The 5/10 score baffled me. By your own policy, that's for games that are "uninspired, unfinished in key areas, or represent poor value for money."

A lot of games journalism disproportionally rewards superficial polish. Bland games that fail to reach audiences score well because they read as prestige. Games that are bold, imaginative, have legitimately interesting mechanical synergies or are just plain "fun" often score far worse because they don't. It's Oscar-bait applied to games.

Some interesting facial proportions there... Eve's lips are half way up her face.

Arkane, historically, resisted corporate interference, PR-heavy staffing and focus-group driven development. Ubisoft is a shrine to corporate bloat, office politics and derivative design. I can't think of a worse pairing - save for maybe Microsoft themselves.

Sad as it is as a fan since Arx Fatalis, much of the talent has moved on. I'm more interested in seeing what Black Pony Immersive and Wolfeye Studios get up to.

@rjejr No worries. There are a couple of impressive Chinese indie devs (2-5 man teams) who would benefit massively from $1M. For larger productions, it's a month or two of operating cost. Tend to agree that's PR only.

@mrbone Possibly, but take something like Wukong. Roughly half of its 30M+ sales are from outside of China. That puts it ahead of many first-party titles.

@rjejr Phantom Blade Zero was about $50m, so pretty superficial for anything approaching Chinese AAA.

The global gaming player base has grown 28% in that period. Steam revenue has roughly doubled, as has mobile gaming revenue. A 1% increase in unit sales is pretty dire when you consider that.

Mixed feelings tbh. Karpyshyn and Hamilton are legitimate selling points, but watching this, I can't help but feel the Hasbro influence.

@whatkindofsick The level design was simplified, multiple areas, enemies and puzzles were cut and the Nemesis was nerfed - all of which was widely criticised. That doesn't mean it's a bad game, though.

@RoomWithaMoose Yeah, that's entirely fair - it's a fairly arbitrary delineation to the consumer and niche implementations go back another 2 decades.

I never worked with either, but from what I've heard, the Saturn was unpleasant. It was a sprite renderer that panic-added 3D through quad warping, required programming the co-processor for 3d geometry, didn't support UV mapping, etc, etc.

The PSX shipped with a purpose built fixed-function geometry engine and hardware triangle rasterizer. It was made for 3d, easier to develop for and most resembles "modern" rendering. Better developer outreach and a $100 lower market price probably didn't hurt either...

@RoomWithaMoose The PSX introduced hardware polygonal rendering to the masses. 3dfx Voodoo did the same for PC in 1996, alongside texture filtering and other improvements. Both were massive milestones in the development of realtime graphics. Foundational, even.

@Wiceheid Cell had multiple SPUs - vector processors, highly efficient at what they did - alongside a weaker PowerPC CPU and GPU. Early generation, existing engines didn't effectively utilise these, resulting in games that used a fraction of the processing power available. Admittedly, targeting multiple platforms was a nightmare, but I wouldn't call it "bad" architecture. It was an imperfect technology, ahead of its time, that forced developers to adopt paradigms that would become industry standard in years to come.

Most of the emulation woes come from its eclectic instruction set, hardware synchronisation and DMA mechanisms and how the SPUs were ultimately used - timing-sensitive synchronisation and self-modifying job systems that are a nightmare for traditional JIT optimisation. Quite different to the CPU limitations of the PS4. Most back-compatibility has also relied on dedicated hardware.

I appreciate "industry insiders" are quick to say EA won't sell off anything, but it's a changed company dealing with two liquidity crises - internal, and its Saudi investment. Isn't it more logical that they're removing potential copyright infringements from surviving franchises in anticipation of selling off the IP? The Amazon series adds immediate value to the franchise - making now the time to sell.

Who knows. Entirely avoidable, tone deaf PR mismanagement is on brand as well.

@extermin8or_ AI assists jailbreaking as much as it does reverse engineering games. Walled gardens are simply becoming increasingly inviable.

Limiting physical media in a misguided attempt to slow this will just result in what happened on PC in the early days of Denuvo. Paying customers had less flexibility and often worse experiences than those playing pirated copies.

Having clocked in 4 hours last night, this score seems ridiculously harsh. I appreciate reviews are subjective, and it seems a polarising game, but if you enjoyed the demo you'll likely enjoy the full game. For what it's worth, it has an Overwhelmingly Positive (96%) approval rating on Steam.

@AaronBayne Thanks. I went and picked it up on PC - performance was stellar and I find aiming easier with a mouse, which with ammo being scarce seems like a good idea. Overall, I really enjoyed the demo and want to support a smaller studio taking risks.

