Josh Sawyer - Breaking the Mold RPG Evolution

[Music] [Music] you [Applause] [Music] all right so I gave this a very fancy title breaking the mold RPG evolution and paradigm shifts and pillars of eternity two dead fire really what this should be called is we tried to do some new things and role-playing games some of them worked and a bunch of them didn't let's find out how all right what in the heck is this talk about this is ultimately it's a post mortem four pillars of eternity two dead fire it's about looking at the assumptions that we made going into the sequel and what proved right about those assumptions and what proved ultimately to be very wrong for those of you who are not familiar dead fire is the sequel to pillars of eternity these are both crowdfunding games that obsidian entertainment made there are four PC they're party based isometric role-playing games in the style of Dungeons & Dragons Arab I Swindell Baldur's Gate and Planescape torment they have a Dungeons & Dragons style character development system so you advanced over several levels with different classes pick various abilities as you go up in level we use 2d backgrounds with 3d characters and lighting systems and they were both made in unity with our own proprietary tools especially for game data editing and dialogue what are we going to be talking about so first we're going to talk about the feedback that we received on pillars of eternity because the feedback that we receive proved to be the foundation for our expectations for the sequel some of that was good and some of it turned out to be very bad changes that we made based on that feedback and the reasons for doing so and how we made those changes I'm gonna talk a little bit about the backer beta which was fairly short the launch itself and what happened after the launch I'm going to go on to some controversial features that we approached during development some of these were controversial within the team most of them are controversial in the community but all of them are worth talking about I think for for the benefit of development in the future and also a long-term post-launch support last week sorry two weeks ago marked the one-year anniversary of dead fires launched we released our 5.0 patch obsidian has become very dedicated to doing long term support for its tight especially pillars of eternity so we did a lot of post launch patching and DLCs and that made a big difference in the long run little note here so I'm gonna be talking about decisions that I made that I'm completely responsible for there are other decisions that were made above my head it might sound like I'm throwing people under the bus I am a little bit but in order to actually do a post-mortem on this and really talk about the decisions that were made and why I felt the need to sometimes refer to decisions that were made above my head because that will give you a complete picture of what actually happened during the development process for those of you who do not know who I am I'm Josh Sawyer I'm the studio design director at obsidian I started at Black Isle Studios in 1999 I worked there until 2003 I worked on primarily the Icewind Dale series of games which is one of the inspiration points for pillars of eternity and dead fire between 2003 and 2005 I was at Midway San Diego working on a gauntlet game that did not go super well then I went to obsidian I've been there since 2005 either as a senior designer a lead designer a game director and now for the past few years the studio design director currently I am between directing projects I'm helping on supporting internal teams as a studio design director and I'm also working on the pillars of eternity tabletop role-playing game I did not expect I would actually ever get a chance to work on a tabletop role-playing game professionally but now I am and it's great so I'm going to talk about some assumptions that we had going into the sequel that's Sven bingo there by the way we are in a space with divinity original sin to kind of spend from Larian it's a very great guy he's a very forthcoming person with any data that he has about the success of divinity original sin and original sin - and one of the things that came up when we were talking with him is he had mentioned that there was actually relatively low overlap between the audience's for original sin and pillars of eternity which we thought was kind of interesting but then again they are fundamentally different games from a combat perspective which is a big deal to a lot of people divinity games are turn Bay pillars of eternity is real time with pause because that's what the Infinity engine games were divinity original sin to divinity original sin to the first game reviewed well sold very well the second game reviewed incredibly well sold phenomenally well so we were like well we reviewed and sold well so if we have a similar approach to refining the formula of the game we will review even better and sell incredibly well Wow wrong that did not happen but it worked out pretty well anyway so now we're going to talk about how we looked at these criticisms and tried to move forward one of the first criticisms by the way when I when I talk about these criticisms I'm only listing the ones that really seem to come up a lot in user reviews and in professional reviews there are many many other complaints that are totally valid that people could make about these games but these are the ones that kept coming up and we thought it was very important to try to address them one of the first is that combat was hard to follow again pillars of attorney was a real time with pause game when Baldur's Gate came out it was kind of a revolutionary that it used real time with pause for RPG combat for most people it was the first game they had seen that used a system like that RTS a--'s were very popular in the late 90s early 2000s not very popular right now so when pillars of eternity came out obviously all the veterans of those games adapted fairly quickly but there were a lot of people that had no experience with this type of game the pace of the combat was very fast and there were a lot of problems that we tried to address one of the things that we approached was how we rendered a visual effects as you can see from the big mass of color in the middle of the screen that's a couple of spells going off what's going on under there who knows um that was an experience that a lot of people had playing pillars of eternity where they would cast a spell it would go off and then they would say like I think someone's dying in there but I don't know so we actually changed our rendering pipeline to reduce that visual noise and helped a great deal we have unengaged a mechanic in the game called engagement this is a crowd control mechanic that tanks can use to lock people down it proved to be noisy in the original game you've had a lot of UI feedback that was really