Larian Studios Forums Baldur's Gate III Baldur's Gate III - Suggestions & Feedback Diary & Suggestions

Hello everyone !

First off, let me give thanks to the hard-working folks behind this game. This is a tremendous effort, and I would like to point out first, foremost and Most Importantly that I am very grateful for having the chance to play it. It's remarkable and brought me a day of joy that 2020 had been lacking so far. All of the things I'll say below are personal opinions, coming from a barely-amateur game designer and a pathetic example for a DM. But my words come from the heart, and my intent here is to add my brick to the wall, in a sense. I don't mean to point out for the sake of showing I've seen something I don't like. I intend on inviting relevant topics to further reflection.

In writing this, my hope is that I can ping-pong with people from Larian on some of the directions that were taken. I would understand not receiving answers however, I'm far from the only one here...
I'm also probably going to ask questions that I feel would better my understanding of the way the game was made. That's because D:OS 1 was the game which inspired me to make my own and there's still a lot I'd like to learn from you guys.

EDIT - DISCLAIMER : Spoilers below. Not many, I'll try to make it readable to everyone, but eh...

First impressions, localization and character options


Alright enough about me. Let me go though my 18 hour "Evil Run" so far... (Yeah, I took a day off to geek out as much as possible, so what ?).

When starting the game, it had it localized in French by our good friends at Steam. Understandable, so before switching to my preferred -english- I decided to check it out. The translations do requires some more work. Option menus' UI breaks significantly in 1920x1080 in French, which makes them unreadable. Some options of the Character Creator are not translated (namely the Tatoo and Makeup intensity bars) and the spells... Bane translation was the drop for me. So I switched it off right after character creation. Note that I would have done that in any case and do not wish to have the game translated for myself. (EDIT : Please, don't waste resource on localized voicing. I know the appeal, but having it at your level of perfectionism would just be too much.)


After some considerations, I went for Mephisto' Tiefling and Fiend Warlock. My goal was to follow Sven's advice and look for specifically evil paths to take and see what happens, so going full-devil would be good. Warlock is a class that's "learned-spells" which means choices in leveling up are important, I knew that. The Invocations and features make it versatile, but you better know what you're doing or you're going to end up with regrets in your build -specially given there's no respec so far. Even without access to the Blade Pact (which from experience is a big-time favorite of most people), the class turns out good once you have meat-shields.

EDIT : Note on the Build. Armor of Agathis and the Blessing of the Dark One do not work together at ALL. This would be one of the events in which you could "break" 5th Edition rules and say that "two different effects can stack temporary hit points, but the same effect cannot". So you can't cast Armor of Agathis twice, nor can you have lots of Temp HP by killing everyone, but you can have BOTH effects on yourself, without ... you know, wasting a valuable slot of Armor of Agathis and then killing someone...

During character creation I appreciated seeing the Trickster Cleric, that's a good sub-class pick. I obviously enjoyed the Tiefling subs that were available. The lack of Variant Human was disappointing, given the feats were already part of the game. Given the sheer variety of subraces in 5th Edition and the fact that WoTC is heading towards races being customize-able, I would argue that it would be a good thing for Early Access to go to the extremes in amount of choices, rather than look for only having balanced choices. Trying to balance a D&D experience is, after all, like trying to drown a fish... Not that you shouldn't try to balance, but maybe not on that point. I understand that the bottleneck is visual design of races however, so I'll be happy with what we can get. One example however : You have Goblins and Gnolls designed already, so give us those as an options from Volo's monsterous races.


Some spells are lacking, but really not many ! Given level cap, the biggest ones are Detect Magic, Knock, Enlarge/Reduce and Locate Object, which feel like they would add a lot to the game (damn breakable thieves' tools). Rope Trick being one of my favorite spells, I would argue there's ways to make it happen and would love to see it. The idea is spawning an "Entrance" to a scene where there's nothing and which only player can access. There are more spells missing, but they are less relevant to my mind. Suggestion and Enchantment I consider to be another topic entierly, which I'm sure will be covered.


Starting to play

Character, check. Awesome cinematics, check. Now to play the damn thing... I wasn't lost when I got to the first room. "Very Lariany.", thought-I. But if you do take the time to look and jump around, a zillion things will already happen to you in that room. I did so... and almost died from hitting a lamp, burned myself from my first misclick, and picked up a brain in a jar by using the "drag shhh from the environment, open inventory, drop" technique from Larian games. Overall, this alone made me very optimistic about the rest of the game. I don't know that everyone will pick up on those things, but you did provide "us Larian fans" with what we wanted from the 5 first minutes I think.

