This morning, news headlines and YouTube videos across the internet declared that Bethesda will not change engines for the upcoming games Starfield and The Elder Scrolls VI, setting off a wave of outrage that’s become oh-so-common in the world of video games. This story, however, is misleading, based mostly on speculation and widespread lack understanding of what a “game engine” actually is.
It started with a November 2 Forbes article that blamed Bethesda’s game engine for Fallout 76‘s technical issues and graphical shortcomings.
That article quoted a June 2018 interview in the German outlet GameStar with Bethesda creative director Todd Howard. Here’s what he said:
For Fallout 76 we have changed a lot. The game uses a new renderer, a new lighting system and a new system for the landscape generation. For Starfield even more of it changes. And for The Elder Scrolls VI, out there on the horizon even more.
We like our editor. It allows us to create worlds really fast and the modders know it really well. There are some elementary ways we create our games and that will continue because that lets us be efficient and we think it works best.
Although the quote itself isn’t particularly controversial, its sudden re-discovery has led to blazing takes everywhere. One article, on the website Push Square, has thousands of shares on Reddit and Facebook with its declaration that “Bethesda Will Keep the Same Fundamental Game Engine for The Elder Scrolls VI, Starfield.”
YongYea, a YouTube provocateur, also talked about this issue to his hundreds of thousands of subscribers. “Fallout 76 in particular highlights more than ever just how utterly inefficient this game engine is with its inoptimal performance and general lacklustre graphical fidelity compared to other titles of its time,” he said.
It’s true that Bethesda’s games have long been criticised for their game-breaking bugs and inability to reach the beautiful graphical standards of other high-end games, but there might be many reasons for that. One of those reasons might be their ambition—few other games offer as much world interaction as Skyrim or Fallout 4.
Perhaps another reason is Bethesda’s internal processes, or programming guidelines, or development timeline, or even some busted line of code buried somewhere in a file that nobody has touched since 2004. It’s hard to say.
Blaming Bethesda’s “game engine” is misguided, however, because the word “engine” itself is a misnomer. An engine isn’t a single program or piece of technology – it’s a collection of software and tools that are changing constantly.
To say that Starfield and Fallout 76 are using the “same engine” because they might share an editor and other common traits is like saying Indian and Chinese meals are identical because they both feature chicken and rice. What we see on the outside, like a game’s graphical style, its animation system, and its physics, can be changed in all sorts of ways without switching to a new engine.
The term “engine” is thrown around often among video game fans and pundits, mostly in a derogatory way. When a game looks or runs badly, people blame the engine, whether it’s through insulting comments about Unity or hackneyed adjectives like “creaky.”
Wrote the Forbes article: “It feels like every month we achieve some new level of detail and beauty with a new release, and yet something like Fallout 76 comes along and it’s just noticeably worse than everything else with an engine that feels like it’s about to crumble into dust, despite bolting on new parts and upgrades to try to keep it going.”
To understand why this trend is so silly, let’s run a quick refresher on what a video game engine actually is.
Say you’ve just made Super Plumber Adventure. It sold a couple of copies, and now you want to make a sequel, which you know will share many of the same traits. You still want your plumber to run from left to right, you still want mushrooms to make him bigger, and you still want coins to disappear when he collects them.
Rather than write new code and create new animations for all of these things, you might take what you built for the first game and reuse it, bundling all those features together as a physics system.
Combine those physics with some other systems—like a level editor and a memory management tool—and you’ve got an engine, a collection of software that you can use from game to game in order to avoid redundant work. Super Plumber Adventure 2 will hopefully take a lot less time now that you’ve already got so much done.
When we use terms like “Unreal” or “Frostbite,” that’s what we’re talking about: a framework for making games. These are not immutable creations, and in fact, a game’s programmers will alter an engine’s features constantly based on what suits their needs. (Most game studios have tools programmers who dedicate their entire jobs to working on these features.)
Often, fans will associate certain engines with specific graphical styles, but that can be misleading, because two games can run on the same engine but have very different art direction. Both the retro-styled Octopath Traveller and realistic-looking Days Gone use Unreal Engine 4. Both the sports series FIFA and the upcoming shared-world shooter Anthem use Frostbite.
Engines are iterative, and any game studio that uses the same engine from game to game will be modifying it constantly, as Todd Howard said in the very quote that’s caused so much outrage.
To reiterate: “For Fallout 76 we have changed a lot. The game uses a new renderer, a new lighting system and a new system for the landscape generation. For Starfield even more of it changes. And for The Elder Scrolls VI, out there on the horizon even more.”
Oftentimes, aspects of an engine will be in development alongside the game. In other words, Bethesda’s engine in 2018 looks drastically different than it did in 2013, and by the time The Elder Scrolls VI comes out (2024?), it will look like something else entirely.
The editor might be similar – as Howard implies in that quote—but that’s just one component of an engine that has been changing for years and years.
This is not uncommon, by the way. As one game developer pointed out to me this morning, even the ubiquitous Unreal Engine 4 is still built on a foundation that started with the first Unreal, which came out in 1998.
