Donate to Codex
Putting the 'role' back in role-playing games since 2002.
Donate to Codex
Good Old Games

RPG Codex Interview: Chris Avellone on Pillars Cut Content, Game Development Hierarchies and More

RPG Codex Interview: Chris Avellone on Pillars Cut Content, Game Development Hierarchies and More

Codex Interview - posted by Infinitron on Mon 30 April 2018, 01:09:31

Tags: Chris Avellone; Pillars of Eternity; Planescape: Torment; System Shock (Nightdive Studios)

[Interview by Fairfax, edited by Infinitron]

I first talked to Chris Avellone about doing this interview back in late 2016. Chris became increasingly busy over time, so the interview took longer than expected. As a result, some answers became slightly outdated. However, Chris went back and updated his older answers, which is why you’ll see “New:” portions in them. Also, the questions and answers are not in chronological order, they were split and rearranged by the editor.

How are you doing, Chris? And how's the search for the studio name? Lots of good suggestions out there, especially in the Codex.

I am doing well! And so's the family. Had a bit of a scare last year with my mother (cancer and a lot of medical bills, which was stressful), but she got the treatment she needed and it went into remission, so I'm relieved about that. She seems to be doing great, I try to visit my family as much as possible. My niece also informed me she wants to start making games (she's 10), so that's been interesting hearing about her game development process. It involves a lot of dragons. A lot.

Got a lot of great studio names, about a 3 page (?) list. Had some I remembered from the old Interplay "let's name the studio" days although some of them probably wouldn't work (example: Colostomy Bag Food Fight). Would always welcome more suggestions - @ChrisAvellone on Twitter is easiest to compile them if anyone has any ideas.

Work's going well. Finished the last bit of work on Tides of Numenera last month (had to do some brief edits to the graphic novel), and Prey is all wrapped up, too. Divinity is still coming along, but need to talk to Swen about dialogue, since I don't believe we ever discussed that, and that's a lot of work vs. working on the story and companion backgrounds. We'll see.

That's it for announced work - other new projects should be announced soon, but the Codex may hear about the announcements before I do - I think the Codex is more aware of what's going in my life and work than I am. There's a lot of stuff in the works, and even stranger stuff - the CEO of Fig contacted me the other day and asked if I wanted to work on PoE2 (I told him I wouldn't be interested, but politely).

Aside from the System Shock Reboot, I started working on a lot of projects with ex-Obsidian employees, some recent departures, some who left many, many years ago, and that's been fun to be able to work with them again.

New: Still struggling with life and family issues. Projects that came out since: Tides, Into the Breach, and continuing to work on Pathfinder: Kingmaker (which should be out this year), did some system work for Burden of Command. System Shock Reboot went back to a remaster (I’m still on it).

There’s many other projects, but they haven’t been announced yet.

Chris Avellone, Tim Cain, Alexis Kennedy at Reboot Develop 2017

I'm glad it's all going well now. I wish your mother an even better recovery.

As for the Codex, yes, we can be very...thorough, but it's all in good fun. Speaking of which, Roguey noticed you listed your role on System Shock Reboot as "Lead Narrative Designer", which sounds bigger than your most recent collaborations. Can you describe what the job entails in this case? For instance, one of the Kickstarter stretch goals was "new crews members and audio logs". Are you the one writing them?

Yes. As for the title, Jason Fader (System Shock Reboot Game Director) did one thing at the start of the project, thank god - he established a hierarchy chart and was clear about where I was on it, which is a relief. FYI, I worked with Jason on the Fallout: New Vegas DLCs (he was our Lead Producer on the run of them), and I'm glad I have a chance to work with him again.

And yes, writing responsibilities have grown beyond the initial push (initially the goal was to flesh out the narrative elements, including extra audio logs and clean up the existing ones, along with the station chronology, world lore, event consistency, etc.). A large part of this expanded role is because the game is a reboot, and it makes you consider the wider universe of (early) System Shock more - which is pretty exciting from a narrative perspective.

Why was that a relief? Is that not common from your experience? I know you did a presentation about this recently, but it's hard to tell how often teams have a structure like the one you advocate.

Established hierarchies, even as simple as a flowchart, has not been common in my experience, no, and it's caused numerous problems whenever one wasn't established (I bring up most of the problems in the presentation). This is often combined with an inability to assign titles (esp. "Lead") to developers. The lack of both causes confusion for everyone. Sure, you might muddle through, but how much time will you waste and irritation will you cause doing so? Is it really worth it vs. a 2 hour task at the outset to organize your own team?

So yeah, hierarchies: Often in the past, I had to push for a chart and responsibilities, and it's usually combined with trying to figure out who the lead is for a department who is making the calls. No one should ever have to ask that - they should know who the Systems Lead, Animation Lead, Art Lead are (and if you have "Directors" like Art Directors and an Art Lead, make their responsibilities and authority clear - who do you go to and who makes the call on X piece of art? Before I spend 60 hours a week iterating on this level, who's feedback should I be following?). Why codify this? The simple reason is you need information, and you need the right information, and you want to make sure you're getting it from the voice of authority in charge of that piece of information.

Co-leads, especially (which I observed at Black Isle) caused problems, and I've never seen a co-lead situation work unless the leads care about different things (one is design, another is art, but same discipline you have to be more careful - say, one systems, and one narrative, etc.). Having them in the same discipline doing the same thing I've never seen work well, it just causes confusion for the team and often for the co-leads as well. Not saying it can't work, but... well, if they both do the same thing, I'd want to know why it was structured that way.

