There’s a particular kind of grimace that crosses the face of a long-time retro enthusiast when they boot up a modern AAA game trumpeting its “16-bit inspiration” or “love letter to the classics.” It’s the same expression you’d pull biting into a supermarket sandwich that promised artisan sourdough. The texture is wrong. The proportions are off. The whole thing has been assembled by committee, focus-grouped into mush, and then garnished with a CRT filter that costs a fortune to license and looks like cling film stretched over a 4K panel.
2025 has been, by any honest measure, a banner year for this phenomenon. We’ve watched publishers throw eight-figure budgets at projects that fundamentally misunderstand what made their source material sing. We’ve seen “retro-inspired” UI design that requires four mouse clicks to start a new game. We’ve endured cutscenes longer than entire NES titles. And meanwhile, quietly, in the margins, a handful of indie developers â many of them working with budgets that wouldn’t cover a single AAA marketing trailer â have produced games that don’t just reference the past but extend it, argue with it, occasionally even improve upon it.
This isn’t a question of pixel count or chiptune fidelity. It’s a question of design philosophy, of understanding what constraints actually did for the games we grew up loving. And it’s a question worth interrogating now, because the gap between the corporate impersonators and the genuine torchbearers has never been wider.
The Crisis of Hollow Nostalgia
Let’s start with what we mean by “retro-inspired,” because the term has been so thoroughly debased it now refers to roughly four entirely different things. There’s the surface-level cosmetic approach â chunky pixels, FM synth, scanlines â which assumes retro is a visual filter you apply at the end of development. There’s the mechanical revival, where studios attempt to resurrect dormant subgenres (the immersive sim, the boomer shooter, the Metroidvania) with varying degrees of fluency. There’s the structural homage, which tries to capture the pacing and economy of older games. And then there’s the rarest beast: design that genuinely thinks in the idiom of an earlier era, accepting its restrictions as creative opportunities rather than obstacles to be smoothed away.
AAA studios, almost without exception, attempt only the first of these. Sometimes they gesture at the second. The third and fourth are essentially the exclusive domain of indies, and even then only a vanishingly small number of them. The reasons for this are structural, not artistic, and understanding them is the key to understanding why this keeps happening.
What the Constraints Actually Did
When Yuji Naka and his team were building the original Sonic the Hedgehog for the Mega Drive in 1990 and 1991, they weren’t trying to make a “retro” game. They were trying to make Sega’s mascot platformer feel faster than anything Nintendo had managed. The game’s level design â the famous loops, the multi-tier route branching, the way you could lose minutes of progress because you took the high path and missed the rings â was a direct response to the Motorola 68000’s capabilities, the Mega Drive’s scrolling tricks, and the cartridge size they were forced to work within.
The constraints didn’t just limit the design. They generated it. The 8×8 pixel grid forced character silhouettes that read instantly at speed. The colour palette restrictions produced the iconic look. The lack of voice acting meant story had to be conveyed through animation, environment, and the occasional terse line of text. Every element of what we now call the “16-bit aesthetic” emerged from specific technical pressures, and the games that survive as classics did so because their designers turned those pressures into virtues.
You cannot recover this by applying a shader. You cannot recover it by hiring a chiptune composer. You can only recover it by understanding why the constraints mattered and choosing to design as if they still applied â even when, technically, they don’t.
Exhibit A: The Corporate Remake Industrial Complex
Square Enix’s ongoing trilogy of Final Fantasy VII remakes is, in many ways, the apotheosis of AAA misunderstanding retro design. Rebirth, released in early 2024 and finally hitting PC in January 2025 at ÂŖ59.99, is a technically dazzling, frequently beautiful, and structurally bloated game that takes roughly 100 hours to cover material the original 1997 release dispatched in maybe 15. It is also â and this is the crucial point â a game that systematically removes the design features that made Final Fantasy VII work.
The original’s overworld map was a brilliant solution to a hardware problem: you couldn’t render the whole world at field-screen detail, so you abstracted it into a navigable diorama. This abstraction created pacing, mystery, and a sense of geographical scale that the fully realised open zones of Rebirth conspicuously lack. When everything is detailed, nothing is special. When every town is a sprawling explorable space rather than a focused set-piece, the rhythm of “town, dungeon, overworld” collapses into a mush of waypoint-following.
