Current Gen Tiering and a Higher Rise Cutoff ("The Hitmontop Issue")

[Meri Berry]

The issue of Hitmontop rising to NU, and now likely to RU, has been widely criticized in an attempt to change the way that our tiering system operates. In particular, the most common solution that has been proposed is a "veto" system, where the tier leader of the tier it is rising into can decide whether that Pokemon is viable enough to rise. However, this is highly subjective and will likely lead to controversial decisions of its own, and these decisions will have a name and face attached. From my discussions, tiering policy prefers to have mechanical solutions to issues instead of the manual ones that have been proposed, which brings us to the main point.

A higher cutoff for rises would help alleviate the issue of anomalous rises such as Hitmontop to NU, which is the main issue that brought this point to the forefront of discussion. It also allows for shorter lived meta trends to not affect lower tiers, which would often cause bans to happen as a direct result. Finally, this method doesn't prevent Pokemon from getting to their appropriate tier if the usage is not anomalous. For example, Flygon has become a very big RU staple and deserves to be in that tier, so preserving rises is in the best interest for builder clarity and tier health.

For the sake of my examples, I'm going to use the percentage 6.6967% to represent what this new rise percentage could be, and look at specific instances to see what might happen. As a disclaimer, we obviously don't know what would happen in consecutive shifts if such a system was implemented, but I'll try my best regardless.

Why 6.6967%?

The current system for both rises and drops has a percentage of about 4.52%. This is based off of a principle that has shaped this percentage much longer than I have been on this site: a Pokemon is considered Overused if it has at least a 50% chance to appear in T randomly selected games.

Recently, when Sword and Shield released, T was changed from 20 to 15, which made this new percentage of 4.52% appear. For 6.70%, I used the same principle, but this time T=10. This number doesn't need to be weighted for quickdrops or anything else, since only rises will be affected by this change. This is not a final number, and it is just what I came to as "reasonable" in testing. However, having a number at all shows how this would work in practice, and I thought it would make a stronger argument.

First, we'll look at NU's rises in the shift occurring April 2022, which is the issue that started this current discussion. In April, both Hitmontop and Machamp moved to NU due to this rise, with only Machamp being on our Viability Rankings. If we look at the usage for this month, we can see that Machamp would rise and Hitmontop would not with this new number.

| 32   | Machamp            |  6.862% |
| 33   | Drapion            |  6.686% |
| 34   | Tauros             |  6.576% |
| 35   | Zoroark            |  6.547% |
| 36   | Toxicroak          |  6.545% |
| 37   | Silvally-Steel     |  6.496% |
| 38   | Golurk             |  6.457% |
| 39   | Guzzlord           |  6.303% |
| 40   | Hitmontop          |  6.224% |

This does not completely prevent the coordinated efforts to try to get Hitmontop to rise to RU, as it will likely surpass this even higher threshold due to outside influence. However, the whole point of tiering is to reflect what people are using, and I think this would be considered permissible because multiple dedicated people will have contributed to its rise. Obviously this is mostly speculation, as we do not have the numbers for May or June.

I'd like to look at Scizor, and to a lesser extent, Mew, in UU. as well. Scizor in particular ended up rising as a small meta trend that stemmed from trying to check Kyurem. These rises occurred in the July 2021 shift, and Scizor's departure massively shaped the tier and was the cause for a few bans. Now that Scizor is back in the tier following the April 2022 shift, this meta disruption seems potentially damaging to the tier, especially given it was the highest used Pokemon in UU in the July 2021 usage statistics.

July 2021
| 35   | Mew                |  5.626% |
| 36   | Hawlucha           |  5.546% |
| 37   | Scizor             |  5.545% |

October 2021
| 27   | Mew                |  7.497% |
| 28   | Scizor             |  6.892% |

Although Scizor would have risen in the October 2021 shift, its very difficult to know whether this was solely because of meta shifts, or whether this was also because of the availability in the builder, but its a good example of rises happening because of a single Pokemon that ended up getting banned, before the check dropped again. This can be seen currently in the May 2022 release of usage statistics with Diancie in RU, which was solely used to check Obstagoon.

| 33   | Diancie            |  6.008% |

The new cutoff would prevent this from happening and potentially disrupting NU, though its not certain whether Diancie will rise at all now that Obstagoon is banned.

These are just a few examples that I found to try and illustrate why such a system would be useful, but its very hard to know what the complete effects of this system would be in practice. However, I think its probably the best solution to the initial tiering problem that has been proposed so far, and it seems to solve most issues around higher tiers taking Pokemon that just aren't that good and disrupting lower tier metas.

Also full disclaimer, I wasn't the one to initially conceive this idea, and I credit Lily for bringing it up in the first place.

 

[MudkipBeans]

S/Os to Dorron for posting this in the NatDex RU channel and the rest of the ND lower tiers people before I start.

I'd like to give my two cents on this topic as well, since this has been bothering me even before "The Hitmontop Incident". Before current gen, we've seen things that shouldn't occur in tiering, mainly the whole Ambipom in UU debacle along with others like Jolteon in RU, with both not being viable at all in those tiers. But currently, shifts are something that most players who currently enjoy National Dex UU (including me) and those who participate in National Dex RU dread to no end.

To speak from my experience, every month we check the NDUU usage stats and are once again not surprised that Donphan still has not dropped down to RU, despite being unranked and outclassed in most aspects since this is a Hippo tier. Here's some gems from last months usage stats for some things that we would randomly take from RU:

| 16   | Whimsicott         | 10.05813% | 1168   |  2.744% | 921    |  2.768% |
| 30   | Donphan            |  6.72501% | 3045   |  7.155% | 2451   |  7.366% |
| 31   | Pidgeot-Mega       |  6.14241% | 1303   |  3.062% | 1040   |  3.125% |
| 34   | Umbreon            |  5.60459% | 2376   |  5.583% | 1764   |  5.301% |
| 35   | Ampharos-Mega      |  5.35452% | 882    |  2.072% | 707    |  2.125% |

Now obviously, ND UU is a much smaller ladder than ladders like SS RU and NU, only having about 21,279 matches in total, less than half of RU's total matches. But solo actors have an exacerbated effect on what moves up and down through these tiers, and can cause incredibly strange disruptions between the tiers and cause things to get banned only for it to suddenly change back and turn the tier back into a state of chaos. This is mostly what happened to Breloom in Gen 7 since Amoonguss kept rising to OU and dropping back down to UU, and the fact that we didn't notice a problem back then is quite confusing to me (although Amoonguss is viable in OU, the constant changing wasn't ideal).

