Reviewers giving high scores to poorly optimised games really grinds my gears

Submitted ⁨⁨1⁩ ⁨week⁩ ago⁩ by ⁨delitomatoes@lemm.ee⁩ to ⁨gaming@beehaw.org⁩

So this problem started back in 2023 as in still occurring. Dead Space Remake, Jedi Survivor and Silent Hill 2 are some of the offenders. All these games were highly rated on gameplay and graphics

Basically, even if you have a 4090, the stutters and poor fps still exist due to the way the game is designed.

Its exceptionally frustrating because open world devs do optimise for performance while corridor type REMAKES like Dead Space and Silent Hill forget the work their predecessors did on worse hardware years ago

source

Comments

Sort:hotnew top

SnotFlickerman@lemmy.blahaj.zone ⁨1⁩ ⁨week⁩ ago
Unpopular Opinion: The last few hardware generations have had diminishing returns while increasing the cost of being a PC gamer drematically. While the DOOM games are generally well-optimized, I just upgraded my whole ass system after 8 years just two years ago and I’m hitting minimum specs to play the game at all.

Seriously, the rig I bought to play fucking Bioshock Infinite kept up for about 8 years. I know I didn’t go all-out but I didn’t 10 years ago either when I put my old box together.

Anyway, kind of feels like a rip-off by the industry to me, and this is the same industry that is pushing for GTA 6 to cost $80-100 because they’re not making enough money somehow.

source
- TachyonTele@lemm.ee ⁨1⁩ ⁨week⁩ ago
  There needs to be better compression for texture and sound files. That’s pretty much the reason for the giant install sizes and RAM requirements. In theory it’s possible, but it hasn’t really changed in decades.
  
  “AAA” companies don’t want to spend q/a time on code. Indie devs don’t seem to have that problem. So there’s a huge gulf between quality.
  
  Personally I think the day one patch excuse for reviewers is bull. Day one patches have been a thing for at least 25 years, everyone should know they don’t fix bad games. If the companies are not being called out for bad practices then they’ll never bother to fix them before going gold.
  
  source
  - Crotaro@beehaw.org ⁨1⁩ ⁨week⁩ ago
    Additionally, if a day one patch were actually enough to fix these issues, then just delay the game by a day. That way, the launch day gamers won’t suffer through a (sometimes) unplayable experience and possibly leave bad reviews.
    
    source
- Toes@ani.social ⁨1⁩ ⁨week⁩ ago
  
  Of course, we also went from 8gb of video RAM being more than enough
  
  After the 1080ti launch, I’ve held the opinion that any card with less vram has been a scam.
  
  source
Poopfeast420@discuss.tchncs.de ⁨1⁩ ⁨week⁩ ago
This is much older than 2023. I remember Fallout 4, the console version was apparently almost unplayable at launch, so Giant Bomb actually lowered the score, compared to the PC version. And even that example isn’t when this started.

Similarly, what if the reviewers don’t get a specific version, that runs like shit? Like what happened with Cyberpunk, where nobody was able to play the XBONE or PS4 versions.

The thing is, as always, a review is subjective. If the game has problems, but the reviewer can look past them or doesn’t care, why should they change the score.

Someone mentioned it already, but review copies might also run outdated code, and reviewers are in contact with the publisher or devs, and they might say some problem is fixed on release. If the reviewer believes them, it probably won’t affect the score.

source
SadSadSatellite@lemmy.dbzer0.com ⁨1⁩ ⁨week⁩ ago
I see this opinion fairly often, and it honestly confused me. My rig is not a showcase by any means, and I had no issues with sh2 or dead space.

I’m thinking more people need to optimize their OS.

source
- bread@feddit.nl ⁨1⁩ ⁨week⁩ ago
  You just aren’t perceiving the issues that are there. SH2 stutters. That’s a fact.
  
  source
Fubarberry@sopuli.xyz ⁨1⁩ ⁨week⁩ ago
In my experience it’s largely been unreal engine 5 games.

