Larian Studios Forums Divinity General Discussion A single sameless map is unnecessary and counterproductive effort

As the title says.

We already discussed this in the past in relation to BG3 AND it's something that came up as passing comment in other threads of this very same subforum, but I think it's a good time to refresh the feedback on this specific topic and give it a dedicated thread.

I think Larian's insistence on trying to cram as much as possible on a single "regional" map is actively harming their ability to build credible worlds.

BG3, even more than DOS 1 and 2 before it, suffered a lot of what some of us on this forum baptised as the "diorama effect", where the environment feels more like a crammed toy miniature constantly wrapping around itself rather than a series of real places.
Which is how we got stuff like the goblin leadership struggling to find the "super hidden druid grove" literally placed 100 meters down the same main road.

While I won't take away from Larian the credit of being able to build great looking locations (sometimes even with very good layouts from a gameplay perspective) I genuinely think their world design would benefit from "fragmenting" their gaming world more, in a series of smaller but more abundant submaps, increasing the number of locations and connecting them through a world map.

Maybe even add some abstract "map travel" through it, eventually opening the way to the possibility to throw in some "randomized" -but not really- encounters to their usual formula.
Then again that would be separate design decision to made that doesn't change the main point of this thread.

I see no problem with Larian's concept. The effect you've mentioned happened not because they crammed a lot of places of interest together but because they've poorly disguised the grove.

Maybe they could space out those locations a bit more though, while keeping the structural integrity of the region. They would need to add some content to compensate for it however. Maybe use verticality and some new mechanics in order to create traversal "puzzles" aka real environmental obstacles. Or add mounts, but sounds too ambitious.

Well, thank god I took time to explain what the problem was, then.

You'd need to ask yourself if that would ACTUALLY be a better solution, though.

Is "we need more empty space on our single, seamless map" really ANY better than "What if we had more maps to begin with"?
What of the two would PACE the game better in your opinion?
What would allow for a more coherent set of different locations?

Maybe even more importantly, what would be more technically challenging to achieve, as well?
I'm going to guess that they are already pushing pretty close to the limits of what they are able to cram on a single map/load in memory in a single passage.
Having more and smaller maps would lower the technical restrctions they'd met while at the same time giving them room to detail and develop every single major location better.

Offsreening travel or location entering is bad immersion and breaks, not mends "PACE" in my opinion.

Remember the infamous Euron Greyjoy fleet from GoT? Or even closer to home, army of the Absolute 'marching' towards Baldur's Gate that arrives at the gates quicker than our small party.

Offscreening events is a narrative challenge, if they can pull it off then ok. I think it's safer to make a seamless map with places of interest that play off of each other in real time. Like it happens with the tiefling party secret back up tiefling scout NPCs appearing if you have a single party member at the grove while killing Ragzlin and the crew at the same time. It is really cool and immersive to see this kind of interaction.

Counterpoint.
Clicking to move from an area to another (and then having individual areas that are packed with interesting content) is FAR better than "walking five minutes through empty parts of the same seamless map only to sell a better illusion of scale", in my opinion.

AND that's saying nothing of the different levels of feasibility of the two solutions, which I already addressed in the final part of my previous reply.



This has literally notyhing to do with the point of this thread, though.
You are talking about narrative expedients here when I'm talking about structural design.

I am talking about implications of your suggestion being realised. It poses certain challenges for narrative cohesiveness.

I revisited your post and it seems to be more about world credibility (say, immersion) than structural design as I understand it. Do you want this thread to be a purely gameplay oriented discussion? If so please provide how Larian's level design falls flat in that regard. Apart from the Hidden Grove issue I've already addressed as well.

It's about BOTH.

More well-paced locations, with room to be properly distanced one from the other and fully fleshed out, rather than crammed in a 8x8 meter corner just because they need to be all part of the same seamless map.

The main issue with Larian's implementation is with scale.