When I broke the news in June that Fallout 76 was an online survival game, one person familiar with its development told me that Bethesda’s engineers had spent years adding multiplayer capabilities to the engine, which was a challenging and complicating endeavour that required rewriting a whole lot of code.
On the outside, Fallout 76 might look similar to Fallout 4, but peeking into its guts would tell a different story. To say they use the same engine might technically be accurate, but it’s misleading.
The concept of a game engine has become a bugbear for fans, and with Bethesda’s longrunning reputation for nasty glitches, it’s always tempting to find factors to blame.
Fans and pundits should absolutely criticise games like Fallout 76 for their ridiculous bugs and graphical failings. But today’s controversy – and the notion that the next-gen games Starfield and The Elder Scrolls VI would use the same “engine” as today’s games – is misguided at best.
Comments
77 responses to “The Controversy Over Bethesda’s ‘Game Engine’ Is Misguided”
Sorry, Jason, but the Gamebyro/Creation Engine is the cause the 90% of the issues with BGS games. Bethesda is only developer left that is still using it. It hasnt stood the test of time Like the Unreal Engine or the the Quake 3 Engine that is still the engine running the latest CoD games.
Lets also not forget that Bethesda also owns ID sOftware and have themselves stated that they got the guys from ID to help them out when it came to the gunplay in fallout4. Its incrediably stupid that they have not gone to ID and asked to use the ID_tech6 engine thats running Doom, Doom Eternal and Rage 2
Spot on.
Rage 2 especially, the game seems like it handles open area very well. It is definitely just them being cheap and lazy.
I’m not sure “lazy” is the proper word here. I understand people wanting better, but when we’re talking about an industry infamous for its crunch time, I hardly can blame people for wanting to use software and systems they know like the palm of their hand.
The problem is that they are still shipping broken half completed games anyway
You only have to look at Fallout 76 and see it looks not much better than Fallout 3, has the same animation issues, has far WORSE lighting and shadow issues, along with Z-fighting making a return.
I dont care how much they want to talk about changes when they have the same basic problems from over a decade ago.
This article is fluffy nonsense looking for nitpicking on nomenclature rather than seeing the truth. That the underlying systems in place in these games need to be redone from the start or the another engine should be picked up. But of course if this place is anything like my current place of work, we will keep papering over the cracks, then fill that paper with sellys, then cover up that sellys thereby ignoring the root cause all along.
God damn right! I don’t understand how such a long article was written trying to deflect blame, only to demonstrate that it is an engine flaw (whether middle wear or one component is irrelevant) and has been present since Oblivion.
ID games feel different to BGS games. There is a completely different focus.
If BGS were to change engines it would set the next game back by years, would end up creating entirely different bugs and the game may end up feeling more like Doom than Fallout.
I would much rather BGS persist with their own engine and make games like no-one else, even if they have similar bugs and graphical presentation from each Fallout to Elder Scrolls and back again.
an engine change would certainly not mean the game felt different. but what it might mean is the BGS need to revisit there scripting and quest system, which lets be honest, would take some time, but the games would be so much better for it.
but I mean they are a trash company anyway who never fixes their games and releases 6 year old games on new platforms that have the same bugs as launch day 6 years earlier.
I would prefer to wait for TES6 and have it feel like a modern game, rather than giving us another game that looks and feels over a decade old.
‘Modern’ feeling games have less interactivity and are less open.
Be careful what you wish for.
Fallout 4 feels way more modern that Skyrim, which feels more modern than Fallout 3 or Oblivion. They iterate with every release but the style of game that they create is way more taxing on the hardware than a corridor shooter or an ubisoft open world game that has limited interactivity and limited memory of world-state (for example, where you left things). No other game engine is making Bethesda style games, so it would be difficult to say what a Bethesda style game would feel like, if it was made with a different engine.
It seems like they’re adamant about sticking with Gamebryo because Todd Howard likes to work with it and doesn’t want to learn a new system!
except the problem is on a fundamental level. to use your example of chinese vs indian they might both be different but, if the chicken is uncooked or the rice inedible to begin with it doesn’t matter how different they are because they have a problem long before the different styles come into play.
I understand full well how a game engine works but, if you keep building on the many programs and just keep making good enough fixes it bogs down the code and efficiency. something that took 100 lines and had a problem is now 1000 lines to fix when if it had been done right from the start it’d only be 90. it sounds wierd but, it is actually something I’ve come accross before. nothing so extreme but, that’s just to illustrate my point.
to say we are all misguided is a fundamentally flawed way of thinking. yes there are things that people have misunderstood about the engine, what it is and how it works but, that doesn’t mean the criticisms of it are all invalid.
Usually a fan of yours Jason, but this article just reads like a 1000 words of splitting hairs. They say engine, you say toolset. It’s the same outcome. Bethesda fans seem to be upset that the use of the same world building, character toolsets and technologies that run the two, are now producing results that are dated and produce the same buggy outcomes in every game they’re used in. Updating the lighting renderer won’t improve facial construction and animation.
Facial animation and construction has almost nothing to do with the engine. It’s the artists. Aaaall the engine actually changes is optimisation, and the coding language.
