All Gens Overcentralisation

[Disaster Area]

So lately I've been thinking about the conundrum that it seems to me that the majority of players with experience playing a given ubers tier (outside of RBY) think that their tier is adequately balanced & competitively deep to be 1U, but that point of view is not commonly held amongst people with limited to no ubers experience. Basically, to determine 1U, players have a lot of priors, which makes it far harder to tier satisfactorially, in my opinion. I've spoken with a few friends about this with a variety of perspectives (thanks Lusch, M Dragon, Cowboy Dan, and Ortheore) and am coming away with some realisations and maybe even some hope that we can find a way to tier 1U satisfactorily.

Now, having said that, I want to primarily focus in this topic on the issue of overcentralisation, what it means, and to what extent it matters. I think the first thing to do is to describe what centralisation is.

A metagame is more centralised if there are fewer viable Pokemon, at various levels of viability (e.g. Pokemon which are at least considered A ranks in a metagame's viability rankings).

Overcentralisation is when one or a few Pokemon have particularly outsized usage in their generational context, such that success in the metagame revolves around using those Pokemon more effectively.

Undercentralisation (one would refer to a metagame experiencing undercentralisation as being decentralised) is when there is such a wide array of viable options that it becomes exceptionally difficult to build teams that aren't reliably beaten by one of the many top-level Pokemon in the metagame.

Undercentralisation is a rare problem, so I will only discuss it briefly. An example of an undercentralised metagame may have been ORAS OU, at least during its more formative stages. A symptom of undercentralisation is the playerbase complaining about matchup. That being said, undercentralisation is not the only possible cause of matchup problems; metagames which revolve around weather-based archetypes often exhibit matchup problems in spite of having a fairly manageable number of viable Pokemon.

One should also note that these definitions do rely to some extent on generational context (making them more subjective too), but I would argue that so does competitive depth: you have far more potential for competitive depth in USUM than in RBY, for example, just due to the number of viable Pokemon, as well as the introduction of various game mechanics.

I would argue that a metagame being overcentralised does undermine competitive depth: there are too few viable options, and that restricts core gameplay significantly more than what is achievable (in a given generation).

I will give some examples of metagames I do or don't consider overcentralised, and use this to inform the different ways we can measure and discuss overcentralisation:

• RBY Ubers is overcentralised: the main focus of the metagame is sweeping the opponent with your Mewtwo before they sweep you with yours. Whilst that's a bit of an oversimplification, RBY Ubers is pretty clearly overcentralised.
  
• In my opinion, RBY OU is not overcentralised (within the RBY context). Whilst there are some Pokemon which should almost always be used, this is the case in every RBY tier I have ever played (which includes 6 different lower tiers as well as OU and Ubers). Hence, individual Pokemon having near-100% usage is in my opinion not enough to consider a Pokemon overcentralising in the generational context of RBY. The best Pokemon in the tier are of a similar power level, and there are numerous viable Pokemon in the tier which can check each of these Pokemon and aren't usually used just because of a particularly good matchup against one of the omnipresent Pokemon.
  
• HGSS OU is not overcentralised. There are numerous different approaches and strategies within each archetype (such as offense and stall) and even subarchetypes (e.g. different sorts of offenses). It's a great example of a metagame which is not overcentralised.
• In my opinion, ADV Ubers is overcentralised. Kyogre, the Lati twins, and Groudon are so good that every team should have one and usually more than one of them. Kyogre enables incredibly niche Pokemon like Ludicolo, Quagsire, Shedinja, and Lanturn (which do not pop up in OU virtually ever) to be considered serious & viable options almost entirely because of their ability to answer Kyogre (and in Shedinja's case, the Lati twins also).
  
• ORAS Ubers is overcentralised. Primal Groudon is such an excellent support Pokemon, as well as significantly better than almost all other Geomancy Xerneas checks, that it enjoys sky-high usage in a generation which includes over 700 Pokemon. It also key in inhibiting Pokemon which were previously top 5 or 10 threats such as non-Primal Kyogre and Palkia from being viable.
• ORAS 1U (as it currently is) is in my opinion not overcentralised. There are a number of different top offensive threats, and responses to those threats, and in my opinion the diversity of options in the tier is adequate in the generational context.

Here are some indicators then for overcentralisation:

- A Pokemon or several Pokemon with particularly high usage.
- A Pokemon having a particularly small pool of checks & counters.
- A Pokemon having checks and counters which are only viable because they check or counter that Pokemon, and which see at all significant usage.

