If you have 40ish minutes, watch all of this.
If you are pressed for time, watch the first section.
Either way, it’s awesome stuff.
Listening to the design behind Skyrim, I think I’ve pinpointed one of the long-term issues I have with the game: the world setting is in the uncanny valley. I’m not talking about its visuals, I’m talking about the simulation aspect. The world feels more ‘alive’ than in Oblivion, things are more interconnected, the NPCs react more to what you do, etc. But it’s not a real world. Not everything is connected. The mage in one city will still recommend you join the mages guild, even though you are not only the guild master, but are actually wearing the guild master’s robes right in his face.
In a ‘normal’ game, that happens all the time and it’s no big deal. Just like in a ‘normal’ game we accept unrealistic character proportions or weapons. But when a game tries hard to be ‘real’, the unreal is far more noticeable. Skyrim tries hard to be a ‘real’ world, and when that illusion fails, it’s far more noticeable. So noticeable sometimes, that it actually detracts from the experience.
And I don’t realistically think the shortcomings of Skyrim can be fixed by tuning the scripts or just making ‘more’ stuff. Yes, you could add a check for that one mage, and when you become guild master, he recognizes it. But there are dozens and dozens of such inconsistencies in Skyrim. And furthermore, while some events change certain locations (the civil war), the rest of the world is stuck in stasis from the moment you leave the intro. Again, in most games that’s normal and not something you notice, but in a game like Skyrim, that places such a heavy emphasis on ‘world’, it is.
What I would love to see the next TES game attempt is an actual world. One where time progresses whether you stand still or clear dungeons. Where the NPCs are driven by needs/wants rather than triggers and scripts. You could still have a central theme/storyline, and all of the side-quests, but they would not wait around for you like they do in Skyrim. If you decide to ignore the dragon crisis, towns/homes/people might suffer. Nothing game-breaking, but perhaps some random quest NPC gets eaten if you don’t kill a certain dragon. Perhaps the mages guild solves its problem before you get there, and instead of that being the path to guild master, it’s something else.
In some ways Skyrim is just too big. By the time you have visited half the towns in the game, the gameplay and awe factor have likely worn thin. The different stories and quests are still interesting, as are the locations, but seeing/experiencing them is no longer as fun. Skyrim really does not need 100 hours of gameplay, because the combat/feel of the game is not that good. I’m not saying it’s bad, far from it, it’s actually a hell of a lot of fun for the first 60 hours or so, it’s just that it wears thin.
So with a smaller world, more time can be spent filling it up, and creating if/when scenarios. Rather than a town having one set of quests, time/effort could be spend creating 3-4 different sets, each depending on when you visit and what has gone on before. And those 3-4 sets don’t all start/stop at the same time, or even on a set schedule. Perhaps the mages guild issue takes 10-15 days to solve, but this varies depending on what you do in the world, and also by some random factors. ‘Missing’ the current issue storyline would not be that big a deal because there would be other content behind it. Becoming the leader would simply open up other possibilities (different mages causing trouble based on who you helped/hindered earlier).
By creating a more believable, living world, the reply value in the next TES game would not come from seeing different content/locations, but seeing similar locations react to your different actions the second time around.
Hardcore Casual 
Thanks for linking that. I expected to watch about two minutes but I watched the whole thing. Well actually I did Erollisi Day quests in EQ2 with the audio in the background and flipped over every time something came up that sounded like it was worth looking at.
Whatever, it was a fascinating speech. I probably disagreed with more than I agreed with, and he made it a lot clearer to me why I have never enjoyed a Bethesda game yet and probably never will, but still there was plenty in there that really resonated.
I haven’t played Skyrim and the more I see of it, read about it and hear about it the less likely it becomes that I will. I very much agree with your “Uncanny Valley” comments in respect of the last Bethesda game I did try, Morrowind. There was a session or two of “gosh-wow this is amazing! This is just like being in an actual world!” and then I realized that most of the NPCs would just trot out the same canned responses and you could take anything with no consequences and the idea that the place had any level of “reality” just crumbled.
I completely agree that the more “real” you try and make a virtual world the less “real” it seems and I think that’s an insurmountable problem at current levels of technology.
