this as a fellow Ubers enjoyer.

Well the point of my method is everyone is allowed, there are no entry requirements. Following I think the conclusion of the first rankings, I added the ability for rankings to be weighted. The Stadium OU VR is weighted, and fairly appropriately I believe. Also I pretty much invited all the top players from that one recent tournament, so I can’t see that being an issue relative to other rankings. The Ubers ones weren’t weighted though.

They weren’t well advertised, but does the method cause that? Also, I’m not specifically aiming to cater for the consensus that is reached in a chat. I mean the chat may not represent the player base (e.g. some could be scared to put their opinions across) and also the chat can agree to dislike something even if it’s an optimal compromise. I’m sure there are flaws that hinder what I’m trying to achieve, but I haven’t been educated about them much so far.

 

[Hipmonlee]

I would allow everyone to submit a ranking until it becomes clear that poor quality submissions are a problem. People submit their rankings even when they wont be counted officially, and most of them seem altogether sensible. Even a couple of bonkers ones wont reallly matter because the system works on averages and prunes outliers.

 

[Plague von Karma]

Well the point of my method is everyone is allowed, there are no entry requirements. Following I think the conclusion of the first rankings, I added the ability for rankings to be weighted. The Stadium OU VR is weighted, and fairly appropriately I believe. Also I pretty much invited all the top players from that one recent tournament, so I can’t see that being an issue relative to other rankings. The Ubers ones weren’t weighted though.

They weren’t well advertised, but does the method cause that? Also, I’m not specifically aiming to cater for the consensus that is reached in a chat. I mean the chat may not represent the player base (e.g. some could be scared to put their opinions across) and also the chat can agree to dislike something even if it’s an optimal compromise. I’m sure there are flaws that hinder what I’m trying to achieve, but I haven’t been educated about them much so far.

The Ubers one is the one that received the most criticism from what I know, so this could be be where the problem comes from. I'd trust your math with my life, so I very much believe the weighting isn't the problem, but I do think there is a problem with poor quality submissions given the opinions that experienced players have laid out. Never did I say you were trying to cater to "chat consensus", as I also trust you to make the VRs in good faith from a neutral perspective. I was saying that the results of the VRs were quite contentious among at least a few good players compared to other VRs that used vapicuno's method, which can mean an all-manner of things. It's worth comparing methods to see which works best, right?

The most educated players should (ideally) have the best opinions, so the results should be largely positive on a voted VR, this is the discrepency that I was poorly trying to get across. If a final VR is not representative of a metagame in the eyes of experienced players, that is a problem.

The advertising, at least to me, could have played a part; I recall at least a few historically notable Ubers players not even knowing it was going on, though at some point you do need to blame the players themselves. I can't remember any names, as this was a while ago. I think further advertising on the forum could have helped a lot, though at the time, this subforum didn't exist, so it may not have helped there. Regardless, we should be looking at the future, as there is a clear case for improvement. RBY advertising channels are largely excellent these days, so this shouldn't be an issue if your method is employed today so long as they are used correctly. Hell, you could probably work with the staff to have it done in the RBY server itself. There's a lot that could be done here if we wanted an open platform VR submission system: play to the openness at that point. Maybe encouraging discussion somewhere could be good, so to get more "educated" decisions made at the time.

In terms of my own thoughts on the flaws in your method, I think the main one is that the ranks seem to go pretty far down. By the time E2 rank on the Stadium OU VR is reached, I don't think any of the Pokemon there would even work outside of like, Nidoking. The Viability % you mention also seems to drop significantly, so maybe there's something to consider there? Maybe there needs to be a stricter cutoff point? The presentation is fantastic and I like the little role mentions, so regardless of what method is employed, I would love for that to get used.

I suppose with that in mind, the discussion is "should submissions be open or limited?". I do like your weighted system; as Hipmonlee says, people make their own VRs frequently - it's almost a subculture in this community these days - so a lot of the time they do end up being fine, and weighting it is probably all that needs to be done. I doubt anyone would abuse the system with spam given it takes quite a bit of effort to do that with VR submissions, and even then, it would be noticed. Maybe a discussion phase post-results where it gets that last bit of pulling around could be worth it. If it doesn't prove to be a problem, let's roll with it. Anyway, I'm rambling at this point...

 

[Ika Ika Musume]

This is my opinion, but I think there is enough known about NU that this tier list:

https://www.smogon.com/forums/threads/rby-nu.3679758/

This is unacceptable to stay in its current form for like 6-12 months which is looking like the case in this proposed timeline. I do think we know more than enough about NU to make an ordered list and split the tiers and one would get the wrong impression of what is generally good and generally bad from that list which I don't think is the case in UU which I'd probably consider the list next most needing an update. How that would be enacted is a different question that I don't really have the answer to but I'd say that if you were purely talking about timeline then that should probably be close to the top. Testing from UU -> NU can probably come later as I'd guess it would probably be best to quickban these Pokémon to NUBL and then deliberate assuming there are drops.