And in the formers case, a good team bypasses it anyway
that’s only partially correct. the artist has control over what they do with the animations, but they still have to work within the limitations of the engine. its the same as any aspect. whilst the engine does not dictate the visual style of a game, it still has limitations on what it can and cant do, and the artists need to work within those limitations.
The animation is definitely handled by the engine and due to the animation being handled by the engine the character construction is limited because they need to build within the scope of the engines limitations.
Yeah… no. Fuck that. This is using semantics to ignore the underlying, critical complaint: SOMETHING (whether you want to call it an engine or not) is setting some games apart from others, and causing shitty results.
It Does. Not. Matter. What. You. Call. It. Go ahead, enlightened game devs – tell us what the fuck the appropriate terminology is for the underlying components of a game that are resulting in the same shitty performance.
What you call the stick doesn’t stop gamers from wanting to beat you with it. THAT is the critical point here.
Weary fans of Bethesda’s games aren’t getting upset about nothing. They’re not getting upset about a technically incorrect label. They’re upset about a trend, a proven (multiple fucking times over) issue in all of Bethesda’s games that share whatever name you want to give for the traits they have in common that fans colloquially call an ‘engine’.
Arguing the name does not ease the concerns about the performance.
I think it matters when the name is the main thing fans are going by. The article isn’t about Fallout 76, it’s about the announcement that Starfield and TES6 are using the Creation engine. But they’re also ages out, fans have no way to know what parts of Creation are going to be completely overhauled between now and then. What Creation looks like in a year’s time (or two, or more) may be completely different to what it looks like today – a different engine in all but name, perhaps.
I think it’s perfectly reasonable to be cautious, but I’d never assume the shape of an engine a year or two out as Starfield is, let alone TES6 beyond that.
They said that for Skyrim and Fallout 4. Nothing really changed.
I can pick out a Creation Engine/Gamebryo Engine game just by how it looks and feels to play, same with Unreal and Lithtech engines over the years. Argue over semantics all you want, it’s still got big issues and should be scrapped.
The difference is the UE titles show a very real scale of progress where as FO/ES titles have looked and played largely the same for ten years.
A lot has changed from Skyrim to Fallout 4, the rendering engine alone is a lot better for an obvious one.
The engine has issues, every engine does. But it shouldn’t be scrapped, that’ll leave all of us worse off.
But it still has the same level of “jank” that makes the games all feel the same. These underlying subsystems need to be replaced.
It’s like the engine that the Arma games use – they all feel the same even if the renderer improves with each release, and they don’t feel good.
I think they’re chipping away at it. They overhauled the UI for Fallout 4 (the actual UI, not the Pipboy interface). There’s something different about the physics in F76 but I can’t quite put my finger on it.
Hopefully they improve the other janky-feeling elements for Starfield and TES6, although if I’m being honest I’ll probably grab them both regardless. Bethesda jank was never a game-breaker for me, to my mind it’s a price worth paying for how much I enjoy the games.
I’ll get them too (well, except F76 at this stage…) but I wish they’d just fix the whole physics-framerate cap thing. It’s really frustrating.
That and the lack of ultrawide support have been my two major complaints for F76. Two minor ones are inability to turn off depth of field and no FOV setting. Hopefully all four of them will be fixed over the life of the game, they’re annoying but not enough to put me off it.
this is a case in point. I started with skyrim, so I do not know how long the frame rate / physics problem existed prior to that. but even with skyrim there is a simple ini setting to change the rate the physics updates so that you can play at higher frame rates like 144 without the physics going wild.
we all know Bethesda cares not to fix the underlying problem, but all they need to do is a simple formula to update that physics value based on the current / max frame rate and everyone is happy. none of this locking people to 60fps crap.
Yes, engines are iterative, but when you have the same physics bug in 4 games it also says you don’t bother fixing things much.
Yes! Exactly.
The problem isn’t that we aren’t expecting upgrades on the build. It’s that the engine Bethesda continues to use and “upgrade” has a proven track record of being broken and now dated.
This is a similar situation to Telltale’s releases.
Until there is history of drastic improvement OR another game uses the engine in a vastly different way, fan’s and pundits have every right to be critical of the engine itself.
63 freaking frames per second. #performancetanty
Yeah nah. I’m just to say their engines are a hulking load of shit, because frankly they are. We’re 2018, nearly 2019 and they still don’t have something that can go past 60 fps because of how poorly it’s coded.
Imagine if Quake or any modern FPS fell in a heap if they exceeded 60fps, yet somehow we are expected to tolerate this from them? It’s lazy a shit programming.
The main problem I think is people demanding Bethesda use a new engine perhaps don’t understand that changing engines generally adds more bugs than it fixes, until a few iterations in. Even a third party mature engine will still produce more bugs as developers unfamiliar with the engine come to grips with it. Almost assuredly, if Bethesda switched to UE4 for example, it’d have more bugs than the engine currently has.
(All the major engines in use today have been iterated on for years, over a decade in some cases. Basically nobody produces new engines any more, and even in-house engines are becoming extinct because of the enormous work required.)
None of this is to say there aren’t limitations that Bethesda need to overcome. In current titles, obviously no ultrawide support is still bad, obviously physics tied to frame rate is bad. Physics glitches don’t rate that high really, most every game has them and they’re more a product of the physics middleware than the engine.
