Interview Divinity: Original Sin 2 Post-Release Interview at PC Gamer

[DraQ]

I hope its still an RPG and not something like Dragon Commander.

Dragon Commander was real fun though

The problem with Dragon Commander is that it's a bunch of individually good to great components joined together in a way where they prevent each other from shining. You have a cool dragon-with-jetpack mode, but it restricts your ability to order your troops and is also on timer, meaning that if you succeed hard in the RTS mode, you won't really get to see it. You have an RTS mode, which is surprisingly competent and works quite well, with good units and combat mechanics (but bad base building and lack of unique maps), but unit production doesn't mesh with ability to manipulate units in TB - with big enough TB army you can blitzkrieg the map without producing anything and whatever you have produced disappears when back to TB. Of course if you blitzkrieg the map, you don't get to be a dragon and play with cool jetpack. And of course you have TB mode, where if you succeed, not only you will have no opportunity to play with your dragon form, due to RTS timer, but you may not even need to get into RTS mode at all.

In short you have three different modes that undermine each other instead of working in concert. And lack of unique maps.

Plus I really miss the early "fantasy Homeworld with dragons and magitech airships" feel - rooting the action to the ground in some way was necessary IMO, but not at the expense of all the cool aerial stuff - it should mainly provide context and support/capture mechanics.

A Dragon Commander with well integrated gameplay modes (I would start with no unit production in RTS or segregated strategic units (which you bring over from TB map but cannot produce in RTS) and tactical units (not tracked in TB map, but produced in strategic units during RTS), always being in dragon form during battle and having the ability to toggle cursor), UNIQUE content, and, for example, surface/cloud mechanics from DOS could be delicious.

 

[Kz3r0]

They don't need you anymore average codexians very vocal minority, they have new friends now.

Actually Larian never specifically relied on the Codex and Codex sensibilities, unlike Fargo.
The Codex just fell in love with them as many others did.

 

[DraQ]

Anyway, my opinion on DOS2, piece by piece (still based on incomplete playthrough):

• Art Direction:
  Great, vastly improved over DOS - gone is heavy-handed WOW-like stylization, current style is not entirely realistic, but much subtler, with a lot of details and much better taste. Things - characters, items, environments look much more interesting and coherent, and some arguably tacky elements are well in the style of good fantasy artwork of old. Things like weapons have varying style depending on their origin - you have normal looking basic stuff, somewhat angular dwarven gear with lots of metal and frequent wolf motiff, lizard stuff with lots of curves, elven stuff largely made of plants intertwinned with mineral cutting surfaces, etc. Overall the game looks very good in more than just strictly technical manner - the game is just gorgeous. Some areas that could use improvement are lizard and elf faces.
  
• Writing:
  Good, vastly improved over DOS - it's much more coherent and interesting. The game has overall much darker and consistent tone instead of constantly whiplashing between extremes and robbing them of effect. Humorous elements are no longer disconnected and dissonant - they are sparser and more subdued but much better integrated and more effective. Pet Pal conversations that were often completely off the wall in DOS1 and simply didn't make sense in the context even accounting for human-level intelligent animals being not unheard of in fantasy now feel much more rooted and, well animalistic. Heavy handed treatment of animal death and suffering in DOS1, that was often dissonant with what you would expect in a faux-medieval fantasy world has been replaced with completely implicit approach that arguably works much better without actually being in your face. Merit of *descriptive dialogue* is mostly matter of taste but I find it a reasonable compromise between keywords and traditional dialogue and it alleviates problems such as not being given an option that fits your particular character by significantly broadening each of the options given.
  
• Content Density:
  Great, vastly improved over DOS (at least based on Act1) - there is ton of stuff to find - quests, places, items, encounters, etc. It doesn't feel like a giant multi-themed arena any more.
  
• Reactivity:
  Mixed Good - there is a lot of alternative solutions and things interacting in logical manner, sometimes blocked by things like your race, sometimes affected by deferred outcome of some earlier quest or interaction. OTOH there are some rather glaring examples where there is no reaction or even wrong reaction to the stuff that should absolutely cause reactions, for example entering the restricted area of Fort Joy from wrong direction after you escape may cause patrols to completely ignore you.
  
• Skill system:
  Good, vastly improved over DOS - skills are generally more consistent thematically, there are more skillsets, including quite interesting stuff like Polymorph, synergistic skills requiring more than one ability are also a welcome addition. The abilities themselves give more general utility bonuses (like Huntsman's bonus for attacks with elevation advantage applying to all ranged attacks) which meshes well with abundance of +ability bonuses on gear and encourages mixing. Separation of civil and combat abilities, each given their own point pool is also a good call in a very combat centric game.
  