 

[phoopes]

Something that got a lot of discussion in the Discord today was the idea of councils, so I'll try and relay some of that information/my thoughts here:

At first, I thought that having a bunch of councils for the lower/other tiers wasn't necessary and that the current RBY Council could just handle everything when it comes to tiering decisions, suspect tests, etc. However, after some discussion I feel like that's not the best idea. For one, it forces the current council to stay up to date and on top of a lot of different metas... OU, Ubers, UU, NU, Stadium, Tradebacks, and possibly even PU should the ball get rolling on that any time soon. I think that's just an unfair ask for them, as well as unfair to anyone who's actually active/interested in those tiers to have the decisions being made for them. In addition, Amaranth for example mentioned he didn't really have any interest in making decisions for UU, so having uninterested people lead a tier's decisions doesn't seem all that feasible either. I say that the following is probably the order of operations:

1. Decide what tiers we want councils for.
2. Come up with a list of qualified and interested candidates for each tier.
3. Go to the current RBY Council for approval.

For number one, I was originally thinking councils for Ubers, UU, NU, Tradebacks, and Stadium. However, after some feedback from the Discord I think the only ones that really need a council are UU and NU. Because realistically, what's going to change with Ubers? And Tradebacks and Stadium don't quite have the playerbase yet to warrant councils (though hopefully we can grow these metagames). For number two, in the Discord there were already ideas floated for interested/qualified people for the UU Council, and I can think of people that would fit on the NU Council as well, so no trouble there. Again, Ubers might not need one, and Tradebacks/Stadium you're really struggling to find people. So that just leaves number three, whether the current RBY Council approves or not of this proposal. Guess we'll see.

---

I'll probably respond to other stuff that was mentioned in this thread later, but for now I wanted to get the council stuff/proposal out there before it gets lost in #general on Discord

 

[phoopes]

Alright, now that I have more time, going to try and tackle the rest of the points here.

---

VR Qualifications:

As I said, I would prefer a more set in stone system than the one we had in place for the 2021 UU viability rankings, which was sort of a "you know it when you see it" thing. I'm hesitant to say there should be hard qualifications to meet because of the example that Volk gave with Thor. But I'm also hesitant to say anyone should be allowed to have their rankings counted - even though outliers and stuff are accounted for in both the vapicuno and Lutra method if I'm understanding them correctly, just in different ways - since you could in theory get some people that have no clue what they're talking about have their rankings counted. So I think the solution is kind of a mix here... I think having hard and fast qualifications as an initial pool of people that get invited or whatever to fill out their rankings is probably a good thing. From there, anyone else could submit their rankings and it would be up to the theoretical council of the tier to decide whose additional rankings count (if any). If the council idea falls through, then it can just be the pool of already decided players.

So like for example, let's say 20 UU players meet the tournament qualifications (don't take this number too literally). Those 20 would be invited to submit their rankings for the 2021 VR. But say an additional 10 players want to submit rankings as well. It would then be up to the council (if we get approved on councils) to decide which of those 10 additional VRs get added to the pool. If the council thing doesn't work out, then the initial 20 would decide whose gets counted and whose does not.

I think this is a good way to incentivize people to participate in tours, because if they prove they can do well then their vote will be counted. However, it doesn't penalize people that are better at supporting, have a fluky bad tour, or are too busy to participate in tours but still know the tier well through stuff like Discord/forum discussions or C&C.

---

Method:

After some more discussion on Discord, my own opinion is that the vapicuno method is a little better, at least for UU and NU. No disrespect to Lutra and his method, because I think it has its merits, but I think for the solution that I outlined above for VR qualifications, the vapicuno method is at least a little better for tiers that have enough tournament results to merit a qualification list. Currently, OU, UU, and arguably NU have enough representation for enough players to qualify via tournaments, plus add in whoever else the council thinks is qualified. These tiers are the ones with a big playerbase, so I think it makes sense to use the vapicuno method for their VRs because it's tried and true with OU and we also used it once for UU with minimal issues.

In the case of a tier where there aren't enough data points for tournament qualifiers, I feel like the Lutra method would work. From what I understand, anyone can submit a list and it's weighted somehow to have the "good" VRs near the top and the not-so-good ones at the bottom of the weighting. I think this is better than vapicuno, which if I understand correctly, is kind of an average of all the rankings, minus outliers. With fewer data points, the outliers would have more of an influence, and as such I think a weighted system is better.