@AaronBayne How reflective of the rest of the game is the demo? More polished, or much the same?

@DennisReynolds That's quite a time consuming ask - especially for an exchange that borders on one-sided. My position is clearly established; what about it do you disagree with? Do you deny that games do this, or think it's acceptable when they do?

@DennisReynolds Largely what would be considered poor writing if applied to things non-political:

• Themes being asserted rather than dramatized
• Characters breaking existing characterisation to facilitate a message
• Dialogue existing only to tell the player the authors position
• Internal consistency being violated to accommodate allegory or contemporary commentary
• Compromising the general verisimilitude of the setting to do the same
• Pausing narratives/undermining pacing to lecture the player
• Supporting cast and antagonists being reduced to ideological caricatures
• Illogical plot beats to facilitate any of the above

@DennisReynolds Because I'm not advocating for excluding politics? That's censorship. I'm pointing out that heavy-handed didactism, especially when at odds with tone or world-building, is a grating creative failure that alienates audiences, is fair to criticise and responsible for much of the demand for apolitical games - more so than any one idea itself.

There's a banal tendency to try and collapse otherwise worthwhile discussions into moral purity tests in lieu of anything of substance. You know, reducing a discussion about craft to "you don't like X group". Let's not do that.

@DennisReynolds Why would I provide distinct topics when my argument is it's treatment that determines reception? There's a massive, unsubstantiated leap between "all art is implicitly shaped by the artists biases" - the meaning of all art is political - and "therefore, the medium must support any amount of ham-fisted, condescending political messaging, even when it detracts from the overall quality".

To be clear, I wholly agree with the former and am not opposed to political themes in games. It's the latter, confirmed by surveys and studies, that the overwhelming majority of players regard negatively regardless of political leanings. It's also just poor writing.

Ignoring the distinction and hiding behind "all art is political" is a lazy motte-and-bailey - a waste of time. You do you.

@DennisReynolds "unrelated to or at odds with", "forced" or "external to the fiction"
Like every creative choice it's a matter of context and execution. Less what, more how and why.

@DennisReynolds Taken to an arbitrary degree, art cannot be art without bias; but there's nuance inane statements like that ignore. A creators life experience and views implicitly shaping holistic worldbuilding is quite different to injecting didactic drivel unrelated to or at odds with it.

In my experience, many people asking for apolitical games just don't want that - to have escapism undermined or immersion ripped out from underneath them because of some forced characterisation, world-building or plot beat that serves only to validate real-world ideology external to the fiction. It's grating, even when it's your own politics.

Strange choice to demo something so early on in development. Roughness aside, they've clearly added quite a bit to combat, so I'll remain cautiously optimistic until release.

@get2sammyb It definitely does look better. I still don't see it sustaining interest long enough to break even.

@Dorkthrone Narrative can detract from a good game. Bad characterisation, dialogue, pacing or just a boring/dumb plot. Not everyone cares about narrative, and not every game needs a strong one, but let's not pretend there isn't a range of quality.

@Cjam36 Platform features rely on server-side infrastructure, and things like achievements are tied to a particular product ID. While simpler fallbacks are usually used during early development, actual testing of platform-specific integration requires registering your game with that platform.

@IntrepidWombat I wouldn't say it's a lost art - it still exists in niche titles - rather a shift in priorities and design philosophy based on market dynamics.

Games cost more to make, requiring more copies sold. This means appealing to a wider range of players, while also competing with more alternatives for their time. "Mainstream" games generally lean towards streamlining mechanics and focusing on player retention at the expense of depth or complexity.

@get2sammyb "shouldn’t the game guide you to the solution organically in a way that makes you feel like the highly intelligent archaeologist you’re commandeering"

I have mixed feelings about that. Accessible game design frequently undermines the intended player experience; at least this can be ignored. The same level of affordance, as environmental signposting, would be so heavy handed it would ruin puzzles for anyone wanting an actual think. Ideally its availability would be tied to difficulty, and difficulty to certain achievements, providing a reward for honest effort.

@AhmadSumadi Visual appeal is interest, impression and memorability. In an exaggerated setting, a vengeful half-cybernetic predator reading as "normal person" is a failure in the same way Anor Londo looking like the average European city or Silent Hill a quaint New England village would be. It's poor design.

Arguing visual appeal isn't important in a visual medium is wild.

Nice to see another well made Barker game. At that price, it's day one.

@AaronBayne How does the sluggish movement feel given the emphasis on combat? It looked a bit clunky in the extended reveal, but I appreciate that's the kind of thing you notice more when watching footage than playing.

