That should at least provide enough time for Bethesda to decide whether or not they want to un-SNAFU themselves. After having played the exceptionally polished The Outer Worlds, the idea of...
That should at least provide enough time for Bethesda to decide whether or not they want to un-SNAFU themselves.
After having played the exceptionally polished The Outer Worlds, the idea of Bethesda pulling off a "space epic" with Starfield is fairly hard not to scoff at. I'm already picturing purple planets, melting faces and spaceships flying backwards.
Yeah, they've gradually dumbed down TES. I've been with the series since TES III, and the amount of RP in the RPG has decreased. Granted, they've done good things to the mechanics (in some areas...
Yeah, they've gradually dumbed down TES. I've been with the series since TES III, and the amount of RP in the RPG has decreased. Granted, they've done good things to the mechanics (in some areas anyway) and Skyrim is a very good game, it's just not long until all that's left is a theme park. Similarly for fallout. FO3 and NV had reasonable amounts of RP elements and FO4 just had.... very little thereof. It's just super hard to make your character memorable in either series' most recent installment. FO4 had some mechanical aspects I really enjoy - settlements show promise but lack a bit of polish, and I just love tinkering with my guns.
I'm really guarded when it comes to Bethesda now. I had preorder levels of good will back when Skyrim was coming out. I waited for reviews on FO4, figured it was fine, and while it wasn't a bad game at all, it just wasn't what I wanted or what I expected at all. I think I'm going to have to wait a fair while with starfield and TES VI, just to make sure it's the kind of game I want Bethesda to make. What kind of a consumer would I be if I were to buy a game not because of what it is but because of what I want it to be...?
If you haven't and you have a relatively beefy PC you should look in to Ultimate Skyrim. It's a pack of 250 mods built around the Skyrim Requiem mod which removes all the levelled enemies (so you...
If you haven't and you have a relatively beefy PC you should look in to Ultimate Skyrim. It's a pack of 250 mods built around the Skyrim Requiem mod which removes all the levelled enemies (so you can't just wander up to a giant at level 2 and have a chance for example) and generally makes the game a lot more realistic. It does incorporate mods which require you to eat and drink and sleep which may be a turn off but you should be able to just disable those during the configuration.
They provide an automated tool which if you have a Nexusmods premium account can just slurp the whole lot down and install them so it's less painful than it sounds.
The scope of The Outer Worlds and a typical Bethesda game are very different. I don't think they are as comparable as people are saying. Environments in The Outer Worlds are completely static....
The scope of The Outer Worlds and a typical Bethesda game are very different. I don't think they are as comparable as people are saying. Environments in The Outer Worlds are completely static. There is almost no physics to speak of, and it's completely missing most of the systems Bethesda builds into their games to create emergent gameplay and narrative.
I don't mean this as a criticism of the game, it just has very different design goals than the games people keep comparing it to. I think the game has much more in common with earlier BioWare games like Knights of the Old Republic. The similarities to Bethesda games are largely surface level (first person view, Fallout-esque satirical dystopia).
I'm really "eh" on Bethesda after Fallout 4 (which was "fine" but forgettable) and Fallout 76 (which I will not spend my money on). I find it hard to express specifically what threw me off, I...
I'm really "eh" on Bethesda after Fallout 4 (which was "fine" but forgettable) and Fallout 76 (which I will not spend my money on). I find it hard to express specifically what threw me off, I think it's just a culmination of things that sort of sucked out the je ne sais quoi and made it harder to immerse myself in it. I don't mind a fun action game or streamlined RPG elements, but FO4 felt to me like diving into the Uncanny Valley of Fallout.
The writing and the dialogue system were also kind of a shame. Although solid dialogue is not something I expect as much out of an Elder Scrolls game, it still indicates more generally that their target audience has shifted. It might not be "for" me any more, which makes me a bit sad.
At the same time, though, there are more open-world games than there were at the time Skyrim was released! Bethesda were pioneers in many ways, but now other companies have iterated on their formula and there's a fuller market for open-world exploration games. I'm not going to swear Elder Scrolls off completely or anything, but I'm also not chomping at the bit to get it like I was when it was first announced. I haven't even started Witcher 3 yet because I'm still working through Assassin's Creed Odyssey, which I only picked up after getting the trophies I wanted from Kingdom Come: Deliverance. Bethesda is going to have to work a lot harder to stand out lately.