Also, the System Shock Kickstarter campaign calls it a "faithful reboot" and a "complete remake", but you make sure to clarify that it's a reboot rather than a remake. It sounds like the team is expanding upon the original, but a reboot usually means the previous continuity is discarded. What makes it a reboot in this case? And since it's a "faithful" reboot, how does it affect your job as Lead Writer?

It's a reboot, not a remake. It means I'm doing more work than originally anticipated in a good way (if I'd only done "remake" work, I'd be done now). Note that if we had done a remake, we still would have had to change some continuity, because of things as minor but as significant as (1) gender mismatches with SHODAN, to larger things such as (2) contradictions in chronology of events on the station.

The words "faithful reboot" are used because (1) it would be wrong to mess with the fundamentals of the System Shock universe, and (2) we're going through all the lore, set-up, and asking the original creators their intentions with the design steps (and we thank them for their time). Also, there are some things that are possible now, narratively and mechanically, that weren’t possible or easy to do then as they are today, and we want to keep that in mind with our design.

We recognize the words have different definitions, which is why we had to make a distinction (remake or remaster means you aren't changing anything, just heightening the quality).

New: System Shock went back to being largely a remake/remaster, although I don’t have many new narrative details at this time. I’m still doing the job I was brought on for, writing SHODAN and tying together some inconsistencies in the logs and lore that largely came about in the transition from non-voiced SHODAN to voiced SHODAN (did you know she used to be male? She’s still referred to as “he” in some logs and references).

Jorge Salgado and Chris Avellone in Road to Eternity

The PoE2 thing sounds strange indeed. Why didn't someone from Obsidian approach you directly?

I have no idea. It may have been the circumstances of the question (I believe he thought I was key to another Pillars crowdfunding success, which I don't believe). Also, I don't know how much Fig itself knows what's going on with PoE2 (ex: Tim Cain will unlikely to be involved considering his new role, the lead story writer for PoE1 left, but PoE1's success can likely stand on its own for another crowdsource, imo). I don't even know if PoE2 is even going with Fig, but if he's asking, I guess it is.

New: PoE2 crowdsourcing seemed to go great, so I don’t think he had much to worry about.

And speaking of PoE, Fenstermaker confirmed in his Codex interview back then that multiple writers pitched a story for PoE and that his pitch sort of won. I've seen people say you pitched a story yourself, but I wasn't able to find a source on it. Was that the case or people (wrongly) assumed you did?

Yeah, I believe I told Crooked Bee I did one of the pitches, yes, and I'd probably broaden the pitch submissions to "designers" not "writers" ("writer" puts you in a corner). My quote:

I also did multiple critiques of the lore and the storyline(s) (including suggesting one framework for the story, along with other designers), but the storyline does not have many of my suggestions or contributions, it is largely the Creative Lead’s voice and direction guiding it, as it should be.

I read the Creative Lead's (Eric’s) interview on the PoE story, and whether true or not (I couldn't speak to the entire team's reception) - if the 1st draft's reception by the team was "tepid" with the second draft being "slightly" + "less tepid" but "could run with it" - well, that's pretty disheartening. The fact it got approved by someone is also disheartening. Sometimes you have situations like that imposed on you (I've been there), but rarely with an internally-owned franchise.

New: In talks, I’ve never got the impression anyone in the approval process was happy with the story (I wasn’t part of that process). I was told that Josh, in his review, wanted more control and oversight over the sequel storyline, and I don’t know how Eric’s review went. I was a little surprised he left Obsidian but I think that was a good move for various reasons.

I tend to go in a different direction with story pitches - do 1 pagers of the major beats, get approval, do a 4 pager, get approval, then keep going. I also tend to weave the system and world lore reveals into the pitch, too, I don't do "Hollywood" pitches - but it makes it important that the lore of the world and the workings of the systems have been set down first, and everyone understands them. Example of how I do game stories sometimes resemble code scripting outlines: (1a) Here's the moment you unlock X power, (1b) here's a quest/event where X power use is proven to be useful and gives the idea of how that gameplay loop will work when used, plus (1c) it makes retraversal in areas B and C more compelling b/c X power can unlock more events there, and (2) here's when you gain Y power/insight, here's how Y power changes the following Z "routine" monster encounters,* etc, etc. I prioritize the elements as well so the reader knows what's most important on the narrative end, and what isn't - it prevents people (including myself) from getting hung up on the details.

If you write too much at the outset and then seek approval, some writers get too attached to the details. For readers, getting a 1st draft that went way too deep ends up being too much to go through before the 1st stumbling block (which is usually evident in a 1 pager).

My advice is do your 1st write up like a player telling another player the story of the game and why the story was the coolest thing ever - starting with what's the one sentence they say about it? (Note arguably, that's not a good representation of real life in game stories - if a game story is good, usually all that I've heard said is, "man, you have to play [X]. No spoilers, but you'll be missing out.") Most stories I've written in the past started with 1-2 paragraphs of the arc.

Other problems I've seen with story iteration is the "well, throw the baby out with the bathwater" syndrome. A writer might get a lot of feedback on a story, receive comments on a lot of strong points that work, and in the next draft (equally long and drawn out as the 1st) the writer chooses instead to write a brand new version of another equally bogged-down idea, but now with none of the strengths the old one had, so others reviewing the work have to start all over in the feedback process, which just wastes everyone's time.