That said, both Starfield and TES6 are ages away. We don’t know what Bethesda has on its engine roadmap between now and then, there’s no reason to panic yet just because they’ve said they’re using the same engine.
I doubt anyone has too much cause to be concerned about what Bethesda will fuck up as they familiarize themselves with a new engine – modders have fixed every Bethesda-created bug that’s able to be fixed and isn’t a limitation of the engine.
There’s no reason they wouldn’t do the same again… That is, unless Bethesda are really determined to limit the possibilities of modding for the sake of controlling how it’s monetized. (*cough*76*cough*always-online*cough*)
The caveat: ‘limitations of the engine’ is where the core problem is.
When thousands of modders collectively try to solve certain problems over ten years and come up with the answer, “This can’t be fixed because of limitations with the engine,” then THAT’S the problem that’s inherited with continuing to use the same engine (or whatever terminology you want to use to describe the overall situation). The problem is what can’t be modded away: the ‘engine’ (or toolset if you’re nasty) limitations that they can’t/won’t/don’t overcome.
That’s what people are asking for them to address and overcome when they ask for ‘a new engine’. They’re tired of the problems that not even history’s most rabidly-devoted crowd-sourced mod community could collectively solve.
What modders are capable of doing is only a small subset of what engine developers are capable of doing, it’s would be silly to think that because modders can’t change it then engine developers can’t either.
You don’t need a new engine to solve the problems that currently exist in Creation, everything I’ve ever seen raised and can conceivably imagine ever would be in future can be resolved in the same engine, iteratively. The only question is whether the effort required is more or less than the effort of switching engine, and in 99% of cases the answer to that is ‘less’.
If it was less, why hasn’t it been done?
The real question is whether they will replace, repair or improve the part of the engine responsible for these effects, when they haven’t been able to/haven’t bothered to in the last decade (more, really) of complaints.
The fact that they haven’t, has made these issues synonymous with the ‘engine’ or toolkit itself. It’s starting to look like a Ship of Theseus argument.
Fact is, when people complain about the engine being the same, when you look at the nature of their complaints, it becomes clear that what they’re really complaining about is the fact that Bethesda has never given a shit about fixing certain broken components of that engine. Because it’s effectively the same thing.
If TES6 or Starfield come out with the same core issues that can’t be solved by modders because of ‘limitations of the engine’, the semantic or philosophical argument about what constitutes an engine and why it should or shouldn’t be changed will be utterly moot.
‘Less’ doesn’t mean ‘quick’, it could still take a lot of time and cost.
While the philosophical discussion of what constitutes an engine is interesting, it’s not what the article is about. What I read from the article, a sentiment I agree with, is it’s way too early to start complaining that a game years out from release is going to have the same bugs as one that came out now.
That’s not to say it will or won’t have X bug and Y bug, my stance is that it’s just too far out to be getting upset already. A lot of the engine changed for F76, who knows how much more will change for Starfield.
I’d argue that now is the time for constructively and clearly communicating why this causes concern, and getting assurances that Bethesda will solve decades-old ‘feature not a defect’ greivances that have historically come down to engine limitations, instead of continuing their legacy of tech-debt-induced engine-inadequacy.
Because if they aren’t making those engine changes/upgrades now, getting angry about it in three years time, instead of now, will be entirely too late for them to do anything about it.
I have no problem with people expressing concern about the problems in the current engine. I’ve done it myself, with F76 beta feedback and I wrote an email to them even with some thoughts. I do have a problem with the assumption that future projects will be practically using the engine unchanged from how it is today. It’s not like I don’t understand people’s concern, and like I said elsewhere, caution is perfectly reasonable. I just think people shouldn’t jump the gun yet.
Shrug. When should they complain? When it’s too late to do anything about?
@transientmind I feel like maybe you didn’t read what I wrote. I said I have no problem with people complaining about issues in the current engine. When should they complain about that? Now, along with me. What I have a problem with is people assuming a project a year or more out will have all the same problems the engine has now and getting upset about that now. Complaining that Starfield and TES6 will be trash because they use the Creation engine? Pointless, this far out.
I think the assumption is reasonable, though. Just from pattern recognition. From precedent. From the fact that they’ve never done anything to indicate we should expect otherwise.
It’s like your mate on the basketball team who’s amazing at defence, passes successfully every time, intercepts like a champ, but for fuck’s sake you just simply cannot stop him from unsuccessfully trying to shoot for 3 if he finds himself in possession at the 3-point line. He’s shit at it. He can’t do it. He fails every single time. But he won’t stop trying and failing, no matter how important it is that he not. He’d be the best player on the team if he could just not do that, but he’s got a private trainer who insists that he do that at every opportunity, and he trusts that trainer because it’s paid off in all his other skills.
And now Bethesda has the ball and they’ve planted feet at the three-point line and everyone is groaning and facepalming because for fuck’s sake, we know, we just KNOW that for all the other good that they bring to their games, all those stories, all that freedom, all that rich detail… we’re also going to have to endure the same ice-skating, colliding off rocks into the sky, head-twisting, plasticine-faced, cell-loading, caves-with-doors bullshit again, like we have for the last half dozen games they’ve made for the last decade and a bit.
