All Gens Your Tiering Philosophy

This thread is more about getting every member of our forums who has some interest in our tiering process to form an opinion on tiering. We work within the tiering system, which provides players with massive flexibility.

Questions you should think about answering:

• To what extent should we adhere / aim to adhere to cartridge mechanics?
• How should we prioritise what gets suspected / banned in terms of: how much does simplicity of the ban factor into the decision, and how much is it preferred to ban pokemon versus banning items, moves, etc.?
• When considering potential (non-Pokemon) bans, how much should (possibly as-of-yet unformed) lower tiers be factored into such a consideration?¹ This is usually more important when considering move, item, etc. bans, and clauses.
  
• What is adequate balance and diversity? Can a tier be insufficiently or too diverse?
  
• When votes should happen on suspects, tiering, etc. what are your opinions on who should be permitted to vote? Should it be restricted, and when and why?
• When is a suspect test of some sort appropriate to assist in determining whether or not a possible rule-change (such as a ban) should be applied or not?
• What is your stance on adaption versus banning? Roughly - how ban-happy do you consider yourself to be? In some respects, this can be in comparison with the rest of the playerbase / community.
• Do you ever consider smaller tweaks to a metagame which are more complex and fix a smaller issue (i.e. one that isn't present in most battles) suitable - what about after a metagame has been around for several years?
• How often should the tiering cut-offs be reviewed (i.e. what is banned at a U-Stage of a Tiering Level for the forthcoming Tiering Level), and does this vary between Generation and Tiering level?
• How would you describe your tiering philosophy or elements of it?
  
• Anything else you care to think of!

¹Note that Pokemon Perfect's Tiering is "transitive" as some players call it. It is also known as ban heritability. This means that any ban, clause, etc. will be applied to all Tiers in lower Tier Stages and Tier Levels than the Tier to which it is applied, in addition to the Tier in which it is applied. To be entirely precise in explaining what ban heritability means: within a Tier Set, everything which is legal (in terms of teams, and actions within the battle) in a tier, given by its Tier Level and Tier Stage, is legal in all higher Tier Stages and Tier Levels. To be very precise but likely confusing to most readers [so unless you have experience working with posets, you should probably skip this part of the note], the set of legal teams in a Tier combined with set containment, and indeed the set of possible battles in a Tier combined with set containment, are partially ordered sets which satisfy the comparability axiom (i.e. these are Chains). This means that any ban, clause, etc. will be applied to all Tiers in lower Tier Stages and Tier Levels than the Tier to which it is applied, in addition to the Tier in which it is applied.

---

Spoiler: My Response

• To what extent should we adhere / aim to adhere to cartridge mechanics?

My view is that it is only acceptable to modify cartridge mechanics when we're out of all other reasonable options. I mean, the game we play is an emulation of what we can play on cartridge at its core. I think that moving towards modifying existing clauses to cartridge-playable clauses is a good idea. In practice I think a cart-legal equivalent isn't possible for Freeze Clause, but Sleep Clause is largely manageable but has some hurdles to overcome.

• How should we prioritise what gets suspected / banned in terms of: how much does simplicity of the ban factor into the decision, and how much is it preferred to ban pokemon versus banning items, moves, etc.?

I think the best thing to do is single-pokemon bans: that's what our tiering is primarily focussed around. That being said, I think sometimes more fine-tuned bans are appropriate when it's better understood what's problematic and collateral can be reduced, but it takes many years of experience playing a metagame to be comfortable in using more technical clauses, such as we currently have for Evasion Abilities in HGSS. On a related note: Sometime, maybe Spring this year, myself and the rest of the ADV community will review the Baton Pass clauses to try and see if there is a better possibility for our clauses: less collateral / more targeted / more simple are the 3 obvious aims, and there's obvious cost-benefit with each.

• When considering potential (non-Pokemon) bans, how much should (possibly as-of-yet unformed) lower tiers be factored into such a consideration?¹ This is usually more important when considering move, item, etc. bans, and clauses.

