One thing I've been quite confused by is how it seems that there definitely seems to be a strange bias I would argue towards fancy new classes and units when they get introduced in patches.
That's just power creep. Its a problem in basically any form of game with a long history.
And in some ways, its almost a requirement. People get set in their ways and what was once cool becomes boring. "Heroic Strike adds 919 damage per swing" was great 4 expansions ago, but you can't just add "Anti-Heroic Strike" which does 191 damage per miss. You need to add a move that pulls and roots and damages the enemy to really add any zest to the class.
And more importantly, you get people to try something. To use Starcraft as an example, if the newest patch did make something OP then the meta wouldn't change and nobody would actually bother using it. And then counterplays wouldn't develop and the scene would remain so unchanged it was basically worthless to have done anything at all.
Shit, its even in card games. In Yugioh a 1900 attack 4 star monster was basically the GOAT and a game winner, no effects needed. If they didn't introduce OP shit with 3 effects and 2800 attack you could summon from the deck that made 1900 cute to try, no one would ever remove Gemini Elf from their deck. It would become a card you were required to own to even remain competitive and then none of their new packs would be as in demand.
Its a balancing act to constantly do it without completely invalidating what came before it. Gentle push and pull. Most companies are not capable of that and instead just triple down on the creep until the vanilla content is so badly outclassed they need to "rework" it entirely to even keep up.
In Yugioh a 1900 attack 4 star monster was basically the GOAT and a game winner
I was like 5 years old when I started playing that, then I stopped for 20 years, and came back to it in master duel.
Holy fuck it is impossible to win now unless you are playing meta. Literally everything revolves around special summons and shutting down the opponents field with ash blossom, etc.
There is a reason why the biggest Yugioh game of this decade was Duel Links. Because it started with the original sets and let you play with your old favorites again in a newish mode.
That's also the reason why the game dropped players hard after the GX era, because they started reaching the "more current" metas that everyone hates.
Personally, I just play Legacy of the Duelist if I don't use my physical decks and stick with the eras I enjoy. It gives plenty of fun even with only like 40% of the game I engage with and me and the boys can have that nostalgic fun.
Back when I played Dota, this was accepted as usual routine. New heroes are overtuned so that people would be incentivised to play them. Having more people play them means more data. More data means the heroes can be tuned more quickly before being allowed in competitive mode.
I think that's why roguelight games are so popular. You need to strategize the weapons you get, and build up based on the randomness. The base characters are all playable, but the strategy at building them up is important.
I guess the game creators either go power creep or strategy creep. Either the high school kids end up owning the world and destroying God, or they strategically defeat the somewhat godlike boss at the end of the game.
Real life seems to have power creep. My Toyota Camry is way more powerful than a 60's Mustang. Back in the day 250 horse power was more than enough. Now we have cars doing 2,800. 5 Gigs on a computer would fill a room. Now it would be a dead potato if you tried modern gaming with it. Even mini switchers which have been around forever have gone from the compiler getting rid of notes because that is too much for them to our modern era with various fruit named devices that can easily emulate an early 00's computer. The switch has upgraded ports from all the previous consoles, and it's a handheld. Because of protein intake being easier in our modern times cultures around the world are taller. I've seen the remains of battles from centuries ago and it looked like children had swords.
I looked into Path of Exile earlier this year and while I didn't really play it more than a few minutes, a lot of opinions about early game seem to match that. Meaning the end game was good, but for a new player getting there was broken. I know several here are into the game so bear in mind to you all this is opinions I read and not personal experience. I didn't end up playing it because it started to look a bit too MMO to me. Not necessarily in the gameplay but the way it will suck you in and become an addiction. I swore off WoW in 2011 for that reason. It wasn't healthy and I stay away from going down those roads again.
I have almost no doubt they make new characters strong to sell them. They need you losing to it so you will want to trade your money for it too. I used to speculate that with things as simple as weapons in an FPS game even, it's just harder to prove there.