All in all, I think we keep vapicuno method for OU, UU, and NU, since those are the ones with good tournament representation at the moment. For everything else that has a smaller playerbase and not a lot of tournament data, we use Lutra's method, with the intention of transitioning to the vapicuno method once there's enough data.

Keep in mind that I was a math major in college but flunked out lol, so I'm not the math/stat expert by any means here, and correct me if I'm wrong about anything I said with the methods.

---

Timeline/Proposed Calendar:

I think I like Volk's idea of staggered yearly rankings, where OU and NU are grouped together, and UU and (theoretical) PU are grouped together six months later. OU is already right after SPL and I think that works great. UU is then six months later which I think should be plenty of time for the meta to settle/develop following any tier drops and subsequent suspect tests (or suspect tests in general). From there, it's six months to NU, six months to PU, etc. However, I do think that this year might have to be a little different because I agree with another point that's been brought up, that the current state of the NU Viability Rankings could be a little better. So here's what I'm proposing:

February 2022:
-NUSD ends

-First official NU Viability Rankings happen immediately after NUSD, using vapicuno method with qualified players from RBYPL II, NUSD, and other players that want to submit their VRs that are included based off my solution above, with either the council or the already qualified players picking them.

-With the first official NU VR established, PU begins

March 2022:
-SPL XIII ends

-OU Viability Rankings are done after in the normal manner, but emphasize a certain deadline rather than just have them be completed whenever in April/May.

September/October 2022 (six months after OU VR):
-UU Viability Rankings happen using the vapicuno method with qualified players coming from the various tournaments UU has had, plus any other players deemed eligible by the council/already qualified players

-Preliminary PU Viability Rankings established. Idk exactly how this will be done but I think it's good to have it on the schedule.

March 2023:
-SPL XIV ends

-OU Viability Rankings are done as normal

-NU Viability Rankings are done as normal

September 2023:
-UU Viability Rankings are done as normal

-First official PU Rankings, with ZU being established if there's any interest


I think this is a good solution for a few reasons. It gets the ball rolling soon on NU Viability Rankings (which desperately need an update), with it only being a year + a month or two until the 2023 Viability Rankings. This also gets PU rolling for the people that want that. Six months later, we have the next UU Viability rankings, which are only occurring a year + a month or two after the 2022 Viability Rankings. We also have an initial PU Viability Rankings... not an official one because it's only been a few months so there won't be any tournament results but maybe this is where the Lutra method comes into play. Six months later, we're really on schedule with the OU and NU Viability Rankings happening as normal, and six months after that, we have UU rankings happening as normal and the first PU Viability Rankings with the vapicuno method. From there, we continue with this system.

As to where to fit Ubers, Stadium, and Tradebacks, I'm not 100% sure, which is why I didn't include them. Open to feedback.

With regards to tier drops, I think that six month gap from OU to UU, UU to NU, etc. should be enough to quickban stuff, suspect test stuff, etc. as long as we have people on top of the game. Which I think we will should the council proposal go through. I think we're getting some answers/feedback from senior staff on that soon, so let's see how it goes.

---

Another long post, but hopefully it's helpful. As always, I'm definitely open to any feedback on my ideas/discussion they generate. Thanks for reading!

 

[vapicuno]

Just want to add that my method can be adapted to include weights as well. Earthworm used it in the GSC Ubers VR in 2019 here -> https://www.smogon.com/forums/threads/gsc-ubers-viability-rankings.3654941/post-8244586

Back then I didn't have code for automatic tier determination and fancy graphs, and everything was done in a spreadsheet. But the code can easily be adapted to include weights.

My opinion on weighting though is that there are too many practical downsides. Meaningful weights can be produced only if those determining the weights understand how to scale them (i.e. decide the numbers) to get the representation that they want. It's hard to be transparent about how one arrives at the weights and the representation they're looking for. For example, I can create the following simple situation of two categories: you have 5 experienced players and 20 less-experienced players; if you want equal representation from these two groups, you would be looking at a weight of 1/4 for the less-experienced players - simple enough. In reality though, you have a spectrum/continuum of abilities/experience/results, which people will inevitably disagree on, and that creates a huge political and mathematical mess for doing meaningful weighting imo. On what basis are we going to assign someone a weight of 1/2, 1/3, 1/5, or 0.14159 to arrive at the representation that is desired? Does the result of the weight-assigning exercise, after considering so many factors, clearly achieve the desired representation? It's a lot more transparent to have an in-or-out system, or at best, an inner circle and outer circle of voters.