The Engine Tax
Part of what’s happening here is what we might call the engine tax. When you’re building on Unreal Engine 5 with a team of 500 people across multiple studios, your fixed costs are enormous. You cannot justify a 20-hour focused experience because the per-hour content cost is too high and the perceived value-for-money calculation has been ruined by a decade of bloated open-world design. So the game expands to fill the budget, regardless of whether the expansion serves the design.
Compare this to FFVII proper, which Squaresoft developed with around 120 people over roughly two years on a budget of about $40 million (huge for 1997, but considerably less than Rebirth‘s reported development costs adjusted for inflation). The original team had to make every minute count. Every screen was hand-built. Every dungeon had a purpose. The economy of expression was forced upon them, and it made the game tighter, weirder, more memorable.
Ubisoft’s Eternal Map-Filling Problem
Ubisoft has spent the last decade nominally rediscovering its arcade roots with games like the 2024 Prince of Persia: The Lost Crown and various smaller-scale experiments, and the pattern is depressingly consistent. The Lost Crown is, to be clear, a genuinely good Metroidvania â possibly the best AAA attempt at the form since the Castlevania DS trilogy. But even here, you can feel the corporate immune system at work: tutorial pop-ups that won’t shut up, an objective marker system that undermines the genre’s core pleasure of discovery, accessibility options that veer into hand-holding by default.
The team in Montpellier clearly loved the source material. You can see Hollow Knight and Symphony of the Night in every screen. But the publisher above them couldn’t quite trust the audience to figure things out, and so they bolted on the same telemetry-driven friction-reduction that has hollowed out every other Ubisoft game of the past decade. The result is a game that’s 80% there and infuriating precisely because of how close it gets.
The Death of the Sharp Edge
One of the most striking differences between genuine retro design and AAA cosplay is what we might call tolerance for sharpness. Old games were full of sharp edges â moments of difficulty, opacity, weirdness, and outright player hostility that modern design dogma treats as bugs to be sanded off.
Think about the original Resident Evil on the PS1. The tank controls. The fixed camera angles. The deliberate scarcity of ammunition. The save system that required ink ribbons. Modern design wisdom would identify every one of these as a “pain point” to be addressed. The 2002 GameCube remake kept most of them and is widely regarded as one of the greatest survival horror games ever made. The 2023 Resident Evil 4 remake, while excellent on its own terms, systematically removes the campy theatrical sharpness of the 2005 original in favour of a more grounded, more universally palatable experience. Something is gained, but something fundamental is also lost.
The Telemetry Problem
AAA development is now driven by telemetry to a degree that would have seemed dystopian even ten years ago. Every quit-out, every death, every menu hesitation is logged, aggregated, and used to inform future design decisions. The result is a relentless smoothing process: anything that causes friction in playtesting gets flagged, and anything flagged tends to get removed.
The problem is that friction isn’t always bad. Sometimes friction is the experience. The reason Dark Souls works isn’t despite its sharp edges but because of them â and notably, FromSoftware is one of the very few large studios that has managed to maintain a distinctive design voice in the face of the smoothing imperative. They’ve done it by being commercially successful enough to override their own analytics team, which is not a privilege available to most.
Exhibit B: The 2025 Indie Renaissance
Now to the good news. While the publishers have been busy strip-mining the past for marketing pull-quotes, a remarkable cohort of indie developers has spent 2025 demonstrating what genuine retro literacy actually looks like. These aren’t homages. They’re heirs.
Lunark and the New Cinematic Platformer
Canari Games’ Lunark, which finally hit consoles in updated form in spring 2025 after its initial 2023 PC release, is the most rigorous Eric Chahi tribute we’ve seen since Chahi himself stopped making cinematic platformers. The rotoscoped animation, the deliberate weight of every jump, the trial-and-error puzzle structure â it’s all there, but crucially, it’s not slavish. The team have extended the form, introducing modern checkpoint conventions without softening the underlying challenge, and the result is something that feels genuinely of a piece with Another World and Flashback rather than a nostalgic shrug at them.