The issues with both Doom DA and Indiana Jones is that they have mandatory ray tracing that can’t be disabled. I generally think that ray tracing is a often a waste, it’s far too resource demanding, other lighting techniques can offer very similar visuals for a fraction of the cost.

source
- mtlvmpr@sopuli.xyz ⁨1⁩ ⁨week⁩ ago
  Sure it wastes your resources but the devs don’t need to do much so it speeds up their workflow. You are a sacrifice they are willing to make.
  
  source
  - Fubarberry@sopuli.xyz ⁨1⁩ ⁨week⁩ ago
    The audience being a sacrifice doesn’t always work out when they’re the ones expected to buy the game.
    
    source
  - ICastFist@programming.dev ⁨1⁩ ⁨week⁩ ago
    “I’m willing to sacrifice potential sales in order to have an easier dev cycle” - I’m frankly amazed that the higher ups ok-d that
    
    source
- termus@beehaw.org ⁨1⁩ ⁨week⁩ ago
  Arkham Knight was the first I noticed it really bad on, no RT present. Since then it seems most Unreal titles suffer from it in some kind of way.
  
  source
Bougie_Birdie@lemmy.blahaj.zone ⁨1⁩ ⁨week⁩ ago
Frickin Dead Space remake. I’m playing through it now and even on the lowest settings it was pretty bad. My computer crashed while the shuttle was crashing, which honestly felt kind of apt

I’ve never beat the original, but my wife wanted to see the game and has never played it. Even after tweaking things to get them running on my computer it’s still not super stable. We might have to switch to 2008

The game looks super tense, gross, and scary. Personally, I think it’d be scarier if it was buttery smooth, but I guess there’s a certain amount of anxiety to be had wondering if walking through a door is going to freeze the game while I’m being chased by xeno horrors

source
- prole@lemmy.blahaj.zone ⁨1⁩ ⁨week⁩ ago
  One of a handful of reasons that I’ve been turning to console for AAA games with this level of graphic fidelity.
  
  source
- DdCno1@beehaw.org ⁨1⁩ ⁨week⁩ ago
  It ran incredibly well on my machine and looked amazing. This is not a poorly optimized game in my experience. Could it be that it also ran fine on the machines of most reviewers?
  
  source
  - Bougie_Birdie@lemmy.blahaj.zone ⁨1⁩ ⁨week⁩ ago
    It wouldn’t surprise me if reviewers have the budget for top of the line computers. Sadly, that does little to soothe my frustration that my computer - which meets the advertised recommended specs - falls short
    
    source
    -> View More Comments
sexual_tomato@lemmy.dbzer0.com ⁨1⁩ ⁨week⁩ ago
I had to refund armored core 6 because the first boss fight stuttered so much. I have a 3080, so not the best but definitely no slouch. No matter what settings I tweaked, it’d always drop to slideshow performance for 1-2 seconds at a time. Acceptable in a cheaper title, maybe, but not on a full price game.

source
Comment105@lemm.ee ⁨1⁩ ⁨week⁩ ago
Start reviewing games based on this kind of performance, or start paying attention to someone who are already doing it.

I don’t know of any.

source
megopie@beehaw.org ⁨1⁩ ⁨week⁩ ago
Often times, the investors or stakeholders at these large video game companies have their backgrounds in Hollywood, or Tech. They then choose leadership who will run the company along the lines of what works well in those industries. This results in optimization being pretty damn near the bottom of the priorities.

What has been most profitable in Hollywood? Not the final quality of the movie, but the marketability. How many people did you get to come see it, doesn’t matter if they loved it, so long as they heard about it, then choose to buy a ticket.

What has worked well in tech? Getting to market as fast as possible with the latest technical developments. Doesn’t matter if it’s a buggy mess and riddled with technical debt, so long as we capture as much market share as possible before anyone else can compete.