Meaning the end game was good, but for a new player getting there was broken
The problem with a lot of long lasting games, especially seasonal ones, is that they stop expecting "new players" at some point. So the game just becomes built around people who already know how to cheese and skip the entire story mode and just instantly start at end game. Often times with the game itself providing the means to do so.
So for the pure brand new players, its a slog where you are basically expected to get someone to Run you to the end quickly to get to the good parts. Which really isn't any different now than it was back when Diablo 2 came out and set the stage for games like POE.
Well, when I played poe (around 8 months ago and prior), the story was basically what most of us just pushed through to enjoy the end game, because we'd seen it a lot. And you don't really get a lot of abilities to explore the builds with enough skill points or gear until mapping.
Power creep is a problem, but the main issue with PoE was they nerfed fun by 99.999% because of the way they handled loot and difficulty.
Using several past WoW expansions as an example, and probably the current one, the concept of OP "Ooo, shiny!" to market new features has been done in Wrath of the Lich King, Mists of Pandaria, Legion, and IIRC in Dragonflight where new classes were added in to the game.
Wrath was the first expansion to add in a new class, the Death Knight. It had 3 specs: Blood, Frost, and Unholy, and originally all 3 could Tank or Dps depending what abilities were picked because the talent trees weren't nerfed to shit back then. While nowadays Blood is the Tank spec in Wrath the best pick for a DK was 2h Frost tanking, something they don't get to do now because Frost is now the dual wield spec while Unholy is the 2h dps spec. DK tanks were next to impossible to kill. If you died it was probably because you fucked up something special, like falling off a mountain. Even later on a well played Blood DK could out last his entire raid group if done properly however that starts to require actual player skill and the issues here are more to do with the inherent class designs.
Then came Monks in MoP who were literally broken as fuck. The healing spec could outperfom dps in damage and was colloquially referred to as "Fistweaving" rather than 'Mistweaving' as it was actually called due to using water themes.
Legion added in Demonhunters which could both Tank and Dps. So they were amazing at both.
In these cases subsequent expansions toned/nerfed these classes down but in their original content they were that OP literal idiots could play them and still do exceedingly well. Much like Hunters and Warlocks in TBC who were so badly designed single button pressing was all you needed so groups filled with any idiot who decided to roll one 🙄 the term "huntard" existed for a reason.
The current expansion, Dragonflight, added in another new class, Dracthyr, who could either dps or heal. Because they are the new kids on the block they are both front and centre at ongoing events as well as class performance because the devs want people to play with the new toy.
Whatever happens next you can be sure the Dracthyr will be toned down to some degree to fit more in line with how other classes perform and if another class comes out eventually you can expect the pattern to repeat itself where the new shiny will be awesome for the duration of its expansion before being firmly resized and stuffed into a box of the devs choosing.
I bring this up because in RPGs I actually enjoy the classic fighter/mage/rogue archetypes. However it seems like more and more with 'modern RPGs' you almost get punished for picking them depending on the situation.
Can you give some examples where that happens, and what you mean by getting punished? Honestly, I can't recall any cRPGs that I've played where fighter/cleric/rogue/wizard (or equivalent, depending on the game) was not perfectly viable for completing the game - heck, Fighty McFighterson is often the first companion you run into in most cRPGs, with Stabby McSneaky and Whitebeard the Wizard coming 2 minutes later.
That said, though, I do admit that combo may not (is probably not) optimal, but there's two things going into that. The first is that, even if it can be achieved, perfect balance is not a state to strive for. Yes, it is perfectly balanced if the fighter's heroic blow", the paladin's "holy strike" and the barbarian's "mighty slam", all do the exact same thing but it's also boring and redundant. If two classes are functionally equivalent in what they can do, then they are the same class and one should be removed. But, that's really a state we don't need to worry about because there is always going to be some imbalance in any reasonably complex system - and that's why fighter/cleric/rogue/wizard is probably not optimal. Take something like Owlcat's Wrath of the Righteous - you've got 26 base classes (each with a half dozen subclasses) and 13 prestige classes without even getting into the epic level paths or mixing and matching with multiclassing. Optimal likely doesn't even exist in that case because changing the class of one party member will add and remove capabilities making the other classes more and less effective.