As for NDUU specific examples, Crobat has a history of dipping from RU and suddenly making speed control a lot less versatile, and the pink blobs in NDOU refuse to drop despite not being that good and also being a crucial part of why a significant chunk of things got banned, with Hydreigon, Thundurus-T and Drought coming to mind.

From this change, I don't see any real negatives coming from a rise in the cutoff to be that bad. All that would really happen is lower tiers get some more things to play with and gain much more stability. This is opposed to dreading shifts and hoping that a good option in the meta doesn't suddenly leave and a wave of bans must then follow, only for it to drop back down with the tier now having to re-evaluate the bans. If something is viable and consistent in a higher tier, it thusly should receive consistent usage, but right now I think the cutoff is just low enough to cause either dishonest actors or people just having fun with their favorite mons to disrupt metagames more than is reasonable.

 

[Moutemoute]

Not a lot to add to what has been already said but :

In particular, the most common solution that has been proposed is a "veto" system, where the tier leader of the tier it is rising into can decide whether that Pokemon is viable enough to rise. However, this is highly subjective and will likely lead to controversial decisions of its own, and these decisions will have a name and face attached.

To be honest I do think this is something doable. That "veto" system could be given not only to the tier leader but to the council of the tier. I tend to believe tier's councils have enough recoil and knowledge of their tier to be able to make those kind of call. Thus the decision would be done by the whole council and not only the tier leader which would alleviate some kind of pressure/guilt that could be put on the TL.

Otherwise I'm fine with everything that I've been said by Meri Berry. Useful post, especially since this kind of shitty events already happened in the past (the whole "Ambipom to the top" in UU was aids af).

 

[sugar ovens]

Can't we just like, accept that lower tiers have shifts and change over time and stuff and stop getting unreasonably annoyed by the occasional unviable mon rising? Or even the occasional viable mon - something changed, tragedy, let's change the percentages, let's remove rises altogether like why.

(and while we are at it, change the cutoff to the original 3,sth% ?)

 

[MudkipBeans]

Can't we just like, accept that lower tiers have shifts and change over time and stuff and stop getting unreasonably annoyed by the occasional unviable mon rising? Or even the occasional viable mon - something changed, tragedy, let's change the percentages, let's remove rises altogether like why.

(and while we are at it, change the cutoff to the original 3,sth% ?)

I'd like to highlight the ending paragraph of my post above, where I reference the negative consequences of raising the tiering cutoff, and currently I see absolutely no negative consequences. More options has never been a bad thing and things moving down is fine and less slightly disruptive than things moving up. Moving up in tiering removes options for everyone on the spectrum on certain ladders, where either their favorite mon is no longer legal in the tier they were playing or an option is forcibly removed for competitive players because of situations similar to Ambipom in Gen 7 with many dishonest actors. Moving down with the OP's cutoff gives people more options and if they're below this cutoff to begin with, they probably should be in that tier by most examples.

 

[EviGaro]

Can't we just like, accept that lower tiers have shifts and change over time and stuff and stop getting unreasonably annoyed by the occasional unviable mon rising? Or even the occasional viable mon - something changed, tragedy, let's change the percentages, let's remove rises altogether like why.

(and while we are at it, change the cutoff to the original 3,sth% ?)

If the argument is that people should "whine less", then logically wouldn't we all be for a solution that reduces the chances of said whining some more? Raising the usage required for rises doesn't eliminate them altogether - an idea I'm frankly not for either - but if it makes it clearer that something is good in a tier before it actually rises then I don't really see how that's a bad thing, or why the older stat was better to begin with.

And sorry, but having shifts that effectively cancel each other because of a "flavour of the month" is a nightmare of tiering. It's "change" but only really artificial, as you're effectively back to where you started. So how do you take that in? If you had to remove two mons by suspects because of a shift that is cancelled a few months later, do you just bring them back? UU wasted some ridiculous time last gen on this, and it showed: the end product of SM was hurried because of all the time spent on having to deal with some very awkward shifts taking and giving back key components of the tier. If your approach to this is like, whatever this is what you signed up for as a lower tier, well, ok, but if there's solutions that still make lower tiers function the way you want them to, while giving the community an additional leeway in tiering stability, it should absolutely be considered.

 

[pokemonisfun]

Perhaps I'm naive, but have our administrators tried to ask the people trying to influence usage? I'm under the impression that it's mainly YouTubers with large followings that can change the tier although I don't know who's behind Hitmontop.

Maybe they just don't realize the disruptive effect they have on Smogon tiering and would stop if we asked nicely.

Smogon could also make a public statement that hopefully other YouTubers could see in case they get a similar idea.

 

[sugar ovens]

I honestly didn't read most of the thread, i've seen various "stop rises" opinions in several places and my post was moreso a sigh in response to all of these, particularly against the "rises bad" and "ambipom bad" sentiments. But ah well, yes, it would be better to actually contribute something, right.. I'll try.

So.. I like playing lower tiers. And there's always this mean tier right above that steals all of our stuff, right? And then they don't give anything useful back, even if it's unviable there! I'm annoyed by that.. i complain about that.. and i like that and i don't want that to change. Lower tiers are not stable, lower tiers change quickly, they get messed up and stabilize again, you lose mons and have to do with what you have avaitable - that's so cool! Isn't it? And yeah, it feels better to get a new toy then to lose an old one, but, at least for me, losing the old ones is just as important. It's not just "removing options". That's when you have to build new teams and figure stuff out again, perhaps find something new! And sometimes the tier ends up worse, sometimes it becomes better. Rises are not an issue, instability is not a bad thing. It's definitely not an *inherently bad thing*.