@DennisReynolds An impressively bad-faith argument. Ignoring the fact that your preferences in no way reflect another's, you immediately reduce appeal - a product of design, costuming, characterisation and performance - to sex appeal and associate it with a minor to try and undermine someone else's personal opinion. Real upstanding.

@ButteredCrumpet I haven't worked with Euphoria, but I'm familiar with related theory - Neuromuscular Control and Physics-based Character Animation. Triggered by game events, a set of defined synthesis behaviours establish goals for a physical neuromuscular simulation. While it conceptually deals in "muscles", they are proportional-derivative controllers with no visual representation.

With sufficient overlap between the logical and animation representation, I can see it outputting activation data (i.e. contract bicep X amount) for synthesized animations. It has no idea of what vertex clusters to displace to actually make a bicep look like its contracted though. That's traditionally the domain of corrective bones and morphs, and more recently neural deformation models. Actual, nuanced muscle sims at scale aren't possible in real-time, but they can be used to train a model that "knows" where vertices should be based on pose and activation. Detailed trajectory and collision aware tissue deformation requires actual simulation, but I don't see anything approaching that in the footage.

Clothing, outside of free cloth (i.e. a cape, loose elements of a dress) is usually bone driven. Bones have explicit, predefined influences, so you'd need a lot of them to adequately represent fine-cloth deformation. Beyond authoring complexity, a large number would slow pose evaluation. Instead, complex material simulations are baked down to textures - parametric wrinkle maps that are driven by joint relationships. Again, while actually simulating cloth that granularly is generally impractical in real-time, such sims are now also used to train models that infer deformations.

A lot of this is pretty well established already. It's just really time consuming, and by implication, expensive to get looking good. That's what I mean by more polish than technical innovation. Don't get me wrong, a studio of their size and expertise no doubt has some interesting ideas under the hood. I just think the fidelity is more a product of craftmanship than technical disruption.

As for NTC - I'm not sure how it'll land. The prohibitive limitations (i.e. no hardware filtering) have largely been addressed by wave intrinsic; it no longer introduces significant temporal instability through stochastic sampling. It still limits pipelining and cannibalises compute to save VRAM at a time where games are starved for compute even at sub-native resolutions. At least some of the VRAM issue is (tech) art discipline, and I would argue texture resolution increases yield diminishing returns at this point, but maybe outsourcing and photogrammetry heavy pipelines amidst a RAM crisis changes the calculus sufficiently?

Neural rendering in general I'm more hopeful for. Real-time has a major under-sampling problem, and the throughput increase required to approach the fidelity of offline renders with brute force is disproportionate. Neural rendering offers alternatives. I just fear both will be thrown in the "AI slop" category because people conflate it with diffusion, thinking it was trained off Artstation.

@ButteredCrumpet No worries. The content was intended for a discussion between non-laymen - based on our fundamentally different opinions on what is occurring each frame. Thought it might be fun to speculate.

@ButteredCrumpet I don't want to take away from the presentation, it is legitimately impressive, but I have to disagree with a number of your reads. Game development is always smoke and mirrors - cheating impractically expensive effects; my guess is this is less novel tech and more an overlap of exceptional execution.

Cloth, outside of instances that benefit from it (free flowing), likely isn't simulated. Helper joints, and potentially corrective morphs, introduce silhouette changes. Belt lines are stable, displacement in anything approaching fitted is not trajectory aware, walk cycles consistently reproduce deformations. Traditionally, going on over a decade, finer creases and folds - what would be impractical to simulate - are baked, the blend parameterised by pose. There's a possibility of a deformation model being used - training pose to vertex displacement - which admittedly would be neat. Similar premise, greater efficiency for a wide range of clothing items.

In terms of actual character meshes - I'm not sure there is soft-body. From limited footage, I don't see anything nuanced enough to demand it. A network of dynamics bones with simple constraints can be plenty compelling. In the unlikely case of soft-body, from experience simulations can be made relatively cheap. A sparse island-based XPBD representation, offline graph-coloured for massively-parallel compute simulation, barycentric skinning of low-resolution proxies onto high resolution geometry or deformation bones. Even then, such effects would never be active beyond the immediate foreground, so scale is less of a consideration.

On the technical front, outside of reflections which have the usual hallmarks of sample limitations, the impressive take away is consistency. GI appears stable in dark scenes, the visual fidelity consistently good. Like everything else, I would imagine this is a product of polish. Many years, countless skilled developers, making an exceptionally refined product.