Maybe they’ll make the shot this time? We have so much pattern-recognition conflicting with that and they’re not saying anything that indicates otherwise, so it’s no wonder everyone is asking why they haven’t got a different trainer.
To clarify: Maybe getting a new engine is not the right answer, but for the love of god, SOME kind of sign that they’re going to address the flaws in their engine would stop people from asking why they haven’t addressed the single point of commonality that is singularly responsible for their games’ persistent and unresolved issues.
@transientmind We’re pretty much on the same page, with your second comment. And I think they have given an indication they’re improving the engine for both Starfield and TES6. I don’t think there’s one kind of singular specific area that contains the bugs (how easy would that be to fix!), and I think they’re also conscious that they can’t promise to fix every bug because if something slips through they’ll get pilloried for it.
That said, I think a generalised “we know there are some issues in the engine, but we promise we’re working on them and we’ve got plenty of time before Starfield comes out to get it right” would be a fine statement.
Me, seeing the article title “Oh no Kotaku, you didn’t go there did you? This is going to be a train wreck”
Reading the comments: “Yep, that’s what I thought”
Wow, that’s a lot of mental gymnastics to try and be right. If an engine has a problem, no matter how small and trivial, then when you use that engine again somewhere else it is going to have those same problems.
To use the ubiquitous car metaphor, if you have a loose part in your car engine that makes thumping noise every so often and then you take the engine and put it into another, newer, car that noise is still going to happen. You might work on that engine and add replace some parts, take some out and add new things but if you don’t fix a problem, it’s going to stay in there or even drag down the performance of everything else because it interacts badly with other changes.
A game engine is the same, the problem may lie in one particular system but the engine is the sum of its whole and that total is going to not be as good as it can be if you have negatives.
This analogy doesn’t really work. If a game engine has a problem, no matter how small or trivial, it can always be fixed before the next game to use the engine, and even on the fly in games that use it currently. That’s the main point in the article – game engines aren’t make-and-forget things, they’re constantly, iteratively being worked on and improved, from the smallest level to the biggest.
The problem Schreier is talking about isn’t Fallout 76, it’s the assumption that in the next year or more until Starfield and TES6 come out that Bethesda won’t fix the major issues present today because they’re somehow intrinsic to the engine – ie. unfixable without replacing the engine altogether. That’s not the case.
Is it though? I mean some of the bugs have existed since Morrowind. Skyrim shares some bugs with Fallout 4. Fallout 76 shares almost all of the same bugs as Fallout 4 – some were fixed, some new ones added but there are a bunch that have been carried over.
Being able to identify a Bethesda game by just looking at its bugs is a bit of a problem. It shows that the problem is systemic and they’re either unable or unwilling to make the fixes that are required.
I’ve never seen a problem in an engine that can’t be fixed given sufficient effort, speaking as someone who has done engine development before.
What bugs from Morrowind are still present?
Physics objects bugging the hell out and flying everywhere? Walking over something and taking damage because it wants to rocket off into space?
Fallout 76 even has a new variant on this…you walk around and seemingly don’t step on anything except solid ground but the game says you’ve just taken damage. Have seen that one several times in videos of the game – once the guy stepped on a nail board and didn’t notice but the rest were bugs.
These are relatively minor issues but they’re ones that Bethesda carries over from one game to the next. I’ve bought Fallout 76 because I like the game but it’s still frustrating that Bethesda don’t fix so many bugs.
That’s why they’re the only dev that gets attacked for the engine they use. If it’s not the engine preventing them from fixing the bugs then it’s their own ability or desire to fix them that is lacking.
I don’t want to sound like I’m making excuses, but this is an area I think gamers don’t understand well: physics is really hard. Nobody simulates it properly, not any of the major physics engines. They all cheat, they take shortcuts to reduce calculation time, and collision physics is one of those areas where shortcuts produce occasional but visible bad results.
I don’t remember seeing a game in the last decade that didn’t have weird physics glitches. Witcher 3 had them, Tomb Raider has them, RDR2 has them in spades. I don’t know what any of these companies’ relationships are with their physics middleware providers, but there tend to be limits on what engine developers can do to tweak the physics layer’s settings to avoid collision problems (which make up the majority of physics glitches in games).
There’s another factor besides ability and desire that’s worth keeping in mind – cost. For the sake of comparison, if we have an arbitrary scale where the cost of a typical bug costs 5-20, properly fixing a physics bug may cost 90. Maybe changing engines costs 120-150, since the entire development pipeline and skill library needs to be replaced, every tool needs to be rewritten, etc. If at the end of the day your budget for fixing things is only 80, you’re going to spend it on fixing the smaller things and leave that 90 (or god forbid the 150) until circumstances change.
I’m not Bethesda. I’ve never worked for Bethesda, I’m not making excuses for Bethesda. I just think that frustration should be directed at the right things, and understanding how engines work, how they’re improved and how game budgets work generally speaking are all important to directing frustration at the right thing.
I think you’ve taken me too literally here. You asked for an example and I gave one. There are others but that was the first one I thought of. I’m not a modder, I don’t live in Bethesda games and play nothing but their games alone, I can’t list 20 common bugs off the top of my head.