Morrowind has the best dialogue system to ever grace video games. You can quote me on that. I'm only partially kidding. The system of being able to ask almost anyone almost anything was really...
Although solid dialogue is not something I expect as much out of an Elder Scrolls game, it still indicates more generally that their target audience has shifted.
Morrowind has the best dialogue system to ever grace video games. You can quote me on that. I'm only partially kidding. The system of being able to ask almost anyone almost anything was really great. The execution required certain tradeoffs at the time, but almost 20 years later, I think we should give that another go. Use machine learning to voice-act text answers (maybe put some clues as to mood etc in there). Use a knowledge base for each NPC to pick what information they put into an answer, then some more ML to turn that into text. The player can select one or two facts/items to ask about and the game tries to tie it into the knowledge base as well as the previous conversation as best as it can. E voila, you have a dialog system for an actual living, breathing word.
NPCs exchange facts about the world, so you can pick up on rumors. Imagine for example what happens if a powerful artifact is briefly seen in a dungeon by a fleeing NPC adventurer. He'll tell people about it, who'll tell other people. Some information is lost in transit, but the player hears about it and can go on a quest for that dungeon, all organically. The player then has to trace those rumors, filter fact from fiction and finally hunt down the artifact. No developer scripted that for you, it all happened as emergent gameplay. Oh, the stories you'll tell your friends.
Is this possible with current technology? I've heard about deepfake videos, but never audio. I'm interested in the other suggestions you make as well, such as generating responses to player input...
Use machine learning to voice-act text answers (maybe put some clues as to mood etc in there).
Is this possible with current technology? I've heard about deepfake videos, but never audio. I'm interested in the other suggestions you make as well, such as generating responses to player input based on a broader knowledge base, although I don't know if even textual machine learning algorithms are smart enough right now to do that without significant errors. It would probably take a lot of tweaking on Bethesda's part.
Did you see OpenAI's Jukebox project that was posted on here? I listened to Ella Fitzgerald singing about how AI was hard to build. It's not there yet, she mumbled through 90% of the rest of the...
Did you see OpenAI's Jukebox project that was posted on here? I listened to Ella Fitzgerald singing about how AI was hard to build. It's not there yet, she mumbled through 90% of the rest of the lyrics before clearly saying "AI is hard to build" (although in her defense, she was coming at it from the angle of "where in the song do these lyrics go" instead of focusing purely on pronunciation), and needs beefy hardware you won't find in any consumer PC or console for a good long while, but dynamically generating the human voice is something that's been achieved with current technology.
Diff has basically covered the feasibility of text-to-speech. We've got that in place. Heck, Siri is a good starting point that's been around and appears to be scalable enough. I don't think...
Diff has basically covered the feasibility of text-to-speech. We've got that in place. Heck, Siri is a good starting point that's been around and appears to be scalable enough. I don't think text-to-speech is a big unsolved problem.
I'm interested in the other suggestions you make as well, such as generating responses to player input based on a broader knowledge base,
That would in my mind work by never even having natural language in there to begin with. The knowledge base has to be structured data to make sense. Think relational database, but slightly different. I'm thinking relational facts. So if you're in seyda neen and ask about balmora, the NPC might divulge some bit of knowledge like house_seat(balmora, hlaalu) or fighters_guild_location(balmora) or never_visited(self, balmora) or something like that. These relations phrases have to be manually created, but the actual data should be doable automatically. Selecting the right facts shouldn't be that hard, you're sitting in a relatively well-structured environment. So you can draw from pre-deep ML methods or just hard-code something. It also seems plausible to me to weigh these facts based on how often the NPC hears them from other NPCs or based on how closely they connect to the previous topic. So if after Balmora you ask about Pelagiad, the AI might pick out something like between(seyda_neen, balmora) or so.
