Free-roaming, either empty or broken

Free-roaming, either empty or broken

Note: This topic primarily discuss RPG's, but free-roaming games in other genrés seems to have the same problem and might be worthy to discuss here as well.

You are in a vast world, do what you like.

That's basically what many games tried to offer throughout the ages. Do not ask me which the first real free-roaming game was. I remember titles like Fallout and Baldur's Gate, but even the ancient Ultima I fits the genré and there are likely many more that do. I will focus on the newer games since this feels like a growing modern problem since free-roaming became popular. What I do remember is that it seems like almost every one of them either fits the description "empty" or "broken" and lately this seems to be the curse of the genré.

The broken part is easy to explain. Here we find games such as Fallout 2, which was initially extremely buggy and many stopped playing it after a patch ripped all old savegames apart. I remember when people of Black Isle started to complain over the games they were making. They were simply too complex and took too much time to be worth it. Complex dialogue and advanced scripts to make unique quests and unique areas that feels different... the more you try, the greater chance it is that something will break. And they are extremely difficult to bugcheck. From Baldur's Gate II and on, all games produced by Black Isle, Bioware and Obsidian have been very linear, save for Storms of Zehir that tries to go back to the roots.
Still, this isn't a problem of the past. We find modern games with this problem such as Gothic 3, which is acording to me THE rolemodel for how to fill a massive world with content (one of the few games that really managed to make the next valley as exciting as the previous one), but wont be recognized at all because the game was broken beyond belief and thus ripped apart by reviews. Many promising games have fit this description, such as Arcanum, Two Worlds, S.T.A.L.K.E.R and Boiling Point. You can really see when playing the game that the developer ran out of time in the middle of the job. Some of these have later been taken up by fans who "fixed" some of it's issues and turned them into great underrated treasures.

The second kind is not as easy to explain, but it seems they are produced when trying to dodge the above problem. Take a game like Oblivion, which by the first glance appears huge, but when you set the quest-list of Oblivion (about 40-50 if you count guildmissions as one quest each) next to the quest-list of Gothic 3 (about 500) you begin to see why Oblivion felt "empty". When you have to fill a gamemap that have a huge map, you will rely more on generated and streamlined content, the game equivalent of "mass-produced and cheap". While easier to keep bugfree, it creates games that feels repetitive, empty and unrewarding. Now these games have bugs, and patchlists are usually long, but those bugs are rare, "hidden" and not rarely so OBVIOUS that every player see them. Because of this, these games tend to score high on reviews, only to be ripped apart by the gamers who managed to play them long enough to see through the first impression. Here are games such as Oblivion, Fallout 3 and Far Cry 2.

So... is this the way things are going to be? Can someone tell me a free-roaming game that is neither ripped apart from gamers for being repetitive/boring/empty or ripped apart for being broken?

I don't see many peoples classified Fallout, Gothic and Baldur's Gate series or any games by Black Isle and Bioware as free-roaming type rpgs. They are crpgs with certain amount of gameplay options, and these choices are part of any rpgs.

However, and as an example, rpgs like TES, Fallout 3 or games by Bethesda have always much more free-form in term of gameplay and design. Some people call it sandbox type crpgs. This type of games always weak in certain aspect (e.g story telling).

Both types of crpgs could be buggy, and i think it depend on the developer capability (or publisher willingness) to polish the game before release.

I found that playing these types of games makes me feel ... 'uneasy'. It's hard to describe, but the feeling of not knowing what I should do or where I should go next is unnerving, to the point of just quitting and uninstalling the game (What I did in Fallout 2).
Many people think that this style of games is how RPGs should be, that's why many times I have to read the reviews, trying to read between the lines, to make sure that a high score is not because of the game being 'free roaming' or the contrary, a low score because of it not being so.
In the style of RPGs I prefer, you always have some reason to do something, to go somewhere. You don't go to a dot in the map just because it's there. FAllout 3 is the perfect example. I finished it in a few hours, my character being lvl 15 or so (and I've seen many people complain that it's too easy to reach the max lvl 20). I just did the storyline. My opinion of the game? just an ok game, while many others and most reviews give it high score for its openness. Well, that doesn't work for me.