MMOs are an entirely different monster, but the good news there is you probably only need to wait a month or two for the next balance pass and the "meta" will change again. That said, though, that is a huge problem with modern gaming which has nothing to do with the developers or even the games. Take any game, give it a week, and some turbonerd on the internet will have done the math to figure out that Class X does 2.3 more damage per second than Class Y, in optimal conditions, and will make a video on it. Then, people will start sharing the video all over the internet. And 2 days later, many people will say "LOL! You're such a noob! Why are you Class Y when Class X is objectively superior? Get out of here, we don't want you and your vastly subpar DPS in our group!" even though Class Y is still more than capable of clearing all the content without difficulty. If anyone figures out a way to actually solve that problem, I will happily support their efforts to become the Supreme Ruler of Mankind in any way I can.
I'm a couple of days late checking this thread but will give some thoughts on the topic in general as well as address some of the specific examples brought up.
I remember how powerful an extra fighter or two could be in Baldur's Gate 2
BG2 didn't function like a traditional D&D game/party. Not only if the protagonist literally powered by the Divine but almost every single companion has some unique twist that makes them stand out from their actual class and subclass.
Minsc is more than just a Ranger, if he even is one because he functions more like a Fighter. His internal party interactions can lead him to adopt Aerie as his new witch after the death of Dynaheir between BG 1 and 2 which will lead to moments where Minsc can enter a berserker rage, despite not being an actual Berserker, if Aerie gets hurt too much. There is also Boo, who turns out to be exactly what it says on the tin. A miniature giant space hamster, and not just a regular hamster some brain adled strongman has been convinced otherwise. Most Rangers don't get played like this, Minsc functions better as a two handed tank/dps than the animal attuned archer/"ranger" stereotype or dual wielding swords type and apart from the protagonist, Keldorn, Sarevok, and a couple of others, Minsc potentially one of the single best melee characters you can get. As a Ranger.
Speaking of Sarevok, he's quite possibly the best melee character in the game again because of unique traits which also require several significant decisions to be made in order to make use of them.
First, you need to resurrect him in Throne of Bhaal and recruit him. Second, you need to give him his sword back, a sword you originally get all the way back in the very first dungeon in BG2, but only if you do the side quest for the Djinn and return its lamp from Irenicus' chambers. Then you have to hold on to the sword for the entirety of BG2, although others can use it in the meantime, because it's not until ToB where giving it back to Sarevok will let him return some of its lost power and upgrade it from a +2 2h sword to a +4 that only Sarevok can then wield. By this point in the trilogy you should be able to take epic levels which means whirlwind and greater whirlwind in the case of Sarevok, and Minsc. Add on to that Sarevok's Deathwalker ability where he can literally one shot almost anything and you can see why he's the best melee character in the game outside of the protagonist who only beats him due to both plot and broken class choices.
And those broken class choices are a multiclassed Kensai/Mage that gain martial abilities from the Fighter subclass in addition to armour bonuses that don't need actual plate, the aforementioned WW and GWW epic feats, and access to various Mage spells to further increase your abilities as well as simply cause damage should you so wish. Or even Wish. Simulacrum, Contingency, and others like Haste, Mage Armour, let you further buff yourself to absurd levels because the game design accidentally permits it, but then you're meant to be a demigod according to plot so acting like one this way isn't really that out of character.
It's the same error Dragon Age Origins made with the Mage subclass Spirit Warrior that permits Mages to use their INT stat to work for both melee weapons and wearing armour when it would otherwise require STR. You can then again couple martial and magical abilities for a synergy that outperforms everything else in the game.
Similar to being powered by the Divine in Baldur's Gate, Mages in Dragon Age really are that terrifyingly powerful so again it's still more or less behaving in line with how the class can be for plot reasons.
