JumpCancel
Banned
You nailed it!
Catherine runs well enough to justify Gamebryo as a functioning engine. It all comes down to what we define a "Bad" engine as.
Un-optimized? Poor performance? Restrictive?
-
Hey Guest. Check out your NeoGAF Wrapped 2025 results here!
Is there such a thing as a bad game engine?
- Thread starter nkarafo
- Start date
You nailed it!
Catherine runs well enough to justify Gamebryo as a functioning engine. It all comes down to what we define a "Bad" engine as.
Un-optimized? Poor performance? Restrictive?
JordanN
Banned
Silicon went on to steal the code from Epic and got defeated by them in court.
So either they were bullshitting or they deliberately wanted performance issues (lol).
If that is a case, the engine is bad for their intended use.
If developers have to make moving train by replacing players arm with train cart, they might have an engine that limits creativity.
DayZ - Instead of using the more efficient arma3 engine, they decided to stick with the buggier arma2 engine to save time. In the end they just spent more time trying to get the engine to do things it couldn't do.
Natural Selection 2 - Instead of continuing development on the Source engine, they switched to a homegrown engine plagued with bugs and performance issues that delayed release by years, and never recovered their player base.
Both instances were terrible gambles by the team leads. It's sad to think what these games could have been with a decent game engine under them.
Natural Selection 2 - Instead of continuing development on the Source engine, they switched to a homegrown engine plagued with bugs and performance issues that delayed release by years, and never recovered their player base.
Both instances were terrible gambles by the team leads. It's sad to think what these games could have been with a decent game engine under them.
Crossing Eden
Hello, my name is Yves Guillemot, Vivendi S.A.'s Employee of the Month!
My mistake. This is what I was referring to when I said they need to ditch and/or overhaul it completely.
Jesus christ.
I'm not sure we should judge engines based on the results, there could be many reasons for a bad game and the engine could not be one of the reasons.
An engine should be judged based on features, ease of use and optimization, various Unreal Engines are probably some of the best, that's why they are so popular, don't blame the engine, blame developers.
An engine should be judged based on features, ease of use and optimization, various Unreal Engines are probably some of the best, that's why they are so popular, don't blame the engine, blame developers.
kruis
Exposing the sinister cartel of retailers who allow companies to pay for advertising space.
What happened was that SK started fixing UE3 first and then they created their own engine using parts of UE3. I assume they had put so much of their own code on top of UE that it became difficult to separate the two when they started to build their own engine. That's what killed their lawsuit in the end. If they hadn't done that and really built their own engine from scratch without using a single bit of UE3, they might have had a case.
I do believe their claims that UE3 was a disaster to work with. They were working with a beta product. At the time Epic was fixing UE3 while at the same time working on Gears of War. That combination was a good way to fix the performance problems in UE3 and when GoW was released in November 2006, most of the problems were gone - if you wanted to create another corridor shooter that looked exactly like Gears. And it sucked for any company who had licensed UE3 in 2005/6, because they were trying to create games with an engine that didn't perform like Epic had promised. Epic gave the source code for Gears to their UE3 licensees as an apology for problems in the past.
PyroPaperPlanes
Member
Even the battles ran poorly for me. The framedrops were incredibly jarring for a handheld console.
All of these except I don't what the hell Bitsquid is. Add Id Tech 5 to the list of shitty engines.
neurosisxeno
Member
It's called the Spark Engine, and their main coder literally wrote it in a weekend. They had a full development blog post about it. Most of the delays weren't engine related, they wasted a lot of time trying to get things to work in Source Engine, then afterwards had to redesign a lot of stuff and rework it to fit within the Spark Engine. As far as I know the game actually plays really well these days--they just rushed to get it out and had to cram to fit the new engine which allowed a lot of promised content to function properly.
As much as I love From Software (and have for 20+ years, they're probably my favorite devs of all time), it's amazing that their games run so poorly on their intended hardware.
From has *always* been like this, going back to even Kings Field days. Their games run like absolute garbage.
Just about every title has had large sub 15-20fps dips, right up to the Souls/Bloodborne era. Honestly, I think they're probably still using a modified version of their original playstation engine to this day..
From has *always* been like this, going back to even Kings Field days. Their games run like absolute garbage.