I don’t think tech is a limiting factor anymore. If a game like UO back in 97 can have a living eco-system, why can’t something in 2012?
I think the real limiter is fear. I think game-devs, and players, fear the unknown or unexpected. “What if a player misses some content” or “What if events line up poorly and make things more difficult” thinking rather than “I wonder what would happen if we just let the AI go, and have it react to the player”.
I think an AI like that is very possible. I’m not sure devs/players are really ready for it.
On AI and gameplay I agree. On realism as in creating a whole city (let alone a planet) where each NPC can react to you as if he/she/it is an individual different to every other NPC, and carry on doing it convincingly every time you go back and speak to them again then no, I don’t think we’re anywhere near that point.
It’s no excuse not to do many of the things you quite rightly point out were being done better over a decade ago, though.
““I wonder what would happen if we just let the AI go, and have it react to the player”.”
Easy… I can point to a stand-alone movie, or a series of them: WarGames….or Terminator.
SkyNet may very well have been spawned on XBox Live… the Terminator movies are deliberate misinformation.
In either case, James Cameron is in league with the Devil. Just saying. lol
It’s my understanding the needs/wants drivers were used in Oblivion (and probably in Skyrim), as it’s actually an easier way to develop AI once past a certain level of complexity. The is a couple of issues with this kind of thing though, that what with production deadlines and all, get over-ridden by case-specific triggers.
There’s a few developer interviews that talk about it – the AI system was called Radiant. Most of them are fluff pieces though.
I can just imagine the production meeting.
Manager: “Where is the mage? He’s supposed to mention the Mage Guild”
Developer: “I don’t know, he might be eating, or sleeping, or practicing magic…”
Manager: “Well, that needs to be fixed, he should be here to talk to the player. Now, find him and show me his lines.”
Developer wanders around and finds the mage. Interacting with the mage gives several responses until eventually the mage randomly cycles out the required line re: the guild.
Manager: “No no not good enough, that line needs to be right up front. Fix it!”
In short – it’s very doable, has been done, but then the player doesn’t feel like the world spanning hero because the AI’s treat him normally, and doesn’t get the design required cues.
Oh, and I meant to mention the case relevant to UO – wasn’t that the one where the players farmed the wolves, so the deer population migrated away from the dragon, now that they weren’t trapped by the wolves… so the dragon migrated to the town in order to eat?
While I agree it could’ve been done much better, I don’t think UO necessary had a living eco-system that was driven by NPC interaction. It had a living eco-system, but that was due to the freedom given to players to act in virtually any role they desired.
As an example, for a time in my early UO days I was a low level functionary for an in-game company called Yarblek Industries. Yarblek “employed” newbie ‘alt’ characters to recall around Sosaria and buy out stocks of reagents. Yarblek would buy them off his employees at a few gp extra, and then sell them en masse to the big PK and PvP guilds for mass profits. These player actions had a massive impacts on the in-game economy, and I’m certain there were dozens if not hundreds of examples of this kind of thing.
But what role did the NPC play? That’s where I bought my Blood Moss.
Skyrim could’ve tightened up their Player-Plot-NPC interactions, but in terms of the NPC side of things, Skyrim felt more alive than UO did. But then I guess UO didnt need to ;)
This is what they did with the original Gothic (1 & 2) , small world, lots of things changing in the small world. You kinda ran back and forth in this small world and got use to the trails and areas and then it changes…. definitely alot more immersive in my oppinion.
The smaller towns/setting also meant less NPCs but more “intelligent” NPCs that actually knew a little more about where you are in the game. Maybe one day someone will get the right combinations together ;)
Don’t you think that when building a game like Skyrim you reach a point of diminishing return, in that any polish you add or changes you make after a certain point can not be justified by projected increased revenue from sales.
The situation with MMOs is perhaps somewhat different in that with so many players competing, making the gameplay tighter and more polished might be of more benefit commercially.
About Challenge in the video:
“We need to put the player in that sweet spot”.
This is the problem. Because what he should have said is:
“We need to allow the player to move towards that sweet spot”
Developers are too afraid to give their players freedom. And, of course, a directed experience is easier to design. But ultimately a player who chooses for himself within a good* ruleset is what games will be.
—
*that’s the hard part.