I think that JemyM wanted point how free roaming is difficult. Either you go for sandbox then you get empty either you go to open western RPG and you get many bugs or unachieved game. Quote that with older Elder Scrolls you get both, sandbox AND bag of bugs!

I haven't investigate the subject but I don't remember editor released to public that include a dedicated tool to not only make scenario but tool(s) to facilitate coherency, cross links and alternates paths. In fact it's since design that such tool would serve but also all along the making of the scenario/game up to tests.

If none of those game choosing this approach didn't start with that no wonder that most fail to produce something deep and with few bugs.

The problem is that a free-roaming game would require a completely different data architecture than a scripted game. The very structure of the scripting engine should maintain the integrity of the overall structure even while permitting you to link events through consequences. This is a hard design problem, and I have a feeling it's one that hasn't yet been seriously addressed, let alone successfully solved. That means that you get either a more or less linear game, with perhaps a few branches or alternative paths, which is relatively easy to deal with using traditional scripting engines, or a "sandbox" with lots of unrelated content that doesn't mess with each other (i.e., Oblivion etc.). These require different types of scripting engines, and you run into serious trouble when you try to do both at once -- which is the philosopher's stone you're looking for.

If I ever make a career change and get into games development, this is one problem I'd set out to solve from the outset, unless somebody else has solved it first.

Indeed. While games have progressed visually quite a bit, the same thing cannot be said about their storytelling narrative. The structure in a free-roaming game is quite different from superlinear experiences.

In Gothic 3, each place had it's own story, it's own issues to be solved and in the end often two ways you could "resolve the city situation". This usually meant "side with the slaves or the slavers" or "the orcs or the humans" kind of choice. Resolving the situation usually boost your standing with the factions, which in turn effected how the next city would welcome you. On top of that almost every city (there were like 15+ of them) had it's own distinct feel to it. The main quest involved tracking down a load of chalices which were spread over the entire world. Trying to adress the situation in each city usually gave you a chalice. Towards the end you could pick with three different endings, also, how you dealt with every companion of yours would be told in the end as well. Gothic 3 was also the first game in which you could feel distinct cultures in the world. The mentality in the far north compared to the far south was very different, including the feel of being in a winter landscape compared to being in a desert landscape. Beyond all of that, the main story in itself was quite good, with a critical take on our modern world, a message quite important to remind ourselves about once in awhile.

Gothic 3 was also a game so buggy that even fans said they wont even play it. Question is if a greater company would have been able to pull the design of Gothic 3 off, or if even attempting that much depth in a free-roaming world, is doomed to fail.

Compromise.

Difficulty-wise unleveled main story, difficulty-wise unleveled side tracks.
Without free roaming, advancement in main story is almost impossible.
Ideally, those side events which you choose would somehow influence the progression in the main story.

Would be nice to have non-combat solutions for some main story parts based on what side quests you´ve done. For example, the reward for finishing some side quest would result in a big boost in diplomacy skill, which could significantly help you in resolving some challenging main quest event, etc.

How I´d like it, main quest should be extremely difficult or impossible to beat without a healthy dose of free roaming. In the same time, some of those free roaming choices should have an impact on how the main story plays out. Also, some story (main or side) branches should be sensibly class/ alingment limited.

Basically, free roaming should be encouraged but in the same time often tied to progess in main storyline.

JemyM said:

You are in a vast world, do what you like.

Click to expand...

Ok. (switching off my computer and enjoying the rest of the world, it´s vast)
I need more fantasy motivation in computer games .
At least something assertive like this: "You are in a vast world, it´s fucked up ad infinitum ... do what you like."

Anyway, I´m christmaswasted so rock on, it´s a very interesting topic.