The fact remains that these bugs have existed for a hell of a long time and never been fixed. There are a number of Fallout 4 bugs that got fixed by modders that are now back again in Fallout 76. These bugs can be fixed. Some hobbyist game modder in his house can solve them, surely a Bethesda programmer with infinitely more experience can fix the same bug in a cost efficient manner.
I don’t expect any game to be bug free, that’s unrealistic. But I would expect that over 15-20 years of making games with a common engine that the common bugs carried over from one game to the next are reduced each time…not just carried over without changes. I mean if they don’t allocate budget to solving legacy issues then they’ve got a pretty poor development process.
I don’t think I was over-reading you but if I did I apologise. You raised a valid problem, I just wanted to lay out why it’s a problem that’s much harder to fix than a lot of people think. I realise that it’s probably because I’ve been a developer myself, but physics glitches are the kind of bug I’m far and beyond the most willing to forgive, even when it keeps coming back, just because I know how hard it is to eradicate them.
I’d say it’s not even necessarily ‘bugs’, but less-than-desirable design quirks that people have tried and failed to fix because of the engine, or perhaps let’s say the tech-debt attached to it. The inability to meet requirements which would be considered ‘enhancements’ that have never come. What my team’s developers would reject if called a ‘defect’ because it was agreed upon as acceptable on release and met the requirements of the time.
The quirks that come at the boundaries of the world’s ‘cell’ structure, the gradual bloat as changes to cells are stored, the load time increases this results in, the weirdness around animation, the rapid deterioration in performance at higher quality settings owing to how textures and objects are handled, the frame-skipping/stuttering, the famously-reported framerate-bound-speeds, just physics in general (fly, my ragdolls, fly! – that might actually count as ‘since Morrowind’ bugs), the ice-skating slip-and-slide movement (hell if I know what the official term is for it), and god knows how much else I’d find if I were up to date on the modding scene’s greivances.
I largely agree with this, although the prevalence of a lot of the physics-based bugs has gone down over the games. People don’t tend to notice the ‘less often’ part, they just notice the ‘still happens’ part.
But I digress a little. Tech debt is a really difficult thing to manage, it’s probably the hardest thing in any long-term project. As tech debt grows, the desire to start fresh grows with it – a greenfields project where all the old ‘bad’ ways are tossed out and everything gets done the ‘good’ way this time.
But from my experience in both industries (game and business), I can say with complete certainty that’s an enormous trap. Greenfields projects (a new engine, in this case) have an absolute shitload of work associated. Every single part of the development pipeline has to change, every tool has to change, a lot of learned wisdoms have to be thrown out. Even the technical element aside, it’s hard on staff too who have been doing things a particular way for a while that have to learn a completely new way, all under the same cost-efficiency gun.
Most programming legends (Fowler comes immediately to mind) agree that tossing out the old project and starting new is almost guaranteed to sink a company. The standing advice, (keeping in mind that these guys live in the business programming space, not gaming) is to always replace componentry piecemeal, iteratively. While doing that, if you find that a large chunk of code has to be replaced then you do it then, but far more often than not it just takes time and willingness from management to pay off the tech debt instead of setting new feature milestones.
Changing engines isn’t necessarily going to fix all those problems if they still want to produce games that can do all the things the current games do though. It’s just as likely to introduce a whole new set of technical debt.
For example, consider Mass Effect Andromeda, where the dev team switched from Unreal Engine to Frostbite at management’s direction, but ended up needing to reimplement much of the functionality that was important to them but missing because it wasn’t needed by DICE’s FPS games. They might have escaped some of the limitations of the old engine, but they were also throwing away solutions to many of the problems encountered while working with the old code.
When implementing those missing features, there’s also the risk that you’ll just reproduce the same problems because you used the same algorithms. Of course, if there is an alternative approach that avoids the pitfalls, you need to ask why that couldn’t have been integrated into the existing engine.
In third person mode, crouch down on the edge of an object (a platform, a rock shelf etc) and using your bow/crossbow/sniper rifle (ranged weapon), aim slightly downwards towards a target you can clearly see, and fire the weapon.
Result: the projectile will collide with the object you are standing on, even if the weapon you are using has a completely clear line of sight to the target.
Now, without moving, switch to first person mode, and fire again.
Result: the shot will work fine.
This issue has been reported to death, has been in every game using this “toolset/engine/middleware” and is super annoying.
What’s more, issues that they fixed in Fallout 4 have returned in Fallout 76 – it seems they didn’t backport fixes.
Another annoying thing with Gamebryo/Creation Engine is that gameplay is tied to frame count, not time. EG: Physics, animations, logic. The creation engine is frame count based and not time delta’ed, for many elements. It’s a big job to overhaul an engine to correct that (rendering speed hasn’t been fixed frame rate for over a decade on most systems now, basically since PAL/NTSC became deprecated).
I fully expect to be playing Starfield and ES 6.0 and still see these same issues.
You are correct though, they could fix them without having to change to UE4 or whatever. But this article about how the issues are not with the “engine/toolset” is just spin.