• Attribute system:
  Still bad, somewhat declined from DOS - attributes are largely divorced from their RL meaning - you know something is fishy when you fight some dumb beast and it has higher INT score. This also applies to dexterity finesse affecting things like bows and crossbow damage and different kinds of gear requiring different minimal attribute score - finesse, strength, intelligence or constitution. Overall effect is that attribute scores make no sense whatsoever and it also makes some of the attributes completely useless for some builds - for example while Warfare can use any skill based weapon, Huntsman and Scoundrel need weapons using finesse. A good option would be to either replace the system with a more abstract one OR one more rooted in reality. In either case I would probably drop or replace intelligence as it is prohibitively hard to do it right even when you do include copious int checks in quests and dialogue, because it should also affects tactics which is player driven and thus cannot be affected by stats. Wits and initiative could be improved by adding some preparation (movement and inventory manipulation) phase before battles on success if forcing player-enemy-player order is necessary. Another option would be dropping the attribute system altogether and offloading its functionality onto abilities if it neither makes much sense on its own nor manages to complement abilities properly (for that it should manage to be mostly orthogonal to ability system which clearly isn't the case).
  
• Armour system:
  Disastrous when paired with cooldowns, merely bad otherwise - as an attempt to curb elemental mayhem and disabling status effects it sort of works and helps encourage limited movement through environment hazards, but in conjunction with cooldowns it makes many skills wasteful to use for a good portion of battle, significantly reducing tactical options. Instead a good portion of each battle is spent reducing enemy armour, often forcing player to focus attacks on one enemy at the time which is further exacerbated by the fact that armour types and skills affecting and affected by them are neatly segregated even when it just doesn't make any sense (for example taunt is somehow resisted by physical armour). The idea isn't entirely unsound and based on actual problem but currently cure is worse than the disease. It also lowers the impact of offensive skill use which makes combat less rewarding.
  
• Cooldown based skill use:
  Disastrous full stop - It's a bad system on its own and combination with armour system turns it into a complete trainwreck. It would be better to just add different circumstantial utility to different skills (and basic attacks which generally should be most effective as far as pure damage is concerned) and further balance using mobility (to uncluster ands escape AoE, for example) and AP cost, maybe pre-cast delay when needed. Maybe add some effectiveness debuff on repeated use (enemy observing what's happening and being less likely to fall to the same trick over and over).
  
• Elevation system:
  Good, could use improvements - It's a welcome addition especially given how it often encourages extensive combat movement. It could use some scaling based on actual elevation difference rather than being a binary thing. Another obvious improvement would be affecting skills differently based on whether the attack originates at the attacker's position (all sorts of physical and magical projectiles) or is summoned directly from the above or surface.
  
• Surface/Cloud system:
  Still great, then improved - It's basically DOS system buffed with ability to bless/curse surfaces and some other stuff. Blood clouds make no sense, though (it should be just steam), and water should either retain its dousing ability or there should be a standard action to drop onto a surface (and standard helping hand short range action).
• Elemental damage system (NOT clouds/surfaces NOR statuses):
  Still bloody stupid - it embodies the idiotic "color coded damage types" antipattern, with multiple symmetrical damage types that differ only in color coding and don't even try to pretend they are making sense - WTF is brown earth damage anyway? There is no need for an element to have it's corresponding (single) damage type. Throwing a massive boulder at someone should deal the same physical blunt force trauma whether it was accomplished via geomancy or in a more mundane manner - not some sort of nebulous earth damage. Similarly zapping someone with a lightning bolt should deal electrical and a dash of fire damage, not some "air damage".
• Stealth system:
  Inferior to DOS1 and pretty much stripped of combat utility. DOS1 implementation was very good and one of my fondest memories from that game was hiding behind a pillar with Wolgraff during a battle (one of the pirate cove ones, where skeletons first kill off orcs), then tripping enemies running past using skill. Probably the most rogue-ish thing I have ever been able to do in a computer game.
  
• Status system:
  Good, but mildly declined - most statuses are very brief instead of having good counters and armour system reduces importance of status effects in early combat. OTOH multi-stage statuses (like shocked and stunned) sem like a good direction to pursue.
• Level progression stat bloat (items and characters):
  Plain awful, further decline - it robs statistics of meaning (there can be no meaning to damage stat when HP pool and damage cause by the same kind of weapon can differ by three orders of magnitude) and discourages non-combat solution due to large level differences potentially making the moment when level up happens a difference between a faceroll and TPK. It also makes some otherwise cool items arguably bad investment. Thankfully it's mechanically separate from pretty much everything else including the rest of itemization, enabling modding.
• Itemization (base stats only apart from level-based bloat):
  Pretty decent, actually, improvement over DOS - randomly generated items are pretty diverse and have wide range of useful bonuses. A welcome addition are quasi-unique names for more powerful generated items. Some items now also have rune sockets allowing further customization. Could be somewhat less abundant to encourage use of uniques.
  