Their areas are very condensed, which creates a narrative dissonance because of the lack of distance between things (Which also gets exacerbated by the zone changing between acts). This isn't exclusive to just Act 1 and the grove. But other things too, for example, why is their a town literally 2 feet outside of Baldur's Gate? In what universe is it logical to make a town a stone's throw away from one of the largest cities?

They have implemented some additions, like how Underdark, Grymforge and Mountain Pass are their own zones rather than cramming them into a singular map for their act... But then these (Especially Grymforge and Mountain Pass) are very small zones because of the lack of content in them (Mountain Pass being literally just the Lathander's Light puzzle, the Creche and that 1 group of undeads on the way to Elminster)

In games like DA:O or PotR, they exclusively do the "Every location is just a teeny tiny location that you travel to via a world map" thing. But it doesn't offer scale either, as the "Travelling" is just a glorified loading screen with the only other aspect being the chance for random encounters (PotR also had some time based interactions, notably early on when there's a scripted attack on the base that also signifies certain other locations being emptied of encounters)

A larger seamless world provides the best feeling of scale, as you actually travel distances personally so you feel the scale.

The major caveat is that in order to do scale, you have to implement ways to deal with the scale because it's not fun to spend hours travelling around a map (Even Eurotruck Simulator, a game about spending hours travelling around a simulated real world, cuts a lot of fluff to make routes shorter so it doesn't take you literally days to drive across the globe)

So things like, mounts, teleports (Like BG3's waypoint runes) and other travel options (In DOS2 your base was a ship, which presumably can sail to different ports. You can also offer caravans to "Fast Travel" between towns)

Of course, with scale also comes the other issue... Which is density. Which is a continual problem for open world games, such as Elden Ring or Ubisoft games. Where they want to make the worlds engaging by having density of PoIs because of the old Skyrim ideology of "You can wander in any direction and quickly run into something interesting" - Which leads to content shortages and thus a reliance on copy/paste (Hence all open world games reusing the same PoIs repeatedly).

So it often becomes a case of pick your poison. Do you want scale? Then you either get barren environements or copy/paste content. Do you want interesting environments? Then you get tiny worlds as everything is crammed into a small area.

World map location systems aren't the solution either as they tend to be small, barren locations and simply cut out the exploration entirely, turning the game into a theme park simulator where you just load into a bunch of unconnected locations, do like 2 things, then leave for the next unconnected location.

I think the single map design is beneficial for multiplayer. In my every coop session and freedom it offers does become helpful - in the very least, that for example a coop partner can hop into town and do some shopping without me having to tag along. It is also a unique feature and design of Larian so I doubt it will go away.

That said! If they stick to it, they could design areas with that structure in mind. Act2&3 of BG3 felt far more natural, then act1. A big city consisting of few large maps with stuff tightly packed together, feels like a fair representation of the setting - a countryside where to factions can't locate each other less so. Perhaps also changing perspective when exploring could also make those smaller areas feel a bit more sprawling.

Rivington is on the other bank of one of Faerûn's major rivers. It was annexed as a useful suburb sometime in the 14th century DR when Wyrm's Crossing was finished, but started outside of the city's jurisdiction as a trading hub for the farmers south of the Chionthar along with all kinds of businesses not wanted inside the city walls - like tanners, fullers, butchers or dyers and everything else considered gross or hazardous. Then you have traders trying to avoid road taxes, city tariffs and trouble with the Flaming Fist. Yet, this way the place is still close enough for the citizens of Baldur's Gate to profit from it, while the proximity to a fortified city still offers some degree of protection. Almost all major walled cities in Faerûn have these "gate towns" or slums that serve the same function, and that's quite the same with many major European cities of the middle and early modern ages. Even today many businesses and people like to move just outside of the major city where prices are lower while you can still use most of the infrastructure. In my area, you're not going to find any non-built up places any more. Note, that the potter's workshop was also on the other side of a body of water in Reithwin since it posed a fire hazard and produced toxic fumes.