Now, lack of options, losing pokémon... The less mons *you* have avaitable, the more mons are viable somewhere.. the smaller the tiers are, the more mons end up in the bottom. And, of course, the smaller tiers are more similar to each other. I don't want tiers to shrink further. The reason to play a lower tier is to have a different set of pokémon to play with. Lower tiers are supposed to be different to each other, not share popular pokémon. They should accomodate as many pokémon as possible, and change over time. Higher cutoffs damage all of this, and... As we've seen, it's not even guaranteed to fix anything. Perhaps it lowers the chances of an Ambipom TTT or gen7 mesprit, but those are not real problems.. just changes. I agree that a rise and a drop in consecutive shifts is an issue though, but.. I don't see how a higher cutoff could help with that. A Pokémon can skip around any percentage, a trend can raise the usage from any.. starting point. Thinking about it now, perhaps a higher cutoff would even raise the chance of that happening, at least "naturally", considering the pokémon would be more likely to become staples in the lower tier and having better niches in the higher tier.. A rise veto is not a solution to that problem either - just a tool to remove rises of tier staples.. aka large changes we don't like.. and hitmontops to the tops. Only the ones caused by youtubers, anyway. How do you determine that something is just a flavour of the three months, or for how many shifts will it stay up. I could agree with a rise veto for the few properly defined cases where this is obvious - when the pokémon rises after or at the same time when the mon that it was used to check gets banned / rises to a higher tier.. but.. Even then it's not *really* obvious if we weren't wrong and our vetoed mon doesn't rise next shift.

Another thing I want to mention.. a messed up tier.. a broken tier.. a disrupted tier.. It's never THAT bad. Sometimes something absurdly broken drops and gets promptly quickbanned, yes. Rises don't make tiers terribly bad.. Common broken/uncompetitive mons that get banned in suspects don't make tiers awful to play. I suppose it sounds kinda bad when written like this.. but i feel it is important to acknowledge.. and that it is not said very often. Tiers have troubles and banned mons for the entirety of the generation, suspects improve them, shifts change them and for the vast majority of time - they are quite fine.


So yeah, tldr, i sort of think that this is what you signed up for. Instability is not inherently bad, rises do not mean "less options" but a change of the tier. Lower tiers should be large, diverse and alive. Raising the cutoff makes them smaller, share more pokémon, leaves more mons at the bottom and doesn't fix the actual problems. Veto could be considered in very specific cases.

 

[pokemonisfun]

I honestly didn't read most of the thread, i've seen various "stop rises" opinions in several places and my post was moreso a sigh in response to all of these, particularly against the "rises bad" and "ambipom bad" sentiments. But ah well, yes, it would be better to actually contribute something, right.. I'll try.

So.. I like playing lower tiers. And there's always this mean tier right above that steals all of our stuff, right? And then they don't give anything useful back, even if it's unviable there! I'm annoyed by that.. i complain about that.. and i like that and i don't want that to change. Lower tiers are not stable, lower tiers change quickly, they get messed up and stabilize again, you lose mons and have to do with what you have avaitable - that's so cool! Isn't it? And yeah, it feels better to get a new toy then to lose an old one, but, at least for me, losing the old ones is just as important. It's not just "removing options". That's when you have to build new teams and figure stuff out again, perhaps find something new! And sometimes the tier ends up worse, sometimes it becomes better. Rises are not an issue, instability is not a bad thing. It's definitely not an *inherently bad thing*.

Now, lack of options, losing pokémon... The less mons *you* have avaitable, the more mons are viable somewhere.. the smaller the tiers are, the more mons end up in the bottom. And, of course, the smaller tiers are more similar to each other. I don't want tiers to shrink further. The reason to play a lower tier is to have a different set of pokémon to play with. Lower tiers are supposed to be different to each other, not share popular pokémon. They should accomodate as many pokémon as possible, and change over time. Higher cutoffs damage all of this, and... As we've seen, it's not even guaranteed to fix anything. Perhaps it lowers the chances of an Ambipom TTT or gen7 mesprit, but those are not real problems.. just changes. I agree that a rise and a drop in consecutive shifts is an issue though, but.. I don't see how a higher cutoff could help with that. A Pokémon can skip around any percentage, a trend can raise the usage from any.. starting point. Thinking about it now, perhaps a higher cutoff would even raise the chance of that happening, at least "naturally", considering the pokémon would be more likely to become staples in the lower tier and having better niches in the higher tier.. A rise veto is not a solution to that problem either - just a tool to remove rises of tier staples.. aka large changes we don't like.. and hitmontops to the tops. Only the ones caused by youtubers, anyway. How do you determine that something is just a flavour of the three months, or for how many shifts will it stay up. I could agree with a rise veto for the few properly defined cases where this is obvious - when the pokémon rises after or at the same time when the mon that it was used to check gets banned / rises to a higher tier.. but.. Even then it's not *really* obvious if we weren't wrong and our vetoed mon doesn't rise next shift.

Another thing I want to mention.. a messed up tier.. a broken tier.. a disrupted tier.. It's never THAT bad. Sometimes something absurdly broken drops and gets promptly quickbanned, yes. Rises don't make tiers terribly bad.. Common broken/uncompetitive mons that get banned in suspects don't make tiers awful to play. I suppose it sounds kinda bad when written like this.. but i feel it is important to acknowledge.. and that it is not said very often. Tiers have troubles and banned mons for the entirety of the generation, suspects improve them, shifts change them and for the vast majority of time - they are quite fine.


So yeah, tldr, i sort of think that this is what you signed up for. Instability is not inherently bad, rises do not mean "less options" but a change of the tier. Lower tiers should be large, diverse and alive. Raising the cutoff makes them smaller, share more pokémon, leaves more mons at the bottom and doesn't fix the actual problems. Veto could be considered in very specific cases.

I agree with most of what you’re saying but I personally frame the “hitmontop issue” as the title puts it in another way.

The problem isn’t instability of lower tiers.

The problem is that a single YouTuber or other person with large following can effectively ban whatever Pokemon they want to from a lower tier. Not even our tier leaders are allowed to do that anymore as all the low tiers have councils now. It’s not proper that any single person has this much power in our system (apart from perhaps Chaos I suppose but we at least accept him as legitimate in part because he doesn’t try to actively tier afaik).

So for me, the question is how to reduce their influence in our system.

 

[Wigglytuff]

Perhaps I'm naive, but have our administrators tried to ask the people trying to influence usage? I'm under the impression that it's mainly YouTubers with large followings that can change the tier although I don't know who's behind Hitmontop.

Maybe they just don't realize the disruptive effect they have on Smogon tiering and would stop if we asked nicely.