At ÂŖ17.99, Lunark costs less than a third of what Rebirth demands and represents, hour for hour, a far more concentrated experience. This isn’t an argument that cheaper is better â it’s an argument that focus is better, and focus is something AAA economics actively prevent.
Mouthwashing and the PS1 Aesthetic Done Right
Wrong Organ’s Mouthwashing, technically a late 2024 release but the discourse around it carried right through 2025, is the cleanest example I can think of of why the PS1 aesthetic works when developers actually understand it. The polygon-wobble, the muddy textures, the truncated draw distance â these aren’t applied as filters but baked into the game’s spatial logic. The lo-fi rendering makes the horror work because your brain fills in what the engine can’t show you, exactly as it did when you were thirteen and playing Silent Hill at 240p on a CRT.
Contrast this with the slew of so-called “PS1 horror” games on Itch and Steam that simply slap a vertex-jitter shader on otherwise modern assets and call it a day. The aesthetic is doing nothing for the design. The constraints aren’t being internalised, just imitated.
UFO 50 and the Anti-Pastiche
Mossmouth’s UFO 50, released in September 2024 and continuing to dominate retro discourse throughout 2025, is perhaps the most important release in this entire conversation. Fifty games, each pretending to be from a fictional 1980s console called the LX, each designed as if it had to fit on a cartridge and sell in a Toys R Us. The point of the exercise isn’t nostalgia â it’s design pedagogy. Each of those fifty games demonstrates a different way that constraints generate creativity.
What makes UFO 50 work where AAA pastiche fails is that the developers (a constellation including Spelunky‘s Derek Yu, Downwell‘s Ojiro Fumoto, and others) actually understand the era they’re imitating because they grew up making games for it, or in its immediate shadow. The historical literacy is total. When one of the fifty games references the conventions of, say, late-period MSX puzzle platformers, it does so with the precision of someone who has actually played those games and thought hard about why they worked.
Animal Well’s Long Tail
Billy Basso’s Animal Well, released in May 2024 on PS5, PC, and Switch for ÂŖ19.99, continued to generate community activity throughout 2025 as players excavated its deeper secrets. This is another genuine heir rather than a tribute act: a Metroidvania that thinks in the idiom of the form’s most experimental moments (the unmappable spaces of Knytt, the parser-puzzle weirdness of mid-80s home computer games) while leveraging modern tech for the things modern tech is actually good at, like the gorgeous lighting model.
Critically, Animal Well trusts its player. There’s no tutorial. There’s no objective marker. There’s no NPC explaining the rules. The game assumes â correctly â that anyone who bought it has played enough games to figure it out, and it rewards that trust with one of the densest layered design experiences in recent memory. A AAA producer would have demanded eighteen onboarding screens. Basso shipped it solo and let the players do the work.
The Question of Authorship
Here we arrive at what I think is the deepest reason for the divergence. Retro games, by and large, had authors. You can name the people behind them. Castlevania: Symphony of the Night has Koji Igarashi’s fingerprints on every screen. Chrono Trigger is unmistakably a Sakaguchi/Horii/Toriyama production. Metroid Prime is Retro Studios under Mark Pacini. Even the more anonymous arcade outfits had distinctive house styles: a Capcom game from 1991 looks and feels nothing like a Konami game from the same year, and both look and feel nothing like a Taito game.
Modern AAA development, by structural necessity, has largely eradicated this kind of authorship. When you have 500 people across four studios working on a single title, the design becomes a managed average rather than the expression of any individual sensibility. There are exceptions â Kojima, FromSoftware, Larian â but they prove the rule by how unusual they are.
Why Indies Can Author and AAA Cannot
This is not a moral failing on the part of AAA studios. It’s a structural inevitability. You cannot run a 500-person team on auteur instinct. You need processes, sign-offs, focus tests, and risk-mitigation strategies. Each of these is individually sensible. Collectively, they squeeze the authorial voice out of the product until what remains is something perfectly competent and entirely interchangeable.