Prime Junta said:

The problem is that a free-roaming game would require a completely different data architecture than a scripted game. The very structure of the scripting engine should maintain the integrity of the overall structure even while permitting you to link events through consequences. This is a hard design problem, and I have a feeling it's one that hasn't yet been seriously addressed, let alone successfully solved. That means that you get either a more or less linear game, with perhaps a few branches or alternative paths, which is relatively easy to deal with using traditional scripting engines, or a "sandbox" with lots of unrelated content that doesn't mess with each other (i.e., Oblivion etc.). These require different types of scripting engines, and you run into serious trouble when you try to do both at once -- which is the philosopher's stone you're looking for.

Click to expand...

Sandbox is an overlooked genre, I bet this is coming from giant hype of GTA. For RPG, where are the sandbox CRPG? I don't see many sandbox CRPG, ok the Elder Scroll and then?

I don't see sandbox and non sandbox CRPG as two different genre of games. It's much more a matter of degree but, it's a curse of any RPG, sandbox or not, to have to manage all at same time, free roaming and freedom of choices with story lines and specific quests and interactions with NPC.

Sandbox means the player don't follow the plot, do his own and interact with the engine. Fine for an action game but for a CRPG it's very fast limited. The problem is that many aspects of a CRPG can't be programmed through an engine:

• Talks with NPC.
• Any bit of story.
• Any mission even if I can imagine that you could have a sort of random generation of missions but quite stupid and basic one.
• NPC behavior up to mimic roughly real human behavior.

All of that can't be generated by an engine and must be written and organized because nobody know make a program that generate itself any of those elements, talks, stories, missions, human behaviors. That let for the sandbox only, free roaming, fighting, actions choices.

For fighting I won't develop it's an intricate debate by itself my point about it is that pure generation and random are unable to match a careful and tuned design of each fight. But obviously many CRPG choose more or less random, it saves work.

Now the two sandbox myths:

• Free roaming: You cannot design a good game if it is fully free roaming, that makes no sense. You go an free roaming the white house or my house? But there are many other good limit, cliffs, sea and no boat, river full of sharks, mountain with a dragon horde, law, and so on. There isn't free roaming and no free roaming but different degree of free roaming.
• Actions choices: Well no freedom here, any possibility should have been carefully designed and program. In GTA, can you become invisible, dig holes in ground, make chemic reactions, freeze time, jump 20 meters high, throw people in air, spit on someone, and so on? I bet not for most. Anyway it's the point, each game offer a different array of general choices, those are choices design and program in the engine and beside some conditions that could allow some of them, they are in general available. There's no sandbox here, it's again just a matter of degree.

To come back to free roaming, this is just a degree between the few CRPG tagged sandbox and other. Plenty CRPG have a lot of free roaming including Gothic 1&2, Ultima 7, Fallout, Baldur's Gate and many other. And the same can be said about action freedom.

To come back quickly to the tool I mentioned, the first target would be only at a design point of view in order to clarify the problems. Quickly and as a draft:

• The starting idea is that under the scene everything could be reduced to something like, nodes, events, options and conditions.
• From this base it's to allow sharing resources, like a NPC used by two different plots or a same NPC in different alternate paths. That can leads very fast to contradiction and broken lines, lack of coherency in dialog or a story line.
• So the tool would allow edit and view in different ways, for example allow to read the multiple possible stories, see at a node line crossing and coherency of the node, eventually notice possible line break and so on.

That's very vague I know, just a rough idea.

DeepO said:

Compromise.
Basically, free roaming should be encouraged but in the same time often tied to progess in main storyline.

Click to expand...

That is certainly a design option that can simplify the merging of a story density and freedom. The current game I'm playing (re in fact) use that approach a lot. It's The Sword Lands Trilogy for the Realmz game. The first module setup that already quite well but the second is just amazing from what I remember (played many years ago) and I have yet to play the third. When you start you have just a vague general quest, "to investigate, on behalf of the King of Bywater, who or what is causing the disruption of trade between Bywater and this region". That's all and that's vague. You have no contact, you know nothing about the region, nobody know you. And here go the adventure. But quite fast you feel that most events are related to a large plan. That goes from dark and suspect knight using a room at the inn close to your own room, to rumors about plans of some lord trying to control the whole region or some other lords from external region that always plan to take one of a main city of the area, missions to escort caravans to protect them again attacks, missions to investigate about goblins raiders and in their leader hut you found a notes, and so on.