confusing to a lot of people so we made changes to how engagement worked to just reduce that overall clutter it also helped to make the tank characters a little more distinctive from the non tank characters and then the most controversial thing that we did is we moved from a 6 character party to a 5 character party this was one of the more criticized things in the game because I Swindell in baldur's gate were 6 character party games and that felt very traditional to a lot of people moving to five characters just irritated a lot of people but ultimately it was an important change for us to make to help make the combat easier to follow and finally we had an AI customization system the customization system really added a lot both for players it allowed them to build their own custom AI sets but it also allowed us to do a lot on the NPC side the monster side so for players it made combat easier to manage because they could automate a lot of things and for us it allowed us to make better AI behavior and that helped a great deal this is just a contrast right here this is a typical battle actually that's not a typical battle that's a big battle from pillars and here is a battle from dead fire the visual effects rendering changes made a big difference there's a lot of visual clutter that's gone from engagement and things like that it did help a lot another criticism is that the stronghold felt very disconnected from the main story i think that was a totally fair criticism we included strongholds because people liked strongholds and it turned out we didn't have the resources to really integrate them as well as we wanted to so we decided to make the stronghold a ship this was something that we argued about for quite a while at first we thought about doing multiple little strongholds all around the archipelago ultimately it seemed like it made more sense to just say look you're gonna have to travel by ship around the game let's just make that your stronghold it's necessary for traveling so you're gonna keep going back there it's a convenient place for all of your companions and other little NPC buddies together so that way if you want to swap party members out you're always going to wind up going back to the ship at various points and it's very appropriate for a pseudo age of sail game pillars of eternity is not it's really not a medieval game it's more supposed to be a early modern game 16th century plus and so going into the age of sail felt like a natural from the original game which was early 16th century another big criticisms that the stronghold mechanics are boring also can't argue with that they weren't that interesting so we added a ship crew and navigation mechanics that was relatively straightforward the crew was an easy place to put NPCs that you found for example if you had a quest that ended an NPC might join your crew as a reward that seemed like a really cool neat thing to do and then we got into the ship to ship combat system which is again one of the more controversial things that we added and we're gonna talk about that later voiceover was very inconsistent pillars of eternity is a big role-playing game with tens of thousands of lines of dialogue and it's very expensive and it's very time-consuming to voice all that dialogue so we had partial vo for some characters almost none of the characters in the game actually had full voice over this was not liked by a lot of people only the chapter breaks were even narrated so any of the prose text that we wrote was just plain text that you had to read streamers which were starting to become very prominent in about 2015 when this game came out they really really hated seeing a lot of unvoiced text and it turned a lot of people off from streaming the game just kind of frustrated people and right in the middle of dead fires development divinity original sin 2 came out and it had Fulvio for everything so that set a new standard I hate them for it they did that though and we had to respond to it we couldn't ignore it and we're gonna get back to that too there were too many trash fights partially my fault actually largely my fault there were does everyone know what a trash fight is it's a fighting role-playing game where basically it just feels like it doesn't offer anything tactically or strategically that's different from the fights around it it feels like a waste of the players time just to drain their resource you go through a dungeon and you'll have like you know every square inch-- is populated by essentially a carbon copy of a fight that you've already seen there were a lot of these unfortunately in pillars so in debt fire we moved so that block outs that people did for their levels had to including counter maps so we could see where the encounters would be documents also had to define those encounters so for example if someone said well I only have three encounters in this level and then we found out that they were all essentially the same encounter right so I cut one of those encounters or change it or change both of those encounters into something else but a lot more work went into the upfront design of saying please don't flood every level with trash fights because the players they just don't enjoy it and then encounters received a system design pass so we have area designers area designers know the systems of the game to a certain extent but the fine tuning of combat was more on the system design side so the idea was when an area designer would put an encounter in a level than we would later have a system designer go through and they would do a pass on tuning that unfortunately there was so much system design work to do that only a small number of encounters received that tuning passed prior to launch and that turned out to bite us in the ass because the game wasn't tuned very well the environments felt too static again this game was a 2-d game with 3d characters in pillars of eternity none of the vegetation was dynamic there wasn't dynamic cloth really except for on the characters the lighting felt a little stilted and often the shadows and the characters didn't integrate into the background it got what I would call the scooby-doo effect where you can see the the moving characters that feel very separate from everything else in the scene we approach this in a couple of ways some of it was graphical and some of it was behavioral so we added a new NPC scheduling and behavior system this allowed characters to sort of have their own lives and routines it made the environments feel more alive because instead of characters just sort of standing in one place all day long they had their own routines that they would go through this was mostly used in cities and things like that but it was also used for monsters and patrols and then we did a lot of rendering work together it's funny because like once the game got to this point its it kind of just looks like a 3d game now that