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

[Prime Junta]

IMO you need someone in charge of projects and hierarchy to keep track of who's doing what and how they are progressing. If someone's not pulling their weight, for whatever reason, there needs to be process to sort out any problems, I know it's probably problematic in creative endeavors as creative fluids might not flow on schedule, but how else can you keep the project running on time?

In my experience, clear hierarchy is necessary to keep it clear who is responsible for what and who reports to whom, of course it doesn't always work like it's planned as there might be personality clashes and whatnot.

As I said, some hierarchy is probably inevitable. The question is how high is the structure, and how do people in it conceive of their responsibilities. Of course situations where someone isn't working out -- not pulling their weight, causing social problems, whatever -- need to be sorted out, and it's maybe a little bit utopian to expect a comrades' court to take care of it.

In my opinion a software house should be organised a bit like a carpentry workshop. Making software -- and this applies to a game just as well as any other software -- is a craft after all. You've got your master craftsman who understands the whole process. You've got your journeymen who are at least competent, possibly extremely competent in one or more specific areas as well as having at least a partial understanding of the big picture. And you've got your apprentices who range from people who are just learning the craft to a majority who's capable of working productively under supervision, to a few that are already very good at some limited areas. The journeymen mentor the apprentices and the master participates in the work, teaches the most promising journeymen personally, and breaks any ties that emerge in the work.

And of course the master and journeymen agree about who takes charge of producing whichever order is being worked on, and how the journeymen and apprentices are assigned to each of the orders. But you don't really need an Assistant Sub-Manager Of Fishtail Joins with a team of Junior Fishtail Join Engineers working in his team, just somebody who's able to make a fishtail join, and some mechanism to make him available where fishtail joins are needed.

 

[Chris Avellone]

Chris Avellone thanks for the honest opinion.

During your time there (at Obsidian) do you feel that the trend of releasing unfinished games was endorsed internally with a "release now and patch later" approach to development or was it simply down to circumstances of working with tight publisher and budgetary deadlines? I ask this because the "buggy obsidian classic" has become an accepted part of gaming culture and obsidian seems to have a track record to back that meme up.

No, it was more "release now or we won't get paid," or "release now or we don't get as much as we would if we fixed everything we thought we should." The royalties for KOTOR2 were definitely a factor in Feargus's decision for the release date, and those royalties would have only been forthcoming and as high as they were if we delivered the game with the original schedule. It was depressing, as those royalties didn't seem to help a whole lot in the grand scheme of things, and KOTOR2 definitely wasn't our finest hour.

 

[<3sRichardSimmons]

Hey Chris, have you played King of Dragon Pass? There was an interview that George gave here a few years ago where he talked about how one of his dream games would basically be a KoDP-clone set in the Fallout universe with a heavy emphasis on C&C and faction interaction, and the player leading a group (BoS, Great Khans, Vault 13, etc.) through the Wasteland. I thought it was an amazing pitch, and when he joined inXile full-time shortly thereafter, I kept hoping he might convince Brian to let him make that game with the Wasteland license, but obviously no dice.

Don't really have a question around it, but you and George should go make that game; I'll buy it as many times as you need for it to be in the black.

One thing I've seen danced around in this thread, but not addressed explicitly (correct me if I'm wrong) is why OEI seems to have so much difficulty retaining writing talent. From my perspective, George Ziets and John Gonzales were punching miles above their competition with their work at Obsidian. Their employment retention should have been of the highest priority for the studio (obviously I don't know what the associated costs were, but George at least seems like such a mellow guy that I have a hard time imagining him asking for an exorbitant compensation package for his talent). Your posts seem to imply that it was primarily a matter of upper-management undervaluing genuine creative talent and over-exerting their own influence where it wasn't warranted. If this is the case, that is immensely sad, and makes me very glum for the future of OEI, but it also makes me wonder: what was different about MotB (and to a lesser extent F:NV)? Were Feargus and Parker just not around due to family/life stuff?

 

[Abu Antar]

There’s still one thing that I don’t understand. Why now contact a lawyer once he got de-ownered and got nothing out of it. It seems like a reason has to be given. There’s also the question why no compensation was paid out. I sure as hell wouldn’t take shit like that lying down. It doesn’t make sense to me that anyone would agree to that.