The sort of cases this question is generally referring to is stuff like: Sleep Clause, potential bans or limitations on moves like Scald or Spikes. A lack of data can make this very difficult to do, but not impossible: we can do sensible estimatates, looking at stuff like move distribution and whatnot, as well as experiences on other sites.

• What is adequate balance and diversity? Can a tier be insufficiently or too diverse?

I think when talking balance: this is a very, very case-by-case thing for me. I want to look at diversity for a moment: I think it's a very rough estimator, but it's not useless. There's a few cases where arguably there is insufficient diversity but sufficient balance leading to certain things being members of 1P versus 1U: take for example, GSC 1P or ORAS 1P. Both of these metagames are fairly balanced, though you can to some extent argue one way or the other. But there is insufficient diversity compared to what one would expect in a generation. Also, take a look at Smogon's ORAS OU: this metagame is more diverse than ORAS 1U, and one can argue that this diversity had some role to play in making the metagame less stable and teambuilding much more difficult (leading, until fairly late into the generation, to frequent claims of losses to matchup). I think in most cases, diversity isn't very relevant, but as a rough estimator it can be used to figure out where there is some sort of a problem.

• When votes should happen on suspects, tiering, etc. what are your opinions on who should be permitted to vote? Should it be restricted, and when and why?

Again I feel this is very case-by-case, and it is different depending on what metagame and the size of the playerbase.

• When is a suspect test of some sort appropriate to assist in determining whether or not a possible rule-change (such as a ban) should be applied or not?

I think almost never. Suspect tests should only be done when understanding the metagame with a rule-change or ban implemented is important in making a decision, which in my view is very rare: when something is broken you normally know it without having to play without it to recognise it.

• What is your stance on adaption versus banning? Roughly - how ban-happy do you consider yourself to be? In some respects, this can be in comparison with the rest of the playerbase / community.

I think this is hard to self-judge on. I think I probably am more ban-happy than most though, being honest.

• Do you ever consider smaller tweaks to a metagame which are more complex and fix a smaller issue (i.e. one that isn't present in most battles) suitable - what about after a metagame has been around for several years?

I think these are often suitable. So long as it makes sense when considering the comparability of Tiers, I think small tweaks make a lot of sense, though a lot of care should be taken to balance simplicity with targetedness.

• How often should the tiering cut-offs be reviewed (i.e. what is banned at a U-Stage of a Tiering Level for the forthcoming Tiering Level), and does this vary between Generation and Tiering level?

I think the answer for right now is not very often at all. This could change in the future, but only if we grow substantially more than I anticipate.

• How would you describe your tiering philosophy or elements of it?

No real way to describe it rn.

 

• To what extent should we adhere / aim to adhere to cartridge mechanics?

To the point where it makes sense not to. If something is easily handled by making a clause / condition not entirely enforceable on cartridge, I have 0 problem myself, especially given that there is precedent set already (Sleep Clause, Freeze Clause).

• How should we prioritise what gets suspected / banned in terms of: how much does simplicity of the ban factor into the decision, and how much is it preferred to ban pokemon versus banning items, moves, etc.?

I feel like this is extremely case by case. In general, it will be easier to just ban the Pokemon with the broken element and call it a day. However, as the game goes on, more and more specialty situations arrive. Most recently, a lot of high level players in S/M have been clamoring for something to be done with Z-moves. As opposed to banning the Z-mechanics outright, or banning every pokemon with a good Z-move set (which is quite a few), the best option would likely be complexity.

• When considering potential (non-Pokemon) bans, how much should (possibly as-of-yet unformed) lower tiers be factored into such a consideration?1 This is usually more important when considering move, item, etc. bans, and clauses.

The metagame in which it is broken should be prioritized far and away. There are pretty few situations in which balancing via transitivity causes a lower tier to be unbalanced (none have occurred to this point to my knowledge).

• What is adequate balance and diversity? Can a tier be insufficiently or too diverse?

I believe diversity is crucial for a metagame. Seeing the same team with a slight twist makes for a very boring metagame most of the time. This, of course, is nearly unavoidable in early gens, (especially RBY, not accounting for Piex) but evaluation of diversity should certainly come into play when a suspect is on the table.