just has an isometric camera but that background is actually all 2d but it has 3d lighting integrated into the 3d characters and everything like that so we move to physically based rendering we moved from unity four to unity five it seems like a no-brainer to go to physically based rendering and start using various tools to just make everything look a lot better on the character side we introduced vegetation in cloth physics all the trees there in the background those are all dynamic and they blow with the wind the wind was the dynamic element that would move cloth and vegetation it added a ton a huge amount of work went into unifying the character shadows and lighting into the world so that when shadow when characters moved through darkness they really felt like they were integrated into the environment instead of being separate from them we had a parallax Lee excuse me we had a parallax layers parallax layers obviously this is very old technology going back to like 8-bit games and even earlier where you have a scrolling layer in the background or multiple scrolling layers that move at different rates of speed this is most useful in areas that have great height where there's a big distance to something down below it was actually a surprising amount of work to get parallax layers to work given our weird rendering stuff but it did add a lot in the areas that we used it and then we had a new water technology the water in pillars one was basic and functional we licensed a water technology called Seto and then we altered it pretty significantly it made the oceans look really beautiful so that when you're on the coast or on a ship which is a lot of the places in the game the water really looked great yeah so this is a criticism of both games we did our best to try to reduce them but it was a very difficult challenge load times were very bad this is especially difficult in areas where you were going in and out of buildings over and over we switched to a hub based loading system so for example in a city or any area that had a large interior with several interiors we would take all the child areas and cache some of the data about those areas that when you transitioned in and out that data was already sort of pre-loaded that made those transitions in big areas a lot easier there was a ton of memory that was being taken up in a ton of loading that was happening with unity data all of our gameplay data all of our spells all of our creature were in unity asset bundles and we would load those dynamically every time we would load into an area that was very time-consuming expensive so we moved all of that out of the unity editor into our own standalone editor that was using XML or JSON so it's text all that stuff would load at the start of the game and that that helped eliminate a lot of the load time as well and then finally the save load system was rewritten this might not seem like that big of a deal but when we rewrote the save load system we also reauthorization trying to reset all these variables and do all this crazy stuff companion reactivity was bad I think people liked our companions overall but there was a lot of plate there were a lot of places where companions either wouldn't react to each other NPCs wouldn't react to them they wouldn't leave the party it was kind of like it felt like they were always along for the ride but they weren't really fully integrated into the game as well as they could have been because obsidian is known for writing this is something that we thought was very important to address so right away we had more in-depth budgeting for writing what I mean by this is that all the writers sat down and we said okay how big our companion is going to be how are we going to break up the writing responsibilities in terms of their hubs which is their main question node their interjections in conversations their banters with each other the relationships you can have with them their personal quests so we broke those down into ranges of nodes nodes are the unit of measurement that we sort of talked about when we talk about how much writing there is to do for a character and we budgeted that very early we also staged the writing so on most of the games I've worked on all the companion writing was done very late in development usually because we reasoned oh well when it's done very late then we can react to everything that we've written earlier in the game the problem was companions are huge and that's an enormous amount of work so after we did the budgeting we also broke up the writing for companions into three phases that made the most sense for development this removed an enormous amount of burden from us because we could focus in each stage development on a different aspect of companion writing and it made the the tail end of a companion writing a lot easier for everyone we also include a companion leave conditions companions were no longer going to follow you to the ends of the earth if you pissed them off they would get into an argument with you they might just get up and leave immediately but they were unique to each character and it added a lot of sort of depth to them that was lacking in the first game and then the final thing is that we added a new topic system with reactions and relationship break points this did not work very well I think it was a cool idea and I'm gonna talk more about it later a lot of people criticized that it had a very slow start pillars of attorney had a very slow start and your motivation was very ambiguous the mean villain of the game was kind of hidden for most of pillars of eternity he was a very cool villain Theo 6r Catan but exactly why he was doing what he was doing and why you were chasing him was not super clear so from the beginning of data fire pacing was a priority the very beginning of the game is your castles destroyed and this God breaks his way out and you're almost killed and then the god of death holds your soul ransom and says please go find athis or I'll just snap your soul into nothing so very clear very clear motivation for why you're doing what you're doing áthis this big big dude with the glowing forehead he is the main antagonist he's obviously the main antagonist right away and we did iterate on the pacing a lot so especially at the beginning of the game there were a lot of times where we would go and we would cut out small loops we would cut out notes of dialogue just to make sure that the introduction to the game was a lot faster and smoother than it was four pillars one the factions were underdeveloped in pillars one for those of you who played pillars one you meet the faction starting at act two and defiance bay and when you leave defiance bay you will never see those factions again and if you play something like Fallout New Vegas the factions are present almost from the beginning of the game all the way to the end so that was a misstep dead fires factions are