Next, we're going to have to make a system that takes a few facts that were considered relevant by the NPC and turn that into text and then speech. So that might sound like "Balmora? Never been, but I've heard it's hlaalu's house seat, and the fighter's guild has a chapter there.". That system, damn I dunno. We don't have the data right now to pull that off, and I'm not sure how hard it would be to gather enough. I'm confident it's doable though, if you've got a bunch of people just making sentences out of a set of facts.
Basically: Instead of facing the rigors of actual NLP (Natural Language Processing), we do as much as possible with as little NLP as possible.
That's an interesting potential from an emergent gameplay perspective but I'm talking more about the very human craft of writing something that is entertaining to read or creating a compelling...
That's an interesting potential from an emergent gameplay perspective but I'm talking more about the very human craft of writing something that is entertaining to read or creating a compelling narrative. We're an extremely long way off from an AI being able to create a comprehensible and funny joke (there have been attempts) or to understand tragedy, irony, etc. If it hits on it, it's generally by coincidence - so while a funny/tragic story that comes from emergent properties might be acknowledged and noticed by the player, it won't be acknowledged by the ingame world.
On another note, I think Starfield is going to be badass. Bethesda Mass Effect? Skyrim on space? I know that's not exactly the idea, but one can only dream. I just hope they fix the engine...
On another note, I think Starfield is going to be badass. Bethesda Mass Effect? Skyrim on space? I know that's not exactly the idea, but one can only dream.
I was about ready to jump down your throat because I've heard "creation engine needs to die" one too many times. No, it's not the engine itself. The engine is custom-made to Bethesda's needs and...
I just hope they fix the engine...
I was about ready to jump down your throat because I've heard "creation engine needs to die" one too many times. No, it's not the engine itself. The engine is custom-made to Bethesda's needs and parts of it are extremely well tested.(1) Also note that the bugs you see in Beth games tend to be different ones every installment. It's not a few bugs lingering forever and ever, it's always new ones because they don't spend the effort to properly test the new stuff they put in. You'd have that with any engine. Tearing out the old parts won't help much, because it'll just mean you'll replace it with bug-riddled new code that is just as prone to breaking as the features they keep adding.
What Bethesda would have to do is to invest into refactoring where the architecture stands in the way of new features, and then test those new features a lot more. Replacing the engine with unity, unreal or cry engine wouldn't help at all, it'd just create work.
(1) I mention this because the people I reply to keep going on about how the fact that the engine has such old origins is to blame. The old parts are well tested, don't replace them unless you need to!
</rant>
I know you didn't say any of what I replied to, I just feel like the "creation engine needs to die" needs to die.
I don't know how they should fix this, but something must change. And I'm not merely talking about bugs, but features as well. It is possible to have better and more vibrant open worlds. Dragon...
I don't know how they should fix this, but something must change. And I'm not merely talking about bugs, but features as well. It is possible to have better and more vibrant open worlds. Dragon Age Inquisition, Witcher 3, and even GTA V are testaments to that.
I agree, but Dragon Age: Inquisition and The Witcher 3 aren't open-world in the same way that Skyrim is. Barring DLC areas like Solstheim, Skyrim's map is one continuous stretch of land from...
I agree, but Dragon Age: Inquisition and The Witcher 3 aren't open-world in the same way that Skyrim is. Barring DLC areas like Solstheim, Skyrim's map is one continuous stretch of land from Riften to Solitude. This means that, conceivably, almost anything in any part of the game can interact with anything else. This has a lot of interesting possibilities for gameplay, and makes the world feel very "real"; there are immersion and worldbuilding reasons that Bethesda seems to like to keep their games continuous, rather than modular. However, it also introduces a huge capacity for bugs. If everything can interact with everything, the number of opportunities for bugs increases exponentially as the size of the world grows.
The Elder Scrolls Online is a notable exception to this design philosophy, being broken up into different regions requiring loading screens to enter, although that was developed by ZeniMax, not Bethesda themselves. In TW3, the game world is divided similarly, with the different areas rarely interacting with one another, barring a few very specific story characters. Most quests take place in only one region, and those that don't typically have no back-and-forth between them. I haven't played GTA V start to finish, but my impression has always been that there is not as much content that has the capacity to break as Skyrim or Oblivion. Unique content and characters are far and few between. This is true also of The Witcher 3, in many ways. Because of this, these games are simply less prone to bugs to begin with.
