Discussion The Precedent of 3rd Suspects of the Same Element

LBN

LBN

is a Community Contributoris a Top Tiering Contributor Alumnus
As I'm sure most are aware, National Dex is currently suspecting tera again in ND OU, and I've seen discussions of the prospects of holding a 3rd Kyurem suspect in SV OU as well. For those unaware, attempting to tier something in this manner 3 times in a row is entirely unprecedented across Smogon's history. The closest that exists are further add-ons to Baton Pass clauses, or BW Latios, who had been put on a slate 4 times back when suspects often had multiple Pokemon on a suspect at the same time, a now seldom-seen practice. (To my knowledge the last multi suspect was SS NU's Bewear + Pangoro). With the advent of this precedent created, I'd like to discuss a hard-line to be drawn as to when a 3rd suspect is too much, what procedures should be followed, and what procedures should be considered to be made to make sure these tests aren't an attempt to force a result through.

Some backstory on these examples are in order. Firstly, Natdex OU's current Tera suspect poses an interesting argument. The case of 2 suspects rule has been made originally before the advent of DLC releases becoming the standard for this franchise. As of now, Tera has been suspected in every single DLC meta. This raises an interesting question as to whether and how DLC releases should be considered in this rule, and should there be a metric to consider on how different a meta must be from the former if at all. For this suspect as well, It's worth mentioning the complete lack of a policy review thread in advance, on a non-Pokémon element at that.

As for score metrics on the prior survey, the enjoyability was a 6.1/10 and the balance was scored at a meager 5.53/10. For some comparison, SV Ubers received an enjoyability of 6.22/10 and a balance metric 5.66/10 on it's survey prior to the 2nd Miraidon Suspect. A tier as volatile by-nature as SV Ubers getting a higher score at a point of record-levels of discontent for the tier is a strong indicator of something being horribly wrong. Closeness of the suspects is also a worthy topic, and the suspects were admittedly quite close both of which being roughly around the 55% range of the needed 60%.

As for the prospects of a potential SV OU Kyurem 3rd run. Some preface, I'm going to argue on the idea that a 3rd SV Kyurem suspect is on the table. If I find out this isn't the case, I would like this thread to continue on the notion that it is, purely for the sake of creating proper procedure. For this one, the outlying points are far more interesting and traditional. A 3rd Kyurem suspect would see that they occurred all within DLC2. This is what I'd consider an important distinction to make, as unlike what you can argue for a DLC-impacted case, SV OU had received zero new abusers or Pokémon into the metagame on the scale you'd expect from a DLC, only what had risen in prominence thanks to alternative bans or meta developments, such as things like Sinistcha's uptick.

For the numbers, I'll list the suspects first. The first suspect received a DNB result of 58.1%, whilst the second one received a suspect result of... 2 separate ratings. The originally considered result was a 61.8%, enough for a ban by 1.8% margin. However, after it was revealed to have the largest instance of suspect cheating this community has seen for the best part of a decade or more, This changed the result from the ban result of 61.8%, to a DNB result of 59.2%, which in context, is a margin of a 1 vote shift away from being a ban. These numbers are quite close, and is also the type of numbers you'd hope to expect to see if consideration of a 3rd try would be in order. It's survey numbers are similarly high, as a 3.8/5 was seen for its 2nd suspect, while receiving a 3.6/5 among qualified in its first pre-suspect survey. These numbers saw Kyurem as the 1st and 2nd most supported Pokémon within their surveys respectively.

With this background information out of the way, I'd propose a roadmap for the direction of the thread, some questions and some things that are irrelevant to what the thread is looking to achieve.

Things to discuss
  • What is the minimum threshold in which a 3rd suspect should be considered? What type of survey numbers, balance score and other factors should be considered?
  • How should DLC/Home be factored into these considerations? Should we consider new metagames as a fresh slate, or should be have something like a limit of 1 suspect per DLC release should we have knowledge of future DLCs upcoming?
  • Should we allow tiers to put things that have failed a 2nd suspect onto surveys until a certain amount of time has passed?
  • Should we require a policy review thread for every 3rd suspect being considered, and should we consider if it's a Pokémon element or not?
  • Should extenuating circumstances be considered as leniency for these cases such as SV OU's cheating scandal?
Things to not ask or things irrelevant to the discussion
  • Whether the examples are truly broken within their specific landscape. It matters very little if Tera is broken or Kyurem is broken, the discussion is not about the individual cases, more what criteria needs to be met to hold a discussion of a 3rd test.
  • I made this thread to be written in a way that excludes myself and my own personal opinions as much as possible, so that this could have been posted by anybody and get the same responses. Please keep that in mind when writing your responses.
 