there pretty much throughout the whole game as soon as you get to ma che island do you meet the wana you eat the prenup or not the Prince PU meet the villian trading company shortly thereafter you meet the principie and then the royal dead fire company so very early on you meet all four of the factions and they're there with you almost literally up to the very end of the game so we did a much better job of including the factions they're also central the political conflict of the dead fire but that did confuse the plot this is my fault I have to take full responsibility for this so we had this very strong start with athis going blank I'm in the knees you know like going across the ocean and the gods like you got to follow that dude and then as you're following him there are all these factions that are fighting over the dead fire and players logically we're like so I'm following that guy it seems like there's a lot of fighting going on here am I supposed to be helping figure out who controls the dead fire or how does this tie together and I didn't do a very good job of actually tying those things together there were two cool plot points that didn't mesh very well but overall we did a lot more planning and plotting about what the factions should do I think that helped a great deal and every faction had a companion champion this was extremely important so all the characters at the bottom are associated with a faction and they have very strong feelings about that factions role in the world depending on the alliances you make some of those characters might leave you for example pala Jeana is extremely sort of hard-ass about being dedicated to the villian trading company but there was actually something that a long time ago Chris Avallone had made a very strong point where it said like you should always have a companion that is a mouthpiece for one of the factions in your game so we did a better job of that in this game we did have a backer beta we had a backer beta on pillars one as well there are some tricky things about backer betas it's very limited content and it's a very limited audience I think anyone who's done early access also knows that there are dangerous to this for those of you who want to do this type of thing understand that your audience is self selected to the most hardcore enthusiastic people they give very good feedback but it is also feedback for psychotic gamers like they have the most strongest opinions they're like they're gonna just grind your game into the dust it is very valuable but it can kind of skew your understanding of how your game plays for a wider audience um also the longer a beta goes on it becomes a well trod path um that content that you put out unless you're gonna put out a bunch of new content every time you update the backer beta people are gonna just run that over over and over and over again and it can lead to a very strange perception of difficulty for example or quality of writing because no writing really seems that great if you've seen it 20 times so there's a little bit of a problem sometimes with content that gets overused and over iterated on for us we didn't have we didn't have any companion or stories elements in the game we wanted to keep that under wraps so we didn't get feedback on that also our this is our earliest content and that's often the the worst content and sometimes it's the hardest to iterate on I know a lot of times I've worked on a game and the very first stuff you do not only is it sort of bad in execution but when you get under the hood it's such a mess that if you try to like pull out and redo it it's a problem so sometimes backers can get the sense that you aren't really doing very much with the content when in fact it's just hard to do anything with it it also personally warped my sense of difficulty because I actually had this opposite effect where because everyone who's playing the backer beta was really good and aggressive and they played it so many times and they said all this is so easy I was like you're just saying that because you're a super crazy gamer and you've already played this 20 times I'm sure the difficulty is fine wrong it was not fine it was actually really easy which we found out when the game came out so what we heard in the back row beta a lot of the new mechanics were confusing or punishing we did get great feedback on that we made a lot of good changes based on player feedback one of the big things is that everyone just moved way too fast this was also a criticism and pillars one that we didn't I didn't really heed as well as I should have we did slow characters down a lot just physical movement speed and the shifter ship combat sucked we're gonna get back to that but people were like this sucks and we're like how about this iteration that sucks - how about at the game launch it still sucks and there are a lot of technical issues so it was worth doing they did really find a lot of stuff both from a gameplay perspective and also a technical perspective that helped the final game in the long run so the game came out yay well I'm not going very fast time I'm oh well so what are we here the game is way too easy there were lots of performance problems the Fulvio was awesome everyone loved it the main plot was too short which is sort of true if you just crew straight through it can be a very short experience the game itself is very big there are ambiguous stakes in the main plot again there is this confusion between the main plot and the sort of the factional plots the companion relationship mechanics were crazy and didn't make any sense at all and a lot of people we had these minor characters called sidekicks they weren't full companions a lot of people they always want more companions so we had added these mini companions we said we can't write full like we can't write 10 full companions so here's a couple of minor companions called sidekicks and people are like we want them be full companions why did you include them at all which you know is questionable I still think that the game is better for having them but people did want them to be fully fleshed out Hey another the ship ship combat is bad the reviewer said it was bad the players said it was bad so in the short term after launching there were a couple things that we couldn't really do anything about anything that had to do with the main plot and the villain you know that's that's my fault that's how I wrote the main plot of the story there's not a whole lot you can do in the short term to fix you know it's just sort of fundamental structural issues with the game and the story ship-to-ship combat we did make changes to the UI we made balance changes that helped again though fundamentally people who didn't like the system just we're never gonna like it the companion relationships seemed all over the place the tuning was really crazy so we did we did some tuning we did some new tagging