So the choice that Bethesda has to make with TES VI is whether to stick with their current design formula and maintain the same level of bug-proneness that their games currently have, or to take a modular approach and reduce the opportunities for bugs substantially. I'm very partial to the way that Skyrim is laid out. I truly love the ability to travel across the entire map in one go. It creates a highly unified feeling of the game that simply doesn't exist in The Witcher 3 and other such titles. I'm somewhat spoiled by the ability to bugfix manually using console commands on PC, but a large part of me would prefer it if they continued in this format. However, I recognize that this is pretty bad form on Bethesda's part, and makes their games a lot harder to play and enjoy.
I have experience with Witcher 3 and its regions are large enough that what you point out doesn't really bother me. I don't know exactly what constitutes a point of breakage, but for the most...
I have experience with Witcher 3 and its regions are large enough that what you point out doesn't really bother me. I don't know exactly what constitutes a point of breakage, but for the most part, GTA works really well, the world is super large and there is no loading.
But I understand your argument and it is a valid one.
Maybe with the next generation of hardware, Bethesda can address some of my concerns while keeping the persistent world.
Regarding what constitutes a point of breakage: Everything that is committed to your savegame is a potential problem because whatever breaks will break permanently. TES has the habit of tracking a...
Regarding what constitutes a point of breakage: Everything that is committed to your savegame is a potential problem because whatever breaks will break permanently. TES has the habit of tracking a lot of things in your savegame. All the items in the world, all the NPCs, a good amount of mobs, dialog states for each NPC, quest states. Try and kill all of the whiterun guard, run to solitude and return. They should still be all dead. Maybe the guard will respawn, but named citizen sure don't. Any time a script accidentally interferes with an NPC, there's a potential problem because accidentally deleting or killing said NPC will soft-break your savegame. So that's why Skyrim has a lot of points of failure. Nevermind that it allows all kinds of interactions between things in the world, so that increases the potential for unforeseen consequences even more.
GTA V for example, in my experience is much more of a theme park. Few things have lasting consequences, the bulk of the world state is not saved (just destroy a neighborhood as best as you can and leave, it'll all be back to normal). The only way GTAV could mess up your savegame is if it somehow messes up storing your relatively simple progress. Things also don't really interact - hence theme park. Most of the open world mechanics have little to do with one another. There's so much to do, but basically everything is living in isolation of all the other things. You can hunt deer. You can play golf. You can not try and kill a deer with a golf ball. The open-ness of GTA is not that open at all. This extends to narrative and progression, where (in my experience with previous GTA games) you have to progress the main story to be able to access new toys and mechanics. San Andreas doesn't want you to have an AK by the time you to shooting practice, because that'd break their carefully crafted illusion of an open world. It's a theme park and they want you to take route A through it, before they set you free to explore.
Witcher is a bit in-between. Not much emergent stuff happening, so that reduces potential for breakage a lot. But the overall amount of stuff the game keeps track of is similar to skyrim. Inventories, dialogue progressions, a good amount of world state. Another big part is the limited role of the physics engine. Physics is notoriously unpredictable (see goat simulator) so the more power you give the physics engine to mess up a game, the worse. An important quest item dropping through the floor in skyrim is a remote, but real possibility. In GTA, not at all. In witcher, save for maybe weapon drops from enemies, I don't think it allows any items to be physically in the world at all. I certainly not remember being able to drop items.
I am reluctant to any major changes, if it means that the community has a harder time making mods. I've played but not completed the three games you mentioned and other games that use other...
I am reluctant to any major changes, if it means that the community has a harder time making mods.
I've played but not completed the three games you mentioned and other games that use other engines like the Unreal Engine (Ark and Conan Exiles for example). While all those games do have mods, maybe with the exception of GTA V, I am not sure if mods in them are really on the same level of freedom that I can see on places like Nexus Mods. I think mods are really important to the longevity and replayability of a game. It is part of why I believe Skyrim has the lifespan it has had and has Youtube Channels dedicated to new mod releases (that and the re-releases of course).