Indies, by contrast, are often single-person or small-team operations where one creative vision can survive intact from concept to release. This is why the games that genuinely capture the spirit of the 8- and 16-bit eras tend to come from these contexts: the production conditions resemble those of the original era far more closely than any AAA studio’s do. Yuji Naka’s Sonic team was about 15 people. Hideo Kojima’s original Metal Gear team was tiny. Suda51’s early Grasshopper output was made by handfuls. These were boutique operations by modern standards, and that’s exactly why the work has the personality it does.
The CRT Question and the Lie of the Filter
No conversation about retro design in 2025 is complete without addressing the CRT discourse, which has reached a kind of fever pitch over the past eighteen months. The community has spent years now debating composite vs RGB, slot-mask vs aperture-grille, the merits of various scaler ICs and the relative virtues of OSSC, RetroTink 4K (currently around ÂŖ650 and worth every penny), and the various MiSTer FPGA cores.
And then a AAA studio will ship a game with a single sad “CRT mode” toggle that produces a uniform scanline overlay applied at output resolution, with no consideration of pixel structure, phosphor bleed, gamma response, or the fundamental fact that CRTs were emissive displays with non-linear luminance characteristics that LCDs cannot meaningfully replicate.
What a Proper CRT Look Actually Requires
The indies again know this. UFO 50 ships with CRT shaders developed in consultation with the Libretro community. Vampire Survivors‘ various aesthetic options reflect a clear understanding of the difference between an arcade monitor and a home TV. The various recent shmup releases from Cave’s catalogue on modern platforms have been ported with care for the source display assumptions.
If you want to see this done properly at the hobbyist level, look at what people are doing with the MiSTer FPGA platform (a DE-10 Nano costs around ÂŖ200, with additional IO boards bringing the total to around ÂŖ350) combined with a quality CRT or high-quality scaler. The image quality you can achieve approaches and in some respects exceeds what original hardware produced, because you can dial in the exact compromises you prefer rather than living with the specific compromises a 30-year-old engineer happened to make.
None of this is the kind of nuance you can capture with a checkbox in an options menu. But the fact that indies even try, and that AAA studios mostly don’t, is itself indicative of the wider cultural gap.
Soundtracks and the Chiptune Trap
Music is another area where the AAA/indie divide is glaring. There’s a particular kind of “retro-inspired” soundtrack that sounds like it was made by someone who has heard about chiptune music but has never sat down with a tracker. It uses the timbres but not the voicings. It has the bleeps and bloops but lacks the contrapuntal complexity that made the best YM2612 or SPC700 compositions sing.
Compare a properly composed Mega Drive soundtrack â Streets of Rage 2, say, or Yuzo Koshiro’s work on Story of Thor â to the average modern “retro” indie soundtrack and you’ll hear the difference immediately. The originals had to coax expressiveness out of a small number of FM channels with strict polyphony limits. The composers learned to use those constraints as instruments in their own right. Modern composers working in DAWs with unlimited resources often produce technically chiptune-flavoured music that lacks the actual musical compression of the originals.
Who’s Getting It Right
Joonas Turner’s sound design work across multiple indies, the soundtrack work on UFO 50, and the genuinely brilliant compositions of Disasterpeace going back to Fez all demonstrate that the form is still alive. So does the continuing work of original-era composers like Yuzo Koshiro, who scored the recent Streets of Rage 4 DLC. The point isn’t that you have to use original hardware (though plenty of musicians do, and you can pick up a Mega Drive with the Mega Everdrive Pro for around ÂŖ200 plus the flashcart’s ÂŖ150 if you want to compose for the actual silicon). The point is that you have to understand what the original form was doing musically, not just sonically.
The Particular Failure of “Pixel Art” in AAA
One of the most reliably bad things in modern AAA is the use of pixel art for stylistic flourish â usually in cutscenes, achievement icons, or “retro mode” minigames. The proportions are always slightly off. The animation lacks weight. The palettes are too uniform. You can immediately tell that someone has been briefed to make something “look 16-bit” by an art director who has never actually studied 16-bit art.