• When votes should happen on suspects, tiering, etc. what are your opinions on who should be permitted to vote? Should it be restricted, and when and why?

I think while the site is still sort of "budding" as it is, voting should be a fairly available right, but the person has to demonstrate some sort of knowledge. Whether it's through just playing,making posts in the threads, or whatever else, we need to make sure some schmuck can't just come in, make 8 accounts, and overrule a suspect just for shits and gigs. (this would be hard to do practically but I'm exaggerating to make a point)

• When is a suspect test of some sort appropriate to assist in determining whether or not a possible rule-change (such as a ban) should be applied or not?

I feel like suspects should be used very very liberally. Honestly, I would make any Pokemon that achieves S-Rank in a tier, even if the tier has been established for a while, undergo a suspect. However, I do not thing suspects should result in a norm of ban, as they do on Smogon currently (something like 90% of suspects resulting in ban, which astounds me). Higher evaluation of Pokemon at the top level of the metagame and what their absence would mean for the meta should be considered in order to achieve diverse and desirable tiers.

• What is your stance on adaption versus banning? Roughly - how ban-happy do you consider yourself to be? In some respects, this can be in comparison with the rest of the playerbase / community.

I would consider myself fairly conservative when it comes to banning. Several suspect bans / quickbans on Smogon throughout the course of XY/ORAS had me scratching my head.

• Do you ever consider smaller tweaks to a metagame which are more complex and fix a smaller issue (i.e. one that isn't present in most battles) suitable - what about after a metagame has been around for several years?

I feel like metagames should constantly be explored as long as the community is there to explore them. Tweaks, small bans, etc. to an existing metagame could be very beneficial in the long run if executed with care.

• How often should the tiering cut-offs be reviewed (i.e. what is banned at a U-Stage of a Tiering Level for the forthcoming Tiering Level), and does this vary between Generation and Tiering level?

Whenever there is a metagame trend that makes for the statements "is this mon REALLY xU?" or "this mon really fits in xU better than we thought" to be said. If we really wanted, it could be a quarterly/monthly thing, but I'm not convinced that will be needed until newer gens.


If I think of anything else to add, I'll do so in the morning.

 

Since I'm one of the main people responsible for tiering I suppose I ought to put my own response! Also note that I'm solely interested in gen 1/2, so I may not always consider later gens in my responses

• To what extent should we adhere / aim to adhere to cartridge mechanics?

I'm very much a cartridge purist- freeze clause ought to be scrapped, sleep clause ought to be modified in order to resemble what you'd use playing on cart, all that

• How should we prioritise what gets suspected / banned in terms of: how much does simplicity of the ban factor into the decision, and how much is it preferred to ban pokemon versus banning items, moves, etc.?

Simplicity is a very high priority, and I think is far more important than collateral, such that the collateral would have to be extreme in order for me to consider it- at that point there's probably something wrong our choice of suspect. Banning pokemon rather than moves/items ties into this into this in my opinion as we should primarily stick to banning pokemon- these are the elements of the game people most identify with and are what we target in our tiering system. They are also generally much easier to isolate than non-pokemon stuff.

• When considering potential (non-Pokemon) bans, how much should (possibly as-of-yet unformed) lower tiers be factored into such a consideration?¹ This is usually more important when considering move, item, etc. bans, and clauses.

It's not so much lower tiers as the fact that any argument for their ban has to be universal and not specific to that tier/meta. Saying you have to also consider the lower tiers focusses on the practicalities of this condition but still violates the condition by being meta specific.

• What is adequate balance and diversity? Can a tier be insufficiently or too diverse?