Smogon could also make a public statement that hopefully other YouTubers could see in case they get a similar idea.

Not likely:

Spoiler

etc etc look up these logs in the RU discord if you want to see full context, the point is made either way

The toothpaste is already out of the tube, but the best thing to do would be to internally find a solution and avoid drawing further attention to the situation. I won't pretend like I know what a solution to this would be, but appealing to the usage based tiering won't be effective to people that don't care a whole lot about usage based tiering in the first place.

 

[Bughouse]

Usage stats is a mainstay of Policy Review with many prior discussions dating back years. This is always a good discussion to have and there are always potential alterations that could be made to stats calculation to improve outcomes (though imo likely too late to implement for this gen and should just happen as a part of Scarlet/Violet.) At the same time, if Smogon tiering is going to continue to be based on usage + banlists, then there should be no separate process outside of measurement of usage that determines the lowest tier a Pokemon can be in, i.e. anything like a TL "veto" of a usage-based rise is wholly inappropriate imo.

We've certainly changed the way we've calculated or interpreted usage stats before, in fact even for this generation. And it's possible that the current perceived problems may merit further alterations to how we handle usage. You can simply interpret the results differently, i.e. alter the tiering change threshold %s for rises/drops in general or for just quickrises/quickdrops. And you can also potentially make changes to the usage calculation formula itself, the simplest method with the most history on Smogon being weighting usage stats differently based on player ratings/win rates i.e. using 1695 stats or 1630 stats, etc., but there's likely some other more exotic options that could involve normalizing usage in some way by user or user+team or user+pokemon so that individual accounts that play an inordinate number of games don't unduly influence stats, etc. Any change here though has to be considered in light of the availability of the actual people who work on usage stats. Simpler solutions like changing the tiering threshold cutoffs or the weighting are generally better here, even if they're more blunt instruments.

But at the end of the day, no Pokemon "deserves" to be in a particular tier. Taking the example to the extreme - If something gets used in sufficient numbers to be in OU but really should be PU by viability, under Smogon's usage based philosophy, it's OU. Simple as that. You can potentially find ways to alter the measurement or interpretation of usage such that the tiering outcomes better comport with community expectation that it "should" be PU. But you don't just say "this is still PU/NU/RU/UU" as a case by case decision for each Pokemon that would be rising to a higher usage tier.

One example that might be illustrative came about with the decoupling of Mega Evolutions from their base in ORAS. When that happened, Pokemon who had previously existed in just their base form alone in XY and only had a mega evolution added in ORAS were quickdropped to the tier they used to be in as their base in XY. i.e. Lopunny dropped to PU. We had some sense of how much usage they might actually get in their base form alone.

But for Pokemon who had always had a Mega Evolution since XY and who thus never really had a chance to be used as only their base form, they were only initially dropped no more than one tier from where the Mega was at, and any further drops would be figured out over time. This meant that for example Manectric and Charizard were only dropped from OU to UU, even though they then continued dropping and pretty quickly ended up in NU. We didn't try to guess at their tiering based on perceived viability. We knew that they didn't get used as non-Megas enough to stay in OU, so they could definitely start off by dropping to UU. But we had no data on how much they might or might not get used in UU, and so there they started and we let usage tiering work itself out. Even though both Pokemon had existed and been lower than UU mons in both DPP and BW. We put them into UU and let them drop over time, both pretty quickly falling to NU in a few months. Even though probably just about no one thought Charizard or Manectric would actually stick in UU, we still put them there, depriving lower tiers of having them, and potentially slowing their development, by several months. Usage tiering means usage tiering.

 

[pokemonisfun]

Not likely:

Spoiler

etc etc look up these logs in the RU discord if you want to see full context, the point is made either way

The toothpaste is already out of the tube, but the best thing to do would be to internally find a solution and avoid drawing further attention to the situation. I won't pretend like I know what a solution to this would be, but appealing to the usage based tiering won't be effective to people that don't care a whole lot about usage based tiering in the first place.

Well in the logs you posted nobody made my claim: the reason this is wrong is because it’s giving too much power to a single person.

I still think we should just ask Freezai (I didn’t know he was the main Hitmontop guy) and tell him why one person shouldn’t be allowed to cause tiering changes and if he/other YouTubers really want to influence the makeup of a tier, explain the proper ways to do so (NP posts, participating in suspects, etc.)

It might not work but we should at least try.

 

[termi]

I have been in favor of a veto system for several years now, and while I support the suggested higher cutoff for rises, I think having a veto system in addition to this would still be a positive thing. While the Hitmontop debacle has been relatively benign since it was never that great of a Pokemon in the tier it initially rose from (though I'm sure ladder would beg to differ), there have been instances of rises in the past that were very destabilizing. PU last gen had to deal first with the rise of its #1 mon in Mesprit, then 3 months later with Froslass, another highly influential mon, moving up to UU. This was near the end of the generation as well, so there was barely any time to respond to these incredibly problematic shifts. I know that the case of Mesprit was the result of primarily one person spamming it, while Froslass' rise also seems to have been the victim of one person or a very small group of people. These cases would have been prevented had the proposed cutoff existed at the time, but we cannot be too certain that similar cases are altogether prevented from popping up in the future. Especially on less active ladders, a sufficiently dedicated group of people could still successfully conspire to make a more or less unviable Pokemon move up and it would be even easier to do so for certain Youtubers with large fanbases. I think it would be a good idea to have a built-in fail-safe mechanism for such extreme cases instead of letting them happen and then having to come back for another PR thread on these issues a few years down the line. It shouldn't be the case that we have no response at all to dedicated trolls, or worse, Youtubers, and I imagine that alternative solutions to a veto system would be much more unwieldy and/or draconian (i.e. you could IP ban people if there is proof of them manipulating usage stats, but this seems undesirable to me).

The main objection to a veto system as far as I can tell is that it would lead to potentially controversial decisions. I think this concern is overstated though, especially if such a system is set up in a way to where a veto has to be coordinated between the council of the tier a certain mon is rising from and the council of the tier said mon is rising to. If both can agree that this rise is illegitimate, there is little room for controversy in my estimation. Moreover, the proposed cutoff would also mitigate this issue, since it already filters out a decent amount of cases where such a veto might otherwise be deemed necessary.