EDIT - Addition: One important thing I need to mention on the negative sides. The Camera is awful. I don't know if my crappy mouse plays a role or not, but still. D:OS 2 was a major improvement in terms of camera, I have not had any issue with it that I can recall, not even in EA. With camera mechanics likely being engine-based, it got me wondering if you had made changes there. It seemed to me than not because let's face it, the two games do look and feel very similar. So this is a "How come?" moment for me. Why no vertical panning ? Maybe terrain is more resource-demanding on this and you need to limit the field of vision for now. I don't know. But with the added "verticality" of the level design (which I love and plan on praising later on) makes for a lot of issues. I'm talking game-breaking issues. Not that the gaming community's not used to broken cameras, but since you've given me a perfect one before, I'm going to need a perfect one on this too !


The first dialogue, with the Brainy, didn't go as well for me ! Oh the dice, how they failed me already. Died a couple of times by initiating combat (Warlock class doesn't fight well alone), then on the third try I ended up with the only option remaining being to "destroy the brain". So no brainy companion for me. Not a big deal, although I'd make this thing hit a little less strongly. Also, its AI can have issues when starting in the initiative, so I'd probably review this particular combat encounter.


I won't go in details for everything, so let's say the rest of the Nautiloid part went fine, a bit linear but it does the job nicely as a tutorial. Sets up the Gith character fairly well too. Not many "Evil" things to do, so I stuck with getting out ASAP from there. Wouldn't want to play Chaotic Stupid. I did try to do a couple of things right before reaching the Helm, but given they had a Goody feel to them, I didn't insist. Tried to kill a Yellow Intellect Devourer and got a bit frustrated that I couldn't figure it out quickly. The combat encounter in the Helm was good with the added sense of urgency. I'll try to fail it on purpose next time just to see. One thing that could be considered by you AI team would be weighing the "Hostility" of enemies, that way you could have both the Devils and Mind Flayers as enemies, but ensure that they fight each other rather than the player... I feel like such a system would allow you to deal with multiple-party combats in an interesting way.


I'll stop there for now, even though I'm almost done killing the Gobbos and I've reached level 4, I need a minute to turn my ideas into somewhat of a Diary.
Hoping you can get something out of this block of text.

Cheers,
Robin

As a follow-up, I would have rather not double-post, but adding onto my first post's edits would also be strange. I will now move onto more of the points I want to touch on, rather than the Diary part, although I will heavily rely on my runs to exemplify. Again the goal here is for me to push some suggestions and for me to further my own critical analysis of a game. Any response is welcome.

Combat

I want to cover several points of Combat and the associated mechanics, because it is probably the most important point for me in the game. What drew me to this project so much is specifically that Larian is the only company in my opinion with the skills and resources to produce a good simulation of D&D combat. That is to say in part that I really am hopeful for a Scene Creator similar to the DM mode we've seen in D:OS2. Let me add Larian should look at this regardless of how well the one in Divinity performed, given they had no DM rules provided with the old one, whereas WoTC provides a lot now. The ability to create small sized scene with creatures, whip up a character build and play that encounter with some people online would be a major addition to the game's set of features, one which is not as far as it would for other companies to reach. This being said, I only have little knowledge on the actual roadmap set for leaving EA in terms of time constraint and so I won't go further into that.

There are plenty to be said about combat and how it moved away from standard 5e rules in favor of other features. This to me is a design choice, one I wish to change Larian's mind about. But I can't do that if I don't acknowledge that the decisions they have taken make sense. Yes, it makes total sense to rely on the Divinity Engine as much as you can. Yes, it would be good that your game is dynamic and feels fast-paced, given the previous two in the series were real-time. This being said, I don't think the directions taken are always the best options for the project. But when thinking about it, with what they have implemented already, reaching a correct 5e simulation would turn out to be "stripping" elements, rather than building some new ones up. And that's a good thing. Let me exemplify with Surfaces, Shove mechanics, Dip mechanics, and Jump.

Surfaces are a tremendous part of Divinity and yes, they do make for interesting combat. In D&D, surfaces that do exist through the Grease spell can be improved upon with Fire interaction. So both can and should work to complement one another. What I perceive now when playing the game is that surfaces were brought in very early in the development and quickly became an important factor in encounter balancing. Which is not a good thing. The clever use of surfaces should be player agency, and not something the AI so clearly relies on. Larian should in my opinion try stripping surfaces to the bare minimum, which is to say remove them from Cantrips and Arrows and decrease drastically the amount of "explosive cans" from the environment. From then, initiate building the surfaces back up, but parsimoniously, with Grease interactions and clever items. As an additional note, although I highly enjoyed the running water problem of our bloody ally, weshould probably get rid of the "Wet" condition and strip conditions to a minimum as well (haven't reviewed them fully, yet). Consider reworking the pathfinding specifically for our ally and have dialogues setup to warn us of danger. This sort of work if implemented should be generalized to other traits such as Sunlight Sensitivity. I've seen him die and struggled getting him back several times now.