Just about every title has had large sub 15-20fps dips, right up to the Souls/Bloodborne era. Honestly, I think they're probably still using a modified version of their original playstation engine to this day..
Gormenghast
Member
At the risk of losing my mind again...
These threads are ridiculous because they have a way of self fueling without EVER reaching any sort of conclusion or furthering the point. They are like "troll, the forum thread".
So. In order to even say a thing about this, we have to DECIDE what a "bad engine" even IS.
What is good and bad re: game engines?
For example:
1- It can be how flexible is the overall toolset, how many different games you can do with it. An engine could be excellent to make a great FPS, for example, but it would be insane to make a 2D sidescroller on it. Or an engine could be excellent to quickly make 2D tiled game, but of course you wouldn't be able to make a 3D game with it.
In this specific case, Unity is EXCELLENT. Because it's powerful, flexible and well documented. It's enough high level to be easy to develop on, but not too low level to require you to put there the same amount of work of starting basically from scratch. On the other side if you were making a purely 2D game it would be still better to lean on something like GameMaker, because it's more suitable for 2D than Unity, and more accessible and faster to develop on.
2- Or maybe we consider a game engine not as a toolset, like Unity, but a complete system that runs a game. Like an open world built by the TOTALITY of its systems. It means we include low latency, responsive controls, great AI with realistic, emergent and simulated behaviors (think Dwarf Fortress), a well simulated open-world, weather systems, day/night cycles, scheduled NPCs that react to everything around them, excellent, responsive UI. And in the end also realistic, life-like, complex graphic fidelity with complex lighting.
From GTA4 to Skyrim or Fallout we actually don't even have that many games that attempt this Big Picture of a good game engine.
3- We purely measure a game engine by its graphic renderer. The amount of complex graphic it's able to put on screen matched with the actual performance. How many shaded polygons with lights and shadows it can move on a fixed hardware.
In this case an engine like Unity provides some good aspects, like physically based rendering, so the potential to match the easy of use with actual advanced, modern graphic. But due to the fact Unity is a shortcut of development and built as a generic engine working on a wide range, it also can't even remotely match the raw performance of a custom-based engine that is accurately optimized for a single game.
The ease of use always comes at the cost of some flexibility of what you can do, and if you spend a lot of time rebuilding Unity own system for more custom purposes then you also lose their advantage.
But in THIS case, if we measure purely the graphic performance and quality Unity isn't even remotely close to an AAA custom engine. You can't make The Last of Us or GTA4 in Unity and expect the same level of performance of their custom engines. The same was with Gamebryo. And both of these have their own problems by having to force co-exist different engine parts not so well integrated. Like SpeedTree being its own engine integrated in Unity or Gamebryo. And so on.
It becomes a really big engine where modifying one part might break something elsewhere. And maybe a specific game doesn't use that part, so it's good, while another suddenly runs like crap because the new change worked not so well with the different structure. Maintaining a flexible engine like Unity is HARD because what would be ideal for "x" game might be not good for "y" game, and your goal is to support both, so you are limited in what you can do, when optimizing.
If pure performance and graphic fidelity is the primary goal, Unity is not a good choice. Gamebryo is/was not a good choice. Middleware always comes at a cost. UE4 too.
A game engine is a tool. Yes. It can be a good tool for a single type of game, and be good at it. or maybe it can be a very flexible engine like Unity.
But Unity, if we consider purely performance and advanced graphic is: "Jack of all trades, master of none".
So was Gamebryo.
P.S.
Catherine didn't put on screen sweeping vistas with a long clip plane, complex lighting and whatnot. It was a relatively graphically simple game. Every single engine in existence CAN run at high fps if you limit what you put onscreen. Catherine and Skyrim, even if they might have engine parts in common, not even remotely have to deal with the same graphic fidelity.
Jacksinthe
Banned
Developer colors outside the lines of the engine:
THIS ENGINE SUX OMG!
Develoepr keeps within an engine's constraints:
Nobody notices.
THIS ENGINE SUX OMG!
Develoepr keeps within an engine's constraints:
Nobody notices.
LordRaptor
Member
I guess you're ignoring that Rift (a once semi-popular large scale MMO), Bully (an open world GTA style title) and Civilisation IV (a highly complex systems driven title) also all used Gamebryo.