* My quickest example of this is Arkham City when you learn the Riddler has spies in every gang, so suddenly all the gang fights you were doing (and were likely getting tired of) suddenly requires you approach these gang fights differently (attacking members in a specific order) to make sure the Riddler’s chosen spy is the last one standing so you can interrogate them. It’s a way of layering challenges into encounters in a narrative way as well.

Part of the Icewind Dale team: Stephen Bokkes, Josh Sawyer, Scott Warner, Chris Avellone, John Deiley, Reginald Arnedo

Would you say your response was "tepid"? I assume you had to avoid addressing that for legal reasons, but people would definitely speculate/assume you didn't like it, so I thought I should ask to confirm.

I don't feel I ever give tepid responses. People I've critiqued are welcome to prove me wrong, though - if anything, I give too many comments. Mostly about cutting adverbs. And odd plot detours. And Why Should The Player Give a Shit About This comments.

The Codex shouldn't speculate, there's no reason to, the story was the story, and I'd argue the Creative Lead had more insight into the whole process than anyone else.

Is that the same approach you use for companion design? Ulysses, Durance and Grieving Mother had cut content (in Ulysses' case, the companion itself) because you wrote too much for them, or at least that's what others have told. Also, in the Codex interview you mentioned, Eric Fenstermaker said Durance and the GM had a lot of "creative energy and research" invested in them. You seem to rely on getting approval every step of the way, which makes me wonder how that happened. Was there a miscommunication on the approval of your drafts and/or outline? And would you agree that the content had to be cut, or would you have tried to save it if you were at the helm? (Including Ulysses in FNV, not just PoE).

Companion Design - not always. Usually, it's paragraph, page, then dialogue. In Durance and Grieving Mother's case - and this I can say - the Creative Lead told me after my departure that the Project Lead had interfered but didn't specify the reasons. I don't even know what the reasons were (although I couldn't tell if I did, to be fair). In short, the Creative Lead said he should have handled the whole matter differently and didn't, regretted it, and apologized for it.

New: Fairfax did provide me the reasons below - these weren’t the same as the others given at the time.

It doesn't matter - what's done is done - cuts don't bother me (they rarely do, if you're a writer in the industry and aren’t willing to make cuts, you probably won't be a game writer for long), but how it was handled overall was an example of how upper management can get tangled up and despite any problems, perceived or otherwise, you can be the last one to know, which makes the whole situation more fucked up than if you'd gotten a direct critique, fixed everything that was requested, and then it's put to bed. It's not a huge deal unless you make it one.

Fenstermaker specified the reasons in his Codex interview:

The cuts came for length. The three limiting factors were time to implement, art resources for the dream sequences, and VO budget. There was a target length we had set upfront for all companions, and we had to stick to it. Otherwise we'd be, for example, voicing maybe one out of every six lines for Durance and the Grieving Mother, and it'd be conspicuously incongruent with the other companions, who had maybe 2/3 of their lines voiced. Unfortunately in this case it meant cutting down characters that had had a lot of research and creative energy invested in them, and there were some good ideas there that it would've been interesting to explore. It was a shitty thing to have to do, but we'd never have been able to implement the original versions in time to ship.

Since you can't comment on PoE's case in particular: is having similar lengths and amount of voiced lines a priority for you in companion design? In Torment and KOTOR2 some companions didn't have nearly as many lines as others, but neither game suffered for it, if you look at how they were received.

I can confirm some of the elements - there’s still people I can ask, so if I can’t remember, it’s easy to find someone who can jog my memory (I can’t check the details on my own). I also managed to get some clarification on some of the points, which was welcome, so…

Of those three reasons, the first was the only one I ever recall communicated to me (I had to get confirmation on it and these other points, since it’s been a long time). The other two reasons weren’t, and I actually got multiple other reasons from multiple people – and some of those people admitted they were just the messenger. This confused things, since they couldn’t articulate what the critiques were since they either didn’t understand them or hadn’t read the material (both our CEO and Parker among them – ironically, after a long speech ending with his admission he hadn’t even read what he was arguing against, Parker did go back and read the companions and found nothing to object to, which cost even more time – to his credit, he did admit his error, but things like that happened a lot).

There were other people who apparently didn’t like Durance’s swearing (easy fix), and the original tie in the GM and Durance backstories were they had violated each other physically and mentally and that’s what broke both of them, which I then cut – although I don’t know if the GM one got removed completely – the intention was the Watcher could fix it mentally by repairing their souls by walking through their minds in stages. I think some of this is still mentioned in the strat guide.

In the end, I just wanted to fix whatever the problems were and move on to the next task, because there were a lot of tasks that needed doing. I had done the best work I could, and it was up to the Pillars team to decide what fit best (which is fine, it’s what vision holders do), but no one was articulating what the problems were.

To speak to the implementation part, it had been promised by multiple people on Eternity (producers and Lead Creative) that they would set aside (their own) time for implementation and make sure it got done.

When the project ran over – and this happens, I don’t blame anyone for that – it was apparent they ran out of time for their own character implementations – and some companions even required two designers to implement. As such, other developers took on what tasks they could to try and make up for lost time. Things that could definitely have helped (hire an editor, like they eventually did for Tyranny) were refused in light of putting more devs from other projects rather than trying to fix the missing personnel. I was later informed that this time was not paid back, which was the hope but not a surprise, and I don’t think Paradox was ever fully aware the Tyranny team had been gutted (in general, publishers don’t like hearing the resources they’re paying for they aren’t getting).

The best-case scenario is you hire the right amount of people for both projects, and if you can’t, then you downscope or argue your case for more funding – I’ve known projects that have gotten an increased budget after showcasing how the additional budget will improve the game (usually with a gameplay sample).