• Races:
  Good, obvious improvement over DOS - it's a no brainer that effectively adding another axis to character customization massively improves the game. Some of the features fall short of the hype though - I remember that lizards were supposed to get a set of bonuses and penalties fitting their cold blood and bad reaction to undead is mostly cosmetic as it isn't really affected by visual coverage of armor as long as all the pieces are present. Abilities like elven memory absorbtion through cannibalism or the fact that undead exist in all racial variants are nice. DraQ bonus for boobless lizards.
  
• Tags:
  Good - a welcome and robust system for modular customization of dialogue options and checks. Could probably use some refinement but it's on a good track.
  
• UI and pathfinding:
  Decent, mostly slightly improved but not nearly enough - main improvement is ability range shadow, OTOH the characters still don't modify their paths when a trap is spotted which is needlessly frustrating, characters in combat still happily run into hazards (default behaviour should be avoidance unless forced by shift-clicking), and paths should highlight segments passing through hazards. Some friendly fire abilities (like crippling blow with which I managed to accidentally off two party members during final Act1 battle) don't show their AoE, similarly bouncing attacks should indicate targets they would affect. White UI overlays are not legible on many backgrounds. Camera in dialogue is wonky - without zoom in option enabled it tends to zoom all the way out, with it enabled the zoom is too high and cuts of heads of any participating female lizard characters. An option to enable hide headgear out of combat is inexplicably missing even though it was in DOS and was arguably the most sensible setting.

I hope its still an RPG and not something like Dragon Commander.

Dragon Commander was real fun though

The problem with Dragon Commander is that it's a bunch of individually good to great components joined together in a way where they prevent each other from shining. You have a cool dragon-with-jetpack mode, but it restricts your ability to order your troops and is also on timer, meaning that if you succeed hard in the RTS mode, you won't really get to see it. You have an RTS mode, which is surprisingly competent and works quite well, with good units and combat mechanics (but bad base building and lack of unique maps), but unit production doesn't mesh with ability to manipulate units in TB - with big enough TB army you can blitzkrieg the map without producing anything and whatever you have produced disappears when back to TB. Of course if you blitzkrieg the map, you don't get to be a dragon and play with cool jetpack. And of course you have TB mode, where if you succeed, not only you will have no opportunity to play with your dragon form, due to RTS timer, but you may not even need to get into RTS mode at all.

In short you have three different modes that undermine each other instead of working in concert. And lack of unique maps.

Plus I really miss the early "fantasy Homeworld with dragons and magitech airships" feel - rooting the action to the ground in some way was necessary IMO, but not at the expense of all the cool aerial stuff - it should mainly provide context and support/capture mechanics.

A Dragon Commander with well integrated gameplay modes (I would start with no unit production in RTS or segregated strategic units (which you bring over from TB map but cannot produce in RTS) and tactical units (not tracked in TB map, but produced in strategic units during RTS), always being in dragon form during battle and having the ability to toggle cursor), UNIQUE content, and, for example, surface/cloud mechanics from DOS could be delicious.

I never once played it in RTS mode though solved all conflicts on the Risk map

And thus about 1/3 of the game mechanics and content was wasted on you. And it's not an RPG where it's par of the course - you'd have to play badly on purpose to even see it.

 

[Valky]

"Vocal minority" cuck my life into pieces, Swen.

 

[DraQ]

Note, I have not mentioned music for a reason. First, it's actually pretty good, second, it's inevitably and obviously inferior to that previous Divinity games. This isn't something that can be fixed, so it's not really fair to judge it as decline.

 

[jungl]

the entire game sucks. Im glad im almost done with it and can uninstall it from my harddrive. The game is riddled with so many fucking stupid design decisions from items, leveling, quests, exploration you name it. it makes arcanum look like a flawless masterpiece that can not be criticized. Larian has not made a single GOOD game since divine divinity.

 

[Tigranes]

the entire game sucks. Im glad im almost done with it and can uninstall it from my harddrive. The game is riddled with so many fucking stupid design decisions from items, leveling, quests, exploration you name it. it makes arcanum look like a flawless masterpiece that can not be criticized. Larian has not made a single GOOD game since divine divinity.