Adequate balance is where every pokemon available has a reasonable number of checks and counters or can in practice be reasonably managed. Idk, providing a definition of this is kinda meaningless since it's so subjective. I don't think diversity is a meaningful criterion for evaluating a tier in that it's not something we should directly be evaluating, only potentially using as an indicator and nothing more. Lack of diversity is not necessarily an obstacle for a tier being good, nor is plenty of diversity a guarantor for the same state. Diversity is however correlated with other things that might be of importance- Snorlax in GSC is a good example, as (practically) every team uses it and runs 2/3 checks to it, which suggests Snorlax might be ban-worthy. By contrast Exeggutor in RBY is on the majority of teams, but it's a really big stretch to suggest it's broken imo, so it's clearly not perfect. There was also a good discussion had a few days ago where we discussed how diversity issues aren't meaningful if they don't arise from a small proportion of the tier's inhabitants- RBY 5U is another interesting example, where there's a relatively small number of viable pokemon, but no major centralising forces that you could ban to increase diversity, in which case I would say not to do anything about it

tl;dr: balance is nice but I'll know adequate when I see it, diversity is a silly criterion and should only be used as a rough indicator

• When votes should happen on suspects, tiering, etc. what are your opinions on who should be permitted to vote? Should it be restricted, and when and why?

This is as much a practical question as anything. At the moment the only requirement should be some experience in the tier, simply because we need to maximise the number of voters. Desirable qualities I guess would be to be an expert rather than merely competent, and also in the case of tiers that are perceived as clashing with those of other sites, someone who cares about the identity of this site as a whole and doesn't want us to merely replicate another site. Also I love the council system, which is a reasonable compromise between favouring desirable characteristics and giving everyone a say

• When is a suspect test of some sort appropriate to assist in determining whether or not a possible rule-change (such as a ban) should be applied or not?

Suspect test as in temporarily banning something to test what a tier's like without it? I've never seen the purpose of those. If something's broken, ban it, this is something you should be able to identify without a full-on test. If anything, it allows an opening for broken-checking-broken scenarios to survive. The exception to this is if you ban for anything other than removing a broken element, which I don't think should be done- an example of this would be if you were to ban Seadra in RBY 5U to improve the tier's diversity as it's not something you'd know well enough from a discussion (you can guess but it's just a guess)

• What is your stance on adaption versus banning? Roughly - how ban-happy do you consider yourself to be? In some respects, this can be in comparison with the rest of the playerbase / community.

I generally consider myself fairly averse to bans, although not always by a great deal. The most notable example is when the legendary birds got banned in RBY 2U, I was opposed to Articuno's banning. Then again, I'm maybe the most vocal in pushing to ban Dragonair in 5U so who knows.

• Do you ever consider smaller tweaks to a metagame which are more complex and fix a smaller issue (i.e. one that isn't present in most battles) suitable - what about after a metagame has been around for several years?

I'm not sure what you're getting at here. Is this stuff like Ingrain Passing in ADV? Firstly, I don't think time should matter- we should treat every tier as though it's current. And I'd generally advocate for the simplest option regardless, a good example being my perspective on trapping in GSC.

• How often should the tiering cut-offs be reviewed (i.e. what is banned at a U-Stage of a Tiering Level for the forthcoming Tiering Level), and does this vary between Generation and Tiering level?

Well, how stable is the metagame? How well explored is it? This is something that should be open to discussion and I really don't think there should be any specific timeline other than "I think a lot of people wanna review stuff"

• How would you describe your tiering philosophy or elements of it?

I would say I stick pretty rigidly to certain rules- simplicity, banning stuff only if it's broken and the like

 

• To what extent should we adhere / aim to adhere to cartridge mechanics?

My view is that it is only acceptable to modify cartridge mechanics when we're out of all other reasonable options. I mean, the game we play is an emulation of what we can play on cartridge at its core. I think that moving towards modifying existing clauses to cartridge-playable clauses is a good idea. In practice I think a cart-legal equivalent isn't possible for Freeze Clause, but Sleep Clause is largely manageable but has some hurdles to overcome.

• How should we prioritise what gets suspected / banned in terms of: how much does simplicity of the ban factor into the decision, and how much is it preferred to ban pokemon versus banning items, moves, etc.?