Take my opinion with a bit of salt though because I can't actually tell you how easy or difficult it is to mod any games. I just think that if users have the ability to add Thomas the Tank Engine, armors/weapons, or brand new quests into the game, then I am bound to find hours of more entertainment. The caveat of course is that Bethesda has apparently been trying to monetize their horse armor, creation club, and shoving cosmetic shop items into people's faces.
While I understand mods are a priority for a sizeable amount of Bethesda customers, the company don't have a moral obligation to keep their games as easy to modify. I'm certainly not against mods,...
While I understand mods are a priority for a sizeable amount of Bethesda customers, the company don't have a moral obligation to keep their games as easy to modify. I'm certainly not against mods, but it's conceivable that they have other priorities. Reducing mod support would certainly displease many people, but other consequences of that decision might please many others. Console gamers, which I am right now, might be okay with it.
Mods are pretty much the only reason Bethesda-made titles have such long tails, and it's enough that Bethesda (to their credit) has pushed for mod support on consoles more than anyone else. I...
Mods are pretty much the only reason Bethesda-made titles have such long tails, and it's enough that Bethesda (to their credit) has pushed for mod support on consoles more than anyone else. I think they realize how important mods are for anything beyond short-term success.
Them dropping support for mod tools would be a seismic event in the industry, and I can't imagine Bethesda are unaware of that.
I'm not sure what other avenue of reducing mod support there is without dropping support for mod tools as one of the early steps. That's what happened with other companies and series that lost...
I'm not sure what other avenue of reducing mod support there is without dropping support for mod tools as one of the early steps. That's what happened with other companies and series that lost modders over time.
That should at least provide enough time for Bethesda to decide whether or not they want to un-SNAFU themselves.
After having played the exceptionally polished The Outer Worlds, the idea of Bethesda pulling off a "space epic" with Starfield is fairly hard not to scoff at. I'm already picturing purple planets, melting faces and spaceships flying backwards.
Yeah, they've gradually dumbed down TES. I've been with the series since TES III, and the amount of RP in the RPG has decreased. Granted, they've done good things to the mechanics (in some areas anyway) and Skyrim is a very good game, it's just not long until all that's left is a theme park. Similarly for fallout. FO3 and NV had reasonable amounts of RP elements and FO4 just had.... very little thereof. It's just super hard to make your character memorable in either series' most recent installment. FO4 had some mechanical aspects I really enjoy - settlements show promise but lack a bit of polish, and I just love tinkering with my guns.
I'm really guarded when it comes to Bethesda now. I had preorder levels of good will back when Skyrim was coming out. I waited for reviews on FO4, figured it was fine, and while it wasn't a bad game at all, it just wasn't what I wanted or what I expected at all. I think I'm going to have to wait a fair while with starfield and TES VI, just to make sure it's the kind of game I want Bethesda to make. What kind of a consumer would I be if I were to buy a game not because of what it is but because of what I want it to be...?
If you haven't and you have a relatively beefy PC you should look in to Ultimate Skyrim. It's a pack of 250 mods built around the Skyrim Requiem mod which removes all the levelled enemies (so you can't just wander up to a giant at level 2 and have a chance for example) and generally makes the game a lot more realistic. It does incorporate mods which require you to eat and drink and sleep which may be a turn off but you should be able to just disable those during the configuration.
They provide an automated tool which if you have a Nexusmods premium account can just slurp the whole lot down and install them so it's less painful than it sounds.
The scope of The Outer Worlds and a typical Bethesda game are very different. I don't think they are as comparable as people are saying. Environments in The Outer Worlds are completely static. There is almost no physics to speak of, and it's completely missing most of the systems Bethesda builds into their games to create emergent gameplay and narrative.
I don't mean this as a criticism of the game, it just has very different design goals than the games people keep comparing it to. I think the game has much more in common with earlier BioWare games like Knights of the Old Republic. The similarities to Bethesda games are largely surface level (first person view, Fallout-esque satirical dystopia).