There’s a big “Maybe” attached to all of that though. Any problem can be fixed yes, but the question is always “What is the cost?”. A fundamental part of the engine may be far too risky to fix because it is used by too many components so you put in a workaround or a hack. Or, you may fix it but then find out it introduces a new issue somewhere else but leave it in because it has low impact. If this continues you keep building in workarounds for those workarounds and hacks upon hacks. Eventually though you have to decide whether you can keep maintaining a big hodge podge of fixes and new code or just create a new engine entirely.
I’m a software engineer myself, specialising in testing and automation, so I have first hand experience with this kind of situation and understand that not everything is an unfixable problem. Sometimes you can keep that ball rolling somewhat smoothly but many projects I’ve worked on have ended up being rewritten from near to scratch at some point because there were too many persistent issues and maintenance became too expensive.
Cost is the ultimate factor, really. Game engines are sufficiently complicated as to make unit testing practically impossible too, so it’s hard to find regression problems. Still, a new engine poses a much higher cost than a lot of people realise, such that fixing what’s broke in what you have is almost always the right choice. Not always, but almost always. And a lot of the time when you think it’s the right choice to change, you’re wrong 😛
Unrelated, what language do you work with? I always enjoy hearing from fellow developers.
Eh… Their games have always looked and felt like shit with lifeless characters and giant worlds with absolutely nothing of interest in them. They have a lot of catching up to do after RDR2, The Witcher 3, BOTW and even the recent AC games.
It’s either a crap engine, or crap development work and ethos…so this article highlights that it is crap development work, and not the ‘engine’. Same bugs for over a decade, so it’s one or the other, you can’t have an each way bet.
Here’s a key point about game engines that is very pertinent here.
Unity is used to make a wide variety of games that look absolutely nothing alike when you compare the two. The games all have bugs. These bugs, when you compare them between even two very similar games are usually different bugs. they’re to do with the coding of each individual game, and not any underlying base structure. Note, I said “usually”, some bugs may be common between both and not due to the specific coding of each game.
That’s where you come to Bethesda. Whether its Fallout 3, 4, 76 or Morrowind, Oblivion or Skyrim each of these games have some very similar bugs. These bugs are not specific to the coding of each game (unless they literally copy and paste the same huge chunk from one to the next), they’re due to the underlying structure, the toolkit or perhaps you might call it an “engine” that is used in all of these games. Yes, the underlying structure may have different bits bolted to it and those bits may completely change from one game to the next but there’s a core section that remains the same. It’s to the point that you can identify Bethesda games by their common bugs.
I have played all of those Bethesda game I just listed, I have bought Fallout 76, gladly. I am aware of the bugs, I take it as granted that they will be there, I will have fun regardless. At some point though a line has to be drawn that Bethesda’s customers expect more than this, they expect these bugs to be eliminated. If those bugs are so far embedded in the engine that they can’t be fixed though then they need to use a new engine.
Where’s the line though? For some they’re refusing to buy Fallout 76, for me I’ve bought it but the next game might be where I draw the line.
It’s the Engine of Theseus. 😛
The games absolutely share code. A lot of the core mechanics in each of those games are almost the same (movement, interaction, dialogue).
It’d be poor business to start from scratch without a significant need or demand (I’m talking from a business perspective not from fans) – which just might be the reason they’re sticking with Gamebryo. They don’t think the issues constantly being raised are equal to the cost of getting rid of them.
Rockstar should do the same. RDR2 is awesome but it’s just a GTA5 Western, plus there will be a remastered RDR2 for next gen 🙂
There’s no good jumping in point, but firstly: check @transientmind and the @zombiejesus thread above. There’s some good technical discussion in there about problems with development that people often overlook or misattribute, which is really the core of what Jason is getting at here.
Similarly, and going back to @thyco’s point: you couldn’t feasibly use the iD tech engine for a game like Fallout. A lot of the interactable iterms that people take for granted in a Elder Scrolls/Fallout game are built into the geometry of Doom’s levels. They don’t have characteristics of their own that would be tracked. Where’s the inventory system?
This was a problem that Bioware ran into with Inquisition. Frostbite looks amazing! It sounds like a wonderful engine. Except when they actually started using it, it had *none* of the tools that were actually necessary to build out a third-person RPG. It wasn’t fundamentally built for the types of games Bioware made. It was made for shooters; it was made by DICE.
People often look at an “engine” and take a super reductive view on what that means. Engines are gargantuan beasts. They’re hard to wrangle. There’s a reason companies employ tool programmers and other specialists that just continually work on ironing out the core parts of an engine, or adding features to it over time. It’s insanely hard, and things break all the time.
I bring all of this up because there are individual bugs – and physics got mentioned earlier – that may not actually be a problem of the engine, but more how certain objects interact and how all of those dependencies clash. @xenoun makes the point perfectly: sometimes it’s not actually the engine, but the exact code.
Now, what happens when people are still dealing with cheese wheels flying around at mach 10?
Obviously, people are right to get a bit shitty at that. Especially if it keeps cropping up time and time again. But the major point here that I think everyone can agree upon is isolating precisely what should be fixed, and making sure criticism is directed at that rather than the engine more broadly.