The Shove mechanic. This touches on both level design and mechanic design. First of all, I highly enjoy the mechanic itself. Shoving is underused in D&D, and the added vertical-side of level design and the combinations it brings to strategy are nice to use. But in this instance too, Larian has taken a direction which differs from D&D and has high impact on game-feel. Also in this instance, the times at which using Shove surpasses the actual D&D options are numerous, which is in part due to the level and encounter design. One quick and dirty way to compensate would be to implement a Dexterity save on top of the current one, to avoid falling from a "cliff" if the shove gets a creature to fall. Setting Shove as a bonus action also seems like a way of making it more likely to be used. Something being underused -specially in D&D- is not always to compensate for. The reward for the success of a Shove is enough. There's also already plenty of situations outside of combat for it to be used, so players will have it in mind. Bonus Actions are set for a very good reason, and I'm curious as to whether Larian discussed this with Jeremy Crawford. Again, the feature is good, just needs to be balanced outside of the spotlights.

Dip - Divinity Is Prevalent... Sorry about the pun. Dip is obviously tied to Surfaces and I've made my point on those above. I do enjoy the fact it's a bonus action and that the weapon coating mechanic is in place, that's great and honestly didn't cross my mind before as a valuable thing to point out to my D&D players. Keep "Coating" in the UI as separate from "Using an item", or don't, to me both are ok. One small point however, "Using an Item" and by extension "Dip" could use breaking 5e rule. Wasting actions/bonus actions is frustrating in the game, and so I believe that if Larian made "Using an Item" as both an Action and a Bonus Action, we'd resolve two issues at once. We could have it item-dependent or not. I just think having the ability to use a bonus action to do something and an action to drink a potion or use some item wouldn't hurt the game.

Jump. Kudos on that. Although again, Disengage and Jump should be standard actions for combat purposes, at least the feature is worth it and used outside of combat, which is great. Interactions with the Jump and Feather Fall spell works, although I found myself more often failing to reach a down-cliff location that I did succeed. This made me consider the level design of the global map. Why do we not have access to Water ? The terrain is vertical now and there are potentially layers involved, so why not add a bottom layer which would contain 120m or so of simply water, that would give a Swimming condition (so Difficult Terrain) and allows us to jump off a cliff, find nothing and just swim back to the closest shore. Something to think on.

My final point would be two UI elements that I dislike : The Chances to Hit display, and the Weapon Attacks thing. The first one should be used, but updated, as I like having the display of what elements are influencing my roll. More details on those would be good too, for instance don't say "Disadvantage on the roll", but rather "Target is prone, ranged attacks made at disadvantage". The percentage is just not cool. It's useful, but it's not cool. Give us an overall idea of the DC or the roll's bonuses, but please, no % in D&D if it's not for the Wild Sorcerer.
The Weapon Attacks UI, with a special new action for each weapon I've simply decided not to use. I like the switching of weapon sets and the ability to pick weapon handling, but the 4 buttons above make no sense to me. I don't enjoy the special actions for weapons, or at least not as default. Why not make a feat that grants those ? I think these would be most useful if they contained the two set's available actions. As in, if you're dual-wielding on the first set, you have Action and Offhand to use on that side. If instead you wield a single weapon, they you get the Action only, and you get the added Great Weapon Fighting action there if you get the Feat. We could have the Booming Blade Cantrips there.


EDIT - Why we don't have a Dodge/Brace action is strange to me.

I generally have little else to say about Combat itself. The parts of the game that are D&D are done very well that I can tell, whether it be Concentration, most spells, the class features and so on. We could use not knowing the HP of enemies but relying on Bloodied status (Green/Orange/Yellow/Red health bars). The Threatened condition I'm not certain fits D&D's flank and I'm not very fond of, but those are optional rules and more minor points.

Let me finish with the fact I still enjoyed the combat tremendously. Surfaces is the only major blocker to me up there, and Larian got pretty close to what I want to be playing. Just would be so lovely to have it perfected.

EDIT - A note on combat resolution. I have the feeling at this time that combat resolution is one-way : Win, or Re-load. It's too bad, not to have NPCs knocking ourselves unconscious. It's also too bad not to have the ability, as you would in a D&D game, to use your Action for interacting socially with a creature, with the option to end combat non-violently. I feel like this will be a part of the development along with the writers, but considering social encounters and dialogues within combat itself would be a great addition. Feel free on this to take liberties with the conversation's duration, as we DMs often do.