I'm really "eh" on Bethesda after Fallout 4 (which was "fine" but forgettable) and Fallout 76 (which I will not spend my money on). I find it hard to express specifically what threw me off, I think it's just a culmination of things that sort of sucked out the je ne sais quoi and made it harder to immerse myself in it. I don't mind a fun action game or streamlined RPG elements, but FO4 felt to me like diving into the Uncanny Valley of Fallout.
The writing and the dialogue system were also kind of a shame. Although solid dialogue is not something I expect as much out of an Elder Scrolls game, it still indicates more generally that their target audience has shifted. It might not be "for" me any more, which makes me a bit sad.
At the same time, though, there are more open-world games than there were at the time Skyrim was released! Bethesda were pioneers in many ways, but now other companies have iterated on their formula and there's a fuller market for open-world exploration games. I'm not going to swear Elder Scrolls off completely or anything, but I'm also not chomping at the bit to get it like I was when it was first announced. I haven't even started Witcher 3 yet because I'm still working through Assassin's Creed Odyssey, which I only picked up after getting the trophies I wanted from Kingdom Come: Deliverance. Bethesda is going to have to work a lot harder to stand out lately.
Morrowind has the best dialogue system to ever grace video games. You can quote me on that. I'm only partially kidding. The system of being able to ask almost anyone almost anything was really great. The execution required certain tradeoffs at the time, but almost 20 years later, I think we should give that another go. Use machine learning to voice-act text answers (maybe put some clues as to mood etc in there). Use a knowledge base for each NPC to pick what information they put into an answer, then some more ML to turn that into text. The player can select one or two facts/items to ask about and the game tries to tie it into the knowledge base as well as the previous conversation as best as it can. E voila, you have a dialog system for an actual living, breathing word.
NPCs exchange facts about the world, so you can pick up on rumors. Imagine for example what happens if a powerful artifact is briefly seen in a dungeon by a fleeing NPC adventurer. He'll tell people about it, who'll tell other people. Some information is lost in transit, but the player hears about it and can go on a quest for that dungeon, all organically. The player then has to trace those rumors, filter fact from fiction and finally hunt down the artifact. No developer scripted that for you, it all happened as emergent gameplay. Oh, the stories you'll tell your friends.
Is this possible with current technology? I've heard about deepfake videos, but never audio. I'm interested in the other suggestions you make as well, such as generating responses to player input based on a broader knowledge base, although I don't know if even textual machine learning algorithms are smart enough right now to do that without significant errors. It would probably take a lot of tweaking on Bethesda's part.
Did you see OpenAI's Jukebox project that was posted on here? I listened to Ella Fitzgerald singing about how AI was hard to build. It's not there yet, she mumbled through 90% of the rest of the lyrics before clearly saying "AI is hard to build" (although in her defense, she was coming at it from the angle of "where in the song do these lyrics go" instead of focusing purely on pronunciation), and needs beefy hardware you won't find in any consumer PC or console for a good long while, but dynamically generating the human voice is something that's been achieved with current technology.
Diff has basically covered the feasibility of text-to-speech. We've got that in place. Heck, Siri is a good starting point that's been around and appears to be scalable enough. I don't think text-to-speech is a big unsolved problem.
That would in my mind work by never even having natural language in there to begin with. The knowledge base has to be structured data to make sense. Think relational database, but slightly different. I'm thinking relational facts. So if you're in seyda neen and ask about balmora, the NPC might divulge some bit of knowledge like house_seat(balmora, hlaalu) or fighters_guild_location(balmora) or never_visited(self, balmora) or something like that. These relations phrases have to be manually created, but the actual data should be doable automatically. Selecting the right facts shouldn't be that hard, you're sitting in a relatively well-structured environment. So you can draw from pre-deep ML methods or just hard-code something. It also seems plausible to me to weigh these facts based on how often the NPC hears them from other NPCs or based on how closely they connect to the previous topic. So if after Balmora you ask about Pelagiad, the AI might pick out something like between(seyda_neen, balmora) or so.