Combine these two approaches and what do you get. The fanciest graphics, huge maps, endless procedural fetch quests to make it look big, all so people will preorder it. Oh and it needs to be done in 2 years or else someone else will beat us to being the fortnight of “live service extraction farming sims”.

So lots of demands on what needs to be in it, and no time to do proper QA, let alone optimize it, that will just have to be done in patches after launch.

The cost of poor optimization gets externalized to the customers who need to buy new hardware or run it on settings so low it could be mistaken for half-life.

source
millie@beehaw.org ⁨1⁩ ⁨week⁩ ago
Not to say this necessarily isn’t the case, but are your drivers all up to date? I don’t know how often I’ve heard people complain about shitty performance or weird artifacts in a game only to hear that the player hasn’t updated their graphics card drivers in 8 months.

source
Ephera@lemmy.ml ⁨1⁩ ⁨week⁩ ago
I’ve heard the reasoning before that reviewers typically only have access to a, well, pre-release version. A day-1-patch is pretty common now.
So, as reviewer, you have to decide whether the performance problems look like they might be fixed on release day, and therefore whether you want to incorporate them into your review/score or not.

source
- Ulrich@feddit.org ⁨1⁩ ⁨week⁩ ago
  
  as reviewer, you have to decide whether the performance problems look like they might be fixed on release day
  
  No you do not. You review what they give you. If it sucks, they shouldn’t have given it to you, and that’s what your audience deserves to hear about.
  
  source
mox@lemmy.sdf.org ⁨1⁩ ⁨week⁩ ago
Throwing another example on the fire: The Last of Us Part I PC port. The people who released that code ought to be brought up on charges for climate destruction.

source
termus@beehaw.org ⁨1⁩ ⁨week⁩ ago
Digital Foundry has complained about the stuttering issues in all of those games. Jedi Survivor still gets brought up.

source
mtlvmpr@sopuli.xyz ⁨1⁩ ⁨week⁩ ago
I don’t even see the optimising problem as bad probably because I’m used to it. Evil Within for example ran terribly on launch but since then I have upgraded my setup and played through it. Checking the benchmarks just is something you should do on PC. Reviewers usually are pretty out of touch with rest of the world or average gamer just doesn’t care about drops and sub 30 framerates.

I feel like a worse problem currently is how games are made with TAA in mind or with forced TAA so you just can’t make them look good even with time and more powerful hardware. It’s either blurry mess with artifacts thrown in or something that doesn’t even look like 1080p on 4K resolution.

source
adarza@lemmy.ca ⁨1⁩ ⁨week⁩ ago
i was half-expecting a rant about PoE2 for some reason…

source
rickyrigatoni@lemm.ee ⁨1⁩ ⁨week⁩ ago
Buddy it’s been going on since long before 2023.

source
malchior@aussie.zone ⁨1⁩ ⁨week⁩ ago
Jedi Survivor…god this game performs so bad. 5800X3D and a 4090 and those stutters. The game isn’t even that graphically impressive so I don’t know wtf is going on behind the scenes, probably some old code that constantly checks for micro transactions.

source
emptyother@programming.dev ⁨1⁩ ⁨week⁩ ago
What should they do about it if it actually runs great on their systems though?

A lot of games only play well if you take some time figuring out a certain combination of graphics settings for your own computer. Then there is bugs and stutters that really is only happening with certain settings. Particularly these days with the four common upscaling models, you never know which games are best optimized for which model, but none of them are optimized for running without upscaling.

So for a regular reviewer to really give a game a fair score, should they run at the default settings? Would be unfair to expect them to know how every weird setting impacts the game. Should they try the game at 4+ different systems to make sure there are no performance issues and stutters dragging it down in certain cases? Leaving performance testing to dedicated performance reviewers and just focus on reviewing the game itself might be the best option.

source
- Boomkop3@reddthat.com ⁨1⁩ ⁨week⁩ ago
  They should have the insight to use a system comparable to the average user. Or at least attempt to understand the perspective of the audience they’re reviewing for
  
  source