-
Welcome to Smogon! Take a moment to read the Introduction to Smogon for a run-down on everything Smogon, and make sure you take some time to read the global rules.
Unpopular opinions
- Thread starter antemortem
- Start date
"I still can't believe my dragons lost to you, <player>! You're now the Pokémon League champion! …Or, you would have been, but you have one more challenge ahead. You have to face another trainer! His name is…"
"Joey! He beat the Elite Four before you. He is the real Pokémon League champion!"
"How are your Pokémon doing? My Rattata's raring to go, just like always."
"How are your Pokémon doing? My Rattata's raring to go, just like always."
"I still can't believe my dragons lost to you, <player>! You're now the Pokémon League champion! …Or, you would have been, but you have one more challenge ahead. You have to face another trainer! His name is…""Joey! He beat the Elite Four before you. He is the real Pokémon League champion!"
"How are your Pokémon doing? My Rattata's raring to go, just like always."
Ok but deadass, Joey using the FEAR strat with a Rattata, then switching to like, lvl 90 mons while you're at lvl 50 would be sooooo funny
Last edited:
Pokemon should learn dramatically fewer moves than they do now. Gen1 was closer to having the right idea on movepools than modern day.
I've traditionally approached this from a competitive play, power creep, option creep angle. Straightforwardly, the more tools Pokemon have, the stronger they become, raising the power level. More insidiously, the more widely that tools are distributed, the more competition there is for roles. Not only is it easier for Pokemon to get outclassed, since good moves like Rapid Spin become less of a unique advantage, but the power level of teams rises in a way less intuitive than the first way. If the only spinner is only good at spinning, using it creates holes that your team comp has to fill. If you can pick from many spinners who have tons of good moves to serve whatever roles you want, the rest of your team can be cranked up in power instead of picking up pieces.
What inspired me to make this post is a new approach, one based on flavor instead of competitive play. The more bloated movepools are, and the more commonly moves are spread out, the harder it is for individual moves to inform a Pokemon's flavor, and the harder it is for a Pokemon's movepool as a whole to inform its flavor.
Those two points are kind of abstract, so I'll do an example. Part of Gyarados's flavor is burning stuff down and shooting beams. If it was the only Pokemon to get Flamethrower, that would follow through on its flavor well, showing its unique ability to burn stuff down with beams. However, so many things get Flamethrower that Gyarados learning it isn't super notable. Separately, because Gyarados learns so many different moves, Flamethrower is just one drop in its sea of options, and struggles to stand out as a part of Gyarados's identity.
My gut instinct is, aside from widely-dispersed TMs and designated-bad-earlygame-moves, most Pokemon's movepools can be condensed to roughly 15 or so real options. This condensing would improve flavor by committing to represent what is important about the Pokemon, and not getting mixed up in a sea of noise. I'll take Gyarados as an example. Gyarados is a challenge and a hard case for my philosophy, since it's trying to squeeze a lot of ideas into one Pokemon.
Gyarados: 20 (plus common TMs like Rest, Hyper Beam, Substitute, etc., and early-game bad-moves like Bulldoze, Tackle, and Dragon Breath)
Aqua Tail
Waterfall
Dive
Bounce (only because Magikarp should have it)
Fly (New)
Earthquake
Double-Edge
Crunch
Outrage
Flail
Dragon Tail
Surf
Hydro Pump
Fire Blast
Ice Beam
Hurricane
Dragon Pulse
Dragon Dance
Roar
Splash (From Magikarp)
Notable Omissions: Thunder Wave, Taunt, Scale Shot, Thunder(bolt), Stone Edge, Power Whip, Ice Fang, Curse
To preserve variety, some moves could be swapped in and out across generations.
I've traditionally approached this from a competitive play, power creep, option creep angle. Straightforwardly, the more tools Pokemon have, the stronger they become, raising the power level. More insidiously, the more widely that tools are distributed, the more competition there is for roles. Not only is it easier for Pokemon to get outclassed, since good moves like Rapid Spin become less of a unique advantage, but the power level of teams rises in a way less intuitive than the first way. If the only spinner is only good at spinning, using it creates holes that your team comp has to fill. If you can pick from many spinners who have tons of good moves to serve whatever roles you want, the rest of your team can be cranked up in power instead of picking up pieces.