I have some MMO examples to go over as well which I'll do in a separate comment.
The first is your nostalgia goggles. You are most likely far better at games now, then you were when you were 10. This will lead to the perception that builds you do now are more powerful than builds back then.
The second is op interactions or combos. Suppose that in a given game there are certain percent of op interactions, say 10%, or 5%, 1%. The specific number doesn't really matter. As a game's lifetime goes on, the number of possible interactions for two effects increases as the number of effects squared. Similarly the number of possible interactions for three effects goes as the number of interactions cubed. Thus if there are only the three archetypes with say five abilities each then there are 225 two ability interactions and 3375 three ability interactions. Now when we increase the number of classes to 10, there are then 22500 two ability interactions, and 3.3 million three ability interactions.
This leads to a situation where you will more rarely get a setup with an op interaction using just the three archetypes, compared to when you mix in one or more of the more exotic classes.
This theory would be more easily seen in tabletop wargames and card games. Think warhammer (40k and sigmar), mtg, yugioh. Where they really want to drive new sales. Not that your RPG ones don't, but the incentives to purchase in the physical games are higher.
But part of it is naturally just powercreep too. There's a natural power creep that comes simply from having additional options. New metas and strategies become available even if the new release is entirely balanced, thanks to there being increased options. That's where looking at card games vs warhammer is useful. There's not much crossover, releasing a new set of space dwarfs doesn't give me any new.
Also, the first release is always imbalanced. As the game matures and the metas settle, it's easier to develop something that's 'solid B tier', rather than the range of F to S+ you initially released. Shiny new things are always going to be at least decent, to drive sales, and because it's now easier to pitch things at that level with expierience.
Yes there's bias to drive sales, you'd be stupid not to. But there's also just the fact that new options create new metas, and new stuff can be more easily pitched at a solid B+ at least.
Also you need to shuffle things to keep it from being stagnant. There should be a natural rise and fall of different armies or builds or characters. Opens up new strategies and interesting interactions.
Almost every incentive is there, even if profit wasn't there, to release stuff at a good B+/A, better than a lot of the OG stuff. When doing so also drives sales, why wouldn't you?
That's just power creep. Its a problem in basically any form of game with a long history.
And in some ways, its almost a requirement. People get set in their ways and what was once cool becomes boring. "Heroic Strike adds 919 damage per swing" was great 4 expansions ago, but you can't just add "Anti-Heroic Strike" which does 191 damage per miss. You need to add a move that pulls and roots and damages the enemy to really add any zest to the class.
And more importantly, you get people to try something. To use Starcraft as an example, if the newest patch did make something OP then the meta wouldn't change and nobody would actually bother using it. And then counterplays wouldn't develop and the scene would remain so unchanged it was basically worthless to have done anything at all.
Shit, its even in card games. In Yugioh a 1900 attack 4 star monster was basically the GOAT and a game winner, no effects needed. If they didn't introduce OP shit with 3 effects and 2800 attack you could summon from the deck that made 1900 cute to try, no one would ever remove Gemini Elf from their deck. It would become a card you were required to own to even remain competitive and then none of their new packs would be as in demand.
Its a balancing act to constantly do it without completely invalidating what came before it. Gentle push and pull. Most companies are not capable of that and instead just triple down on the creep until the vanilla content is so badly outclassed they need to "rework" it entirely to even keep up.
I was like 5 years old when I started playing that, then I stopped for 20 years, and came back to it in master duel.
Holy fuck it is impossible to win now unless you are playing meta. Literally everything revolves around special summons and shutting down the opponents field with ash blossom, etc.
There is a reason why the biggest Yugioh game of this decade was Duel Links. Because it started with the original sets and let you play with your old favorites again in a newish mode.
That's also the reason why the game dropped players hard after the GX era, because they started reaching the "more current" metas that everyone hates.
Personally, I just play Legacy of the Duelist if I don't use my physical decks and stick with the eras I enjoy. It gives plenty of fun even with only like 40% of the game I engage with and me and the boys can have that nostalgic fun.