Next, we're going to have to make a system that takes a few facts that were considered relevant by the NPC and turn that into text and then speech. So that might sound like "Balmora? Never been, but I've heard it's hlaalu's house seat, and the fighter's guild has a chapter there.". That system, damn I dunno. We don't have the data right now to pull that off, and I'm not sure how hard it would be to gather enough. I'm confident it's doable though, if you've got a bunch of people just making sentences out of a set of facts.
Basically: Instead of facing the rigors of actual NLP (Natural Language Processing), we do as much as possible with as little NLP as possible.
That's an interesting potential from an emergent gameplay perspective but I'm talking more about the very human craft of writing something that is entertaining to read or creating a compelling narrative. We're an extremely long way off from an AI being able to create a comprehensible and funny joke (there have been attempts) or to understand tragedy, irony, etc. If it hits on it, it's generally by coincidence - so while a funny/tragic story that comes from emergent properties might be acknowledged and noticed by the player, it won't be acknowledged by the ingame world.
On another note, I think Starfield is going to be badass. Bethesda Mass Effect? Skyrim on space? I know that's not exactly the idea, but one can only dream.
I just hope they fix the engine...
I was about ready to jump down your throat because I've heard "creation engine needs to die" one too many times. No, it's not the engine itself. The engine is custom-made to Bethesda's needs and parts of it are extremely well tested.(1) Also note that the bugs you see in Beth games tend to be different ones every installment. It's not a few bugs lingering forever and ever, it's always new ones because they don't spend the effort to properly test the new stuff they put in. You'd have that with any engine. Tearing out the old parts won't help much, because it'll just mean you'll replace it with bug-riddled new code that is just as prone to breaking as the features they keep adding.
What Bethesda would have to do is to invest into refactoring where the architecture stands in the way of new features, and then test those new features a lot more. Replacing the engine with unity, unreal or cry engine wouldn't help at all, it'd just create work.
(1) I mention this because the people I reply to keep going on about how the fact that the engine has such old origins is to blame. The old parts are well tested, don't replace them unless you need to!
</rant>I know you didn't say any of what I replied to, I just feel like the "creation engine needs to die" needs to die.
I don't know how they should fix this, but something must change. And I'm not merely talking about bugs, but features as well. It is possible to have better and more vibrant open worlds. Dragon Age Inquisition, Witcher 3, and even GTA V are testaments to that.
I agree, but Dragon Age: Inquisition and The Witcher 3 aren't open-world in the same way that Skyrim is. Barring DLC areas like Solstheim, Skyrim's map is one continuous stretch of land from Riften to Solitude. This means that, conceivably, almost anything in any part of the game can interact with anything else. This has a lot of interesting possibilities for gameplay, and makes the world feel very "real"; there are immersion and worldbuilding reasons that Bethesda seems to like to keep their games continuous, rather than modular. However, it also introduces a huge capacity for bugs. If everything can interact with everything, the number of opportunities for bugs increases exponentially as the size of the world grows.
The Elder Scrolls Online is a notable exception to this design philosophy, being broken up into different regions requiring loading screens to enter, although that was developed by ZeniMax, not Bethesda themselves. In TW3, the game world is divided similarly, with the different areas rarely interacting with one another, barring a few very specific story characters. Most quests take place in only one region, and those that don't typically have no back-and-forth between them. I haven't played GTA V start to finish, but my impression has always been that there is not as much content that has the capacity to break as Skyrim or Oblivion. Unique content and characters are far and few between. This is true also of The Witcher 3, in many ways. Because of this, these games are simply less prone to bugs to begin with.
So the choice that Bethesda has to make with TES VI is whether to stick with their current design formula and maintain the same level of bug-proneness that their games currently have, or to take a modular approach and reduce the opportunities for bugs substantially. I'm very partial to the way that Skyrim is laid out. I truly love the ability to travel across the entire map in one go. It creates a highly unified feeling of the game that simply doesn't exist in The Witcher 3 and other such titles. I'm somewhat spoiled by the ability to bugfix manually using console commands on PC, but a large part of me would prefer it if they continued in this format. However, I recognize that this is pretty bad form on Bethesda's part, and makes their games a lot harder to play and enjoy.