What inspired me to make this post is a new approach, one based on flavor instead of competitive play. The more bloated movepools are, and the more commonly moves are spread out, the harder it is for individual moves to inform a Pokemon's flavor, and the harder it is for a Pokemon's movepool as a whole to inform its flavor.
Those two points are kind of abstract, so I'll do an example. Part of Gyarados's flavor is burning stuff down and shooting beams. If it was the only Pokemon to get Flamethrower, that would follow through on its flavor well, showing its unique ability to burn stuff down with beams. However, so many things get Flamethrower that Gyarados learning it isn't super notable. Separately, because Gyarados learns so many different moves, Flamethrower is just one drop in its sea of options, and struggles to stand out as a part of Gyarados's identity.
My gut instinct is, aside from widely-dispersed TMs and designated-bad-earlygame-moves, most Pokemon's movepools can be condensed to roughly 15 or so real options. This condensing would improve flavor by committing to represent what is important about the Pokemon, and not getting mixed up in a sea of noise. I'll take Gyarados as an example. Gyarados is a challenge and a hard case for my philosophy, since it's trying to squeeze a lot of ideas into one Pokemon.
Gyarados: 20 (plus common TMs like Rest, Hyper Beam, Substitute, etc., and early-game bad-moves like Bulldoze, Tackle, and Dragon Breath)
Aqua Tail
Waterfall
Dive
Bounce (only because Magikarp should have it)
Fly (New)
Earthquake
Double-Edge
Crunch
Outrage
Flail
Dragon Tail
Surf
Hydro Pump
Fire Blast
Ice Beam
Hurricane
Dragon Pulse
Dragon Dance
Roar
Splash (From Magikarp)
Notable Omissions: Thunder Wave, Taunt, Scale Shot, Thunder(bolt), Stone Edge, Power Whip, Ice Fang, Curse
To preserve variety, some moves could be swapped in and out across generations.
I don't think I agree. I think pokemon having wide movepools is one of the reasons modern pokemon is much more flexible with building and allows you to actually use your favorite pokemon, without it being a dead weight good for nothing party member.
it also makes pokemon a lot more shallow as creatures. Yeah, gyarados dex entries talk a lot about hyper beam, but its pretty clear its a violent pokemon that makes use of its body to destroy things too? removing stuff like ice fang and scale shot, which make thematic sense, is kind of silly to me.
as for power creep... man we're 1000 pokemon in lol. This ship has long since sailed, power creep is always going to be part of this franchise. Just ban pokemon
it also makes pokemon a lot more shallow as creatures. Yeah, gyarados dex entries talk a lot about hyper beam, but its pretty clear its a violent pokemon that makes use of its body to destroy things too? removing stuff like ice fang and scale shot, which make thematic sense, is kind of silly to me.
as for power creep... man we're 1000 pokemon in lol. This ship has long since sailed, power creep is always going to be part of this franchise. Just ban pokemon
I think power creep is more a thing in singles while in VGC it's a lot slower.
For a long time I've been of the opinion that in VGC, everything is inherently more balanced (Pick 4 meaning that you can create more lopsided MUs against broken things and vice versa + Double targeting, Protect, Fake Out, other disruption) and on top of that, a lot of mons still find more niches despite being very old just for VGC.
Ofc it also has a lot of new viable mons, but all the time you see old mons rise up to cool new niches, usually because of these combinations of traits.
Gen 9 is probably the most power crept it's been since Gen 5, with many of the best Pokemon being from Gen 9, because a lot of them are built around VGC design wise, but even then there have been interesting shifts like Magmar and Electabuzz as support options, Amoonguss continues as always, Talonflame has an niche, Articuno was found to have a cool use with modern Snow, etc. etc. etc.
For a long time I've been of the opinion that in VGC, everything is inherently more balanced (Pick 4 meaning that you can create more lopsided MUs against broken things and vice versa + Double targeting, Protect, Fake Out, other disruption) and on top of that, a lot of mons still find more niches despite being very old just for VGC.
Ofc it also has a lot of new viable mons, but all the time you see old mons rise up to cool new niches, usually because of these combinations of traits.
Gen 9 is probably the most power crept it's been since Gen 5, with many of the best Pokemon being from Gen 9, because a lot of them are built around VGC design wise, but even then there have been interesting shifts like Magmar and Electabuzz as support options, Amoonguss continues as always, Talonflame has an niche, Articuno was found to have a cool use with modern Snow, etc. etc. etc.