Back when I played Dota, this was accepted as usual routine. New heroes are overtuned so that people would be incentivised to play them. Having more people play them means more data. More data means the heroes can be tuned more quickly before being allowed in competitive mode.
I think that's why roguelight games are so popular. You need to strategize the weapons you get, and build up based on the randomness. The base characters are all playable, but the strategy at building them up is important.
I guess the game creators either go power creep or strategy creep. Either the high school kids end up owning the world and destroying God, or they strategically defeat the somewhat godlike boss at the end of the game.
Real life seems to have power creep. My Toyota Camry is way more powerful than a 60's Mustang. Back in the day 250 horse power was more than enough. Now we have cars doing 2,800. 5 Gigs on a computer would fill a room. Now it would be a dead potato if you tried modern gaming with it. Even mini switchers which have been around forever have gone from the compiler getting rid of notes because that is too much for them to our modern era with various fruit named devices that can easily emulate an early 00's computer. The switch has upgraded ports from all the previous consoles, and it's a handheld. Because of protein intake being easier in our modern times cultures around the world are taller. I've seen the remains of battles from centuries ago and it looked like children had swords.
I looked into Path of Exile earlier this year and while I didn't really play it more than a few minutes, a lot of opinions about early game seem to match that. Meaning the end game was good, but for a new player getting there was broken. I know several here are into the game so bear in mind to you all this is opinions I read and not personal experience. I didn't end up playing it because it started to look a bit too MMO to me. Not necessarily in the gameplay but the way it will suck you in and become an addiction. I swore off WoW in 2011 for that reason. It wasn't healthy and I stay away from going down those roads again.
I have almost no doubt they make new characters strong to sell them. They need you losing to it so you will want to trade your money for it too. I used to speculate that with things as simple as weapons in an FPS game even, it's just harder to prove there.
The problem with a lot of long lasting games, especially seasonal ones, is that they stop expecting "new players" at some point. So the game just becomes built around people who already know how to cheese and skip the entire story mode and just instantly start at end game. Often times with the game itself providing the means to do so.
So for the pure brand new players, its a slog where you are basically expected to get someone to Run you to the end quickly to get to the good parts. Which really isn't any different now than it was back when Diablo 2 came out and set the stage for games like POE.
Well, when I played poe (around 8 months ago and prior), the story was basically what most of us just pushed through to enjoy the end game, because we'd seen it a lot. And you don't really get a lot of abilities to explore the builds with enough skill points or gear until mapping.
Power creep is a problem, but the main issue with PoE was they nerfed fun by 99.999% because of the way they handled loot and difficulty.
My issue was the shitty drops and lack of progression in a way that made sense.
Every mechanic they had to craft items got nerfed, good loot drops got nerfed. I like grinding, but fucks sake
Using several past WoW expansions as an example, and probably the current one, the concept of OP "Ooo, shiny!" to market new features has been done in Wrath of the Lich King, Mists of Pandaria, Legion, and IIRC in Dragonflight where new classes were added in to the game.
Wrath was the first expansion to add in a new class, the Death Knight. It had 3 specs: Blood, Frost, and Unholy, and originally all 3 could Tank or Dps depending what abilities were picked because the talent trees weren't nerfed to shit back then. While nowadays Blood is the Tank spec in Wrath the best pick for a DK was 2h Frost tanking, something they don't get to do now because Frost is now the dual wield spec while Unholy is the 2h dps spec. DK tanks were next to impossible to kill. If you died it was probably because you fucked up something special, like falling off a mountain. Even later on a well played Blood DK could out last his entire raid group if done properly however that starts to require actual player skill and the issues here are more to do with the inherent class designs.
Then came Monks in MoP who were literally broken as fuck. The healing spec could outperfom dps in damage and was colloquially referred to as "Fistweaving" rather than 'Mistweaving' as it was actually called due to using water themes.
Legion added in Demonhunters which could both Tank and Dps. So they were amazing at both.