Real pixel art is a discipline. The masters of the form â Paul Robertson, Henk Nieborg, the team at Vanillaware â work with extraordinary intentionality about every single pixel. They understand sub-pixel anti-aliasing, dithering patterns as texture, the way that strategic absence of colour can imply form more efficiently than its presence. This is craft that takes years to develop, and it cannot be faked by hiring a junior artist and pointing them at the SNES pixel grid.
The 2025 Pixel Art Standouts
The pixel art renaissance in indies continues to produce extraordinary work. Crow Country, released last year and still finding its audience through 2025, demonstrates how to do “low-poly with pixel art textures” in a way that respects both traditions. Tactical Breach Wizards, while not strictly pixel art, has the kind of compositional clarity that the best 16-bit games had. And the various ongoing pixel-art shmups (the recent ZeroRanger follow-ups, the new wave of Cave-inspired work from small Japanese teams) demonstrate that this is a craft tradition that’s far from exhausted.
Buy these games. Tell people about them. The discipline survives because there’s an audience for it, and that audience needs to remain vocal.
The Boomer Shooter Lesson
If you want a perfect case study in how this can go right and wrong, look at the boomer shooter revival of the past several years. The genre â fast, sprite-based or low-poly, complex level geometry, weapons-as-puzzles â has been thoroughly revived by indies. Dusk, Amid Evil, Prodeus, the ongoing Selaco work, and most recently Cultic‘s second chapter have all delivered experiences that don’t just imitate Quake and Blood but extend them.
Now look at AAA’s various attempts to recapture the same energy. Doom Eternal is brilliant on its own terms but is fundamentally a different kind of game from what id was doing in 1993 â more an action-puzzle than a navigation-puzzle. The various other publisher attempts at fast-paced FPS revival have largely failed because they cannot bring themselves to ship without the modern infrastructure of progression systems, narrative throughlines, and seasonal content roadmaps.
Quake from 1996 was a game. Quake from 2024 in its updated form is also a game, distributable for around ÂŖ8 and playable in glorious 4K with modern source ports. The boomer shooter indies understood that the value proposition of these games was complete in the box, not a service to be sustained.
The Collector’s Perspective
For those of us who’ve spent decades building up libraries of original hardware and software, all of this has a particular resonance. The original artefacts continue to appreciate â a sealed copy of Earthbound on SNES will set you back four figures, loose carts of Panzer Dragoon Saga are now routinely north of ÂŖ600, and don’t even get me started on Neo Geo AES prices â but more importantly, the originals continue to play. They continue to teach. They continue to reward attention.
When I want to remember what a tightly-designed action platformer feels like, I don’t replay a modern tribute. I put Ninja Gaiden on the NES into the Everdrive and rediscover, for the hundredth time, that there’s not a wasted frame in the entire game. The level design teaches you its rules and then asks you to apply them under pressure. The enemy placement is mathematical in its precision. The whole thing is two megabits of code and assets and it remains one of the great action games ever made.
The Preservation Angle
The other reason the AAA failure matters is preservation. When a publisher remakes a classic, they often delist or de-emphasise the original. The 2023 Resident Evil 4 remake’s existence makes it harder for newcomers to access the 2005 original. Square Enix’s FFVII remake project has muddied the question of what Final Fantasy VII even is. This is not preservation; it’s replacement.
The indie community, by contrast, has been extraordinary at preservation. The work of the various emulation teams, the MiSTer cores, the archives maintained by hobbyists, the documentation work being done by the likes of the Video Game History Foundation â this is where genuine respect for the medium’s past actually lives. Donate to them. Support them. Buy their merchandise. They are doing the work that publishers should be doing and aren’t.
Practical Recommendations
If you’ve made it this far, you probably want some concrete advice for what to actually play and how to spend your time and money in 2025. Here it is.
The Essential Indie Buys of 2025
- UFO 50 (ÂŖ19.99) â Non-negotiable. Buy it. Play it for a year. It will teach you more about game design than any university course.