I had a longer response written out to address the individual bullet points. Then pretty much realized/decided it was probably too strict for where there needs to be some leniency here and there this gen due to the increased power level of everything, where even now everyone is still trying to figure it all out. I trust tier leaders enough in how they want to run their own formats even if I may not agree in all instances (normalizing 5s and 6s as acceptable scores, third suspects, etc.). A third suspect is pretty wild though, and I think at minimum some sort of PR thread with a smaller pool of people tagged should be the start before an actual test.
 
Suspecting a pokemon for a third time in a generation, spread out, due to very clearly and significant changed circumstances, is perfectly fine so long as the council clearly delineates what those significantly changed circumstances are. I don't really think any PR thread there is necessary, even if it's suspecting that pokemon for a third time since circumstances can, due to the volatile nature of DLCs, change considerably and we should (hopefully) have councils competent enough to address this appropriately and do what is necessary.

Suspecting a mechanic or other non-pokemon element (like tera) for a third time in the same tier, without any sort of PR thread, is pretty crazy. You're going to probably need nearly the entire community to outcry against the mechanic to justify something like that.
 
BigFatMantis said:
Suspecting a pokemon for a third time in a generation, spread out, due to very clearly and significant changed circumstances, is perfectly fine so long as the council clearly delineates what those significantly changed circumstances are. I don't really think any PR thread there is necessary, even if it's suspecting that pokemon for a third time since circumstances can, due to the volatile nature of DLCs, change considerably and we should (hopefully) have councils competent enough to address this appropriately and do what is necessary.

Suspecting a mechanic or other non-pokemon element (like tera) for a third time in the same tier, without any sort of PR thread, is pretty crazy. You're going to probably need nearly the entire community to outcry against the mechanic to justify something like that.
Click to expand...

I was gonna make a long post detailing how context can be needed in these situations, but this post just made it better and shorter, If it is a mon, I don't think a PR thread is really necessary

If it is a Move, or an item, you have to do a PR thread for it

Then there's the generational mechanic, the mother of all boogeymen, I don't wanna say anything in this one, since dynamax was banned no issue, and tera has only been trough 1 suspect, and please forgive my language, but you can't expect to trust whatever the fuck the embarrassment of humankind that is gamefreak is cooking for the next gens. But If you really need an answer right now, I think hard cap it to 2 suspect tests no more no less not even PR thread unless one of them got hijacked like the kyurem one


And one more thing, for the sake of full completion. What happens when the gen stops being current gen?
 
A third Kyurem suspect should be on the table if, and only if, there’s clearly sufficient support for it and it is spread out from the prior suspect (ideally at least a few months, but hopefully a little more). I do not see it happening next or soon, but I absolutely will not dismiss the whole idea either as both suspects were close and the last one left a bad taste in everyone’s mouth.

I do not think the parameters for a third suspect are much different from that needed for a second suspect personally, but with any additional suspects there probably needs to be a bit of extra burden of proof as to why it is being done again — be it pointing at major metagame circumstances or closeness of prior vote(s). This was true for prior repeat suspects during my tenure on OU Council, too. We do tier to make the best metagame rather than to honor things from old metagames after all, so if there has been a sizable shift, then it’s our job to respect the current format’s situation over anything else when applicable.

I have no comment on the National Dex situation as I do not play the tier, I do not think they are held to the same standard as official tiers (their quickban track record stands out far more than this, for example), and I do not know the underlying data well. I do think tier leadership and councilmen would have to go above-and-beyond for a second or third suspect of a generational mechanic to even be on the radar in official format though, and that’s been my mindset when we previously considered a second Tera test in OU (which hasn’t happened for a long while now — so do not take this as that discussion being open).
 
I think with the big changes that the DLC model brings, repeated suspect tests can be run with a little bit more leniency. A lot can change when you add an extra one or two hundred Pokemon and a bucket full of TMs. For example, Ursaluna was banned from Doubles OU between Home and DLC1, then was unbanned at the beginning of DLC1 and has caused no more issues since. While there are a few different reasons for its unbanning and continued legality, the big ones are the introductions of Sinistcha and Ogerpon and the return of Grassy Glide Rillaboom. Meanwhile, we also tested Flutter Mane a total of three times - it was banned before Home and then unbanned when Home came out, it was suspected alongside Basculegion during DLC2 and not banned, then tested again on its own and banned. In each test, the circumstances were significantly different, whether that was the pool of legal Pokemon or the nature of the test itself.

I think if a Pokemon were tested three times in the same format, then there would be a really valid reason to call foul, but sometimes the circumstances change so much that you can justify a new test.
this is basically Mantis's post but with examples
 
I don't think resuspecting the same Pokémon again and again is an issue. When a suspect happens you're inspecting if that Pokémon (that element of the metagame) is too much for the format or not at that time. If the playerbase wants a third suspect then I don't see why not.