I'm just at the beginning of my replay but if I remember well that raise more and more and overall, beside curiosity, most action or information isn't required but just setup more deeply the web of events around the main story.

But I think that there's a variation that also increase story density, it's cross events. There's two way for them:
- Anticipation: You learn something, it could be vague, and later you discover it and what's behind the information or rumor.
- Flashback: You do something and later in another place or at another time someone mention you something about what you have done.

What's interesting with those sort of mechanism is that they can be very simple, no need of a complex series of events. And by building a web of events you get story density and bring life to the game world. The difference with what you mention is that no link to the main quest is required.

JemyM said:

Gothic 3 was also a game so buggy that even fans said they wont even play it. Question is if a greater company would have been able to pull the design of Gothic 3 off, or if even attempting that much depth in a free-roaming world, is doomed to fail.

Click to expand...

I don't think it's doomed to fail; it just requires a different design approach. Sort of like the move from procedural to object-oriented design in general software.

It is not doomed to fail. But it needs a looooooooooooot of resources. PB just did not have enough resources to pull it off.

Everytime you have an option you create an exponential number of oppertunerties more making it impossible to test all possiblities. Combine that with a big open ended world and you have a nightmare to program!

But in MMORPGs they pulled it off!!!! since they had a loooot more resources, and they didn´t have that many choices and consequences.

It takes a lot of imagination and some clever thinking to view the real world in terms of strict cause & effect, but imaginary worlds are a whole different story because their realities are asserted.

Weird science isn't only acceptable, it's preferable. RPG worlds can be made to respond in reaction to the player, giving them a lot more "wiggle room" in terms of making sense. And that's a huge advantage. It would also provide collaboration, and that's more fun, isn't it?

Why design worlds to work like a boomerang when they can be designed to work like a bird instead, free to make choices of its own? There's less need to be concerned with things like the laws of physics when what you're playing with reacts like a dog or a cat, not a yo-yo or a ball.

IMO, the problem with sandboxes has been their lack of narrative vision. Instead of striving toward a complete goal, sandboxes tend to ramble forward haphazardly, and in the end their narrative pieces don't fit together very well.

If it could be implemented in software, mandalic expressions could make sense out of the sandboxiest of RPG worlds. That way their narrative wouldn't be just a bunch of fragments. Each could fit into something cohesive and unique.

Re-reading this, I really need to elaborate (so what else is new?).

Very simply put, karma is an expression of strict cause-and-effect where your world is a reflection of the state of your life. So by changing your karma, you're making a cause that will effect your world. Buddhism examines that concept by defining characteristics of life and imagining how they might exist in relation with one another, resulting in different states of life.

RPG could have game worlds that did something similar. They could be designed to identify and react to the player's choices, like the character he chooses and the way he plays it. Narrative could unfold in tandem with those choices with a "state" always in mind. Each "state" could correspend with a total narrative that's unique and satisfying.

GothicGothicness said:

It is not doomed to fail. But it needs a looooooooooooot of resources. PB just did not have enough resources to pull it off.

Everytime you have an option you create an exponential number of oppertunerties more making it impossible to test all possiblities. Combine that with a big open ended world and you have a nightmare to program!

Click to expand...

Again, a CRPG isn't an action game, in CRPG character skills progress so that make no sense to have a really wide open world. That's the first point and that's what should be used to manage the main quest.

The second point is that apart the main quests it should be series of independent webs, independent in a sense that a web won't depend of the other webs. It' exactly how it works with a design like explained Deepo.

That doesn't mean that a same node, for example an NpC isn't shared by multiple webs. On this sort of link the coherency is put only on the node, it's not combination on the whole. For example the node will have or not one additional dialog option depending of web 1 and two other dialog option depending of web 2. No full combination on the whole world, it's only the combination on the node itself, that's much more reduced. No node will ever concentrate all the webs of a RPG, that would be bad design.