So why did you waste your life playing it till the end? Are you that desperate or can you not make simple decisions?

 

[Ausdoerrt]

I wouldn't be surprised if Larian's next project is an Oblibion clone.

But they already made Divinity 2...

 

[Xeon]

I kinda liked Divinity 2, was pretty good overall iirc, didn't play the expansion tho, iirc people said it was a lot better than the base game or something.

 

[DraQ]

Both Divinity2 and the expansion are must plays as far as I am concerned. They also demonstrate good use of Oblivion's engine.
And so far it seems to me that DOS2 is much closer to Divinity 2 than DOS1 ever was.

 

[Absinthe]

Actually Larian never specifically relied on the Codex and Codex sensibilities, unlike Fargo.
The Codex just fell in love with them as many others did.

The Codex is not a hivemind. It's just that megathreads can breed shitty polarized subcommunities/circlejerks.

What a load of shit, not one DOS 2 defenders bothered to tell me how wrong I am in http://www.rpgcodex.net/forums/inde...2-release-thread.118260/page-143#post-5327153, unless 2 butthurt ratings count as defence. Luckmann released his vast inner autism and went full sperg on how DOS2 attributes are banalshit compared to DOS1 and it too went pretty much without echoes. Multiple people went on and on how the armor system is decline without anyone bothering to even play devil's advocate. You can hardly have a sensible discussion on the Codex if you're a 2017 who's just talking shit.

One liners get you cool ratings, sperging over video games gets ignored.

Don't post that shit in a megathread you idiot. Start a new thread if you want to start a discussion on some analysis of mechanical failings in the game. I don't check out the D:OS 2 thread but if you started a new thread on this I probably would've had a look at it. I'm sure I'm not the only one like this.

 

[Ismaul]

"Oh noes, combat is ambush after ambush."
"Oh noes, the armor system prevents disabling every character on the first turn."
"Oh noes, the shit initiative system prevents one side to fully disable all their opponents."

If that's true, then fights beginning with an ambush aren't a problem at all, because you don't get disabled before being able to react. If the criticism is the reverse, that your party can't use initiative and CCs to fully disable the enemies before they act, then isn't that ok?

I agree that Initiative is less good a stat because of the way it's implemented, but doing it the other way has problems too. I also agree somewhat that there's a little numbers bloat on Tactician mode, but still, git gud: build versatile characters that can attack both physical and magic armor when needed, target low magic armor enemies with magic attacks, same with physical armor, learn how to maximize damage to get through armor (build, positioning, buffs), and use attacks that CC regardless of armor (teleport, oil...). The numbers only really feel bloated on Tactician if you don't optimize characters and have shit strategies.

What is true is that enemies have maybe a tad too much armor at the beginning of the game in Tactician, their weaker armor could be lower. Haven't finished the game yet so I can't comment on the end.

 

[lukaszek]

deterministic system > RNG

 

[imweasel]

"Oh noes, the armor system prevents disabling every character on the first turn."

Are you not aware of the fact that D:OS 1 has saving throws and even hard counters, or did you just very conveniently forget that?

 

[lukaszek]

deterministic system > RNG

 

[imweasel]

Are you not aware of the fact that D:OS 1 has saving throws and even hard counters, or did you just very conveniently forget that?

If you had trouble getting your skills effectiveness over enemies saves then you were playing it wrong

Case in point. A minority had problems with them, so saving throws and hard counters were removed in D:OS 2 to make the game more accessible.

 

[Fenix]

Basically people were happy playing glass cannons in dos1 with initiative to the roof. Started every fight by ccing whole army, now this tactic doesnt work anymore.
If you read older threads with dos1 build tips, most will be about neglecting your defenses and offense being best defense.

Basically people were happy to bit everything to bloody pulp without being retalated.
If they really were happy to play glass cannon, they could play like I did - rogue for pure damage, mage for buffs.
But I see, they complain that magic armor doesn't let them cc mobs.

Also, they can play glass cannon in DOS2 as well.

 

[jungl]

Beat the game yesterday, holy shit was the last puzzle and encounter annoying. I just realized that lucian was the main character from divine divinity. That is how much the story sucked. Playing nioh itemization so much better. Original sin 1 had better itemization also. Original sin 2 I had level 14 gear that was better then anything late game. +4 con armor=1000 extra hp I take that over 500 more armor level 20 armor provides.

 

[lukaszek]

deterministic system > RNG

 

[imweasel]

Case in point. A minority had problems with them, so saving throws and hard counters were removed in D:OS 2 to make the game more accessible.