That said, for the Eternity companions, I didn’t mind any cuts for length (it only took me a few hours to fix, which is nothing). I still don’t – it’s part of being a narrative designer.

As for VO, I’d agree that was an issue, but not in the way it’s explained in the answer. The problem was there were incongruities at other points in the game that suggested there wasn’t a pattern to adhere to (simple example: Wael conversation later on after you do his side quest).

I’ve asked about this incongruity, and the answer I got was, “I think it’s intentional.” No one I’ve spoken to recalls if there was an established VO vs. not VO standards in a PoE1 writing standards doc. I was told, however, that each writer chose the important lines for their own characters to be voiced, which without a global standard, could make them sound jarring as well since no two writers may have the same choices or consistency (at that point, that’s what standards are for).

So in short, I didn’t understand the whys and wherefores of why certain characters in PoE1 had voice over. When playing it, my only reaction was it felt random, as I couldn’t understand the reasoning behind the choices – it didn’t feel like there was a pattern. I also had the problem (this is developer bias, because players rarely have to experience this issue) in that I’d played the builds without VO for quite some time before the voices came in. Then when I heard the VO, it felt jarring because of how I had imagined them speaking from the prose, and it didn’t always mesh with the actor, so it took some characters I was cool with and made me dislike them (or distance myself from them a bit). In other instances, it did make the characters better, though, so it’s a double-edged sword.

PoE did make me realize it was damaging to have prose placed before any VO’d line, which I pointed out and eventually became part of the official procedure.

Note this process and the voice delivery may have improved in the expansions (I haven’t played them). I do recall wanting to know the pattern in PoE1 b/c I realized whatever pattern was in PoE1 would likely be expected in Tyranny, but we were pretty far from voice acting withTyranny when I departed. Even asking now, I haven’t gotten an answer (mostly b/c people don’t remember, not out of malice).

If it had been up to me, I’d have asked them to find another consistent solution for voice acting consistency in PoE1 and save a lot of time in the studio, even if it meant doing BG and Torment-style interjections, vs. inconsistent voicing. The added benefit of Torment and BG-style “openers” is they can tell you something else beyond the character vs. the intro text (it’s like a bonus line, and you don’t have to match the opening text). Also, now that I’m aware of the “voice acting limitation,” I’d have asked them not to voice GM (she gets along fine with a whispery mind effect, especially since it’s supposed to be telepathy – why not make the strengths of the franchise work for you and save resources? Win-win), and of the excess text, there’s no reason for Durance to speak in the dream sequences that were cut (I’m not sure he did much anyway). In the end, they were not a point for discussion, and it was at the stage where the team had to take stock of where things were, and figure out a way for the game to get done. I do feel there’s other ways to accomplish the implementation, but you have to think outside the box for ways that save resources and maintain consistency. My only goal was to write Planescape-style companions and try to do something interesting that would set them apart from being solely traditional fantasy.

The art for the dream sequences would have required maybe 12 images, and it was designed so you could do 1 large image and split the scene, so each image could be repurposed. I will say there were other cuts (the GM’s special Planescape-y weapons and Durance variant equipment) that were just simply dropped and would have required little in the way of resources – it was more a matter of will vs. time in some instances.

I agree with your assessment in that you don’t need equal lines across all companions. It’s important only in that (1) the voicing of lines should be consistent vs. prose, (2) the voicing of lines at key moments for each category of characters and what the player can ask them be consistent, and (3) if a character has less lines (ED-E, Rex, Dogmeat, etc.) it’s pretty clear to the player as to why.

There are other factors as well - there are some companions that will have more lines if they are introduced earlier in the game, for example, or if you’re gearing them toward some end role (ex: Ignus and Vhailor were eventually meant to be threats, Kreia is a different kind of companion beast, etc.).

Some standards you may want to set, for example, which are pretty cheap:
  • Intro Line
  • Opener, Normal (1-3)
  • Opener, Serious Reaction (Anger/Love, varies)
  • Any interjections in another fully voiced dialogue (which you should avoid).
  • Banter
  • Voice Sets (imo, this is where you get the most feel for a character, and it’s where you should concentrate most of your VO efforts) Maybe 40-50 lines.
  • Key cut scenes (ex: While we didn’t do this for Torment, ideally all the end cut scenes would be voiced), including goodbye moments.
Kevin [Saunders] and George [Ziets] did something I really liked in NWN2: Mask of the Betrayer – they had a few companion lines that would be randomly chosen when entering an area, which I thought was great: It showcased the companion, it showcased reactivity, and it also made the companion seem more “aware” of the world.

Lastly, it is jarring, not only to lead a voiced line with a prose piece, but for a largely VO conversation, having non-voiced companion interjections is jarring, too (speaking to Wael example I mentioned earlier if you have Kana and Eder in your party).

So even if some of the KOTOR2 characters didn’t have voice actors, they had simulated VO, so it still felt consistent. And even when we knew we couldn’t do VO for every character’s written line in Torment (note that at the time, that idea would have been considered impossible logistically – it wasn’t until KOTOR1 did it that they set a benchmark for other games to do the same, despite the word count), Torment was designed to have the “opening line” for most of the major figures, and at key moments, they would also have VO if we thought the line would add impact to the scene. I don’t think it was 100% consistent, but I never felt like it was out of place while playing it (then again, I’m biased, so if you disagree, feel free – either here or in the comments).

But in the end: VO is expensive. Inconsistent VO is also expensive and should be evaluated because it might benefit you to do less, especially if you’re already over budget and over scope in other areas (writers who spend less time in the studio can fix more bugs).