Obviously no banners might feel like the council is trying to force a specific way, but I think it helps to see suspects as inspecting if a specific element of the metagame is broken at that time or not. Pokémon / the surrounding meta can change behaviour a lot and honestly at that point it feels like a different suspect to me (e.g: Miraidon suspects you mentioned).

People shouldn't vote on a specific pokemon but rather the element in the metagame it represents. The fact that it's the same pokemon is honestly not that relevant IMO, like what if metagame / DLC shifts happened that bring one pokemon to then fulfill a near identical role to something that'd been suspected in the past?
 
If an element in a tier is considered suspect-worthy, then it should be able to be suspected regardless of whether it was suspected before.
Obviously, though, having successive suspects in quick succession for the same thing isn't OK, but as long as there's a decent amount of time or a very strong reason for doing it quickly, then it should be fair game. Anything less is basically depriving a tier of the ability to reach a balanced and enjoyable state if a suspect is held too early or inappropriately beforehand.

There's a lot of factors for people to consider voting for status quo. An unfortunate one includes "I wish action would be taken on this instead, so I'll vote no". We can't stop people from voting like this, but then when the time rolls around that it's clear that said action comes around, we get stuff like this. This sort of thing also becomes more likely the more "brokens" are in the tier. A huge part of the discourse that I've seen in ND and non-ND OU is people complain about a swath of different stuff and then disagreeing on what needs to be banned. And then when no action happens, everyone is left unhappy.

In the case of Tera, it's a question of whether the tier should ban a huge swath of Pokemon that are pushed just over the edge by Tera combined with the threat of other sets, or whether it is tera itself that is the problem. As suspect tests continue between these Tera suspects and more Pokemon get banned with little change to the tier itself, I can understand why people would continue to revisit this decision.

As new information along these lines come out, people change their minds and may regret their status quo vote. And then we're left in this mess. So yeah, it's fine. Doing anything to stop it is basically robbing the tier of its ability to regulate itself.
 
All of this is my opinion.

A tiering system's ultimate aim should be to make a tier better. That should be its ultimate goal, and all actions should facilitate what a majority of the playerbase sees as a "better tier". If we go off of that interpretation of tiering, then it doesn't make sense to limit how many times or when a suspect occurs.

Of course, if we continuously resuspect something that has failed to be banned numerous times without any significant metagame shifts or tiering decisions beforehand, that goes against what the majority of the playerbase wants. But if the playerbase has shown (through surveys and general dissatisfaction) disinterest or dislike with the state of the metagame, then a suspect should be allowed, no matter how many times the element or Pokemon in question has been tested.

That's just my two cents on whether or not a third suspect should ever be allowed. Now for the questions posed:

What is the minimum threshold in which a 3rd suspect should be considered? What type of survey numbers, balance score and other factors should be considered?
I believe that a major fall in survey numbers from the previous survey should be a warning sign. There shouldn't really be a lower limit to how low the survey results have to be exactly - a 7/10 satisfaction rating is fairly good in a vacuum, but when that result is a drop from a 8.5/10 without any major shifts or releases, something's potentially gone wrong in the tier. As for balance score, I'd argue that it should be held to a slightly higher standard - resentment from the earlier suspects over a perceived "incorrect" decision could drive scores up, though to what extent I don't know.

How should DLC/Home be factored into these considerations? Should we consider new metagames as a fresh slate, or should be have something like a limit of 1 suspect per DLC release should we have knowledge of future DLCs upcoming?
A DLC release or the return of Home in a new generation should absolutely be factored into these considerations - indeed, I'd argue that as a council retests contentious Pokemon from the previous meta's banlist, they should also reconsider the non-banned contentious Pokemon. A new metagame should be considered a fresh slate.

Should we allow tiers to put things that have failed a 2nd suspect onto surveys until a certain amount of time has passed?
I'd say this should be gone about in the opposite way - the only things that should be put onto surveys after failing two suspects are things that had a very close suspect vote (parameters can be ironed out), and they shouldn't be put on surveys until they become repeatedly mentioned in the "Other Things to Look At" section or in community discussions.

Should we require a policy review thread for every 3rd suspect being considered, and should we consider if it's a Pokémon element or not?
No and maybe. For the Pokemon element argument, I'd say that requirements to put the mechanic on a survey should be higher than putting a Pokemon on a survey; however, voting thresholds and the threshold on surveys to cause a suspect test should stay the same.