 

[Lutra]

For the Stadium OU VR, I recorded ELO, and rounded the ELO ratings to the nearest 100. I then did, 1x, 1.4x, 2x, 2.8x, 4x weightings (roughly multiply by root 2) as players go up 100 points each. Not a precise way to output, but follows the gist of higher rated players having more voting weight corresponding to what proportion of games they are expected to win. Including more tournaments should help.

I don’t believe in any kind of outlier removal. I think people are likely to copy each other if they lack information and it could be possible that someone just knows the metagame better than the consensus of players do. It’s possible the outlier is more accurate for some aims. I can’t really think of a way to have clarity about it, so I just leave them be.

Formulae are generally quick to change in this kind of stuff. For me, dealing with all the input, like descriptions etc, was the most challenging thing.

 

[vapicuno]

For the Stadium OU VR, I recorded ELO, and rounded the ELO ratings to the nearest 100. I then did, 1x, 1.4x, 2x, 2.8x, 4x weightings (roughly multiply by root 2) as players go up 100 points each. Not a precise way to output, but follows the gist of higher rated players having more voting weight corresponding to what proportion of games they are expected to win. Including more tournaments should help.

Okay, I'm going to nerd out here, since using ELO for creating weights is an interesting mathematical problem to think about.

My understanding of your premise is this - a pool of voters plays a best-of-x (where x is a large number) round robin against every other voter. For every game you win, you get a vote share. If you lose, you don't get that share. Simply put, the weight you get is proportional to your ability to win against the whole field of players.

It looks objective, but there are several issues, and I'll outline what I think is the biggest issue in its current form compared to some kind of qualitative VR methodology for smaller volume tiers and an in-out system for larger volume tiers. It is that one still does not know what the representation is, and that the representation depends on more than the elo-determined weights.

To tie back the weights to representation, which everyone is ultimately interested in when assigning these weights, I'll borrow a concept from economics - the Lorenz curve - which describes income distribution. The x axis represents the fraction of voters in increasing order of merit (in this case, elo), and the y axis represents their weighted fraction of votes. I have plotted the curve below from the spreadsheet here that you created for the RBY Stadium OU 2021 VRs.

How to read this: the top 20% of voters (0.8-1 on the x axis) take up 40% of the weighted vote share (0.6-1 on the y axis), meaning they yield 40% of the influence. Follow the dotted lines if this is confusing. Likewise, the bottom 80% of voters yield 60% influence. For reference, the dashed straight line represents full equality in representation (one person, one vote).

The question that ultimately needs to be answered is: Given a weighting system, is the community willing to accept such a representation, as visualized by the Lorenz curve above? In my opinion, the community opinion should drive the weights, not the other way round based on logic that gets lost in an abstract basis that nobody really understands (feel free to disagree with me on this point). If the community is unwilling to accept such a representation just based on common sense of knowing who the players involved are and their skill levels, then that's a problem.

Another issue is that this question has to be asked every time the VR is tabulated. The curves will always look different, even if you use the same weighting methodology, because at the sample sizes we have, there is no way you will get close enough to a logistic distribution of elo. With 9 voters in this sample, the distribution looks like this, which in doesn't look like a bell curve (this opens up another can of worms which I'm not willing to go into right now). I'm pretty sure bringing it up to 20 voters isn't going to change anything.

TLDR: Math should serve, not dictate, the community; to do so, weights used should be tested by looking at the representation curve.

 

[phoopes]

First of all, I just want to say I appreciate all the math above. There's definitely some more discussion to be had here, but after discussion on Discord I think I want to amend my previous statement about using Lutra's method until we get enough tournament results for vapicuno's method. The intention that I was going for here was actually using a weighted method until we get enough data for an unweighted method, since it looks like both Lutra and vapicuno's methods can be either weighted or unweighted, as described above. For me at least, a small pool of players means a small amount of rankers, so we're really scraping for data points. Thus, we should allow everyone who wants to to submit a VR, but then weight it so that the better/more experienced players' votes count more. This is very subjective and is going to be different every time we do it, but I believe in us as a community to make those determinations when the time comes. I'm willing to bet there could be a hiccup or two along the way, but I really do think that's the best solution right now for Stadium, Tradebacks, and - when we get there in about 6/7 months - PU.