Voice acting is one of the most expensive parts of a game. Even if non-union actors are used – if you can minimize and standardize VO, you can spend it on other parts of the game (or hire staff you need) and make the core gameplay better and fix the problems you do have (editing) vs. ones you’re creating with inconsistent VO. That’s not a slam on Pillars of Eternity, that’s a general statement I stand by for any project.

In addition to communication problems, it seems a lot of it came down to the priorities the Project Director and the other leads had in mind. When you're the one at the helm, what do you prioritize when it comes to making sacrifices?

When looking at the cut content from the projects you directed and previous interviews, my impression is that you favoured cutting items and areas, and went out of your way to preserve reactivity and characters, especially companions. It also sounds like VO isn't a particularly high priority for you.

It’s done by priority and you set up things to be sacrificed before you need to do the sacrificing. I do it by ranking it A-priority, B-priority, C-priority (the definitions for those are in the Appreciation Thread, but I can copy them over if you’d like). Those items can be characters, areas, quest lines – anything, and it all should have been prioritized in advance. That way, when something needs to be cut or downscoped, then the decision is already made for you. If all you’re left with is A-priority material to cut, then the game has really gone off the rails.
  • A-Priority: If this isn’t in the game, the story or game would be obviously unplayable or broken
  • B-Priority: If this isn’t in the game, the player will notice it, but the game will be playable, just not as good as it could be.
  • C-Priority: Something the player wouldn’t miss if it wasn’t there (thus, cutable) but if it was there, it would make the overall game experience better. (The talking appliances in Old World Blues.)
Then I try to break the content into 50% is A-Priority, 25% is B, and 25% is C. This will help prevent headaches later on because if you need to cut something, you know what to cut.

The Fallout: New Vegas team

Ideally speaking, do you think a team should have design and writing departments, with Lead Designer and Lead Writer as separate roles? I ask this because both Planescape: Torment and KOTOR2 had highly praised narratives and writing without a specific narrative design department.

Depends on the team and genre - for RPGs, I think it’s a good idea.

I once asked you if you wanted to be Lead Designer again in the future, and you said: "I would take on that role, but I'd rather be a Project Lead, since those seem to turn out better for me overall."

Project Lead is different – in my experience, Project Lead is above everyone, including the Lead Designer. Sometimes it's the same person sharing both titles (and doesn’t have to be Lead Designer, can be Lead Artist, Lead Programmer, etc.), which can work. Sometimes, it can even be one of three titles, and that's where it starts becoming a problem - part of the role of Project Lead is to oversee and provide goals, not getting too deep in the weeds or spreading themselves too thin - I argue this happened on Planescape: Torment for certain, and I've seen it happen to others with the same results).

As an example - Tim Cain is a Project Lead now, and he has a Lead Designer (an excellent one) reporting to him.

Which project(s) did you mean by that? I couldn't find a project in which you held that specific title. Did you mean Torment and KOTOR2? KOTOR2 doesn't have a Project Lead in its credits, and in Torment's case, Guido Henkel and then Kenneth Lee did hold the title of "Project Director" before the game came out. However, both ended up listed as "Producers" in the game's credits. Was that the team's way of acknowleding that you were the de facto Project Lead?

I'd argue I was Torment's de facto Project Lead, yes, at least for part of the time (Guido was Project Lead while he was there, I believe). We didn’t have a Lead for a while at the start, so it kind of fell to me since I was the first one on the project.

Would your ideal hierarchy be something like this?

That's it exactly. Note that the term "Project Director" varies from studio to studio, but in this case it's the chief vision holder who can guide the course of the game (and at times, they may share one of the Lead or other roles). They make all the final calls on the content, although ideally, they will have assembled leads that work with each other and production to settle on the best issues in advance before anyone needs to "intervene" in a stalemate (usually any reasonable conversation between leads will reveal the best course of action, usually when put into context by the producer's metrics - "well, we really only have 2.5 weeks to do this and two animators free, which means we could do X and Y or Y and Z in the time provided, but not X, Y, and Z unless we got more resources or time.")

Did any of your projects have this structure?

Yeah, the New Vegas DLCs was the 1st experiment.

And how many/which of these roles do you think a designer could have at once without holding the project back, like you argued in one of your presentations?

For instance: I assume Project Director+Lead Designer+Lead Systems Designer is too much, but Project Director+Lead Narrative Designer is alright? Although I imagine that could be weird, since the designer would be both below and above the other leads.

It's always hampered by anyone having (1) more than 2 roles, imo, and (2) any other issue where the dev team realizes someone isn't getting back to them quickly because they have too many other responsibilities (if you do it, you can't be a chokepoint, if that makes any sense - you need to delegate/let go to keep things moving).

Is that what some devs call "blocking"? I saw it in Jason Schreier's Blood, Sweat and Pixels:

“I think to get more specific on the complexities at this scale of game development, it’s the dependencies,” said Aaryn Flynn. “It’s the things that have to happen for the other things to work and be successful.” The common term in game development circles is “blocking,” which describes when a developer can’t get work done because he or she is waiting for someone else to send over some integral art asset or piece of code. “‘OK, well I was going to do this today, but I can’t because we have a crash, so I’m going to go to this other thing,’” said Flynn. “Good developers are constantly juggling these little tasks on a daily level.”