of relationships I'll go into that the most important thing I think we did is that we displayed a record of changes between your companions relationships because a lot of times people would get to a point where they'd say like why do these characters hate each other I have no idea why so the companion relationship page would actually show all of the steps that had led to that point there wasn't any silver bullet for performance optimizations we did what we could but we found that in a lot of cases our AI especially in big areas like NACA taça the main city was just really atrocious there wasn't an easy way to split those levels up we didn't want to pull a lot of NPCs out and potentially screw things up so that was difficult to deal with and then difficulty there were two things that we did if we found what I call the tall blades of grass we would cut those so people hate this people you everyone knows that people hate this so do it as early as possible if there's one character build that is just ludicrously powerful and like nine out of ten people use it because it's just such a joke while at the same time saying I hate how easy this game is nerf that build like very quickly just change it cut it out as quickly as you can the other stuff that we did is we started doing all the encounter tuning that we should have done earlier so now we're going to talk about controversial features the five character party again we had feedback that combat was hard to follow larger parties usually mean that there are more enemies in the battle makes sense pillars characters are more active than characters in Baldur's Gate and I Swindell on average characters like fighters and rogues and Rangers in the old infinity engine games didn't have a whole lot to do in pillars potentially they have a lot more to do that means that you're spending a lot more time managing every character we found that our UIs were cramped by the in the HUD area and in the inventory they're really cramped by having six characters also we followed the same sort of approach that Baldur's Gate and torment did where you start with the main character and you add more characters over time one of the struggles with a system like that is that you have to work very hard to keep the players attention with regard to pacing you're starting with one character doing a very small number of things then you have to then you have three than you have for then you have five as you're getting more characters they're advancing in levels and they're getting more things to do per character so does this really crazy explosion of potential maintenance that you have to do so this is a big challenge in pillars of eternity you would usually get a dare as your first companion followed shortly by a LOF or vice versa the last companion most people got was sigani and because she was a ranger she actually added two characters so your party was 7 which was just a lot to manage and a lot of times we heard people give feedback they said I always forgot about one character there was always one character in a battle that I just like I just forgot what they were doing they were just Auto attacking over in a corner you know like shooting their wand or whatever and that didn't seem great so implementing this was actually really easy we were already planning on rescanning the HUD and UI anyway lowering the party limit was not hard the battle sizes came down we had fewer enemies per fight because there were fewer characters to fight against and with all the other changes that we had made based on previous feedback the combat had become a lot easier to follow the reception to this a lot of people grumbled about it grumbling though here's the thing it's ok like if someone's like oh that's dumb I don't like it but they're not like really angry if it actually is making a good change in the game just just do that just keep going there were some people that really really really were mad like I'm not playing this game it has fewer than 6 characters which is you know a weird thing in this world to be that mad about but there were some people that really didn't like it and that's just how they how they felt um the logic was like well baldur's gate nice when Dale had 6 character parties Ergo and pillars won at a 6 character party ergo kind of ignoring the fact that we had received all his feedback like wow this is a lot to manage overall people said that the combat was easier to follow so what I do it again yeah probably I mean a lot of it depends on the way your combat is structured if we didn't have real time with pause combat maybe six characters are fine maybe maybe 10 characters are fine go play some battle brothers get 12 characters in there you know like who cares you could have huge battles if you really want to but a lot of it depends on the context but for this game I would totally make the choice again the companion relationship mechanics I'm gonna dive a little bit into the weeds here just because there's a murky system here that was very confusing and frustrating for people what are you talking about the companion relationship mechanics so the way that companions would change their relationship to you and to each other was largely through the system called topics so you see these little icons around palutena here they represent things like being prevailing Republic or pro-ana man see or anti-religious these are things that she likes she likes people that are anti-religious she likes people that say Anna man she is cool every time a character says something that's tagged with that topic she goes yeah and they're sort of like in a procedural and not on a moat but basically a reactive line or pal Geno's like yeah and and then her reputation goes up with that character and then over time that builds toward a break point either positive or negative and then once you hit that break point you know there's an emotional sort of explosion the character is like I think you're awesome thanks for being so great and hating religion or they say like you know like I love you you're the best and then you as the player can either interact with that or just say mm-hmm sorry I don't want to deal with it which actually a lot of people did there were certain conflicts where players are just kind of like back away and let things take their course so what problem was I trying to solve I believe that there is a space between doing purely hand scripted reactivity for characters and things that feel more topical and I don't want to say procedural but a little more mechanically driven if you look at something like fallout 4 you have companions that will react to things like picking locks or stealing which are sort of devoid of context and that's mechanically driven on the other hand if you have a thing where characters only respond to very specific lines that every character says that's a lot of work and it feels very special Casey like you