Concerning the question of maps, I really can't make up my mind. I didn't expect full realism from BG3, so I just went with the flow, while knowing that it's probably a few days' journey into the Underdark and not a short elevator ride. I knew everything would be less complicated and flashier than the AD&D tabletop game I was coming from. Thus, I took a Mountain Pass in a region completely devoid of mountains, fast-travelling and a condensed world as part of the game flow which worked for me. I stop describing miles of empty road to my players as well, when I think they've got the point. A casual gamer may have never found the grove if it was buried in a maze of forest. Druid groves are also usually very well hidden with magic in the original game lore, so it was no immersion breaker for me

I guess any solution to travelling the map can serve to frustrate different gamer types, getting lost in a frameless pretty painting, paralysed by possibilities or endless loading times/other performance issues. For me, the most important thing would be that it suits the game's narrative flow which in BG3 works for me.

Agree with the above . When you talk to Minthara in the Goblin camp and she interrogates you about the grove's location, So storywise, the region is much larger than shown on the map, and you'll have to use your imagination to make things fit.

Yes, businesses like to move to the outskirts, especially industrial ones. But that's the thing, they're on the outskirts of the city limit. They're not a separate town outside of the city limit.

Businesses and residents will always want to be part of the city to benefit from its infrastructure. With it being more notable for a place like Baldur's Gate that literally has a gate that can and does seal off access to the city proper.

I get that there will always be people setting up outside the city, to profit from the traffic despite their inability to enter it. But this doesn't usually turn into a full on town.


Not really. Could easily have been lead by Aradin's group the same as the goblins were. Much like how approaching the Grove as it is currently implemented leads to your characters hearing their ruckus and deciding to go check it out.

Heck, you don't even necessarily need to actually encounter Aradin's group directly... A trail of scuffles and splatters of blood could serve as a guide.

It's pretty basic design to facilitate lore friendly guides for players (That doesn't involve copious amounts of yellow paint...). If it's something important to allow players to find, you can make it easy for players while maintaining the logic of it being difficult for other people who didn't have such serendipitous circumstances to aid them.


Ehh... Not really. Like, most of the issues you bring up are not really to do with map design and are more problems due to a fixation of graphical fidelity (Which can make routes blend into the background and increase the performance cost of loading)

The only real hurdle for maps is travelling fatigue. That's the core of map design challenges.

Keeping players engaged with the game while travelling is the key.

The most obvious design for this is PoI density. If there's lots of things to encounter, then travelling isn't boring because you're not sat around doing nothing for long periods of time. The caveat to this is the amount of resources it takes to incorporate scales with map size.

The other prominent design choice for this, is fast travel. Options to "Teleport" around the map so you don't actually have to travel (Or alternatives like making "World Map Travel" designs where travel between areas is done as a dotted line moving on a map over the course of a few seconds). The caveat to this is that it doesn't always fit into the lore or can feel jarring with how disconnected you feel with the world when you just zip around all willy nilly.

A somewhat popular design choice is also to make travelling fun. Stuff like having solid vehicle mechanics, or fun things like grapple hooks and gliders can make the act of travelling enjoyable so that people don't mind travelling long distances because they still have fun doing so. The caveat to this is it's hard to pull off well and is generally only really possible in ARPG's (Or actual driving games like Driver or Burnout: Paradise) where there's more control over character movement and mobility.

A lesser used design choice is also autotravel. Things like Flight Paths in World of Warcraft, the autopath mod in Skyrim, the autodrive mod turned feature in CP2077 (Along with its Metro system), Taxi's in GTA etc. The ability to select a destination and have the game do the travelling for you is a peculiar option for the challenge. Since it doesn't provide player engagement, but rather provides players a breather. They can sit back and take in their surroundings, or even step away to get a drink or go to the bathroom. The caveat is that this doesn't help people who are unwilling/unable to take a breather and can thus contribute to travelling fatigue in people who want to continue to actively be playing the game (So it's usually best implemented as a secondary/tertiary option for travel rather than the primary means of moving through the world)