Other objections concern the subjectivity of a veto system or some variation thereof, i.e. that no Pokemon really "deserves" to be in one tier or the other. Such an argument misses the mark entirely. As far as subjectivity is concerned, it is worth stressing that usage-based tiering is itself not objective. Especially in cases where an individual or a small number of people influence the usage of a certain mon to a great extent, we are essentially dealing with the opinion of one group versus another, the difference being that in order to become a tiering council member you need to have been recognized by your peers as a sufficiently worthy player and contributor whereas no such criteria exist for ladder players. Moreover, council members can't make decisions based on arbitrary reasons, whereas ladder players can coordinate the rise of a certain mon for no reason other than that they simply want to.

The argument that no Pokemon "deserves" to be in one tier or the other misses what I believe to be the greater point of a veto system, which is that it can serve to maintain a degree of stability in a tier in cases where it would otherwise be destabilized for no good reason. Obviously there are cases where an influential Pokemon rises from one tier to the one above it because it just so happens to be really good there as well (i.e. Talonflame from PU to NU this gen), which is something we simply have to accept. On the other hand, the sudden surges in usage of Mesprit and Froslass in tiers above PU last gen did not happen for great reasons but they made an indelible impact on the PU metagame. Obviously it is preferable if such things are prevented "mechanically" but this should not mean that we have to put up with unwarranted shifts that are highly destabilizing to our metagames if the mechanism fails to prevent them. Since we are playing and tiering competitive metagames, we should be concerned first and foremost with the competitiveness and stability of our metagames. If there is a way for us to improve on this without fundamentally messing with the usage based tiering system in such a way that it could have negative downstream effects, it should be implemented imo.

 

[Jaajgko]

I think a veto system can be a good fix as long as the system is only used for very particular cases. For example if a Pokémon has the usage to rise/stay in a tier where it's unranked, the council can take action, and I don't think this would be controversial, even if a lot of players deem an unranked Pokémon to have a slight niche, it'd be C rank at best, which is a rank usually filled with lower tier Pokémon. In cases like Diancie in RU, where it became popular just to check Obstagoon which got banned, the council could decide to wait and see how its usage develops in the next three months before letting it rise. There again, I think most players would agree that a Pokémon shouldn't rise because of a dead trend. I think the council voting over every rise would be too much though.
Cases like Amoonguss in Gen7 UU wouldn't be fixed with this system, and they are obviously problematic for the stability of the tier, but I don't think the OP's solution would be preferable, as it also stops good Pokémon with 5-6% usage from rising (which exist). On the matter of Hitmontop, I think players are overreacting a bit, it's dumb but it's not like it ruins RU, NU or PU. However it does show that the current system is flawed and it's a good opportunity to try and fix it.

 

[shooting star]

In particular, the most common solution that has been proposed is a "veto" system, where the tier leader of the tier it is rising into can decide whether that Pokemon is viable enough to rise. However, this is highly subjective and will likely lead to controversial decisions of its own, and these decisions will have a name and face attached.

I wholeheartedly agree with a veto system. In fact, it is actually easy to detect which mons are genuinely good in higher tiers, and which mons are merely the "flavour of the month", being used in only 1 team either by the larger player base (due to events like the ladder tournament), or by only 1 user who just simply loves laddering. Just go to https://www.smogon.com/stats/2022-04/moveset (you can specify the month in the URL) and click on the tier to look at the commonly used mons and their teammates percentages. If a mon is genuinely good, it will naturally find itself in a variety of teams, and not only on ONE team. Just look at Landorus-T's teammates in OU for April 2022. The glue mon of OU, its 11 most common teammates percentages are very balanced with a small range, from 28% to 14%. Now, let's look at other undesirable rises that have happened over the past 3 generations, and you can see how the statistics make the claims of usage manipulation more credible.

Hitmontop 1630 gen8nu Mar 22 stats

Froslass 1630 gen7uu Sep 2019 stats

Mesprit 1630 gen7nu Jun 2019 stats

Mesprit 1630 gen7nu May 19 stats

Azelf 1695 gen6ou Sep 16 stats

If you look at all the hide tags, the teammates stats have some things in common: the top 5 teammates percentages are unusually high at about 40% or more, and the difference in percentage between the 5th and 6th most common teammate is awfully stark. This is plain evidence that the mon is being spammed on only ONE team, so much so that it rose to the higher tier. The council of the higher tier simply needs to do extra homework digging into the stats before arguing their case that the mon does not belong here, and thereafter exercising their veto power judiciously.

 

[Arcticblast]

Well in the logs you posted nobody made my claim: the reason this is wrong is because it’s giving too much power to a single person.

I still think we should just ask Freezai (I didn’t know he was the main Hitmontop guy) and tell him why one person shouldn’t be allowed to cause tiering changes and if he/other YouTubers really want to influence the makeup of a tier, explain the proper ways to do so (NP posts, participating in suspects, etc.)

It might not work but we should at least try.

To be completely honest, if I was in freezai's place and someone tried to have this conversation with me, I would just double down. From my hypothetical standpoint, it's funny as hell, and what are you going to do to stop me anyway?

I can see the video title now: SMOGON MODS TRIED TO STOP THE TOP!!

This sort of thing has been happening for years and is going to happen no matter what you do. Trying to stop it on an individual level seems pretty pointless.

 

[EviGaro]

To be completely honest, if I was in freezai's place and someone tried to have this conversation with me, I would just double down. From my hypothetical standpoint, it's funny as hell, and what are you going to do to stop me anyway?

I can see the video title now: SMOGON MODS TRIED TO STOP THE TOP!!

This sort of thing has been happening for years and is going to happen no matter what you do. Trying to stop it on an individual level seems pretty pointless.

To be completely honest, this is not what happened lol. I said it's stupid not to have a hope to end it and I really don't care what freezai decides to do or whatever video he would get from it, I just said it's stupid because I find it to be that. It was dumb as hell to have plenty of conversations solely about Hitmontop in the RU discord the same way it was dumb as hell for the UU channels to be spammed with Ambipom last gen. It has never been restricted here and my approach to those issues tends to not be that, but hardly any regular contributor found it an positive addition before I said that and it's not like the added traffic will reliably boost the tier anyway.