have to take certain routes through the game to hit a certain breakpoint with a character so what this was trying to do is say hey there are all of these topics that people talk about throughout the world let's use the fact that the player makes hundreds of dialogue choices that sort of intersect these topics and companions make tons of offhand comments and use that to do mechanical work behind the scenes so the player will feel that it's a more organic way of building a relationship so that when two characters decide that they're buddies or they hate each other's guts the players like make sense because every time this character keeps saying something that's anti-religious this other character keeps going mmm until they just explode and go crazy and the players like yeah I saw that happening makes sense so the implementation was actually really easy the mechanics of this were very straightforward um but the designers had a difficult time really understanding it part of that is because I designed a system that was probably one layer of complexity beyond what it needed to be you had a lot of a lot of adjustability and fine-tuning that ultimately proved to be very difficult to wrap your head around it tuning was very difficult for a number of reasons so again the goal of this is to produce something that feels natural and organic so that's a feeling and trying to get that result with a mechanical system can be very tricky but how nodes are tagged as arbitrary for example anti-religious what counts as anti-religious that's a subjective thing racist what counts as racist that's also a subjective thing we also had strengths to that this is a minor racist comment this is a major racist comment to different designers can interpret those very differently it's not resulted in a lot of ambiguity between those things the values on individual topics were largely arbitrary so you know you'd say like ah well this topic comes up about 15 times in the game so it's 8 points per thing this topic comes up five times so it's 20 points per instance that was done for flexibility but ultimately it was just very difficult for people to wrap their heads around an organic system is hard to test that's part of the point is that if you take a different route to the game you can arrive at that relationship through a different way and what feels good is ultimately subjective so someone might say yeah this feels totally great and another person might say this feels completely rushed or fast or why is it taking so long for this character to react to the choices that I've made so the reception not good this did not go well there were lots of behaviors that were just way out of whack um Alif was one of them al Hoff came across as this supercritical like he would just he was constantly frowning and like being mad about almost everything you did and it was because his topics were marked very correctly but the instances of that were just so frequent that you were just constantly seeing a lot like threat about everything you did which was not really the intention and then the other one that came up is that show D and palutena had these like terrible conflicts um show D is a priestess pala Geena hates the gods and all religion and there were a couple of places where you could like have showed in your party get pala Geena and then go into her hub and be like tell me about yourself palette you know and she's like well I hate the gods also I think all priests are stupid morons they're complete idiot it's unlike Bing Bing Bing Bing Bing and so over and over again show Dee is just like going crazy and then after five minutes of going through palo jeana's hub you could exit and showed he would immediately explode on this person that she met five minutes ago that was not intended but that's how it worked so the goal is to be more organic and natural but these failures felt especially artificial they felt like really really bad so it was the complete opposite of what I wanted would I do it again only if I simplified the system I do think that there is a cool thing that was going on there but it was just too complicated for people to use it's entirely my fault I did want to try a new system I think we learned important things from it and it could be valuable in the future if it were tuned much better ultimately I feel like sometimes you have to go too far to know where the limit is I found it now we're going to talk about ship-to-ship combats this is very controversial this is the UI for ship-to-ship combat so instead of having like a thing on the open seas where ships would fight when two ships encountered each other you would do this turn-based minigame that was abstracted you had your ship position relative to the other ship at the bottom you had turns that went in sequence this was very abstract this is very bored gamey also very disliked by a lot of people I thought it was fun but I could tell that a lot of people are not like it as soon as the ship became the stronghold I remember very clearly I said hey guys let's not turn this into the pirate game and that is exactly what it became so it's a it's a ship minigame where you maneuver around and blast each other it seems very logical I have a ship there are other ships sailing around I should be able to shoot cannons at them okay I guess that makes sense it's a cool idea maybe for some people if you like this style of presentation a lot of people really did not so how do we implement this it really is a full game within a game like you could have we could have had like a whole indie dev studio just make this game it seems very simple but there is like so much stuff going on in it it's a ton of stuff it really could just be a game on its own not a very fun game for a lot of people but a game it took us a huge amount of resources to get to alpha feedback was very negative on the team they were like we don't like this there are a couple people that did I actually thought it was had promised even more resources went into the first iteration on it and the feedback was still extremely negative so at that point I said goodbye this feature has to go away it's it's quicksand it's if you've been in development long enough and you see like people just keep marching into this pit and disappearing and then you just keep sending more people into the pit and nothing's getting better that's quicksand it's a quicksand feature and so I cut it and my boss brought it back as a crowdfunding goal this is where throwing people under the bus happens I said hey man I think this is going to be really expensive and really difficult and I don't think a lot of people are gonna like it but he brought it back as a crowdfunding goal and so we were gonna do it it was the most expensive feature in pillars of eternity two dead fire in terms of dev time and money spent it was a drain on every department especially testing but every other department had