 

[Quillon]

And of course the master and journeymen agree about who takes charge of producing whichever order is being worked on, and how the journeymen and apprentices are assigned to each of the orders. But you don't really need an Assistant Sub-Manager Of Fishtail Joins with a team of Junior Fishtail Join Engineers working in his team, just somebody who's able to make a fishtail join, and some mechanism to make him available where fishtail joins are needed.

I don't know man. Sounds too fishy.

 

[Jaedar]

(For some reason this doesn't seem to work with Chinese men. They immediately start looking for the top dog in the system, and once they've decided who he is, they start undermining him in order to take his position.)

Interesting story about chinese corporate culture(relayed by people at my work who sometimes cooperate closely with chinese): For almost all managerial positions, they hire two people for the same job and keep them in constant competition to try and get more/better work out of them. To keep the amount of people down in important meetings, they apparantly often divide it in two, the first part everyone is present and is essentially just a presentation of the issues. Then they'll go have lunch (or something), and only invite the people who had something useful to say during the first part of the meeting, and it's during this lunch the actual decisions will get made.

I know American companies love their hierarchies, european ones tend to be way flatter. I suspect that like the chinese, it might come down to cultural differences that make the optimal configuration different in different places.

In other words Chris might need a clear and somewhat rigid hierarchy to be at his best, whereas this would be counter productive in other places.

Also, Chris Avellone Kotor2 was one of obsidian's finest hours imo.

 

[Chris Avellone]

Hey Chris, have you played King of Dragon Pass? There was an interview that George gave here a few years ago where he talked about how one of his dream games would basically be a KoDP-clone set in the Fallout universe with a heavy emphasis on C&C and faction interaction, and the player leading a group (BoS, Great Khans, Vault 13, etc.) through the Wasteland. I thought it was an amazing pitch, and when he joined inXile full-time shortly thereafter, I kept hoping he might convince Brian to let him make that game with the Wasteland license, but obviously no dice.

Don't really have a question around it, but you and George should go make that game; I'll buy it as many times as you need for it to be in the black.

One thing I've seen danced around in this thread, but not addressed explicitly (correct me if I'm wrong) is why OEI seems to have so much difficulty retaining writing talent. From my perspective, George Ziets and John Gonzales were punching miles above their competition with their work at Obsidian. Their employment retention should have been of the highest priority for the studio (obviously I don't know what the associated costs were, but George at least seems like such a mellow guy that I have a hard time imagining him asking for an exorbitant compensation package for his talent). Your posts seem to imply that it was primarily a matter of upper-management undervaluing genuine creative talent and over-exerting their own influence where it wasn't warranted. If this is the case, that is immensely sad, and makes me very glum for the future of OEI, but it also makes me wonder: what was different about MotB (and to a lesser extend F:NV)? Were Feargus and Parker just not around due to family/life stuff?

Yes, I enjoy that game very much, and am working on something similar for the same reason: I think it's a fun style game, and I wanted to work on something like it. It was a complicated decision to say "yeah, I'm interested." No details I can provide yet, though.

 

[VentilatorOfDoom]

Lead Writer: Well, this is the shitty thing about doing the job: AP needed a lead designer, and I was chosen by default just as Chris Parker was chosen as Project Director b/c we were the two best options at the time.

So... who's to blame for the dumbass minigames every 5 meters?

 

[Chris Avellone]

I can’t find the post, but the gist of it was “did you see yourself distancing from Obsidian with FTL, Wasteland 2, Tides, etc. and did that impact Obsidian?”

Those projects did open my eyes to other possibilities, but FTL:AE was a few weekends of work I did for free, WL2 was contracted and directly benefited Obsidian and was given guidelines determined by Feargus and I that were then broken (see below), and Tides was to be done in my own time and Feargus agreed to let me be a part of it as long as Obsidian got the tool payment – he did try to cut my personal amount for contributing in half, not realizing I’d already talked to Fargo about the amounts (I didn’t get angry about this, but I did counter his offer and told him I’d already discussed my terms without him and Brian and I had discussed a much different price for my involvement).