I guess you know better what sven wanted. Some would say that he was annoyed by majority of squishes save scumming as their tactic was to cc the boss on round 1.

Some would say that Swen is simply trying to make his game more appealing to a wider portion of the market.

Judging from D:OS's high ratings and sold copies (over 2 million) they are probably right, because there really wouldn't be any other reason to butcher the mechanics.

TBH I am perfectly fine with developers making a game more accessible for casuals by adding a story mode. Just don't butcher the mechanics while doing so.

 

[lukaszek]

deterministic system > RNG

 

[imweasel]

its not the case with saves. It was replaced with deterministic system. Its new and different. Not worse/easier.

Not really. Deterministic systems in turn-based RPGs feel predictable and samey.

An overreliance on the RNG, however, is not good either. There should be some kind of middleground.

 

[Ismaul]

There should be some kind of middleground.

Wouldn't encounter variance overcome the samey feel that can be had in a deterministic system? You've got different enemy composition, different enemy abilities, different terrain, different enemy positioning. Also, you've got variance on the side of the player, with increasing abilities and equipment. Plus, the AI will act differently given those parameters. Is that not enough?

Deterministic systems are good for turn-based because they favor planning. Predictability is not a liability, but an advantage to be used by the player to plan his moves ahead of time, to make a chain of actions that synergizes, without getting fucked by RNG. You know it's time to get the CC ability out, because the target does not have the required armor to block it. You know that making two successive attacks that shake the enemy will stun him if he does not have magic armor. You know you have to boost your defenses, or you'll get CC'd. No reason to play the RNG / save scum to evade a CC or put one on the enemy. Make plan -> get result. Shit plan -> think it over and make a better one.

 

[DraQ]

Some would say that Swen is simply trying to make his game more appealing to a wider portion of the market.

Judging from D:OS's high ratings and sold copies (over 2 million) they are probably right, because there really wouldn't be any other reason to butcher the mechanics.

while some areas were butchered, its not the case with saves. It was replaced with deterministic system. Its new and different. Not worse/easier.

It's not worse because it's a deterministic system, although I will always maintain that the best way to combat too much reliance on RNG's clemency is adding more rolls - it's worse because it effectively makes large number of your abilities worse than useless for about half of each battle, because using them early not only will not have desired effect, but will deprive you of them when they will finally become useful (because of disaster that is cooldowns), so you end up spamming basic attacks and rare direct AoEs over and over which is boring.
It's also worse because physical and magical armors are completely separate and entire ability trees usually target either one or the other, so, unless you have hybrids, half of your party will effectively be useless to the other half. Not to mention that some status-armor pairings are just plain baffling, like taunt being resisted by physical armor.


It's not hard to suggest improvements, even keeping in vein of a deterministic, ablative armour.

For starters, let's consider what this "armor" is actually supposed to represent. It can't be actual armor, because it no longer degrades, so wrong metaphor for wrong thing. What it could look like is ability to defend yourself - largely facilitated by your gear's defence score, but not the defence score itself - which will progressively degrade as you are attacked to the point where you are overwhelmed and have to soak in damage and statuses from precise blows. So far it's identical to the system in game, so now the difference - the defence rapidly regenerates between turns - if you are no longer piled on by enemies, you can get your shit together and your defences back up. Of course the defence pools would have to be much smaller, switching from a resource that gets depleted on per-battle basis to a resource that gets depleted on per-turn basis, allowing use of CC early and often at the price of having to concentrate your attacks much more fluidly and dynamically in order to make them work (and it would add more tactical value to AoE, as it would allow to keep more enemies down at once instead of just offering choice whether to whittle down enemies all at once or sequentially). The % regen rate could be stat or skill dependent so a good fighter could effectively hold up to a lone crappy one indefinitely, regen would also stop when incapacitated. You could even bring back armour repair without making it too much hassle by letting armour health (small number!) degrade only when character's health gets damaged.

Second, mix magical and physical damage types across trees (hurling big rock at someone - magic or not - should definitely cause physical damage) and/or tie both armour values together - percentage for percentage. Say you have 20 physical and 5 magical. You get hit for 4 magical, you are left with 1 magical (20%) so your physical armour also gets reduced to 20% (4). If instead you get hit for 4 physical, you are left with 16 (80%), so your magical armor also gets brought down to 80% (4). This still makes some characters endure physical or magical damage better, but allows physical and magical attackers to support each other instead of merely avoiding getting in each other's way.

Finally, cloud/surface damage and statuses should be dealt per AP spent while occupying affected area (so you don't benefit from keeping still while standing in flames, but want to get out of them ASAP).

 

[lukaszek]

deterministic system > RNG