to contribute to it it never really seemed to satisfy players I think this was the thing is that there were players that really liked it I liked the system I thought it was pretty fun but there was a big portion of our players who were like I hate the whole system they hate the presentation they just they just hated the whole thing quicksand yeah it just it always seemed to need more stuff so the reception to this was technically mixed because again there were people that really did like it most people hated it lots of reviewers it was the most criticized feature in the game it was a very abstract system with limited mechanics would I do it again no way like again it's not that I don't think that I don't it's not that I personally don't think it's fun is just that for something that most players will have to engage with knowing that many of the people playing the game were never going to like it I would not do it again because ultimately this is the fantasy we needed to deliver on which was deck to deck combat we already had that remember this this is what we showed players you'd be doing you would be shooting people in the face at point-blank range like I did really want to get tentacles attacking your ship I never got that so sorry about that but ultimately fighting on the deck of the ship this that's what we essentially showed the player that we were going to do and that's what we needed to do the ship to ship stuff I think just wound up being a lot of time full voiceover alright I'm rushing through these guys so as I said before divinity original sin to set a new standard we had a lot of internal debate about Fulvio I thought it was too late because it was halfway through development and we did not start off thinking that we were going to be able to that we were going to do Fulvio we were going to do a lot more vo but not fooled you I was the only person really on that team that had previous experience working on a game of Fulvio which was Fallout New Vegas and that was 65,000 lines of dialogue and it was a ton of work so what I said is let's do full view in the early game and not do side characters in the later game let's like you know sort of hedge our bets and just and manage that I was overruled that the owner levels owners believe that we really really really needed Fulvio so we did it Fulvio implementation critical role came out to do the major character voices which was very very cool when we had to do prep that was with to production milestones left so people are going hey why aren't the writers writing anymore I'm like because they're recording vo and that takes all of their time for a game that has again tens of thousands of lines of dialogue we were recording across three time zones including two am recording sessions which were very very bad if people missed sessions either due to being very sleepy or overlapping we would get casting and recording errors and then we would need pickups or we'd have to just live with bad vo initially the owners would not move the schedule I will go on record here as saying this was the most intense personal pressure I've ever had on a project and that includes shipping Fallout New Vegas in 18 months so this was extremely extremely fatiguing and draining took me to the breaking point this was very very difficult to get through eventually the schedule was pushed by two and a half months we needed that there was just no way we were gonna get the game done without that extra time so we did get that extra time and we got through it how did it go over I went over very well people loved it streamers reviewers players loved it critical role added a huge amount is this a hard limit am I gonna die if this timer runs out I guess I'm gonna find out would I do it again absolutely not under these same circumstances even knowing how much Fulvio added to the game knowing that we were not going to get the extra time initially I would never do that it was extremely demoralizing for the narrative team and it put everyone under a huge amount of stress I would I would still advocate the same partial VL solution with proper time yes because now it's an expectation which kind of sucks but that's what it is oh my god many times up post-launch the dlcs seeker Slayer survivor beast of winter and forgotten sanctum I think the DLC teams did a really fantastic job these games are much more focused than the original game awesome teamwork they did awesome job I just want to call out all the work that they did is being very very very good there was internal debate about doing a big expansion versus three DLCs owners wanted to do three DLC so that's what we did or rather what the expansion team did it was better than doing the split x-mansion but it was still very stressful doing three DLCs on a tight schedule is very hard lots of overhead for every single launch because if you do three DLCs in a row you're essentially launching three products and then you're supporting each one of them as you go on alright in the long term difficulty we made lots of balance changes over months and months and months of time we also tuned almost every crit path in major side area and the game encounters that's work that should have been done early on we had a god challenges these are optional gameplay modes that made the game more difficult for example ma grens challenge made it's that you had limited time to pause the game and in turn based mode which we added later your turns would actually run out and just skip Macbeth God challenge any character knocked out for more than 10 seconds would die so these are pretty significant gameplay challenges we added 11 overall and then an ultimate that wrapped it up that was very cool we continued iterating on the ship UI we did actually include narrative changes to explain some of the narrative shortcomings these were included in the 5.0 patch I think people have received them very well again it's my fault that it wasn't clear to begin with but we did try to address it in the long run and then finally those sidekicks those miner companions they did each get expanded into sort of a mini companion with the DLCs again I have to thank the DLC team for doing great work here Edwin Edwin Constantine and Ficino each got expanded into a mini companion with their expansions or with their DLCs the biggest thing that was added post-launch that made the biggest difference over was turn-based combat this was the thing that we didn't do from the beginning because the Infinity engine games are real-time with paws but about a billion people bought divinity original sin too so let's give it a whirl um this was not my idea this was the idea of Nick Carver working with a programmer named Brian McIntosh I initially resisted it because it did not accurately preserve the action economy of the real time with pause combat system the real time with pause combat system was based on seconds and tenths of seconds and