But its Gamebryo that was shit, not Bethesda.
None of the things you described are inherent traits of the engine. It's just how developers choose to do things. Everything looked like it was covered in slime because the industry always mindlessly exploits whatever new visual tech they are given like a kid who just discovered the lens flare in their cracked Photoshop. Shiny, normal-mapped surfaces is what looked "next gen" at the time.
TheDinoman
Member
As someone who played tons of TF2, I think Source can be pretty bad.
Wonky hit detection, wahey! I love trying to hit someone with the Eyelander but it didn't damage them even though I'm like right near them. It's like everyone has these invisible force fields around them that keep flickering on and off.
Wonky hit detection, wahey! I love trying to hit someone with the Eyelander but it didn't damage them even though I'm like right near them. It's like everyone has these invisible force fields around them that keep flickering on and off.
Zombegoast
Member
The engine Darksiders use is pretty bad. There are zero graphic options and Darksiders 2 had stuttering issues. The "Deathinitive" edition was even worse with the Ps4 and XBO locked on 30fps and the PC version is practically broken and haven't updated the game since November.
Gormenghast
Member
Again, if we measure graphic performance I don't know how you can consider Civilization IV an example.
"System driven" doesn't mean much, engine-wise, since Gamebryo can't write game rules. It doesn't come with a "strategic tool". You spend time in Civ4 with the game mostly paused, while you interact with its UI.
The engine doesn't do much. It's mostly UI. The reason why the game is so moddable is that its engine does essentially very little.
And neither Rift nor Bully match graphic fidelity and scope of Skyrim. Not even close.
The more moving parts your game has, the more the chance to break somewhere. Skyrim has to have actors navigate ALL meshes and conditions with an universal AI and animation system. That's why it glitches. It doesn't almost never have custom specific animations to use.
TheMan
Member
the vast majority of neogaffers (myself included) don't know enough about how game development works to say what is the result of a bad engine vs what is the result of poor optimization or design choices. I'd bet that even engines regarded as "bad" have been used to make games that run just fine and people here don't even realize it.
Pantaghana
Member
While it looks bad, it's just rather clever asset reuse and something that developers do all the time.
After all, if it's a one off event in a 200-hour game, does it really make sense to make an entirely new object class whose only job is to play a single animation?
No, make the object a helmet, put it on an invisible NPC and have them run around, silly but efficient.
While I'm not a big fan of Gamebryo or its successor Creation Engine, they are an excellent platform for mods, that alone justifies their use in my opinion.
When commenting on New Vegas, a dev from Obsidian commented that despite its problems, no other engines are suited to being modified to the extent that Elder Scrolls and Fallout fans push them to.
LordRaptor
Member
You are frankly delusional if you think an MMO is less complicated than a single player title.
Rift has to handle everything that Skyrim does, but for hundreds of players concurrently.
MMOs are - by far - the single most complicated and difficult to develop type of game available on the market.
It's easier to talk about these things if we differentiate between tool-sets and engines. Then, whether an engine is good or not depends solely on the quality of the finished product. Of course that depends on how well the engine satisfies the needs of the game code which is arbitrary across all games. And that in turn depends on the programmers ability to understand the problem he is trying to solve.
So we have a situation where a generic engine is never going to offer good efficiency across all types of games. A possible solution is to modify the engine source code. This however creates another problem, you now need a solid understanding of the engine and it's possible the way it was designed will work against how you need to implement your code for it to be efficient. Shoehorning one system into another system that isn't designed to support the structure you need can often be more complicated than writing something from scratch. It's going to depend a lot on what the game needs.
It doesn't matter how well designed and optimized unity or ue4 is, they won't be able to solve these problems for you.
Basically the point is that writing software is a craft that is difficult to master, not a high level fantasy land with abstract platforms where things get done for you and magically work well. Code runs on hardware and its up to the developers and programmers to solve that problem.
Now, it's perfectly ok to take a hit on performance if things are already fast enough but most of the time with games that isn't the situation we find ourselves in today.
On a side note, I personally think a lot of indie devs could benefit from doing their own engines provided they have a decent programmer on their team. Most indie games aren't going to need 95% of the features and tools a generic engines provide and with just a few members on the team the overall understanding of the engine is going to be much higher with something built from the ground up.