Yes. They are also sometimes called "dependencies" or "being blocked" (ex: "I'm blocked on adding audio to the level because Feargus has the level checked out, so I guess I'll work on weapon SFX.") Blocking ultimately means a developer can't do the task they should be doing at the time - it can be due to a code error or dependency on another developer.

The more responsibilities someone has, the more blocking can occur - which is why I was always an advocate that the owners and project directors should not (1) be responsible for any content to the game, but they should direct and review it as appropriate, and (2) if they do, it should never be an A-priority critical task because they have too much on their plate already - and it's worse if they get yanked away for a few days on a convention or marketing trip, because that content may have bugs or issues that need to be resolved.

Leads (Lead Designer, Lead Animator), etc. should also be careful about taking on too many A-priority content tasks, as they can cause similar blockages (and they should be spending half their time reviewing other's work).

Swen Vincke and Chris playing Divinity: Original Sin II

Makes sense. Still, don't you think it may depend on the developer? Swen and Miyazaki (FromSoftware) seem to be positive examples of heavily involved managers/directors who also create content.

It depends on the fact you’re on a schedule and your resources are limited. If you aren’t, you really aren’t blocked. It's a nice place to be in, I've just never been part of a game like that. ;)

New: So now I’ve been on two games like that (Into the Breach and an unannounced title), and it’s been pretty great. “It’s done when it’s done.” And when it’s done, then the release date is announced. The creative process is much more iterative than I’ve had the chance to do, and the end result is better for it, I think.

Swen and Fargo definitely get involved with their titles, that’s true, although they have numerous folks that help implement their high-level ideas and broader designs, then they critique those designs (at least with the design elements I saw them involved with – for example, Swen was heavily involved in story meetings, but he would critique stories, not write sections – at least from what I saw – he would indicate what he liked and what he didn’t, and changes to address those critiques were discussed. And Fargo did area design for the Rail Nomads for Wasteland 2, but he didn’t do all the implementation for it, etc.). They may have been involved differently in other areas of the game, however, but I couldn’t speak to it.

Project Directors are best at setting goals, setting expectations, and having the authority to make the final call on parts of the game – they keep the vision, and they can dictate the vision. They can be involved in development as much as they want, but these efforts should always be respectful of their team and be to the betterment of the game - noodling over details or micro-managing can make someone miss the bigger picture, or even damage caused by indecision or lack of commitment – for example, if a PD knows they don’t like the art style and want to see changes for a project but the art team is currently laboring 60-70 hour weeks to get environments done (environments that will likely have to change or be redone b/c of the PD’s attitude that they haven’t quite set a goal for b/c they’re too busy with other things), then the PD's messing up. It’s the Project Director’s responsibility to step in and say, “stop, artists, don’t spend more time on this, and here’s why,” immediately and be ready to accept the cost of that decision – or commit to the art style already in place as a necessary decision based on time and resources. Two of the worst things to do, however, are (1) not to do anything, all the while realizing changes will be made (this is effectively an anti-blocker, because you’re not only stopping work, you’re reversing people’s work and making them do it twice) or (2) stall and ponder - “everybody, I know we have only a month to do this task or we'll have to let everyone go, but I need to think about how to articulate my opinion for 2 weeks first.”

By content, I do agree that managers and Project Directors should be familiar with the pipelines but this familiarity should be reasonable (you don't need to write a companion to understand how the dialogue system and reactivity works), and if they create content, they should be wary about creating critical path content (that’s why I suggest B and C-prioritization tasks b/c these, by design, aren’t blockers and B or C-priority content could be cut from the game). The more nebulous part of the answer comes in asking, "while they are doing this content (even if they can technically find time to do it), what isn't getting done or what could be done even better at the PD level?" PD's have the authority to make a game great and demand fixes and changes that other developers don't have the same authority to make - especially the hard decisions. Often, the reason PD's create content is because that's why they are in game development - there's parts of the game they love to work on. This isn't a problem unless it creates problems for others - or even themselves.

So if you want to become a Project Director, my advice – start finding people to team up with (programmers, other system designers, artists).

New: Subset Games is two core developers and contractors as they need them, and it works pretty well. They don’t sell 500 million units across 50 million platforms, but they make a game they like and sacrifice (either time and/or money) to do it.

It doesn’t have to be a huge studio-size group, but few Project Directors can do it on their own – they need to build a solid team around them. Start small – there’s no need to design a huge game at the outset – just design a game around one system you think would be fun, polish it, then in the next game, use that same system but add another system, etc, etc. (BioWare did this process for years, adding about 20% new material to each game and keeping the systems and pipelines they’d already set in place for the other 80%.)

And you can wear multiple hats, but use others to help you with your discipline – in your specific example of being a System Designer, a PD involved with systems can outline goals a system design should support and define player-facing experience, but for the fine details, the heavy grunt scripting, and the skeleton of the system/prototyping, having a detail-oriented assistant systems designer/gameplay programmer for you who can build the framework on your specs, and then you and that developer can iterate on it until you were happy with it, if that makes sense.

Also - part of being a PD or a Lead is also finding the right people who share your aesthetic not just the implementation work – for example, it doesn’t help a Lead System Designer if the only system designer he relies on doesn’t agree with the design aesthetic, because then you may spend more time debating implementation or simply being dissatisfied by not seeing eye-to-eye on elements. Same is true for animation, environment art, level design, programming, etc.

The Planescape: Torment team

How do you avoid a situation like that? How do you lay the cards on the table and make sure everyone's on the same page before it becomes a problem? Have you ever found yourself in a position where you and the Project Director (or one of the Leads) didn't see eye-to-eye during production? If so, what did you do?