The main conversation with freezai there anyway isn't about what he should do, it's about what we should to with shifts that are just troublesome. RU is experiencing them in a particular way, but when that convo popped off a lot of tier leaders brought their experience and we tried to find a solution for most people. This thread is the tentative result of that discussion, and what I said to freezai is what I would hope to push forward. I wouldn't really qualify these logs as anything more than my opinion and the opinions of regular people that were being a bit antsy about this, but besides the way tricky veto power, the main proposal here isn't about Hitmontop at all. Would not see this as pretty pointless if we actually improve a system that is less than ideal.

 

[Marty]

Going to play devil's advocate here since this:

if a Pokemon has a probability of at least 0.5 of being present in one in [X] teams, then it is considered OU; otherwise, it is not

has been the case for longer than I have been on the site, too. It's the reason the rise/drop cutoff has always converged to the same number every 3 months, so people could say that a continuous group of Pokemon down to a certain usage threshold is definitely OU, and everything below that usage is in some other tier. If we go with this proposal, what then is the definition of OU? For sure there will be several Pokemon with usage between 4.52% and 6.70% that are not OU because they dropped below 4.52% at some point and never made it back above 6.70%. So you can't say >=6.70% is OU because some are below that, and you can't say >=4.52% either because some are actually UU or RU or Gastrodon or Quagsire.

Are we okay with throwing away the clear definition now? Having staggered cutoffs was briefly discussed before Gen 8 too but it was shot down pretty much immediately and we went with a higher one with equal weights per month instead.

 

[termi]

Going to play devil's advocate here since this:

has been the case for longer than I have been on the site, too. It's the reason the rise/drop cutoff has always converged to the same number every 3 months, so people could say that a continuous group of Pokemon down to a certain usage threshold is definitely OU, and everything below that usage is in some other tier. If we go with this proposal, what then is the definition of OU? For sure there will be several Pokemon with usage between 4.52% and 6.70% that are not OU because they dropped below 4.52% at some point and never made it back above 6.70%. So you can't say >=6.70% is OU because some are below that, and you can't say >=4.52% either because some are actually UU or RU or Gastrodon or Quagsire.

Are we okay with throwing away the clear definition now? Having staggered cutoffs was briefly discussed before Gen 8 too but it was shot down pretty much immediately and we went with a higher one with equal weights per month instead.

What is your practical concern here? So Quagsire might be PU and also get a decent amount of usage in NU or UU or wherever, what of it? If anything, the cases you're bringing up are perfect examples of Pokemon that under the current system tend to upset the balance of tiers by being defensive staples that are constantly at risk of flip-flopping between several tiers. The practical benefit of the proposition at hand for all lower tiers clearly outweighs some concern about the clearness and rigidity of the definitions of our tiers as far as I'm concerned. The rules, definitions, and mechanisms by which we tier should serve the community and not the other way around. So my question is: what would the concrete negative impact of the issue you're bringing up be for the community?

 

[Tuthur]

If I understood correctly, the OP is looking for a solution to prevent a user or a group of users to manipulate usage stats by getting a shitmon a huge amount of usage by spamming the ladder. The problem with rising the cutoff is that it also affects Pokemon that would "legitimately" rise. For instance if I take Meri's example of a 6.6967% cutoff in NU, Silvally-Steel would never have risen from PU to NU and it would have taken 3 more months for most Pokemon that rose from PU to NU to make it. These are quite big side effects on tiering that go beyond preventing Hitmontop, Mesprit, and Ambipom to rise to a tier where they are (close to) unviable. I am not completely against making rise harder for every Pokemon, but this is not the topic of this thread, so I will make the assumption we still want a tiering process where Pokemon like Silvally-Steel, Vileplume, Passimian, Guzzlord, Toxicroak, etc would have risen to NU the same month they did.

In my opinion, a veto system seems like a better approach as it has, as far as i know, no repercution on "legitimate" rises. However, who is getting the right to veto rises? Taking the instance of NU Hitmontop, is it NU council, NUTLs, PU council, PUTLs, a mix of them, or something else? The PU players have no legitimity deciding about a Pokemon's place in NU, while I don't think NU players want to take the responsability to take a decision that doesn't affect the tier they are playing, even if they have a concrete idea of the Pokemon viability. This is an issue Meri raised; every rise isn't easy to sort out between "it is a metagame defining Pokemon" and "it's unviable", every rise isn't as simple as RU Flygon and UU Ambipom. It is quite hard to fix objectif criterias as the viability of certain Pokemon within a tier is subjective. I've seen people talk about Viability Rankings, however I think it is a fake objective criteria, as the VR is decided by a few selected users, who could manipulate the rankings to affect tier shifts. I can't think of any objective criteria that could be used to determine whether a rise is legitimate or not.

I think the both solutions suggested in the op are unpractical, SS Hitmontop is fun tiering oddity and it doesn't signficantly hurt any tier. I don't think these issues are a big deal and in my opinion, we can keep the system as it is. SM Mesprit is an unfortunate rise, but afaik it is the instance of such "illegitimate" rises having big consequences on the tiers they rose from.

 

[Plague von Karma]

If you want a veto system, legit question: why not employ viability-based tiering like RBY through ADV for future generations?

Now before you rev up those haha reacts...
Yes, I know this'll be taken as the "baby with the bathwater" approach.
Yes, this may be "out of the scope of the thread". If you think so, I could write more about this in its own post!

Shush, hear me out. I think this problem is a lot wider than people think and I want to see a long-term solution that reduces potential tiering errors.

The problem identified here is outside influence that's magnified by large community figures. This has happened many times before and as PokeTubers grow with ScVi this will only grow larger. Every time a new game comes out, a few people will spark off their channels and rise to the forefront, and some may have the insanity to repeat what's happened here. This is unavoidable and history has shown it's happened multiple times. However, this is one among many other problems: I'm sure you all remember the famous Metang fiasco among other "incidents". As Antar has said in the past, Molk would have risen Metang whether the weighting was there or not, and while alterations have been made since, this is still repeatable. I think an issue that people sometimes forget is the way Showdown presents a tier's Pokemon. I'm sure you know of a few specific Pokemon beginning with A or B shitting up the ladder for ages because of this. I've criticised this quite heavily here with a proposal to switch to Viability Ranking presentation, and strongly believe this would assist with the issue you have right here. I'm 99.9% sure that there are a few more incidents and issues than what I'm saying here, but this is off the top of my head, it seems Tuthur listed a few others.