I'll be honest, I don't think you read my post that closely.
I'm not sure where you got the dead weight idea from. No Pokemon needs > 15 viable move options to be usable in game. I preserved basically everything good about in-game Gyarados; the omitted moves are usually too inefficient to gain or use, so they're mostly not much help in-game. I briefly expressed measured sympathy for Gen 1 design, sure, but that doesn't extend to making Pokemon dead weight.
Nowhere did I suggest stopping Gyarados from violently using its body to destroy stuff. That's why I gave it a bevy of thrasing, aggressive contact moves. Double-Edge, Outrage, Flail, and Aqua Tail are all about that, and maybe Crunch too. I don't interpret Ice Fang and Scale Shot as following this theme particularly closely, at least from a Gyarados lens. I'm a bit peeved with this critique, because I explicitly said Gyarados was a challenge because it has a lot going on, and then I put more moves about physical violence than about beams. Making Pokemon shallow by removing important material is the opposite of my goal; I want to remove unimportant material so the important stuff can shine, free of muddying and distractions.
About power creep, I don't agree with you at all. We ban Pokemon all the time, but people still think the modern metas are worse than Gen 6/7, which themselves are thought worse than Gen 4 (and maybe Gen 3). Even if the default trajectory is increasing power creep, we can remove existing creep to improve the meta, and put in measures to slow future creep and preserve a good power level for longer.
Obligatory note that I claim no knowledge on anything VGC / double battle related, just that I believe GF has a ***limited*** responsibility to balance for ***certain dimensions*** of singles play.
this is what i dont get. There's nothing distracting about gyarados moveset? I don't think someone realizing gyarados leans thunder, taunt, scale shot etc is forgetting gyarados is a mon that thrashes and shoots beams. Especially because these moves are move relearner or tms, which already signal them as side moves. If anything, what might delay people from remembering thats the flavor of gyarados is that it takes an extra 20+ levels for it to get hyper beam.
and while there are ways to stop power creep, i think theres more effective ways to do that without removing moves from pokemon, that would go a longer way really. like not making every new pokemon min maxed LMAO.
then again I will always prioritize flavor and pokemon design over the competitive, so thats a priority thing
If you want to bring another mon as an example feel free, but gyarados is the only one you brought so its the only one i can use to explain my issues
Uhhh
It can't learn Flamethrower Gen 1. That was Gen 2. It can learn Fire Blast from TM though
Heck, it doesn't learn Waterfall in Gen 1 at all, has to be transfered from Gen 2. That was originally exclusive to Goldeen line, despite them not being found near waterfalls at all cuz they're sea mons. They made it into an HM field move that required more mons to learn it for overworld. And this all ignores that Gyarados still was suffering from no STAB off its attack till Gen 4
Cardass already showed that having a signature move doesn't mean it isn't ass, nor not generic. I don't think GF ever cared too much of it actually mattering until Gen 7, same for abilities
I think that this mentality doesn't work in modern generations due to the way the games have progressed in design. More than Power Creep, Scope Creep is the real killer here - because the games have made a very conscious effort to "yes, and" existing game elements and balancing decisions, (and I've seen enough people erroneously complain about Game Freak removing Toxic from Gen *9* movepools to understand this call) there isn't really room to walk a lot of things back besides blazing a distinct trail forward.
There's also the fact that, in practice, competitive movesets often do boil down to a pool of 20 moves or less.
In most situations, Pokemon are going to stick to using their 10 best moves, upwards of 15 at most, and even that will often contain universal tools like Protect, Substitute, Return / Hidden Power / Tera Blast, etc. Limiting the set of tools to a total of 20 is a way to force pokemon into a specific line of play which can be more interesting, but even ignoring the impossibility due to backlash, it isn't necessarily a way to guarantee a stronger design.
There's also the way you have your list set up, which I think doesn't help. You picked a really tricky case to try with out and I don't think it worked.
Starting out, you have multiple examples of Redundancy - Waterfall and Aqua Tail, Fly and Bounce, to a lesser extent Surf and Hydro Pump - how granular you want to be in deciding what counts as overlap is up to you, but these I'd say at a minimum could be condensed.
There's also the moves that don't exist. Splash is the easy target, but I'd also say all of the Special moves aside from maybe Fire Blast and Hurricane do not contribute anything to this kit over 99% of the time. The overwhelming difference in attacking stats and Dragon Dance on top simply concentrates things on the physical side too much - even in a setting devoid of EVs or optimal natures, it's impossible to ignore the difference outside of exactly Ferrothorn.