Saying the engine itself is crap or not worth it is too reductive. And it also ignores the immense work it bores that other engines simply cannot do efficiently without a massive amount of behind-the-scenes work, or just aren’t suited for in the first place. There’s always a good reason or ten why studios prefer in-house engines – it’s not a graphics or performance thing, but gamers tend to see everything through the lens of those two areas first, because it’s what is most immediately apparent.
Or put another way: using the Quake 3 engine made sense when everyone was riding the mass-market military shooter wave. But does that mean you could easily retrofit it to work for a third-person RPG? Probably not.
When people are handing over coin, and see the same, the exact same, problems time after time, it isn’t good enough. Particularly when the head developer makes condescending jokes about the bugs. No amount of making apology covers Beth’s attitude.
I think half the problem is that the term “engine” can apply to many different things. I’m used to terms like “Rendering engine”, “Physics Engine”, and “AI Engine” which are the engines responsible for those particular aspects of a game because they are complex beasts unto themselves. When I see the term “Engine” on its own though I take it as the game engine, the core set of systems and logic that tie game state to player interaction to rendering to physics to AI.
Other people though see the term “Engine” as describing all of those individual engines as a collective.
While I agree, because engine dev is hard and off the shelf solutions don’t really work for these kinds of games, none of that changes the fact that the Creation Engine is inheriting bugs with each iteration and has some fundamental problems that need to be addressed. Yes, it’s hard and costly – but physics being tied to a strict expected frame rate (and still bugging out in odd ways) etc should be fixed. Oblivion launched in 2006 – it’s 2018 and similar bugs are still manifesting, the same limitations exist. At what point do they admit that it’s now fundamentally flawed and dig in to fix it?
Yes, it’s hard etc and yes it might change by TES6 but Bethesda have a poor track record over the past decade(!). And the article trying to use semantics and split hairs is a bit silly.
The Bioware example is why I think BGS should persist with their own engine.
I want their games to be different from ID / Epic / Dice.
You can add a spoiler and new paintjob to a 73′ Datsun, but it’s not going to turn it into a Tesla. The actual engine is still the same under the hood. You might have added a new steering wheel, new suspension and maybe even new filters and spark-plugs – but it’s still the same engine. It will perform essentially the same.
The bugs, glitches and issues players are experiencing with F76 have been present for what, 3 or 4 different Beth games now? This means the underlying issue – the ACTUAL engine coding and construction, isn’t being fixed. Primarily (I assume) because of cost – both to fix it, make a new engine, or bootstrap existing in-production projects to whatever the new solution is.
Rolling out a new engine is a huge, expensive and time-consuming undertaking, I completely understand that (and have worked in that area – it’s a big challenge.) It’s still flabbergasting however that Beth isn’t using the dev cycles of their two huge new games to roll out something new and optimised for the current tech climate and importantly, something they can clearly and loudly say “We’ve addressed those problems, and we’re excited about the new opportunities!” Instead they seem happy to continue bolting on spoilers and paintjobs and ignoring that weird knocking noise coming from under the hood.
A very, very clear example is Beth limiting frame-rates to address performance based physics. This is something I would’ve scolded my game dev students on, because it’s not addressing the actual problem – it’s just a band-aid and not a player-focused one at that.
Getting down to the code that actually governs that is deep and specialised, but it’s also something Beth have had years and years to examine.
Other companies like Epic, Unity, Crytek, iD and EA DICE seem to understand the need for innovation. Take a look at how silky smooth DOOM is on just about every platform it appears on. Now take a look at the terrible performance issues of F76.
There’s no excuse when they’re helming near-billion dollar franchises to ignore a need for engine development to keep pace with players expectations, and competitor products.
Thank you. People keep saying that fixing the bugs if they’re so core to the engine would be expensive and difficult, and that switching to another engine altogether wouldn’t be viable. Most likely all true, of course. But goddamn it, as a multi-million dollar company that is expected to compete against other multi-million dollar companies that manage to tackle the issues in their games, without the same ones reappearing consistently?
The expense or the effort doesn’t matter. It’s just what’s expected, especially after the efforts of other developers. People can excuse it all they want, but the fact is that that just isn’t gonna fly in a market so oversaturated with functional, quality games.
Engine is terrible most of the times. Even skyrim on ps3 was ass. Terrible models, effects. Fallout 4 looked ok at times, but alot of it was ugly and only saved by the missions of getting from a to b, as motivation.. I really am concerned now for elder scrolls 6. Who wants to play the same games again.
The problem is havoc engine
Nothing wrong with creation engine the problem is mainly to do with havoc engine just need to replace havoc.
Bethesda has to much invested in creation engine.
6 games across 2 IP’s and people debating fix vs clean slate GTFOH. Almost 13 yrs since Oblivian released, plenty of time and revenue to address fix ghosts in the machine. But hey they just upped stash box even tho previous limit blamed on “engine limitations”, and what do ya know, new bugs.
If they can’t get Gamebryo/Creation engine right by now, imagine how much worse trying to break in a new and untested engine, or even an established engine their engineers aren’t familiar with would be. I say for all it’s flaws, they should stay with what they know, between patch cycles and modder intervention it all winds up essentially playable in the end.