Should extenuating circumstances be considered as leniency for these cases such as SV OU's cheating scandal?
Yes - a suspect with as much manipulation as the Kyurem suspect test is illegitimate in capturing the voices of those who legitimately got reqs and voted for which side they wanted. While there should not be an immediate resuspect (leave the timing up to the council), the margin of the Kyurem vote and a glance at OU discussion makes me think Kyurem is still a fairly contentious topic.
 
As a natdex mainer id like to speak on our situation a little bit, we've had an incredibly volatile tier since the release of SV and neither quickbans nor sus tests have seemed to have a significant effect on raising the competitiveness of the tier. Tera has been a defining factor in the majority of our bans in combination with the fact we have z moves legal making set variety even greater than in official tiers. Were going into a third yet unprecedented suspect test now however our previous two fell marginally short of a super majority and its no secret that tera has been consistently singled out as the "broken" part of the tier amongst our tournament community for a long time now. Despite the results of the last two tests it has been a majority opinion for a long while that tera is a detriment to our tier and a significant strain on the builder hence the third suspect test.

it would be nice to have clarity on where tiering policy falls on the concept of a third suspect test and how this would apply to both "official" (not the real official metas like VGC and BSS but the "muh cartridge" metas that smogon pretends are official) and "unofficial" metas too
 
On pure policy grounds, trying to ignore my subjective views:

I don't think the current NatDex suspect should be looked at as a "third suspect" because both Home and the two DLC releases are distinct metagames. I would use "Was there a vote to drop mons from the banlist?" as a guideline, with the emphasis that this isn't meant to be a hard and fast rule, just a good sign for when there's been enough shakeup to call it a distinct metagame. To date this standard would only be reached by an updated version (BW2, ORAS, USUM, etc.) or DLC (including Home), but going forward anything that causes a massive shakeup on this level would qualify as well.

The hazard for this standard going forward is GameFreak deciding to try rolling updates every few months, but that'd wreck such merry hell on tiering in general that any such future would need much deeper changes.

For a third test without such a fundamental change, I think there should be specific guidelines, something like "Six months have passed since the last test OR multiple bans since the last test." I'd support that between the first and second test too, even - requiring significant time or changes between tests ensures that there isn't a council trying to force through a change (which has happened in the past, if only rarely), while giving a specific standard will hopefully reduce the whinging demanding an immediate retest.
 
Would any of this discussion extend to suspects that are on the same element but from different directions?
Like for example Melmetal in SS OU was quickbanned (not a suspect), tested for an unban once (suspect 1) and tested later again for a ban (suspect 2). In theory if the tier leaders were to decide to suspect Melmetal for a ban again would that be considered a third suspect?

In general multiple unban suspects seems implausible but combinations of ban and unban suspects have happened. I think this is relevant because direction can affect the % needed for a particular outcome. The % balance is already something that becomes a little counterintuitive since an element can have 50% of people saying something is a 5/5 broken and the metagame is 0/10 enjoyable, but the other half of the players would keep it from hitting 60% for the ban to happen. Luckily we have tier leaders to take survey results with a grain of salt but its clear widespread hate for an element doesn't necessarily mean it will hit 60% ban, but could lead to multiple suspects
 
Last edited:
LBN said:
For the numbers, I'll list the suspects first. The first suspect received a DNB result of 58.1%, whilst the second one received a suspect result of... 2 separate ratings. The originally considered result was a 61.8%, enough for a ban by 1.8% margin. However, after it was revealed to have the largest instance of suspect cheating this community has seen for the best part of a decade or more, This changed the result from the ban result of 61.8%, to a DNB result of 59.2%, which in context, is a margin of a 1 vote shift away from being a ban. These numbers are quite close, and is also the type of numbers you'd hope to expect to see if consideration of a 3rd try would be in order.
Click to expand...
I don't have much to say on the rest of the post, but one thing I've always thought is that it feels wrong to look at the vote% when arguing about a resuspect, unless we re-define thresholds in a way that makes sense. Keeping something unbanned with a threshold in the 55-60% range seems to mean "you're safe for now but we'll keep trying until you actually do get banned", and since these %s are already above a majority of players and close to a supermajority, it's inevitable that more people will push for a resuspect until it happens. This makes a DNB vote less valuable and suspects non-definitive, which runs counter to their intended purpose.

I'm not suggesting this as a proposal, moreso as a hypothetical, but what if when ban vote reaches majority (althought 50% may be too low, so 55% could be a good middle-ground) but not supermajority the suspectee enters a "resuspect state", where it remains unbanned for a few weeks/months before getting suspected again? This should remove ambiguity for contentious suspects and standardize the process for most resuspects, at least in theory.
 
You must log in or register to reply here.
Back
Top