Thus, that means for Ubers, OU, UU, and NU, I think we should be using an unweighted method, where people qualify through certain objective requirements to create an initial pool of voters. As listed above though, I would encourage anyone who wants to to submit a VR. It's just that it would be up to a council or the initial pool to decide whether or not which (if any) of those additional VRs would be counted. As such, I've come up with a list of requirements that would give us an initial pool for UU and NU. I'm not touching OU because they seem to have that on lock as far as what works, and Ubers I don't quite know enough about to make a determination on that (more on Ubers late). So anyway, here's what I got:

UU Initial Voting Pool Requirements:

• Top 4 from RoA Olympics 5 (kjdaas, FriendOfMrGolem120, Torchic, Amaranth)
• Top 4 from RoA Olympics 6 (Torchic, FriendOfMrGolem120, Koalacance, Unowndragon)
• Top 2 of RBY UU Spotlight Tour #1 (Volk, Sevi 7)
• Top 2 of RBY UU Spotlight Tour #2 (Mikon, chuva de perereca)
• Won at least 2 games in RBYPL (Thor, kjdaas, Sevi 7, iKiQ)
• Won at least 2 games in RBYPL II (phoopes, Volk, Lusch, Shellnuts)
• Won at least 2 games in UUFPL (Lusch, Holly, Shellnuts, phoopes, AM, pac, Justamente)
• Won at least 2 games in UUSD II (Koalacance, Torchic, Amaranth, pac, phoopes)
• Some cutoff from RBY UU invitational (yet to be determined)

So now here's the "why." The above are all the major tournaments that UU has had in 2020-2022 (that I can think of anyway), and show that the qualified players have some skill in and know the tier. As you can, a lot of these players have qualified more than once, showing that they really do deserve to be here. You may have noticed, however, that I didn't include players that qualified from the Dragonite suspect tours. This was kind of controversial in the Discord, but I have my reasons. First of all, is that these were live tours. Live tours means not everyone has the chance to participate in them, for a number of reasons, including but not limited to poor timezone compatibility or just being busy during said live tours. The other reason is that these were single-elimination tournaments with best of one series. Maybe I'd be more inclined to include the Dragonite suspect tours if it were best of three, but I think pretty much any RBY tournament that doesn't do best of three isn't rigorous enough to be included in this list of tournaments. Besides, of people who qualified through these live tournaments, some if not all have been invited to the invitational anyway. I feel like they can prove their skill there. And even if for some reason they still don't qualify, they can always submit a VR to be potentially approved by a council/the initial pool of players, like I mentioned above. With that said, our initial pool of players is as follows:

UU Initial Voting Pool:

1. kjdaas
2. FriendOfMrGolem120
3. Torchic
4. Amaranth
5. Koalacance
6. Unowndragon
7. Volk
8. Sevi 7
9. Mikon
10. chuva de perereca
11. Thor
12. iKiQ
13. phoopes
14. Lusch
15. Shellnuts
16. Holly
17. AM
18. Justamente
19. pac

I think that's a pretty solid initial pool. And there will be more added (potentially) through the invitational and the other submitted Viability Rankings that are taken into account. Keep in mind though that this is not a final decision, just my proposal. I'm open to any/all feedback from the community before coming to a decision! Besides, in my proposed timeline the next UU VR update wouldn't be for another half year+ anyway, so there's time to change things around if need be/something comes up.

Something that's a bit more time-sensitive, however, is NU. NU is an official tier but does not have official Viability Rankings yet, which is why I'm hoping to get those done shortly after NUSD is over. Now the same question comes up as it did for UU: who qualifies? There haven't been as many NU tournaments, but I still think we should use some tournament data for this. However, I'm going to be a little more lax on requirements. In addition, I'm going to add members of the QC team that haven't met tournament requirements to give us a little bigger of a pool. Ideally, for the 2023 installment, there are going to be enough tours where we can have a similarly rigorous list as we did with UU. But for now, I think the following works:

NU Initial Voting Pool Requirements:

• Played in at least 3 games in RBYPL II (Khaetis, Serpi, Enigami, shiloh, Oathkeeper)
• Played in at least 3 games in NUSD II (Serpi, Ika Ika Musume, Khaetis, Thor, Oiseau Bleu, Enigami, Koalacance, May, pac)
• Top 2 in RBY NU Spotlight Tour (Earthworm, Hipmonlee)
• Member of the NU QC team (Volk, Enigami, Shellnuts, Sevi 7, S1nn0hC0nfirm3d, SBPC, phoopes)

Some of you my balk at the fact that there's no win requirement for team tournament participation like there was in UU. However, I feel that in this case, participation in at least three games is enough. First of all, being drafted in these tours is proof that there was at least some prior-known skill in this new tier. Secondly, it shows that the team that drafted these participants trusted them enough to start for multiple weeks, even if they did not win a game (which is the case for some). With the lack of other tournaments, I'm deeming this to be decent enough. This is the same reason why I'm reaching back to 2020 for the RBY NU Spotlight Tour: lack of anything better. I know the meta was completely different back then, but again, we need data points. This still only brings us up to 13 people, however, which is why I'm including the QC team as well. Clearly, the people on the QC team have enough knowledge about the metagame to be on the QC team, so even if some don't have tournament participation/results, I'm willing to let them into the initial voting pool. Thus, our initial voting pool is as follows:

NU Initial Voting Pool:

1. Khaetis
2. Serpi
3. Enigami
4. shiloh
5. Oathkeeper
6. Ika Ika Musume
7. Thor
8. Oiseau Bleu
9. Koalacance
10. May
11. pac
12. Earthworm
13. Hipmonlee
14. Volk
15. Shellnuts
16. Sevi 7
17. S1nn0hC0nfirm3d
18. SBPC
19. phoopes

This brings us up to 19 voters again, same as UU, which I think is fair considering the approximate size of the playerbases. This also doesn't include any additional VRs submitted that may be approved by a council/the initial pool, just like UU, so there's that. You might argue that we're trying to push out official NU Viability Rankings too soon which the kind of requirements that are being laid out here, but I'd argue that it's better to get an official one out now than wait another however many months/year to do it. This will get us on a timeline sooner rather than later as well as let people start looking at PU, which I know has a sizeable demand from talking to people in the Discord. Then we can hopefully line up a preliminary PU viability rankings with the aforementioned UU rankings, going off the timeline I proposed earlier. Again, as stated in the UU section, this is not a final decision, just a proposal. So please, let me know your feedback!

---

Finally, I want to bring up other tiers... Ubers, PU, Stadium, and Tradebacks. I think it's a fine idea to get yearly viability rankings for these tiers as well, I'm just not 100% sure how to approach them quite yet...

Ubers: It looks to me like there may be potential for Ubers to have voting requirements similar to NU, as there's already been an unofficial Ubers Viability Rankings (thanks Lutra!) as well as its inclusion in RBYPL and UWTT. However, I'm not the expert on that which is why I was hesitant to cook up requirements/a voting pool. I could probably do it, but I'd like some other input from the community. Please let me know what you think!

PU: I think after the establishment of NU Viability Rankings in the coming month we can kind of feel it out for the first few months and then see what we come up with once it's time (in the proposed timeline that lines up with UU anyway). I know that's kind of just kicking the can down the road a little bit, but I still want it to be on the radar. Finally, Stadium and Tradebacks: it's cool that there's Viability Rankings for both of these right now but

Stadium: Again, I think it's great that we have an unofficial Stadium Viability Rankings (thanks again to Lutra!) but I'm kind of unsure how to move forward with official Viability Rankings because of the lack of tournament representation. That could just be on us to host more tournaments and stuff like that to promote interest in the tier. We'll see, but as always, feedback is requested.

Tradebacks: Basically what I said about Stadium lol.

---

Anyway, that's another long post but I'll try to give a tl;dr

tl;dr: Even though I came up with UU and NU VR voting requirements and pools, those are just proposals. I'm open to feedback on requirements/who should get a say. For other tiers, I could definitely use a little more help, especially Stadium and Tradebacks.

Thanks so much for reading!

 

[AM]

I agree with not including those from the Dragonite Suspect more so because there is already a larger more relevant group to pick from with the increase of RBY UU Tours and its placement in team tours. Honestly you can just not include the ones from RBY UUFPL (I'm aware I would be cutoff) because that felt like awhile ago compared to everything else and I think most of the people in that pool are already or going to be part of the RBY UU Invitational, so they would fall under that criteria as far as I'm aware. Worst case scenario you just pick and choose who you want a ranking from I don't think it's too big a deal to have some subjective criterias here and there in terms of picking people.

 

[Plague von Karma]

So now here's the "why." The above are all the major tournaments that UU has had in 2020-2022 (that I can think of anyway), and show that the qualified players have some skill in and know the tier. As you can, a lot of these players have qualified more than once, showing that they really do deserve to be here. You may have noticed, however, that I didn't include players that qualified from the Dragonite suspect tours. This was kind of controversial in the Discord, but I have my reasons. First of all, is that these were live tours. Live tours means not everyone has the chance to participate in them, for a number of reasons, including but not limited to poor timezone compatibility or just being busy during said live tours. The other reason is that these were single-elimination tournaments with best of one series. Maybe I'd be more inclined to include the Dragonite suspect tours if it were best of three, but I think pretty much any RBY tournament that doesn't do best of three isn't rigorous enough to be included in this list of tournaments. Besides, of people who qualified through these live tournaments, some if not all have been invited to the invitational anyway. I feel like they can prove their skill there. And even if for some reason they still don't qualify, they can always submit a VR to be potentially approved by a council/the initial pool of players, like I mentioned above. With that said, our initial pool of players is as follows:

While the reasoning is otherwise sound, I strongly dislike the bolded one. "People couldn't attend for x reason" is not and should never be a reason to exclude a tournament, otherwise you may as well throw out a bunch of UU/NU tournaments for not having Volk, Sevi 7, or Shellnuts attending every time. It doesn't matter if a tournament didn't have all the notable competitive players in it, it would only matter if there was a significant absence of them, which is simply not how it played out. There were plenty of notable players in attendance and just giving the brackets a passing glance would tell you that. Additionally, the tournaments were played on various time zones at the end of the week SPECIFICALLY to make them as open to players as possible, and people made excellent use of this. We went with what has historically worked for RoA Tour Nights, and we got rewarded with a diverse range of players. At the time there was a significant demand to just get the Dragonite test done already, and we followed through with that strong, persistent, talked-about-daily demand. When running these tournaments, I got one single complaint from a user regarding being able to attend, and that player had already qualified and was already notorious for being unable to attend despite their skill anyway, so they didn't need to join. The "stackedness" of the events was fine, there were enough strong players in the bracket to remove the ones who didn't deserve to qualify, it played out just fine. The implied notion that the single-day nature of these events somehow hurt the player pool is thoroughly disingenuous, disregards the results, and only serves to hurt the overall argument. When determining notability of a tournament, a significant and often sole determining factor to consider is the player pool, and here I would absolutely say it was fine for the time.

Best of 3 could theoretically have been possible in these tournaments, but was excluded at the last minute due to time concerns. I think you could argue either way for whether this notion was wrong: the round robin qualification finals could have been much shorter had they been Best of 3, but all-in-all, I think that the tournament would have gone for significantly longer otherwise. Keep in mind, these were one-day tournaments, we didn't have time for loser's brackets and Best of 3 alone could have made the tournaments take an exceedingly long time. We were after a good test period that allowed for players to qualify without arbitrarily lengthening the test for those who already qualified; remember, this test took ages to even get off the ground. The players who qualified as a result of this were all respectable and deserved it. The voter pool prior to the tournaments was small and could thus have been easily swayed, and I believe that said pool was better off from being increased above 20. You don't need to try and project that the player pool was somehow diluted, because in practice, it wasn't. The only one I would even begin to argue was somehow worse player-wise was the third one, which still had Oiseau Bleu, Amaranth, you(!), egalvanc, and Ruft in them. The first even had one of the best players of this metagame of all time in attendance, who did so solely to kick off inexperienced players.

If you want to argue for their removal - which I see the argument for - the correct way to go about it is to remove it on account of it not being Best of 3, but even then, it seems silly. Why? Because you end up saying "if for some reason they still don't qualify, they can always submit a VR to be potentially approved by a council/the initial pool of players". In that case, what's the point of excluding them at all? The players who would have qualified were fine, they had perfectly good bearings of the metagame, and I don't see their VRs being cut. This is just so odd. If they all have a high chance of their VRs being accepted, really, what's even the probem? This feels like red tape being there for the sake of it.

Stadium: Again, I think it's great that we have an unofficial Stadium Viability Rankings (thanks again to Lutra!) but I'm kind of unsure how to move forward with official Viability Rankings because of the lack of tournament representation. That could just be on us to host more tournaments and stuff like that to promote interest in the tier. We'll see, but as always, feedback is requested.

Tradebacks: Basically what I said about Stadium lol.

Stadium has had a Global Championship and Spotlight Tour in recent times, which should already set up a small but relevant voter pool. Stadium simply needs more tournaments, and setting up individuals on the forum feels like a strong way to build this up. I strongly believe New Frontiers-style tournaments like what Pokemon Perfect did would help with this, being a simple Round Robin people can enter and leave at any time. Perhaps this could be given a little tweaking (eg. maybe adding a hard limit or modifying join/leave rulings) but it's absolutely the best way to go for tiers that people continually call "underexplored", "untested", etc. I absolutely adore the "join or leave at any point" concept since it acts as a kind of matchmaking system while adding stakes, it's just a really nice way to encourage actually testing the tiers.

Also, bring back Triple Threat OU, that tournament was so great and could be used to form qualifications for both of them really quickly.

Ubers: It looks to me like there may be potential for Ubers to have voting requirements similar to NU, as there's already been an unofficial Ubers Viability Rankings (thanks Lutra!) as well as its inclusion in RBYPL and UWTT. However, I'm not the expert on that which is why I was hesitant to cook up requirements/a voting pool. I could probably do it, but I'd like some other input from the community. Please let me know what you think!

There is plenty of tournament history over the past 2 years, and even a little more going back further, thanks to Lutra and RoA's continued support for the tier. Voting requirements are easily drawn up.