Cowboy Dan suggested that we collect usage stats for New Frontiers (and Ortheore suggested that furthermore that we should both have it split to week by week as well as having an overall total); it is both useful for measuring indicators of overcentralisation, and for purposes such as having experts help identify Pokemon which are seeing inadequate usage given their prospective viability.

And finally, I would like to remark that sometimes a ban will decrease competitive depth (e.g. one can definitely argue that banning Kyogre from ADV 1U would make the tier less competitively deep). As multi-suspects are out of the question (although I am fine with them, most players it appears are not), one should note that what matters is not whether the metagame is more competitively deep after a single ban, but whether the metagame before the ban is adequately competitively deep. Furthermore, the Pokemon which should be first targeted for a ban is the one most likely to be undermining the tier's competitive depth. At the very least, any Pokemon which is suspected should be one of a number suspected of undermining competitive depth. So, whilst for example ADV Ubers is balanced and decentralised to a certain extent, and banning a single Pokemon from the tier will lead to a less balanced and less decentralised tier, we would still be right to ban a Pokemon from the tier if we believed that the tier was not adequately balanced and decentralised before a ban.

 

[Ortheore]

Most of this post I think looks good. I disagree on rby 1u in that I think it is overcentralised even after you compare with other rby generations- the only one that comes close is RBY 5U, but even then I think that 1U's even more overcentralised in that there's less flexibility in how you use the big four (5U has a big 4 + 2 others, 1U has big 4 + lead + 1 other, at least most of the time anyway). That said, I don't think it's a terribly important thing to point out lol since it's just a difference in opinion.

I was going to write up a note in my own guidelines explicitly describing centralisation and how it fits in with my framework, specifically the part where I link traditional definitions of brokenness to the idea of adequate competitive depth, but then I got to thinking- to what extent can centralisation be dissociated from brokenness? We know overcentralisation by its effects when teambuilding, requiring certain pokemon be run, while niche checks that are otherwise useless may get a boost in viability just as some otherwise promising pokemon may be adversely affected. But what about if a pokemon doesn't directly have that impact, but instead just raises the power level of the tier to a point where few of the remaining pokemon can compete? Actually, the more I think on it, the more this seems like something one pokemon/element isn't sufficient to produce, so idk. I guess an example would be dropping Ho-oh into GSC OU. Based on my ubers experience, Ho-oh doesn't feel at all broken, there are plenty of checks, most of which are viable regardless and so on. But it raises the overall power level of the tier considerably, due to the need to threaten a defensive behemoth. To me, that's something that is awkward to fit in to most ideas of centralisation (note I'm still 100% in favour of GSC Ubers)

That said, the question of centralisation does not alter my position which is that my guidelines are the way to go, as I believe centralisation is merely an indicator that is suggestive of potential issues, but not entirely reliable as tiers can be healthy with some degree of centralisation, and unhealthy in its absence. The reason for writing the note is to simply clarify how common phenomena fit into my framework (although I'd argue that the traits defining centralisation are already incorporated into my guidelines)

Undercentralisation is a good thing to bring up imo. I don't think it's an issue in any of our current or prospective tiers, but it can be an issue and I think it requires a novel solution. The obvious one is to unban something, but I don't think it's reliable in that there would need to be a viable candidate to unban, which in theory there shouldn't be, as we should only ban things if they're known to cause problems, in which case that pokemon should stay banned. Nothing else springs to mind immediately.

 

[Disaster Area]

Most of this post I think looks good. I disagree on rby 1u in that I think it is overcentralised even after you compare with other rby generations- the only one that comes close is RBY 5U, but even then I think that 1U's even more overcentralised in that there's less flexibility in how you use the big four (5U has a big 4 + 2 others, 1U has big 4 + lead + 1 other, at least most of the time anyway). That said, I don't think it's a terribly important thing to point out lol since it's just a difference in opinion.

Yeah I mean I think arguing over whether a tier is overcentralised is a healthy one and, well to me personally, RBY 1U is not overcentralised in the context of what can be achieved in competitive RBY, but even if it were (which is an understandable position), I think that the tier is about as competitively deep as it gets in RBY (most lower tiers besides in my opinion maybe RBY 5U are less competitively deep, for example, and I think a lot of that is due to the prominence of recover and, to a lesser extent perhaps, explosion users in RBY 1U).

I was going to write up a note in my own guidelines explicitly describing centralisation and how it fits in with my framework, specifically the part where I link traditional definitions of brokenness to the idea of adequate competitive depth, but then I got to thinking- to what extent can centralisation be dissociated from brokenness?