I think the best thing to do is single-pokemon bans: that's what our tiering is primarily focussed around. That being said, I think sometimes more fine-tuned bans are appropriate when it's better understood what's problematic and collateral can be reduced, but it takes many years of experience playing a metagame to be comfortable in using more technical clauses, such as we currently have for Evasion Abilities in HGSS. On a related note: Sometime, maybe Spring this year, myself and the rest of the ADV community will review the Baton Pass clauses to try and see if there is a better possibility for our clauses: less collateral / more targeted / more simple are the 3 obvious aims, and there's obvious cost-benefit with each.

• When considering potential (non-Pokemon) bans, how much should (possibly as-of-yet unformed) lower tiers be factored into such a consideration?¹ This is usually more important when considering move, item, etc. bans, and clauses.

The sort of cases this question is generally referring to is stuff like: Sleep Clause, potential bans or limitations on moves like Scald or Spikes. A lack of data can make this very difficult to do, but not impossible: we can do sensible estimatates, looking at stuff like move distribution and whatnot, as well as experiences on other sites.

• What is adequate balance and diversity? Can a tier be insufficiently or too diverse?

I think when talking balance: this is a very, very case-by-case thing for me. I want to look at diversity for a moment: I think it's a very rough estimator, but it's not useless. There's a few cases where arguably there is insufficient diversity but sufficient balance leading to certain things being members of 1P versus 1U: take for example, GSC 1P or ORAS 1P. Both of these metagames are fairly balanced, though you can to some extent argue one way or the other. But there is insufficient diversity compared to what one would expect in a generation. Also, take a look at Smogon's ORAS OU: this metagame is more diverse than ORAS 1U, and one can argue that this diversity had some role to play in making the metagame less stable and teambuilding much more difficult (leading, until fairly late into the generation, to frequent claims of losses to matchup). I think in most cases, diversity isn't very relevant, but as a rough estimator it can be used to figure out where there is some sort of a problem.

• When votes should happen on suspects, tiering, etc. what are your opinions on who should be permitted to vote? Should it be restricted, and when and why?

Again I feel this is very case-by-case, and it is different depending on what metagame and the size of the playerbase.

• When is a suspect test of some sort appropriate to assist in determining whether or not a possible rule-change (such as a ban) should be applied or not?

I think almost never. Suspect tests should only be done when understanding the metagame with a rule-change or ban implemented is important in making a decision, which in my view is very rare: when something is broken you normally know it without having to play without it to recognise it.

• What is your stance on adaption versus banning? Roughly - how ban-happy do you consider yourself to be? In some respects, this can be in comparison with the rest of the playerbase / community.

I think this is hard to self-judge on. I think I probably am more ban-happy than most though, being honest.

• Do you ever consider smaller tweaks to a metagame which are more complex and fix a smaller issue (i.e. one that isn't present in most battles) suitable - what about after a metagame has been around for several years?

I think these are often suitable. So long as it makes sense when considering the comparability of Tiers, I think small tweaks make a lot of sense, though a lot of care should be taken to balance simplicity with targetedness.

• How often should the tiering cut-offs be reviewed (i.e. what is banned at a U-Stage of a Tiering Level for the forthcoming Tiering Level), and does this vary between Generation and Tiering level?

I think the answer for right now is not very often at all. This could change in the future, but only if we grow substantially more than I anticipate.

• How would you describe your tiering philosophy or elements of it?

No real way to describe it rn.

 

Fwiw I strongly dispute the idea that GSC ubers is too lacking in diversity. That's just an excuse we made to rationalise our concession to the various players who don't want to see anything besides gsc ou, despite the fact that it's glaringly inconsistent with any sort of sense in tiering. If GSC ubers were lacking in diversity, why would we pick gsc ou's rule set? If anything, Lax would be one of the first to go. I mean if we weren't just appeasing the gsc ou players and didn't have that as an "alternative" what would we do about this supposed issue?

 

-To what extent should we adhere / aim to adhere to cartridge mechanics?

Cartridge mechanics are a mere starting point for tiering, from there, ban shit, clause shit, and move on from there. We have already left the realm of "cartridge accuracy" with the ability to see past turns, chat with your opponent, see the opponent's spent PP, etc. I have no objections to out-and-out modifying game mechanics if they'd make the tier better, lowering the PP of one-turn recovery moves, making Scald utterly useless, for example. Etc etc.