So we have a situation where a generic engine is never going to offer good efficiency across all types of games. A possible solution is to modify the engine source code. This however creates another problem, you now need a solid understanding of the engine and it's possible the way it was designed will work against how you need to implement your code for it to be efficient. Shoehorning one system into another system that isn't designed to support the structure you need can often be more complicated than writing something from scratch. It's going to depend a lot on what the game needs.
It doesn't matter how well designed and optimized unity or ue4 is, they won't be able to solve these problems for you.
Basically the point is that writing software is a craft that is difficult to master, not a high level fantasy land with abstract platforms where things get done for you and magically work well. Code runs on hardware and its up to the developers and programmers to solve that problem.
Now, it's perfectly ok to take a hit on performance if things are already fast enough but most of the time with games that isn't the situation we find ourselves in today.
On a side note, I personally think a lot of indie devs could benefit from doing their own engines provided they have a decent programmer on their team. Most indie games aren't going to need 95% of the features and tools a generic engines provide and with just a few members on the team the overall understanding of the engine is going to be much higher with something built from the ground up.
As others have said, I think a lot of so-called "bad" engines only have that reputation because sometimes developers try to use the engine in ways that were never really intended.
For example - a lot of people shit on idTech and point at The Evil Within being a technical mess, ignoring the fact that Wolfenstein looks and runs great on the same engine. TEW was a mess because they took an engine that was mainly designed for FPSes running at 60fps and tried to shoehorn in their crazy lighting system while locking the framerate to 30. It's like trying to fit a round peg in a square hole.
Gamebryo is a weird case, largely because it sounds like Bethesda has made so many internal modifications and changes to that engine on their end that its hard to say whether it's the core of the engine at fault or Bethesda's iterations on it.
The only engine of the top of my head that I can think of that is truly, genuinely shitty without any caveats is The Telltale Tool. It's a buggy, janky mess that appears to be put together with duct tape and Kevin Bruner refuses to entertain the idea of a new engine because it's his baby. For all of the complaints about Unity, I would much, much rather TTG use that than their current engine.
For example - a lot of people shit on idTech and point at The Evil Within being a technical mess, ignoring the fact that Wolfenstein looks and runs great on the same engine. TEW was a mess because they took an engine that was mainly designed for FPSes running at 60fps and tried to shoehorn in their crazy lighting system while locking the framerate to 30. It's like trying to fit a round peg in a square hole.
Gamebryo is a weird case, largely because it sounds like Bethesda has made so many internal modifications and changes to that engine on their end that its hard to say whether it's the core of the engine at fault or Bethesda's iterations on it.
The only engine of the top of my head that I can think of that is truly, genuinely shitty without any caveats is The Telltale Tool. It's a buggy, janky mess that appears to be put together with duct tape and Kevin Bruner refuses to entertain the idea of a new engine because it's his baby. For all of the complaints about Unity, I would much, much rather TTG use that than their current engine.
The Baron Count
Member
Yeah, well they kind of needed to.
From what I recall, Tim Sweeney initially wanted to go with a fully real time lighting system like Id Tech 4. I've read interviews from 2004 when UE3 was first being shown where he said there was no more light maps.
Clearly that changed at some point, though you could still choose to use real time lighting with the engine if you wanted, most opted for lightmaps again, including Epic.
The thing is, their lightmap system only took into account direct lighting, which lead to developers using alternatives like Beast so they could get global illumination.
This was remedied by Light Mass though. And it did make for some gorgeous results.
God, yeah. I forgot about Tell Tale's junky tech. It's unbelievable that they stubbornly cling to it.
The Id Tech 5 thing with TEW is still confusing to me. Obviously it was designed with a 60FPS game with baked lighting like Rage in mind, but Carmack was talking about Doom 4 being 30FPS with more dynamic lighting back in the day so I would think it would be more flexible than it's been shown to be.
You're damning an entire engine because of one game? Do the developers bear no responsibility?