I don't think brokenness really matters in the context of centralisation. In my opinion, if a metagame is overcentralised, you need a very good reason to argue why, in spite of that, doing bans would only reduce or maintain a similar level of competitive depth, even if a resulting metagame was less centralised. For example, prominence of users with moves or abilities that add to competitive depth (such as recovery moves, weather abilities, and so on) could provide a compelling counterargument.

We know overcentralisation by its effects when teambuilding, requiring certain pokemon be run, while niche checks that are otherwise useless may get a boost in viability just as some otherwise promising pokemon may be adversely affected. But what about if a pokemon doesn't directly have that impact, but instead just raises the power level of the tier to a point where few of the remaining pokemon can compete? Actually, the more I think on it, the more this seems like something one pokemon/element isn't sufficient to produce, so idk. I guess an example would be dropping Ho-oh into GSC OU. Based on my ubers experience, Ho-oh doesn't feel at all broken, there are plenty of checks, most of which are viable regardless and so on. But it raises the overall power level of the tier considerably, due to the need to threaten a defensive behemoth. To me, that's something that is awkward to fit in to most ideas of centralisation (note I'm still 100% in favour of GSC Ubers)

I do think that in some cases maybe a single element doesn't produce it, but maybe a "small" list of Pokemon may. Again, I think it's an argument for the possibility of suspecting multiple things. Like I think banning any single element from ADV Ubers will make it less deep, but if you were to ban Kyogre, Soul Dew, and Groudon, you would at least tone down the power level quite considerably (though I think a number of Pokemon would still need bans, there's far more Pokemon that can come close to matching the power levels of Lugia, say, than Kyogre). I do think overall that it is possible for a Pokemon or small group of Pokemon to restricts the viability of a number of Pokemon simply because their power level is just on another plane, and I think in that circumstance a ban is a good idea. (I wonder if maybe Kabutops is an example of that in RBY 6U?)

That said, the question of centralisation does not alter my position which is that my guidelines are the way to go, as I believe centralisation is merely an indicator that is suggestive of potential issues, but not entirely reliable as tiers can be healthy with some degree of centralisation, and unhealthy in its absence. The reason for writing the note is to simply clarify how common phenomena fit into my framework (although I'd argue that the traits defining centralisation are already incorporated into my guidelines)

Well my view of this was that it was meant more to complement the guidelines / focus on a specific aspect of it, rather than supplant it. I think there is a healthy window for the amount of centralisation and if something falls just a little outside of it (like maybe ADV Ubers or RBY 1U could be argued to be on that edge) then u should consider other traits of the tier as well as look at the pool of legal Pokemon to form a more nuanced picture of whether the tier's competitive depth is adequate.

Undercentralisation is a good thing to bring up imo. I don't think it's an issue in any of our current or prospective tiers, but it can be an issue and I think it requires a novel solution. The obvious one is to unban something, but I don't think it's reliable in that there would need to be a viable candidate to unban, which in theory there shouldn't be, as we should only ban things if they're known to cause problems, in which case that pokemon should stay banned. Nothing else springs to mind immediately.

Yeah, I think undercentralisation is a really tricky problem to deal with. It's something that we should acknowledge exists and be glad that it will most likely rarely afflict our tiers.

 

[Ortheore]

DA I honestly have no idea where this discussion is leading. From what I can see, most of the defining traits of overcentralisation are already outlined in my guidelines but it's clear you think otherwise. If this has something to do with me arguing against using overcentralisation as an argument in the past, then it's possible that it may have been when I first proposed banning based on competitive depth, otherwise idk, but again, I think it's a moot point since most of the defining traits have already been incorporated into my guidelines. So what do you think I'm missing?

Also brokenness is definitely linked to overcentralisation, since they share many of the same traits and the former is highly likely to cause the latter.

 

[Disaster Area]

To be honest all it comes from is me trying to understand why people think the Ubers tiers are fine for 1U and just trying to get more of a handle on what competitive depth really means, and centralisation is something that's much easier for me to work with than competitive depth. I think all that I want is just some more explicit mention of overcentralisation and the argument of why that can be argued to be restricting competitive depth. I guess yes overcentralisation and brokenness are linked but I also think it's far easier to argue that something is overcentralising than that something is broken; I think overcentralisation is a relatively more objective than brokenness and it's easier to help people escape their perspectives by talking about things in that way.