Well, ideally you hire for it, and the design test proves it out. That doesn’t always pan out if people were hired for different projects at different times and because of differing specialties.

My approach is when you present a design/art/programming idea, that’s it’s agreed to by the leads before the rest of the team is exposed to it. This is not done to restrict information, but at first, it’s more the LEADS have to understand what you’re doing, debate the obvious problems, then agree on an approach before they can defend and speak to their team about why a decision was made – if they see issues with it, have critiques, have better ideas, or worse, they don’t think it’s a good idea, maybe you should re-evaluate it yourself. I have in the past – usually when you get 3+ people objecting to an idea, it’s time to take another look at that idea (that happened as far back as KOTOR2).

So ideally, if there’s a problem, the leads and PD discuss it amongst themselves until a solution is reached. Again, that discussion isn’t exposed to the team (which can cause confusion). You do this in the Leads meeting, and make sure the reasons for why you’re doing a specific approach (this isn’t design, it can also be the animation lead explaining why certain animations won’t be done or will be done differently). If everyone agrees and understands, it means they can communicate the reasoning to their departments. If you keep the conversation respectful, and you genuinely listen to everyone’s concerns rather than continuously saying “no” without reason, I’ve never had an issue presenting an idea.

This doesn’t mean the idea won’t ever change, but it gives a good compass point for the conversation. It also doesn’t mean conversation doesn’t stop on the item, but it can focus it and limit it in more constructive ways.

If you’re the one objecting to an idea as a Lead to your Project Director, then you raise the issue as a critique – once. If your boss still says, “do it,” then hierarchy demands you embrace the idea and go with it. This is because your boss ideally is the one holding the vision for the project, and it’s important that that vision be consistent. Everyone is there to help the PD realize that vision – it’s the job.

I also remind folks it’s the same attitude they’d want if they were Project Director – the less time you spend arguing, the better. It doesn’t help anyone if you spend 3x as long looking for ways to object to an idea vs. trying it in the game or giving it a shot once. Look for ways to make the idea work before simply objecting.

Ideally, if the single critique you raise is a good one, then your boss should be good enough to listen to you and reflect on it. That’s what a good boss does – and a good boss would also thank you for raising the issue vs. getting defensive, because they should know you’re doing it to help the project be a better game, not to attack them personally – which you shouldn’t be doing anyway.

Lastly, recognize when you’re not a good fit for a role. I didn’t take the role of Lead Designer on Dark Alliance 2 because I didn’t think I’d do a good job, and Dave Maldonado had a better sense of how to make a game like that great – he should be in charge, not me.

You mention Planescape: Torment as an example of spreading yourself too thin, but don't you think it paid off?

That’s a complicated answer, and I’d say “no.” Did it pay off for story players? I hope so. Am I proud of it? Sure. Do I think it could have been better had I not spread myself too thin? Absolutely. Did I grow as a designer by doing weeks of a certain task (ex: plotting the coordinates of pathing routes)? After the first few days, not measurably.

And when we question how it paid off, that’s a good question - because who reaped the benefits of PST’s development challenges? It wasn’t Interplay, the company who had asked for the game in the first place and needed a return on its investment in the license. inXile certainly reaped the benefit, many, many years later, when the game was regarded as a cult classic that deserved (?) a sequel. Yes, I put a question mark there (not a slam against inXile, but it was good they took the general “Torment” premise vs. doing a direct sequel to Planescape Torment). I will say for crowdsourcing 13-14 years later, it definitely paid off – but that was never the intention.

I do think players who played it and enjoyed it saw the benefit, but that wasn’t a ton of folks compared to other RPGs coming out at the time. And many who bought PST didn’t finish it or want to, but I don’t begrudge them that.

Again, this doesn’t mean I’m not proud of the title or that I regret working on it, but yes, I was absolutely spreading myself too thin – I was Lead Designer and Creative Lead (and at points pseudo-Project Director). I’d argue part of the reason was it was a small team and I didn’t know any better (did I really have to map out the pathing of every Dustman on every Hive map along with every citizen in the Hive, too? Probably not. Should I have been drawing out area concepts? No, and even if I enjoyed it, that’s the artist’s job.). It’s one thing to be aware of how a system works as a lead, but that doesn’t mean it should consume weeks of your time that could be better spent overseeing the project or be a chokepoint for those tasks… and it would have meant a better project if I had been familiar with one or two instances of the process, how long it took, the demands on an implementer’s time, and at the end of the day, I hadn’t done those things.

Some of Chris' concepts and the finished areas

It seemed like the project had a lot going against it: small team, short development cycle, timid marketing, terrible box cover, Lead Producer quitting after beta stage, strong competition, dead IP (Planescape), among other things. Your own lack of experience in that role was a challenge as well. All things considered, I believe it's a remarkable achievement, and the game's towering legacy is also a considerable benefit, even if it wasn't planned.

As for "who reap reaped the benefits?", I'd also add all the developers who have been inspired by it and went on to make their own games, including fellow Codexer MRY (Mark Yohalem).

The question concerns if spreading myself too thin paid off and imo, no, it would have been a better game had I not. Inspiring others to go on to make their own games as a benefit is true, it’s great (plus you get to play their games), and you do always hope for that, but in the context of the company and people and the time spent to make Torment (it was very late, even using another developer’s engine), it was problematic. Also, Interplay wasn’t doing well at the time (Fallout 2 had needed to be released to prevent layoffs) so there was that hanging fear of dread as well – not necessarily in Black Isle, but the fact your friends in other divisions could get laid off if the company didn’t start seeing a return on their games.