On a side note, I find the substance analogy a little confusing. The PS6's material functionality is neural acceleration for compression and inference. Authored materials, encoded into neural appearance models for reproduction, rather than variation. Rather than rely on large textures, BRDF response is inferred from a lightweight network for a texel; the opposite of consuming vast amounts of VRAM and time reassembling a dependency graph of texture transformations as an input to a BRDF.

@Weez Agreed - I'd be playing this on PS if it wasn't for the modding support and having bought the original on PC (don't think you can cross-platform upgrade).

I do find the subtitles annoying though. I game on both, playing different types of games on each, and have to constantly look up whether or not a game is multi-plat.

@BearsEatBeets It's one of the more misleading instances of an annoying habit they have. That footage was captured on PC which looks quite different (extensive RT and path-tracing are PC only).

Doesn't land like it did a generation or two ago. Saros and Astro Bot are the two exclusives I'd actually miss if I didn't own a PS5.

Both look good, but I prefer the previous version; it fits the setting better.

@get2sammyb Thanks. Not overly stoked about "foreign dev trying to do Marvel script", but I guess it can't get everything right. I really like the look of the gameplay and art direction, and am starved enough for character action to pick it up day-one regardless.

@get2sammyb Thanks for the preview; I'm really looking forward to this. A quick question about the writing though?

Quips and cheese are par for the course in character action, and modern SSM and Naughty Dog is a pretty broad descriptor. Does it lean heavily into the Marvelesque writing that's becoming popular in western releases? You know - insincere, constant sarcasm & self-aware dialogue, pathologically needing to undercut tension, trying too hard to be "quirky".

@Wilforce It's a remake - the characters and tone are already established. Gena is a wild, thrill-seeking petrolhead in a punchy, absurd franchise. Even if one assumes the original style wouldn't land, which I don't believe is true, there's no reason to default to strait-laced hipster.

Her and Axel's new designs are just bland. I legitimately hope that isn't a reflection of changing times.

Makes you wonder what they'll do to the upcoming Golden Axe remake.

@Random_play05 To be fair, many of those "failures" are only deemed so because of massive production costs. Needing to clear 5M+ units to break even necessitates safe design. Those costs primarily come from obsessing over graphical fidelity, insisting on huge scopes (open-world or 20h+ playtime) and introducing hours worth of cinematics into an interactive medium.

I'd be happy to see more games sacrifice some of the above if it meant the freedom to try new gameplay ideas.

I get dogpiling isn't fun for the developers, but that strikes me as more of an algorithm problem. Social media (like PushSquare) monetizes controversy. The handful of examples posted, though, seem pretty tame and reflect a growing dissatisfaction with derivative design and Marvel properties.

@colt45z It was. I also really enjoyed the soft, painterly style - probably why I'm here whinging about it.

There's still fun to be had with the remaster. It just deserved more care than being outsourced to a middling third-party developer.

@colt45z Personally, I feel the opposite. They butchered the art direction for moderate gains in fidelity. The colour grading is muddy and desaturated, ancient cities are pristine and unweathered and the creature redesigns are consistently less imaginative. All of the original gameplay gripes are still there, with a bunch more added. It just feels lazy.

I eventually settled on playing the original - heavily modded - on PC.

@PuppetMaster I get where you're coming from, but those artists are working around the clock on SB2 and have a different skillset. Most AAA studios (outside of the few with dedicated CG departments) outsource promotional CG to shops like Goodbye Kansas or Blur. Doing so comes with a lead time of 3-6 months from pitch to product, and the adage of "million dollars a minute" is often not hyperbole. It's too early for game footage, so without AI this would have just been promoted with cover art - which... likely would have been better.

The silver lining, as I see it, is if the AI division has time to make slop like this, they hopefully aren't too involved in actual production.

@get2sammyb I'll reserve judgement. This falls into the category of things that wouldn't have existed if not for AI (for better or worse). They wouldn't have spent literal millions on high quality CG for a music video.

Same thing with the promotional video. Placeholder content shouldn't have been publicly demoed but there's nothing indicating it'll be in the final. If by release there is nothing obnoxious that ruins immersion or world building, I have no complaints.

@Swandivetotheasphalt Those numbers don't account for the changes in install base. Working on attachment rate, it looks quite different:

9.7% of PS3 owners bought God of War 3 in its first year. In 2026, that equates to 9.1m units.
12.3m to reach parity with God of War 2018 (13.2% ownership in its first year).

It's not a perfect metric, but far more accurate than the alternative. The franchise has typically been a strong seller.

@Papa_Godzilla Far less so in regards to renderer/engine development - where staffing is highly selective and generally proven.

Hugely disappointing, regardless. idTech is great, and it's sad to (potentially) see another proprietary engine shelved.