Open World Role Playing Games
A Rant About Ever-Increasing Scale
2015 saw the release of The Witcher 3, an undeniably great game (not without flaws) whose effects still ripple through the genre today. It had a huge open world for the time that was just utterly stuffed with meaningful optional content - a compliment many have regarding CD Projekt RED is that they often have side questlines that can match or even exceed the main story in terms of quality.
Since then, countless others have begun scaling up the worlds in their games - see Bethesda with Starfield, or Ubisoft with the newer Assassin's Creed formula. However, many of these games lack the meaningful side content that CDPR had in The Witcher 3, and again in Cyberpunk 2077.
The problem of the ever expanding open worlds has reached a stage of community exhaustion where people often describe "lifeless open worlds that are too big for their own good" as "Ubisoft open world". Though Bethesda's Starfield suffers from a similar, albeit still very different problem - procedural generation.
Procedural generation or "proc gen" as it is often called is not a bad thing in and of itself, it can provide theoretically infinite substance to a world with otherwise limited content. However its real problem is when a developer relies too heavily on it, Starfield has approximately 1,700 planets (and moons) that can be visited - of which it is estimated that only ~10% have life generally. You can land pretty much anywhere on a planet but unless it's a city or quest specific area, it will be procedurally generated, which causes the content to become stale very quickly if the player wants to go sightseeing between quests.
A game that (I think) does procedural generation well is Shadows of Doubt, an indie 1980s-cyberpunk detective immersive sim. It relies on procedural generation to form the murder cases or side jobs which the player can take on without it feeling too stale or overwhelming. The reason why it works despite relying so much on it is because the entire gameplay loop is designed around it - finding the evidence, murder weapon and arresting the killer feels so exhilarating every time because of the fact that it all forms at the whims of the game. The murder is always solvable if the player can put in the effort to follow all possible leads. Rather than using procedural generation for everything else in the world, Shadows of Doubt uses it where it can shine brightest and developed the gameplay loop to perfectly compliment it.
With each paragraph getting significantly longer than the last, as I let myself get further off-topic to explain the individual problems and the counterpart games that solve them, let's get back to the original idea - What is the problem with Open World RPGs and what can solve this.
Open world RPGs right now suffer from A) ever increasing world scale and a lack of content density to fill it, and B) reliance on procedural generation in these ever expanding worlds to fill out the side content with quantity rather than quality.
The solution to this would be for studios to stop for a minute and take a breath, bring the scaling back down a little, we don't need maps that encompass half of Britain or a thousand soulless and empty planets to "explore". The open world doesn't need to be gargantuan, it specifically shouldn't be that big unless there is enough content to make it truly enjoyable to roam. Scaling down the worlds to a reasonable size would also allow for procedural generation to compliment the world, in the form of radiant quests which nowadays can have astonishing amounts of content and diversity in comparison to the radiant quests in 2011's Skyrim.
I'm no industry expert but having played each of the listed games to death, I would like to think that I know what I'm talking about regarding the most prevalent problem in the genre.
Games Listed
Honourable Mentions (other open world RPGs off the top of my head that don't suffer from the overarching problem)
Baldur's Gate 3
Disco Elysium