- Animal Well (ÂŖ19.99) â The most rewarding Metroidvania since Hollow Knight, played with no spoilers if at all possible.
- Lunark (ÂŖ17.99) â For the Another World heads. The real deal.
- Mouthwashing (ÂŖ13.99) â Short, devastating, the high watermark for PS1-aesthetic horror.
- Crow Country (ÂŖ15.49) â Survival horror with the courage of its convictions.
- Tactical Breach Wizards (ÂŖ17.99) â Not retro per se but possessed of the same focused-design virtues.
The AAA Releases Worth Engaging With
To be fair to the publishers, not everything they do is hollow. Prince of Persia: The Lost Crown remains worth playing despite its rough edges. Capcom’s continuing remake program is, if nothing else, technically impressive. The various FromSoftware releases continue to demonstrate that AAA scale and authorial voice are not entirely incompatible. And the new Mario & Luigi entry late last year proved that Nintendo, at least, still understands how to make a focused 20-hour game without padding it out to 60.
Hardware Investments Worth Making
- MiSTer FPGA setup (~ÂŖ350 all in) â The single best investment for serious retro engagement.
- RetroTink 4K (~ÂŖ650) â If you’re running original hardware on modern displays, accept no substitute.
- A decent CRT â Free to ÂŖ200 on the second-hand market, increasingly hard to find, increasingly worth the effort.
- Analogue Pocket (~ÂŖ220) â Genuinely the best way to play handheld classics, with growing core support throughout 2025.
What Comes Next
The trends I’ve described aren’t going to reverse. AAA economics will continue to push publishers toward bigger, smoother, more risk-averse products. The indie scene will continue to be where genuine design experimentation happens. The gap will widen.
What might change is the audience’s tolerance for the divergence. There are early signs â the commercial success of UFO 50 and Animal Well, the continuing strength of the boomer shooter scene, the visibility of the CRT and FPGA hobbyist communities â that a meaningful slice of the gaming audience is increasingly aware that what they’re being sold by the major publishers is not what they actually want. The question is whether this awareness will translate into commercial pressure significant enough to change publisher behaviour.
I’m not optimistic on that front. AAA publishers are responding to a market that includes hundreds of millions of casual players for whom the kind of design literacy we’re discussing here is genuinely irrelevant. The smoothed, bloated, focus-grouped product is what works for the median customer, and the median customer is who AAA exists to serve.
The Sustainable Future
What this means in practice is that the future of retro-literate design is, and probably will remain, indie. This is fine. This is, in many ways, how it should be. The original 8- and 16-bit eras were dominated by small teams working with passion and constraint. The current best successors are also small teams working with passion and (self-imposed) constraint. The form has returned to its native scale.
The role of the serious enthusiast is to support this ecosystem. Buy the games at full price. Talk about them on whatever platforms remain functional. Mentor newcomers into the rich back-catalogue. Maintain the hardware. Document the history. Pressure the publishers when they get it wrong, but don’t expect them to fundamentally change. They are not the keepers of this flame and never really were.
A Closing Thought from the Margins
The greatest single thing about retro gaming as a culture is that it has always been a counter-current. When publishers were chasing CD-ROM full-motion video in the mid-90s, the people making the best 2D games were already being treated as anachronisms. When publishers were chasing motion controls in the late 2000s, the people preserving and extending traditional design were doing it on the fringes. Now, when publishers are chasing live-service models and AI-generated content, the keepers of focused, authored, complete-in-the-box game design are once again in the margins.
The margins are a good place to be. They’re where the interesting work happens. They’re where the design literacy actually lives. And they’re where, in 2025, the actual heirs of the games we loved as children are quietly making the games we’ll still be loving in 2055.
The AAA studios will keep failing at retro-inspired design because they cannot afford to succeed at it. The indies will keep succeeding because they cannot afford not to. The chasm between them is now the most interesting feature of the entire gaming landscape, and as enthusiasts, our job is to look across it clearly, name what we see, and back the right side with our attention and our money.
The future of retro gaming, paradoxically, has never looked healthier. You just have to know where to look â and increasingly, where not to.