With those factors considered, your 20 move list is functionally cut in half out the gate. The types of movesets it can create is also pretty limited - Dragon Dance is usually very centralizing on Gyarados anyway, but with only Roar for utility, it's even more the case here. The culmination of this process, then, is basically just Gen 4 Gyarados without the support sets. It doesn't work as well as you'd like.
Meanwhile, a movelist that replaces the aforementioned moves with some of Gen 9 Gyarados's best options might look something like this:
And that's... Basically Gyarados's movepool in Gen 9. Which isn't surprising, except that Gyarados really only just got over the hump of having 20 good, non-generic, non-redundant moves as of Gen 9. And it's pretty easy to argue the relevance of moves like Dive, Bounce / Fly, Scale Shot, Icy Wind, etc on such a list, but they're definitely more likely to see actual use compared to Surf.
A relatively small amount of effective moves applying to Pokemon could work well in some limited context - for the RPG gameplay, for the strategy game ROM Hack stuff, or for a modded gamemode that redesigns around that focus, I could see it being really fun, but I'm out of brain power to elaborate any more than that.
Edit: Reading your later post, I guess you were more on about a single 20 move moveset that applies to both the RPG gameplay and competitive / PVP gameplay. Can't agree at all there, part of the reason movepools ballooned in size so much was trying to account for handling both of those sets of players at once.
Last edited:
I feel like trying to limit movepools to the core flavour would run into the same issue D&D-equivalents often have with wizards. It "makes sense" for something like a dragon to have a wide movepool: they're almost always depicted as having powerful natural weapons of both the physical and energy variety, and semi-frequently as talented sorcerers on top of that. So even in a more narrow focus, Dragonite would probably be keeping on just fine with its Extremespeed, Dragon Dance, and special coverage diverse enough to make you think you can get away with running three of them. In contrast, a mon like Vivillon is kinda just some animal, so even if you keep the powder gimmick its signature move highlights, the likes of Psychic, Draining Kiss, Weather Ball, Light Screen, and Hurricane still stand out as "unrealistic" for it to have. The frequent end result is an obvious lack of balance between options because one group can interact with the game in a larger variety of ways than another. And, unlike the traditional wizard setup, the more mundane counterpart isn't guaranteed to have better base stats before the caster starts setting up (it's the opposite with my current examples).
As primarily a monotype player, I would also be concerned about homogenization within a type. Compared to the average mon, Nidoking is a physically imposing mon with venomous spines. Compared to the average Poison-type, Nidoking is a desperately needed source of special coverage. A movepool update that drops Ice Beam but keeps Poison Jab would go against an important aspect of the mon's identity even if it looks fine when zoomed out.
As primarily a monotype player, I would also be concerned about homogenization within a type. Compared to the average mon, Nidoking is a physically imposing mon with venomous spines. Compared to the average Poison-type, Nidoking is a desperately needed source of special coverage. A movepool update that drops Ice Beam but keeps Poison Jab would go against an important aspect of the mon's identity even if it looks fine when zoomed out.
I have to say this, Pokemon NOT being like the typical RPG where team members are limited, and can be permanently dropped for something better due to move complexity and initially field moves is why we can't just mindlessly copy paste RPG standards
Gen 1 tried to be closer to DQ3, but the types being done last minute and enormous roster severely shifted dev. It's why we had a lot of redundancies for similarly typed mons that are awkwardly still hard to get, and map design that is honestly extremely boring visually, which FRLG and LGPE exposes. And no rose tinting of "but they were referencing Japan's urbanization" will fix that, it was beyond undercooked and lacked cohesion for some mons, and the 1996 Pokedex book by staff and Johto's version of Kanto confirms there's no ulterior theme
Gen 2 dev the SW97 map was scrapped largely cuz it fell into the same trap; they were treating it more like a generic RPG, which made game balance unfeasible for the planned new mons. Even final has major issues with it that led to overly specific teams in casual playthroughs, but SW97 was worse that front
3-4 still has traces of attrocious coverage for some early route mons (ft. Nuzleaf), but map design drastically improved cuz they stuck to way better mon leveraging
I can't in good faith say the hyper coverage late gens is entirely badbar Grass Knott Infernape, especially when a lot of it is post game TM/TR. People were pissed at the Move Dexit largely cuz it removed so many options in pvp and some in game challenges. I enjoy watching WeedleTwineedle vids cuz of some niche as fuck strat for Doubles/VGC relying on uncommon moves like Round, or Stockpile, or Sparkling Aria. Moves that are mostly useless in Singles!