I wouldn't say Luminous is in the same state as Panta Rhei. We've actually had a Capcom dev working on PR say that whatever problems they've tried to fix, it's a case of going 1 step forward 2 steps back for them. Both Deep Down and XV are being used as the base game to develop their respective engine, but one is a f2p random dungeon crawler with corridor pathways, the other is a massive open world game. It's been 2 years since they've shown anything from Deep Down whereas we're seeing continual development progress on Luminous which is actually shipping that huge game this year.
CC2 using UE4 for VIIR makes sense, they can use that experience to help them on their own future titles using UE4, Luminous won't be an option to them. And If Luminous were to be co-developed alongside KH3, FFXV and DQ11, they'd be making the same mistake twice and we'd end up with another Crystal Tools situation.
Square on Crystal Tools said:
SoulUnison
Banned
Gamebryo/Creation Engine.
Possibly Unity.
Crystal Tools and Luminous seem to take far too much manpower and time for a final product to come together.
Possibly Unity.
Crystal Tools and Luminous seem to take far too much manpower and time for a final product to come together.
Dictator93
Member
We know why that happened. Consoles happened.
Lys Skygge
Member
Whatever engine Destiny uses. The graphics look nice, and performance isn't an issue, but I recall a developer saying how much of a bitch it is to work with and causes a lot of headaches.
The Baron Count
Member
Yeah I get that but it seems weird to me that Epic would over estimate what the next generation of hardware could handle by that much in 2004.
It's not like the 360 wasn't powerful, it was plenty powerful for it's time. But having a real time lighting system like they were proposing on the scale that they were proposing it at seems pretty nuts.
Brawly Likes to Brawl
Member
I don't even want to know how much money SE blew on this shit
Dreamworld engine, inhouse engine by Funcom, used in Age of Conan, Secret World, Anarchy Online etc.
Based on Secret World they seemed to have cleaned it up a lot, but Age of Conan seemed to suffer to no end with it and they basically froze AoC development for nearly a year one or two years after launch in order to hunker down on engine improvements.
IIRC there had been comments made by the AoC devs about their issues working with it.
Since it was the Anarchy Online engine I have to presume a lot of the trouble just came from endlessly adding to an engine that probably was not cleanest/most well thought out to begin with and developed in house.
Based on Secret World they seemed to have cleaned it up a lot, but Age of Conan seemed to suffer to no end with it and they basically froze AoC development for nearly a year one or two years after launch in order to hunker down on engine improvements.
IIRC there had been comments made by the AoC devs about their issues working with it.
Since it was the Anarchy Online engine I have to presume a lot of the trouble just came from endlessly adding to an engine that probably was not cleanest/most well thought out to begin with and developed in house.
There are many aspects to an engine which could go "bad". Just as important as performance, for example, is usability. In that department, Destiny's engine is notoriously awful.
The Telltale Tool seems to suck on both of those fronts
The Telltale Tool seems to suck on both of those fronts
FTFY. Most of its issues come from its age and how much shit they've piled onto it.
Well "bad game engine" is just too general.
I choose to criticize engines based on how they allow the developers to preview and test content in real-time, as well as how good that content looks and plays.
-Destiny : Heard that to move a simple asset in their editor would require the entire map or chunk of the world to be reloaded.
-IDTech 5 : It takes hours to compress megatextures. Same for loading a map.
The megatextures tech wasn't mean for the GPU generation it was released on (Streaming issues-texture pop-ins and Editor issues )
I choose to criticize engines based on how they allow the developers to preview and test content in real-time, as well as how good that content looks and plays.
-Destiny : Heard that to move a simple asset in their editor would require the entire map or chunk of the world to be reloaded.
-IDTech 5 : It takes hours to compress megatextures. Same for loading a map.
The megatextures tech wasn't mean for the GPU generation it was released on (Streaming issues-texture pop-ins and Editor issues )
I agree. People are confusing bad code or bad stylistic choices with bad engines.
beat
Member
Bully was originally developed on a modified GTA3 engine, which was Rockstar proprietary. I hadn't known till now that the ports were on Gamebryo.
Vital Tundra
Member
The engine thy used for The Walking Dead was abysmal.
LordRaptor
Member
I was always under the impression that GTA3 was Renderware
MarcoZombieCannon
Member
Gamebryo, hands down. We've been dealing with its issues ever since Morrowind. And don't buy their bullshit, Skyrim and Fallout 4 just iterated on it. Same shit in a different PR box.