In these cases subsequent expansions toned/nerfed these classes down but in their original content they were that OP literal idiots could play them and still do exceedingly well. Much like Hunters and Warlocks in TBC who were so badly designed single button pressing was all you needed so groups filled with any idiot who decided to roll one 🙄 the term "huntard" existed for a reason.
The current expansion, Dragonflight, added in another new class, Dracthyr, who could either dps or heal. Because they are the new kids on the block they are both front and centre at ongoing events as well as class performance because the devs want people to play with the new toy.
Whatever happens next you can be sure the Dracthyr will be toned down to some degree to fit more in line with how other classes perform and if another class comes out eventually you can expect the pattern to repeat itself where the new shiny will be awesome for the duration of its expansion before being firmly resized and stuffed into a box of the devs choosing.
Can you give some examples where that happens, and what you mean by getting punished? Honestly, I can't recall any cRPGs that I've played where fighter/cleric/rogue/wizard (or equivalent, depending on the game) was not perfectly viable for completing the game - heck, Fighty McFighterson is often the first companion you run into in most cRPGs, with Stabby McSneaky and Whitebeard the Wizard coming 2 minutes later.
That said, though, I do admit that combo may not (is probably not) optimal, but there's two things going into that. The first is that, even if it can be achieved, perfect balance is not a state to strive for. Yes, it is perfectly balanced if the fighter's heroic blow", the paladin's "holy strike" and the barbarian's "mighty slam", all do the exact same thing but it's also boring and redundant. If two classes are functionally equivalent in what they can do, then they are the same class and one should be removed. But, that's really a state we don't need to worry about because there is always going to be some imbalance in any reasonably complex system - and that's why fighter/cleric/rogue/wizard is probably not optimal. Take something like Owlcat's Wrath of the Righteous - you've got 26 base classes (each with a half dozen subclasses) and 13 prestige classes without even getting into the epic level paths or mixing and matching with multiclassing. Optimal likely doesn't even exist in that case because changing the class of one party member will add and remove capabilities making the other classes more and less effective.
MMOs are an entirely different monster, but the good news there is you probably only need to wait a month or two for the next balance pass and the "meta" will change again. That said, though, that is a huge problem with modern gaming which has nothing to do with the developers or even the games. Take any game, give it a week, and some turbonerd on the internet will have done the math to figure out that Class X does 2.3 more damage per second than Class Y, in optimal conditions, and will make a video on it. Then, people will start sharing the video all over the internet. And 2 days later, many people will say "LOL! You're such a noob! Why are you Class Y when Class X is objectively superior? Get out of here, we don't want you and your vastly subpar DPS in our group!" even though Class Y is still more than capable of clearing all the content without difficulty. If anyone figures out a way to actually solve that problem, I will happily support their efforts to become the Supreme Ruler of Mankind in any way I can.
I'm a couple of days late checking this thread but will give some thoughts on the topic in general as well as address some of the specific examples brought up.
BG2 didn't function like a traditional D&D game/party. Not only if the protagonist literally powered by the Divine but almost every single companion has some unique twist that makes them stand out from their actual class and subclass.
Minsc is more than just a Ranger, if he even is one because he functions more like a Fighter. His internal party interactions can lead him to adopt Aerie as his new witch after the death of Dynaheir between BG 1 and 2 which will lead to moments where Minsc can enter a berserker rage, despite not being an actual Berserker, if Aerie gets hurt too much. There is also Boo, who turns out to be exactly what it says on the tin. A miniature giant space hamster, and not just a regular hamster some brain adled strongman has been convinced otherwise. Most Rangers don't get played like this, Minsc functions better as a two handed tank/dps than the animal attuned archer/"ranger" stereotype or dual wielding swords type and apart from the protagonist, Keldorn, Sarevok, and a couple of others, Minsc potentially one of the single best melee characters you can get. As a Ranger.
Speaking of Sarevok, he's quite possibly the best melee character in the game again because of unique traits which also require several significant decisions to be made in order to make use of them.