the turn-based system that they could implement reasonably couldn't sustain that it had to kind of break the action economy but I realized that it was an argument of perfect versus good yes we could not faithfully adapt the original system but tons of people who were turned off by the real-time of pause combat would be drawn into the game even with an imperfect implementation so we did it by far the most significant implementation are a thing that we added to the game this was don't get me wrong this was a lot of work but compared to things like the ship combat system this was like nothing like this was a very sort of straightforward implementation there were lots of bugs to work out but the reception - it was almost universally positive it was a great thing for us to add and we added it you know nine months or so after the game launched and it people responded very well to it all right I actually made it through so some lessons learned here we made some assumptions about real time with pause and turn-based combat that maybe don't apply anymore it's not the late 90s or early early 2000s I know somebody's gonna read this and say oh my god obsidian is never gonna make a real time with pause game again I'm not seeing that but what I'm saying is that I've always personally really liked turn-based games I think we're now finding that there are audiences that also really appreciate them I think the fact that our turn-based motorists issue so well means that we don't we don't have to just constrain the piller series to doing real-time with Paul's combat in the future don't make any assumptions about that but it's just to say that turn-based is a totally viable thing that we can do now easier difficulties are very important but hardcore RPG players they really do need a challenge when the game launches keep in mind the first people to buy the game are the ones who are most excited about it often they're the most passionate they've played your back or beta they played early access all that stuff so if the game is not challenging its gonna be really really disappointing I'm not saying to make the game incredibly difficult for everybody but if you neglect the difficulty too much as I did it can bite you in the ass and that's not good when you see a quicksand feature you got you'll know when you see them you just you just see resource after resource going to them try to cut them as soon as possible fight like hell to get rid of them because they're quicksand they will just consume resources until the game launches and ultimately if they don't really work well if your design it's not gonna be a good fit and unfortunately I would say now the expectation of Fulvio is now very real in these games it's not just because of divinity it's also because of us we contributed to it I actually saw reviews of owl cats Pathfinder kingmaker where all these people were going like why doesn't set up full Bo this is such BS dead fire and divinity original sin to have Fulvio all cat was working on a much smaller budget and it's it's it's kind of a ridiculous expectation but now it has become this expectation where players just think that everything will have full view it sucks umm I don't really know what to do about other than to plan plan plan Fulvio and a big game is extremely expensive and it takes so much time you have to lock your dialogue so much farther in advance than you're expecting to and I mean again I think it was 62,000 lines of dialogue for dead fire like it's just it's a logistical nightmare so if you if you go down that road just understand that you're gonna have to plan a lot for it and ultimately again the plot should have been about one struggle not two we had two very cool neat ideas that didn't really quite converge entirely my fault but lesson learned also I should probably direct a different type of game next I mean I've worked on I Swindell Icewind Dale two Neverwinter Nights 2 and then pillars of eternity one into these are all party based fantasy real time with pause games I'm a little burned out on them dead fire was again the most stressful sort of directing experience I've had so far so hopefully the next thing I direct will be a little bit different and then maybe I I still like these games a lot but I need some time perspective to come back to them you can burn out you can burn out on directing the same type of game over and over again and then you're not really doing good work because you're just burned out thank you very much thank you very much is there a question time bombs one or two if it's a right okay hello fact thanks for the great lecture thank you I was wondering because a year ago I was also working on RPG game and was big game and I was wondering how many times have you actually played the game from beginning to the end like in one one size file because I actually never did it so I think it's important personally because I'm director and often the lead designer I always try to play through the game at least once from beginning to end every game that I've worked on in a significant capacity I've always worked on I've always played through the whole game with one character the whole way through I think it's very important to get that experience on dead fire I had one full playthrough and three half playthroughs I think total so one of the problems is that dead fire is an enormous game there's also some debate I think which is a separate topic should directors design content or should they purely just play the game and get feedback I I want to work on the game like I also want to make things in the game so like in pillars one I did almost all the system design on pillars do I did some system design that was the co narrative lead with Carrie Patel but I also then played the game and gave feedback it's a difficult balancing act but I think if a person has responsibilities on the game it's very difficult unless the team leadership makes time to play the game so one thing that we did do a little bit better on pillars then on debt fire was we have team play days and they're enormous ly valuable for getting people's eyes on the game because sometimes you're so deep in implementation you just have no idea what's really going on so but the thing is people camp I mean they can make time on their own but that's a lot to ask of people usually what we try to do is schedule time for people to play through the game so they get more exposure to it it altum Utley makes the game better in the long run okay a thing we have to and here so in order to prepare for the next speaker right all right thank you thank you very much [Applause] [Music]

Channel: Digital Dragons

Views: 72,977

Keywords: gamedev, gamedev conferece, game industry conference, art in games, game design, business in games, digital dragons, dd2019, digital dragons 2019, krakow, lecture

Id: xChOXFJ83-g

Length: 53min 53sec (3233 seconds)

Published: Thu Jun 06 2019