I believe that the Smogon userbase is perfectly capable of tiering Pokemon via Viability Ranking slates as they currently function instead. Smogon has changed a lot in regards to how it handles Viability Rankings these days to the point that I think this is, well, a "viable" solution: I think we can respect the council and tier leader's intelligence with regards to how the ranking slates operate. What you get here is far more stability and the ability to make tier shifts more predictable, while also ensuring every Pokemon rises appropriately. While you end up magnifying criticisms towards tier leaders and get potential backroom arguments, I think there is a net positive with the stability factor. This allows tiers to get precise Pokemon pools that reflect how the tier operates, while reducing the potential for hype phases when a Pokemon is "discovered" or unjustifiably gassed up. Sure, this can be achieved simply with the ability to veto rises on their own, but this doesn't achieve the potential of dropping Pokemon in a more timely manner, such as when Tyranitar persisted in early SwSh OU over the "Save Tyranitar" movement, despite being terrible during that period. Pokemon can be forced to rise, but they can also be forced to stay.

"So what you want is the veto power but a bit more?"

Yes.

How many times have you heard the phrase "the usage doesn't reflect the tiering of this Pokemon"? Probably quite a few, especially if you're a False Swipe Gaming viewer or have stuck around in old gen communities enough. They're wrong sometimes, but I feel that there are more benefits to this approach beyond the ability to veto rises.

Tiering by usage is usually lauded as this "democratic" approach to things, but if it's easily influenced by large figures and dedicated groups, is it really working as intended? Easy answer: no it's not. The rational solution is to take steps to "improve the democracy", but in this case, what you end up doing is the opposite as-is. The steps taken to improve usage-based tiering appear to fail at least a few times every generation, and I personally believe it's shown more in SwSh than any other. Encourage those invested in the tier to join the forums and lend their voice to what they want risen. These people could potentially become big contributors to a forum whose advertising appears to be growing worse with every passing generation. Bring them into the community. You could argue that viability-based tiering would replicate usage-based tiering's flaws in a different form - hype could still lead to a mistaken rise - but my point is error reduction, not prevention. A more controlled environment will usually have less errors, and so long as a competent leader is in place - which we can respect and expect on a site like this - I don't see why this couldn't work.

You see, old gens have this issue with Pokemon frequently getting trapped in OU as a result of the generation ending. However, RBY through ADV can avoid this because they've been tiered by viability from the start and, thus, it is continuable despite the depleted sample size. If you're after a case study, RBY's tiering exploits this well and has been astonishingly successful. If you considered the original OU list to be the closed usage stats, Persian, Golem, and probably Lapras would be "OU by Technicality" by now, trapped in the tier with no place to use them with an expected good experience. Cloyster would also be stuck in a lower-tier despite being a massive presence. Thus far, no rise or fall has been "kneejerky" thanks to the people running the show being intelligent, competent people. With this, lower-tiers are capable of evolving further while you get to see the premier ones represented properly. You end up future-proofing your tiering this way, allowing the communities that stick around to experiment further with their metagames to see if it can be made even better once it comes to a close. It's way more fun, encourages community engagement, and allows for more educated faces to strut their stuff.

At the end of the day, a tier should be fun above all else. An important Pokemon leaving because of someone's idiocy makes it unfun. A Pokemon overstaying its welcome means the lower tier has to wait longer for itself to be shaked up, which is also unfun. Why not move towards the system that, while limiting towards outsiders, also results in better tiers?

For the record, this doesn't mean killing usage stats as a "thing", they exist for the boomer gens too. They just wouldn't be influencing tier shifts with their inherent inconsistency.

Anyway, you get to stop Ambipom's ugly ass from invading people's games in tiers it has no place in! That's fucking awesome! Let's do that.

---

Now, I already know this would be shot or otherwise argued down. That's why I'm gonna translate this into the typical ad hoc compromises that people love so much!

• The veto system should absolutely be a thing if a case can be made for usage stat manipulation, especially if it is active and persistent.
• Drops should also be counted in fringe cases where a Pokemon is clearly not viable enough for the tier. Eg. If it's inches above the threshold and users are just going "just drop already ugghhhh"
• I strongly believe that my Viability Ranking presentation proposal should be implemented. This alone would make rises and drops occur much quicker by presenting what's actually good to the layman user. Less Ambipoms sticking to a tier like a tumor, more gamers rising up.

This would result in a hybrid that achieves something similar but still less stable than it could be. You remove the "Hitmontop Issue" and "Save Tyranitar" possibilities at once while getting a better interface that lets drops and rises happen more easily. TL;DR tighten the edge cases up, get some oil in your door hinges. But you still don't get the useful future-proofing viability-based tiering gets which makes me sad...

Anyway, pelt me with angries instead of hahas, they look cooler.
Or message me calling me cringe, whatever tickles your pickle.

 

[quziel]

Had a bit to think about this. Issue really comes down to the fact that as the gen gets later ladder activity dies down, and the ladder becomes far easier for single actors to influence. We saw this in SM NU with Mesprit rising to NU due, in a large part, to the efforts of one user; we're seeing this now in SS NU / RU with Hitmontop rising to NU / RU in a large part due to the efforts of one user (note a lot more people are running it now). I've computed the effective number of battles used for tiering for the years 2022, 2021, 2020 in the months April, March, February. I should note that NU was released in February 2020, so the usage stats there are obviously inflated, but just want to show how the number of battles above 1630 has varied with time (treat Feb 2020 as an outlier).

​	April​	​	​	March​	​	​	February​	​	​
​	Tot_Battles​	Avg_Weight​	Use_battles​	Tot_Battles​	Avg_Weight​	Use_battles​	Tot_Battles​	Avg_Weight​	Use_battles​
2022​	52912​	0.041​	2169.392​	43286​	0.038​	1644.868​	35882​	0.066​	2368.212​
2021​	48059​	0.09​	4325.31​	53850​	0.068​	3661.8​	44499​	0.091​	4049.409​
2020​	55625​	0.124​	6897.5​	47422​	0.116​	5500.952​	107885​	0.113​	12191.005​

I also apologize for the mediocity of the table formatting.