First, you need to resurrect him in Throne of Bhaal and recruit him. Second, you need to give him his sword back, a sword you originally get all the way back in the very first dungeon in BG2, but only if you do the side quest for the Djinn and return its lamp from Irenicus' chambers. Then you have to hold on to the sword for the entirety of BG2, although others can use it in the meantime, because it's not until ToB where giving it back to Sarevok will let him return some of its lost power and upgrade it from a +2 2h sword to a +4 that only Sarevok can then wield. By this point in the trilogy you should be able to take epic levels which means whirlwind and greater whirlwind in the case of Sarevok, and Minsc. Add on to that Sarevok's Deathwalker ability where he can literally one shot almost anything and you can see why he's the best melee character in the game outside of the protagonist who only beats him due to both plot and broken class choices.
And those broken class choices are a multiclassed Kensai/Mage that gain martial abilities from the Fighter subclass in addition to armour bonuses that don't need actual plate, the aforementioned WW and GWW epic feats, and access to various Mage spells to further increase your abilities as well as simply cause damage should you so wish. Or even Wish. Simulacrum, Contingency, and others like Haste, Mage Armour, let you further buff yourself to absurd levels because the game design accidentally permits it, but then you're meant to be a demigod according to plot so acting like one this way isn't really that out of character.
It's the same error Dragon Age Origins made with the Mage subclass Spirit Warrior that permits Mages to use their INT stat to work for both melee weapons and wearing armour when it would otherwise require STR. You can then again couple martial and magical abilities for a synergy that outperforms everything else in the game.
Similar to being powered by the Divine in Baldur's Gate, Mages in Dragon Age really are that terrifyingly powerful so again it's still more or less behaving in line with how the class can be for plot reasons.
I have some MMO examples to go over as well which I'll do in a separate comment.
There are two things you are missing here.
The first is your nostalgia goggles. You are most likely far better at games now, then you were when you were 10. This will lead to the perception that builds you do now are more powerful than builds back then.
The second is op interactions or combos. Suppose that in a given game there are certain percent of op interactions, say 10%, or 5%, 1%. The specific number doesn't really matter. As a game's lifetime goes on, the number of possible interactions for two effects increases as the number of effects squared. Similarly the number of possible interactions for three effects goes as the number of interactions cubed. Thus if there are only the three archetypes with say five abilities each then there are 225 two ability interactions and 3375 three ability interactions. Now when we increase the number of classes to 10, there are then 22500 two ability interactions, and 3.3 million three ability interactions.
This leads to a situation where you will more rarely get a setup with an op interaction using just the three archetypes, compared to when you mix in one or more of the more exotic classes.
It's just math.
This theory would be more easily seen in tabletop wargames and card games. Think warhammer (40k and sigmar), mtg, yugioh. Where they really want to drive new sales. Not that your RPG ones don't, but the incentives to purchase in the physical games are higher.
But part of it is naturally just powercreep too. There's a natural power creep that comes simply from having additional options. New metas and strategies become available even if the new release is entirely balanced, thanks to there being increased options. That's where looking at card games vs warhammer is useful. There's not much crossover, releasing a new set of space dwarfs doesn't give me any new.
Also, the first release is always imbalanced. As the game matures and the metas settle, it's easier to develop something that's 'solid B tier', rather than the range of F to S+ you initially released. Shiny new things are always going to be at least decent, to drive sales, and because it's now easier to pitch things at that level with expierience.
Yes there's bias to drive sales, you'd be stupid not to. But there's also just the fact that new options create new metas, and new stuff can be more easily pitched at a solid B+ at least.
Also you need to shuffle things to keep it from being stagnant. There should be a natural rise and fall of different armies or builds or characters. Opens up new strategies and interesting interactions.
Almost every incentive is there, even if profit wasn't there, to release stuff at a good B+/A, better than a lot of the OG stuff. When doing so also drives sales, why wouldn't you?
You'de be better off asking something like this on the Codex.