I have experience with Witcher 3 and its regions are large enough that what you point out doesn't really bother me. I don't know exactly what constitutes a point of breakage, but for the most part, GTA works really well, the world is super large and there is no loading.
But I understand your argument and it is a valid one.
Maybe with the next generation of hardware, Bethesda can address some of my concerns while keeping the persistent world.
Regarding what constitutes a point of breakage: Everything that is committed to your savegame is a potential problem because whatever breaks will break permanently. TES has the habit of tracking a lot of things in your savegame. All the items in the world, all the NPCs, a good amount of mobs, dialog states for each NPC, quest states. Try and kill all of the whiterun guard, run to solitude and return. They should still be all dead. Maybe the guard will respawn, but named citizen sure don't. Any time a script accidentally interferes with an NPC, there's a potential problem because accidentally deleting or killing said NPC will soft-break your savegame. So that's why Skyrim has a lot of points of failure. Nevermind that it allows all kinds of interactions between things in the world, so that increases the potential for unforeseen consequences even more.
GTA V for example, in my experience is much more of a theme park. Few things have lasting consequences, the bulk of the world state is not saved (just destroy a neighborhood as best as you can and leave, it'll all be back to normal). The only way GTAV could mess up your savegame is if it somehow messes up storing your relatively simple progress. Things also don't really interact - hence theme park. Most of the open world mechanics have little to do with one another. There's so much to do, but basically everything is living in isolation of all the other things. You can hunt deer. You can play golf. You can not try and kill a deer with a golf ball. The open-ness of GTA is not that open at all. This extends to narrative and progression, where (in my experience with previous GTA games) you have to progress the main story to be able to access new toys and mechanics. San Andreas doesn't want you to have an AK by the time you to shooting practice, because that'd break their carefully crafted illusion of an open world. It's a theme park and they want you to take route A through it, before they set you free to explore.
Witcher is a bit in-between. Not much emergent stuff happening, so that reduces potential for breakage a lot. But the overall amount of stuff the game keeps track of is similar to skyrim. Inventories, dialogue progressions, a good amount of world state. Another big part is the limited role of the physics engine. Physics is notoriously unpredictable (see goat simulator) so the more power you give the physics engine to mess up a game, the worse. An important quest item dropping through the floor in skyrim is a remote, but real possibility. In GTA, not at all. In witcher, save for maybe weapon drops from enemies, I don't think it allows any items to be physically in the world at all. I certainly not remember being able to drop items.
I am reluctant to any major changes, if it means that the community has a harder time making mods.
I've played but not completed the three games you mentioned and other games that use other engines like the Unreal Engine (Ark and Conan Exiles for example). While all those games do have mods, maybe with the exception of GTA V, I am not sure if mods in them are really on the same level of freedom that I can see on places like Nexus Mods. I think mods are really important to the longevity and replayability of a game. It is part of why I believe Skyrim has the lifespan it has had and has Youtube Channels dedicated to new mod releases (that and the re-releases of course).
Take my opinion with a bit of salt though because I can't actually tell you how easy or difficult it is to mod any games. I just think that if users have the ability to add Thomas the Tank Engine, armors/weapons, or brand new quests into the game, then I am bound to find hours of more entertainment. The caveat of course is that Bethesda has apparently been trying to monetize their horse armor, creation club, and shoving cosmetic shop items into people's faces.
While I understand mods are a priority for a sizeable amount of Bethesda customers, the company don't have a moral obligation to keep their games as easy to modify. I'm certainly not against mods, but it's conceivable that they have other priorities. Reducing mod support would certainly displease many people, but other consequences of that decision might please many others. Console gamers, which I am right now, might be okay with it.
Mods are pretty much the only reason Bethesda-made titles have such long tails, and it's enough that Bethesda (to their credit) has pushed for mod support on consoles more than anyone else. I think they realize how important mods are for anything beyond short-term success.
Them dropping support for mod tools would be a seismic event in the industry, and I can't imagine Bethesda are unaware of that.
At no point I said "dropping support" or anything to that effect.
I'm not sure what other avenue of reducing mod support there is without dropping support for mod tools as one of the early steps. That's what happened with other companies and series that lost modders over time.
I'm not sure either.