Free-roaming is difficult to properly implement, but well worth the effort if done right. BG would qualify as free-roaming (IMO) because once you are out of Candlekeep, there are only a few areas that are not immediately accessible. You walk to the edge of one area, and then you can access the area next to it. I love that! One thing I didn't like about BGII was that you had to open up areas through quests or other dialogue. You couldn't just go out an explore.

Although Ultima VII is a classic, the sequel, Serpent Isle was free roam, and man, if you did some things out of order you would have to start over.

Serpent Isle is a classic example of free roam being "broken".

Free roaming isn't that difficult to implement. In my opinion Bethesda has laid out the BEST models for free roaming worlds so far with Morrowind & Oblivion.

The worlds were big enough (Morrowind was just perfect), and you have to admit the gameworld has great lore and a very diverse and interesting set of factions/races.

The problem is they dropped the ball big time with the thing that matters most to a lot or RPGers: choice & consequence. The other problem being the awful level scaling (Oblivion was the worst of the two).

The level scaling issue could have been very easily fixed. They could have just made it so the higher level creatures/enemies were out in the wilderness, furthest away from towns as most games tend to do.

And then perhaps allowed SOME of the human NPCs to level since if you're levelling others SHOULD naturally get better at skills as well. But instead of have it based on YOUR level it should be a natural time based progression, perhaps a level for every 1.5 hours of gameplay or something like that and of course put a cap for the type of enemy/NPC it is (naturally guards shouldn't outlevel their superiors).

Choice and consequence couldn't be too hard to implement either. Of course you can't have it run throughout the ENTIRE game world - well you could but it would be very difficult and time consuming to design/program - BUT at the very least the main quest should have many instances of C & C, and there should be some level of C & C for side quests, if not for all then for some.

The main quest should also trigger several plot points depending on your choices and make it so that at some point of the game you can branch off and end up with an entirely different ending than a different player.

And I don't mean ending as in which slide show or movie plays at the end. I mean the last couple of hours of gameplay/quests should radically alter. If you can have dozens of side quests in a game, why not have at LEAST half that many branches off the main quest so that you have a very different experience depending on your choices?

I would rather have an open world where an intricate main quest with lots of C & C is the focus and the side quests are just that, mini quests that are done relatively quickly (with high difficulty of course, and with rewards to justify the difficulty) and then have insanely high difficulty enemy encounters for those who like to explore.

Relayer said:

Free roaming isn't that difficult to implement. In my opinion Bethesda has laid out the BEST models for free roaming worlds so far with Morrowind & Oblivion.

Click to expand...

I agree about Bethesda doing the best job so far but with Daggerfall, not Morrowind (and certainly not Oblivion).

IMO, the fatal flaw of Bethesda's depiction of a free-roaming world has been its committment to an unremitting first-person perspective, two things that are somewhat at odds. With each new game in the series, the first has been gradually sacrificed in lieu of the second.

What it boils down to this: Depicting worlds in more and more detail leaves less and less to the imagination. So an imaginative world requires more work, since it has more to depict while a shallow one takes less, creating opportunity to go into greater detail.

I've thought about it a lot, and I think the best way to depict and experience a free-roaming RPG world would be similar to how great fantasy-adventure and science-fiction authors depict theirs -- more like experiencing an adventure and less like steering yourself around.

Clever intuitive fast-travel, selective partial graphic depiction, modularity and collaboration between the player and the game itself are the things I think of whenever I imagine a working free-roaming RPG.

I agree there needs to be a set of organizing principles and architecture to enable the development of tools to insure consistency within a large evolving free-roaming world with choices and consequences.

The complexities are too combinatoric to handle with only free form scripts.

I agree with wolfing - I find it hard to simply explore a world and "do whatever I want". That's why I prefer story driven games, where I'm so immersed in the story that I can hardly stop playing because I simply HAVE to find out more about Doctor Mysterious and why he keeps leaving little notes on every crime scene, indicating that my life might not be what it seems to be.

If I wanted to just run around, I could simply take a walk in the mountains.