Basically the number of battles that matter for tiering has been dropping, precipitously, as the years go on, the gen becomes older, and interest in laddering seriously drops. My proposal to help fix this is one of the following:

1) Do not conduct rises during the final X months of a gen.
2) Rises should account for more than 3 months of usage during the final X months of a gen.

Option 1 prevents late-gen rises screwing up lower tiers (forcing them to entirely retier right before a new gen is out), eg imagine if Passimian had been very slightly higher in usage during the closing months of SM? SM NU would have been such a different meta post-gen to how it was during current gen it would be completely insane.

| 62 | Passimian | 3.203% | (3.41% cutoff)

Option 2 is a bit less nuclear, but makes it harder for a single dedicated actor to force a tier shift during the late periods of a gen.

Edit: phrasing

---

I don't particularly like the implications of a veto system. One of the major advantages of usage-based tiering is that it appears objective to an outside observer. Being able to veto rises directly impinges upon the apparent legitimacy of the usage-based tiering method, and could really hurt the perception of the most "objective" parts of our tiering method.

 

[Rabia]

was talking w/ meri earlier today a little about what quz posted more or less re: late-generation shifts, and I agree 100% that we should not be doing tier shifts up until new releases. the risk of fucking a lower tier up irreparably because of how little time left there is to do actual tiering is too great and I don't think we lose anything; I actually think we stand to gain a fair bit because you are then given roughly 3~ months to do any last-minute tiering that is guaranteed to not be affected by shifts dropping some new problem in or rising a meta staple out

 

[Plague von Karma]

Option 1 prevents late-gen screw ups, eg imagine if Passimian had been very slightly higher in usage during the closing months of SM? SM NU would have been such a different meta post-gen to how it was during current gen it would be completely insane.

| 62 | Passimian | 3.203% | (3.41% cutoff)

This screw-up has actually happened, and in this case, it's quite tragic.

Right at the end of ORAS, after being a very useful UU Pokemon, Azelf got enough usage to be ranked OU. It's C- on the Viability Rankings and was seen as "not OU" even at the time of its rise. I think this was an OLT thing or something? I forgot. Anyway, because this specifically happened the month before release, there was no time to stop Azelf from getting trapped there, and now it's not just there, but near the top of the teambuilder, which is probably why its usage has been so high ever since.

Src for graph

This is one very, very major case where locking tiers near the end of a generation could have made for a much better metagame once it closed.

 

[Marty]

So Quagsire might be PU and also get a decent amount of usage in NU or UU or wherever, what of it? If anything, the cases you're bringing up are perfect examples of Pokemon that under the current system tend to upset the balance of tiers by being defensive staples that are constantly at risk of flip-flopping between several tiers.

Well, yes, that's exactly why I mentioned them. In this hypothetical higher rise/lower drop scenario, it's completely possible that things in UU or RU, or Gastrodon or Quagsire, have higher usage (between 4.52% and 6.70%) at the time of any given shift than Pokemon that are literally in OU. But they wouldn't become OU themselves.

So, what is the meaning of OU in this case? It's hard to tell if you didn't understand the point I was making or maybe it doesn't matter to you, but personally I'm not interested in having TLs field questions about it for the next 3 years when I don't have an answer.

Since the beginning of usage-based tiering, anyone can look at the current usage stats and know what's OU just by finding the cutoff and seeing all the Pokemon above it:

 | 22   | Tyranitar          |  7.788% |
 | 23   | Blissey            |  7.750% |
 | 24   | Magnezone          |  7.683% |
 | 25   | Blacephalon        |  7.542% |
 | 26   | Pelipper           |  7.360% |
 | 27   | Zeraora            |  7.332% |
 | 28   | Volcarona          |  6.952% |
 | 29   | Slowking-Galar     |  6.741% |
 | 30   | Zapdos-Galar       |  6.367% |
 | 31   | Victini            |  6.063% |
 | 32   | Volcanion          |  5.972% |
 | 33   | Barraskewda        |  5.859% |
 | 34   | Ninetales-Alola ^  |  5.831% |
 | 35   | Mew             |  |  5.672% |
 | 36   | Regieleki       |  |  5.640% |
 | 37   | Bisharp         OU |  4.820% |
 ---------------------------------------
 | 38   | Nidoking    not OU |  4.380% |
 | 39   | Rotom-Wash      |  |  4.314% |
 | 40   | Slowking        |  |  3.908% |
 | 41   | Hippowdon       v  |  3.896% |
 | 42   | Scizor             |  3.856% |
 | 43   | Moltres-Galar      |  3.632% |

There are 8 Pokemon between 4.52% and 6.70% here, but unless it's the very first shift of the generation (when none of them would be OU) how can anyone tell any of these Pokemon are actually still in OU unless they already know? At the end of a usage cycle, all of these Pokemon stay where they were before because the rise threshold is higher than the drop threshold. Volcanion has more usage than Regieleki but the former stays RU and the latter stays OU? How do you explain this to a new player, given that Smogon claims to tier by usage?

 | 22   | Tyranitar          |  7.788% |
 | 23   | Blissey            |  7.750% |
 | 24   | Magnezone          |  7.683% |
 | 25   | Blacephalon        |  7.542% |
 | 26   | Pelipper        ^  |  7.360% |
 | 27   | Zeraora         |  |  7.332% |
 | 28   | Volcarona       |  |  6.952% |
 | 29   | Slowking-Galar  OU |  6.741% |
 ---------------------------------------
 | 30   | Zapdos-Galar    OU |  6.367% |
 | 31   | Victini         UU |  6.063% |
 | 32   | Volcanion       RU |  5.972% |
 | 33   | Barraskewda     RU |  5.859% |
 | 34   | Ninetales-Alola OU |  5.831% |
 | 35   | Mew             UU |  5.672% |
 | 36   | Regieleki       OU |  5.640% |
 | 37   | Bisharp         UU |  4.820% |
 ---------------------------------------
 | 38   | Nidoking    not OU |  4.380% |
 | 39   | Rotom-Wash      |  |  4.314% |
 | 40   | Slowking        |  |  3.908% |
 | 41   | Hippowdon       v  |  3.896% |
 | 42   | Scizor             |  3.856% |
 | 43   | Moltres-Galar      |  3.632% |

I imagine this would be pretty awkward for Kalalokki to illustrate in usage graphics as well. If nobody else cares about the optics then that's fine, I just want everyone to understand what this proposal means in practice.