Gen 1 tried to be closer to DQ3, but the types being done last minute and enormous roster severely shifted dev. It's why we had a lot of redundancies for similarly typed mons that are awkwardly still hard to get, and map design that is honestly extremely boring visually, which FRLG and LGPE exposes. And no rose tinting of "but they were referencing Japan's urbanization" will fix that, it was beyond undercooked and lacked cohesion for some mons, and the 1996 Pokedex book by staff and Johto's version of Kanto confirms there's no ulterior theme
Gen 2 dev the SW97 map was scrapped largely cuz it fell into the same trap; they were treating it more like a generic RPG, which made game balance unfeasible for the planned new mons. Even final has major issues with it that led to overly specific teams in casual playthroughs, but SW97 was worse that front
3-4 still has traces of attrocious coverage for some early route mons (ft. Nuzleaf), but map design drastically improved cuz they stuck to way better mon leveraging
I can't in good faith say the hyper coverage late gens is entirely bad
Not to beat on a dead horse here but man, I really dislike the SV models. And I don't mean in the "graphics bad" kinda way, I just feel like no thought was put in as to how to represent the Pokémon at all.
The problem really comes down to the designs having a strong cartoon appeal, but the art direction in SV doesn't seem to be concerned about keeping that appeal and only slapping a realistic material on top of them. Torkoal and Weezing have rigid clouds that look like they're made of play-doh, and Quaxly has a realistic sheen to its hair despite also having a stylized shine drawn in the design. Instead of making the Pokémon look like living, breathing creatures, it just ends up looking like one of those Minecraft HD texture mods.
Now compare this to Legends Arceus. Pokémon like Chimchar and Gastly have particle effects added to them, so they feel like they're being affected by the environment. And more importantly, that game actually has dynamic lighting, so the models feel like they belong in the environment, unlike SV and its tacky highlights . Even New Pokémon Snap was relatively subtle when adding details to the Pokémon textures, relying a lot more in, again, strong lighting to make the Pokémon look good.
This is probably the meanest post I ever made here but I've always loved the art in Pokémon games, but that's an aspect of SV that I feel was completely neglected, even before the performance issues. Ugh
The problem really comes down to the designs having a strong cartoon appeal, but the art direction in SV doesn't seem to be concerned about keeping that appeal and only slapping a realistic material on top of them. Torkoal and Weezing have rigid clouds that look like they're made of play-doh, and Quaxly has a realistic sheen to its hair despite also having a stylized shine drawn in the design. Instead of making the Pokémon look like living, breathing creatures, it just ends up looking like one of those Minecraft HD texture mods.
Now compare this to Legends Arceus. Pokémon like Chimchar and Gastly have particle effects added to them, so they feel like they're being affected by the environment. And more importantly, that game actually has dynamic lighting, so the models feel like they belong in the environment, unlike SV and its tacky highlights . Even New Pokémon Snap was relatively subtle when adding details to the Pokémon textures, relying a lot more in, again, strong lighting to make the Pokémon look good.
This is probably the meanest post I ever made here but I've always loved the art in Pokémon games, but that's an aspect of SV that I feel was completely neglected, even before the performance issues. Ugh
No actual Pokemon RPG on Switch could look like New Pokemon Snap or anything close to it, just to be clear
Oh yeah I'm only mentioned Snap because it goes to a similar aesthetic, I'm not using it as my standard
pokemon snap does have the benefit of being a cutscene you can look around and interact very vaguely.
honestly im mixed on the sv textures, the metal ones are completely fine, even cartoons can have pretty metal rendering, but the fur and scale ones annoy me. pokemon have very cartoony fur spikes and shapes, putting realistic fur on top of that makes it look like a toy, as said above. I feel like pokemon would look much better with stylistic rendering, 2d shaders, etc
honestly im mixed on the sv textures, the metal ones are completely fine, even cartoons can have pretty metal rendering, but the fur and scale ones annoy me. pokemon have very cartoony fur spikes and shapes, putting realistic fur on top of that makes it look like a toy, as said above. I feel like pokemon would look much better with stylistic rendering, 2d shaders, etc
My personal opinion is I think SV does a good job with the Pokemon models themselves, I like some of the texturing, don't like some o the other texturing.