If we gave too much of a damn about what that infernal company thinks, you people would all be playing VGC, stall would be deader than Harambe because you were playing VGC, and fuck cartridge accuracy stop letting stupid dogma stop you from making the tiers better.

-How should we prioritise what gets suspected / banned in terms of: how much does simplicity of the ban factor into the decision, and how much is it preferred to ban pokemon versus banning items, moves, etc.?

Ban simplicity is irrelevant. Choose whatever ban makes the tier better. Ban Scald/Brightpowder/Snow Warning/whatever makes the tier better etc etc.

-What is adequate balance and diversity?
Have your tiers be different. Have actual choice in said tiers. Relative playstyle strength in tiers can vary, and that's fine. More choices for everyone that way. One tier may be ideal for stall-vs-stall hazard wars, another may be more offense-vs-offense, etc. I don't complain about hyper offense being unviable in RBY/GSC Ubers, but you shouldn't complain if a tier ever comes along where stall is unviable. It happens.

Ban things and clause things to make the tier optimized to its strengths, rather than some overarching dogma of how "all tiers should be". You guys tweaked ADV 2U until it was a balance-stall fatfest; however for a meta that clearly refuses to let stall be viable without several dozen bans you just say "fuck it this is an offensive tier, offense vs offense vs balance here we go" and ban according to that, rather than attempting to force stall onto every single tier. If every playstyle needed to be viable in every tier, we'd have to ban lots of shit in RBY and GSC to make hyper offense viable.

-Can a tier be insufficiently or too diverse?

This is relative. Dropping the whole "cartridge accuracy" bullshit will flatout allow you to add more mons in old gens, or thin out the movepools of mons in later gens, for example.

-What is your stance on adaption versus banning? Roughly - how ban-happy do you consider yourself to be? In some respects, this can be in comparison with the rest of the playerbase / community.

Just preemptively get rid of the game element if you think it may be an issue in the future; I'm not the kind of person to just dance around the issue of Mega Sableye only to ban it in the end anyways, as I would've just slapped that fucker on the initial banlist so I can actually have counterplay through that TauntWispSplit whatever that alleviates stupid matchup issues.

-Do you ever consider smaller tweaks to a metagame which are more complex and fix a smaller issue (i.e. one that isn't present in most battles) suitable - what about after a metagame has been around for several years?

Sure.

-How often should the tiering cut-offs be reviewed (i.e. what is banned at a U-Stage of a Tiering Level for the forthcoming Tiering Level), and does this vary between Generation and Tiering level?

Case-by-case basis.

-How would you describe your tiering philosophy or elements of it?
I don't let dogma or cartridge accuracy or whatever shit stop me. I don't let anything stop me. I play what I want to play, and you guys should play what you want to play, without letting dumb shit stop you.

I may or may not answer the other questions later.

 

You do realise that literally all of the features you described as abandoning cart accuracy can be done while playing on cart right? Tracking past turns and PP are accomplished by good old pen and paper, while there's nothing stopping you from talking to an opponent irl. The only genuine mods to cart mechs off the top of my head are sleep/freeze clause (which should be changed or scrapped), exact % HP, and the removal of some glitches, such as desyncing in RBY (breaks the game) and Acid Rain in gen 4 (impossible to implement fairly)

I tend to agree that the whole "all playstyles must be viable" thing isn't super great. This is more based on the fact that I don't really like the idea of attaching labels and categories for these things, so talking about stall, bulky offense, HO, whatever I don't find very meaningful.

 

(Note that we should really have it measured in pixels, split into 48ths) 48ths vs percentages for HP | Smogon Forums

... maybe Enigami can get this implemented on the server

(Counter desync in RBY is avoidable, I think Psybeam should be banned since you basically RNG a desync so it's unavoidable

All Gens - Acid Rain and host cartridge related play issues. | Pokémon Perfect

 

Yeah I meant Psywave