In regard to "lead producer quitting after beta": Guido Henkel was the one who said that, but considering some interviews with him, I'm not sure he's reliable. Was that what happened?

I think Guido left about halfway through the game’s production, although during that time, he spent most of it on Neverwinter Nights 1. Guido was involved heavily in the game’s marketing, however (logo, cover). Kenneth Lee came in after, and he was focused solely on Planescape (and he designed a lot of the spells).

[Fairfax: NWN1 was originally going to be published by Interplay under the Black Isle division. However, after a legal dispute with Interplay over unpaid royalties, BioWare was able to get a new publisher: Infogrames/Atari.]

You've said in other interviews that you've been working on your current projects from home. What are the pros and cons of doing that? And does that change with the new studio? If so, how?

Yep, wake up, roll out of bed, start typing. Keep typing. Get designer’s block. Switch to another project for an hour, keep designer block on other project in mind, suddenly find a solution while your unconscious percolates on the problem, switch back to other project. Keep doing this until it’s night.

The negative is I stopped working out b/c I couldn’t stop, I felt like I was back on Torment, which while good isn’t great for one’s health. I also miss some of the Obsidian devs, but I still see them on occasion (and see a number of them now at other jobs).

It’s weird – I used to have to get away from work to get work done (if you ever wondered why I used to travel), but now I don’t travel much b/c I’m finally able to write at a natural pace – and on a variety of different writing and design tasks, some broad, some very specific.

Differences from working in a studio – so at home, there’s less meetings. I don’t have to wear pants for meetings (which was a hard enough struggle to begin with). If family is sick or in trouble, your schedule allows you to move quickly. Other studios have different and better tracking software. Strangely, non-studio contract payment is more reliable. Don’t have to clear it with everyone and their mother if you need to go to the doctor. A huge variety of other things you can do design-wise is available (interface, system design, different genres, different dialogue mechanics, etc.) rather than doing one specialist role. Lastly, I never, ever realized exactly how much time is wasted driving into work (even a small commute) not to mention getting ready for work – cut that out of your day, and it sets countless hours free. Family circumstances forced me from taking a few jobs in the past 2+ years, but it may have been for the best – this work style suits me much better.

Brian Menze, Anthony Davis, Chris Avellone
Plenty of studios are having success with remote staff, and you sound like you don't want to work in an office ever again. Is that the future for you and the new studio?

My issue isn’t being in a studio, it’s my life. I can’t be full-time because of family concerns. I do appreciate that when Obsidian assigned writing tasks to me, they did allow me to do work via remote, and I got much more writing done as a result (before then, I had to work nights and weekends when everyone else had left so I could concentrate, and by that time, you’re usually exhausted - it's even worse when you're on two projects, and one is close to shipping).

I do think remote work has a lot of pluses (and drawbacks, too). There’s plenty of ways to have internet meetings if necessary, or keep Skype running with select teammates if working on a shared problem. Google Drive also makes sharing and editing docs painless – I far prefer it to other feedback systems we used in the past.

Overall, I think in-office work is great for brainstorming, getting incidental ideas from others, doing high-level concept or gameplay presentations, and coordinating some aspects of development (esp. if you work in a “pod” with 3-4 other developers on the same game level, for example), but remote work does save a good chunk of the day for work. And if you’re writing companions, you need that uninterrupted headspace to keep writing or their tone can get distorted over time (I imagine programmers writing complex code may feel the same about keeping the momentum going).

Also, when you’re remote, you tend to have less “fill me in on what you’re working on in a large 1 hour meeting” vs. “hey, I sent you a 5 minute email updating you on my completed and next tasks.” If you multiply that saved time by 5 (at least), that saves you so much time in any given work week, since in-office meetings can have multiple “give me info” meetings over the span of a week, and it gets worse the higher up the management chain you go. You can put a price tag on meetings like that, and it’s more expensive than people may realize (I'd be afraid to calculate what owner meetings cost a company per year - which is made even worse if there happens to be no resolution at that meeting, or the meeting makes a detour from the goal, requiring a second meeting to address what should have been done in the first meeting).

Another benefit of remote (it’s harder to do in-office, but you can try), is if you insist on a meeting agenda and goals, you can often answer agenda points in advance and either cut the meeting time in half or remove the need for the meeting entirely – and often you can respond faster (before the meeting) and give data and updates much, much sooner than trying to herd everyone in the same room, trying to find the right meeting room, double-checking where the doc you'r referring to is on the server, who lost the company meeting room computer's login credentials, etc., etc.

Whether you are working remote or in an office, btw, you always should insist on an agenda, but in an office, I often found it was skipped for expediency (or met with anger b/c the organizer couldn’t be bothered), when in fact, not doing an agenda wasted more time. I consider agendas to be a producer’s/lead’s way of organizing and taking notes before a meeting begins, if used properly.

(Also, in any properly managed system, you shouldn’t need to be sending any task updates to your boss – the system should be designed so they can glance at the tasking system at any time and see for themselves without even asking you. If people don't use it, then you give them feedback, then adjust their raises next paid period as necessary, either positively or negatively.)

Chris, thank you for putting so much time and effort into this, and best of luck with your current projects.

There are 5344 comments on RPG Codex Interview: Chris Avellone on Pillars Cut Content, Game Development Hierarchies and More

Site hosted by Sorcerer's Place Link us!
Codex definition, a book manuscript.
eXTReMe Tracker
rpgcodex.net RSS Feed
This page was created in 0.13319611549377 seconds