I hope the extra time for Gen 10 lets them cook an aesthetic more because it's also hard for me to judge the Pokemon because the game's visuals in general feel underbaked. Does Pawmot's texturing look weird, or is it that it's sitting in an underbaked weird environment, kinda deal.
I'm a weirdo who liked how SWSH looked. I thought its battles were very pretty with the HD models. It was the empty environments that made the game feel empty to me, which does effect how I see the aesthetic but in general I thought people saying the game looked bad was strange.
I hope the extra time for Gen 10 lets them cook an aesthetic more because it's also hard for me to judge the Pokemon because the game's visuals in general feel underbaked. Does Pawmot's texturing look weird, or is it that it's sitting in an underbaked weird environment, kinda deal.
I'm a weirdo who liked how SWSH looked. I thought its battles were very pretty with the HD models. It was the empty environments that made the game feel empty to me, which does effect how I see the aesthetic but in general I thought people saying the game looked bad was strange.
ME TOO yeah the trees were rough or whatever but 1. models are super cute, love the lineless look, 2. the background for battles looked really nice!!
swsh was gorgeous imo, the wild area was underbaked but places like ballonlea make me crazy. I almost wish they just left to figure out open world on sv itself so they had more time to make bigger cities and routes to explore tbh
To me SV and SWSH are two coins of unfinished:
SWSH was much more polished but the content wasn't there, while SV had content (one of the longest campaigns in a mainline game, if not the longest) but unpolished as hell.
It does also help that SWSH is just also an easier game to develop too, I imagine lol
I know a lot of people see SV and hope for a Return to Monke but honestly I love the new dev team's ambition. People used to say every Pokemon game is the same and that is 1000% not true anymore
Funny enough SV's lack of art focus can be seen on human models. Despite *trying* to do more standard lighting, human faces are forcibly flat for shading due to tweaking normals. This is a technique modern "anime" games do to have extremely basic shading
Granted I hate how 99% of the time they make the face itself flat even in anime games, but for Pokemon it sticks out worse cuz...why are you even doing this? It looks like a weird gradient cuz of the lighting
I also think doing 2D tone shaders won't solve anything. LA tried...and it looks ugly still, despite better particles some mons
The other issue, a lot of times games with 2D shaders really limit dynamic lighting. I see this issue in Dragon Ball Sparking Zero, despite it being a 3D arena ish fighter. Auras don't actually light chars most of the time, nor do they like having shadows from the environment or other objects impact chars. It's very baked on samey lighting that for a sprawling world WILL feel generic once you notice
BotW meanwhile, despite over abuse of post processing like Bloom, tries to be dynamic. Though ToK does some weird things like some particles very much not looking two tone shaded (like the smoke in the Goron area), looks like clashing styles. Still, an actually dynamic cel shaded game that respects two toning is rare
Oh man don't get me STARTED on the human models. The way Geeta looks in game is an actual insult to take's art.
Tbh I wouldn't feel so strongly about this if were not by every nerd going like "oh Pokémon have scales now! They figured out materials lol" when the game came out
Tbh I wouldn't feel so strongly about this if were not by every nerd going like "oh Pokémon have scales now! They figured out materials lol" when the game came out
Could you explain the joke here to me? The Capcom series I interact the most with by far is consistently cartoony.
Devil May Cry, Resident Evil, and Monster Hunter...
Like next to Square Enix, they have stylized, aesthetic, or even grotesque "realism" down pat. They've shattered the wrong notion that realism means not stylized or lacking expression
Like Peacock colored swamp monster toad platypuses? Heck yeah
Regardless, Pokemon won't want to do this cuz it frankly isn't cute or marketable to casual audiences for the most part. Detective Pikachu was hit miss in attempting this, but people really want mons to look like toys. There are a lot of Pokemon fans that aren't aware Lugia is canonically feathered cuz the 2D drawings and current models are very flat even with SV detailing, so...
All I want for GF to actually get consistently good for...any style really. SwSh is jank for most areas, SV is very grey and muddy and really bad for light exposure, and LA is Unity 2013 failed attempt of celshading
New Snap and Pokken look good, LGPE I'm not a fan of the style admittedly but it's competent and polished. They have examples of what they can do before diving into more experimental styles