RoAO
There's some history in RoA Olympics and the like here, and if you're using it for UU, you should use it here.
RoA Olympics 5 would be; kjdaas, Luirromen, SMB, and Frrf
RoA Olympics 6 would be; 64 Squares, ACII, FOMG, and Fc
Every player here is known to be pretty fantastic, really, I have no issues here.

If you want to go back further, these go into the previous decade, and there is merit in allowing these players based on their extensive history. However, you would likely want to extend UU's qualifications, but I remember there being minimal pool difference from there.
RoA Olympics 4 (2019): Lunala, Lusch, Hipmonlee, Nails
RoA Olympics 3 (2018): parivard, Caetano93, Ranshiin
3 was 3-way finals.

RBYPL
Players that played at least 3 games in RBYPL I: HSOWA, 64 Squares, Lutra, Mr.378, Ranshiin, Oiseau Bleu
Players that played at least 3 games in RBYPL II: 64 Squares, Ctown6, Mr.378, Oiseau Bleu

Under the NU Requirements, Sevi 7 is excluded, but HSOWA qualifies. However, if you use the UU requirements, the reverse happens, where Sevi 7 qualifies from his RBYPL II results. I believe that an "or" clause to qualify them both would be perfectly fine, both are fine players that would give quality feedback. Plus, you increase the player pool.

So, with an or clause, you would end up having it look like this:
Players that played at least 3 games OR won at least 2 games in RBYPL I: HSOWA, 64 Squares, Lutra, Mr.378, Ranshiin, Oiseau Bleu
Players that played at least 3 games OR won at least 2 games in RBYPL II: 64 Squares, Sevi 7, Ctown6, Mr.378, Oiseau Bleu

RoA Spotlight Tours
Top 2 in RoA Spotlight Tour 2020: Peasounay, queso
Top 2 in RoA Spotlight Tour 2021: kjdaas, Khaetis
kjdaas double-qualifies given RoAO, all the other players are quite good. I don't remember if queso interacts with RBY these days but the replays were solid.

Overall
If I'm to list them now;

• kjdaas
• Luirromen
• SMB
• Frrf
• 64 Squares
• ACII
• FOMG
• Fc
• HSOWA
• Sevi 7
• Lutra
• Mr.378
• Ranshiin
• Oiseau Bleu
• Ctown6
• Peasounay
• queso
• Khaetis

Overall player count: 18, 25 if including RoAO 3 and 4.
Incredibly diverse.

 

[phoopes]

In response to May's post...

Live Tours/Dragonite Suspect Tours/UU qualifications:

Alright so in short, after this and some talking on Discord, I concede the point. The six people that qualified in Dragonite suspect tour tests should also probably qualify for the UU Viability Rankings. I'll admit that I was completely wrong about not liking that they were live tours due to what May said above. I still don't like that they were best of 1 tours, but it's come to my attention that Ice Yazu, Lily, meloyy, Jyuux, May, and Inmundo are still active in the tier (either through tour nights, participation in the invitational, etc.) so I was wrong in that regard too. So adding them, that brings us up to 25 people that would participate in the UU VR. This might seem a little high considering the last OU VR had 22 people, but I think that all the players have proven their worth through tournament qualifications. To AM's point about maybe cutting UUFPL results... I think I'd rather keep them in. Cutting UUFPL cuts AM, Holly, and Justamente. Holly and Justamente are playing in the invitational so it's clear they still keep up with the meta, and I know AM does too (even if they have spicy VR takes I've always found them very helpful). So if a majority of people agree that keeping the UUFPL results in is good, then I am as well.

More Stadium/Tradebacks tournaments:

I agree with this. While I personally would most likely not participate because these metas aren't my cup of tea, I think anything that grows the wide variety of tiers that we have is good. Establishing official Viability Rankings for these tiers can only be helpful, and we may see a kickstart in activity that grows the playerbase, much like we saw with UU and NU. Who knows, maybe Stadium UU/Tradebacks UU will become a thing eventually. We'll see.

Ubers Qualifications:

I like the list of 18 that May drew up. I think it's a diverse pool of good players, and the requirements from RoA Olympics 5 and 6, RBYPL I and II, and the Spotlight tours make sense. I'd hesitate to reach back any further than that to 2019 and 2018 just for the sake of getting more players... you're already looking at six tournaments from the past two years. That's pretty good for an old gens non-OU tier. Not trying to exclude anyone specifically because from what I understand, the folks from RoA Olympics 3 and 4 are good players, but reaching back to 2018 and 2019 seems less than ideal to me.

---

Again, thanks for the feedback! I'm glad we're getting community discussion on all this stuff.