Wasteland 2 I was contracted out for by Obsidian, and all the money for working on that project went straight to Obsidian (this went to the company account to help with the debt, though, not the employees – for the employees, however, I did take the money I received for the All Roads graphic novel from New Vegas and donated all of it to get everyone FNV shirts even though Feargus had told me “you don't need to, you can keep the money for yourself”). I don't think people were fully aware the WL2 days were to help Obsidian with debt.

Anyway, for WL2, Feargus and I agreed on a schedule and # of days I would be an inXile based on the amount Fargo and inXile were paying for my time and involvement. However, halfway through our agreed-on schedule, Feargus ended up changing his mind and canceling the WL2 time (although I don’t know if he spoke to Brian about this – all the WL2 designers knew was that I wasn’t there anymore, and I wasn’t sure I was allowed to talk about it, since there wasn’t anything positive to say - it was pretty depressing). What was really depressing is we'd agreed on a plan, we were adhering to the plan, and even after I showed him what we agreed to, calculated it, and showed him the breakdowns of time I had spent and had left to go, it didn't matter b/c it wasn't "well, that's not how I feel today," experience all over again - he just wanted me back vs. adhering to our agreement.

Feargus didn't usually like me doing anything outside of the Obsidian roof (although he did give me permission to work by remote when needed when writing for Eternity became important, although I don't think this was communicated to the teams).

 

[Jim the Dinosaur]

Chris Avellone maybe you've talked about this elsewhere, but were you despondent/enthusiastic/indifferent about the early-00's transition to 3D, other than it being a sort of sink-or-swim type of response to the troubles for PC-only gaming at the time? I mean in hindsight it seems like a horrendous transition, but maybe you guys were excited initially about the flexibility it would afford developers?

 

[Jarpie]

IMO you need someone in charge of projects and hierarchy to keep track of who's doing what and how they are progressing. If someone's not pulling their weight, for whatever reason, there needs to be process to sort out any problems, I know it's probably problematic in creative endeavors as creative fluids might not flow on schedule, but how else can you keep the project running on time?

In my experience, clear hierarchy is necessary to keep it clear who is responsible for what and who reports to whom, of course it doesn't always work like it's planned as there might be personality clashes and whatnot.

As I said, some hierarchy is probably inevitable. The question is how high is the structure, and how do people in it conceive of their responsibilities. Of course situations where someone isn't working out -- not pulling their weight, causing social problems, whatever -- need to be sorted out, and it's maybe a little bit utopian to expect a comrades' court to take care of it.

In my opinion a software house should be organised a bit like a carpentry workshop. Making software -- and this applies to a game just as well as any other software -- is a craft after all. You've got your master craftsman who understands the whole process. You've got your journeymen who are at least competent, possibly extremely competent in one or more specific areas as well as having at least a partial understanding of the big picture. And you've got your apprentices who range from people who are just learning the craft to a majority who's capable of working productively under supervision, to a few that are already very good at some limited areas. The journeymen mentor the apprentices and the master participates in the work, teaches the most promising journeymen personally, and breaks any ties that emerge in the work.

And of course the master and journeymen agree about who takes charge of producing whichever order is being worked on, and how the journeymen and apprentices are assigned to each of the orders. But you don't really need an Assistant Sub-Manager Of Fishtail Joins with a team of Junior Fishtail Join Engineers working in his team, just somebody who's able to make a fishtail join, and some mechanism to make him available where fishtail joins are needed.

Yeah, obviously, I think there is the balance (Insert Sawyer-meme in here) between very strict hierarchy and very loose organization, no matter what's the structure, the shitty people will fuck it up for everyone. If you have clear/tight hierarchy, you have to make sure that the higher ups won't get involved with the nittygritty things, a great example of this is in the finnish documentary about the rise and fall of Nokia. If you've watched it, you might remember the bit about the slogan "Connecting people", the marketing people had to give the board of directors the list of the slogans, and the board would then choose which they'll use. The marketing department put so utterly shit slogans in the list, so that the board would have only choice, Connecting People.

No matter what's the organizational structure, people in top needs to leave the professionals to do their work. Based on what Chris Avellone has said, this doesn't seem to have been case in Obsidian. Also based on my own experience, some people works best when they have clear instructions what are expected of them and some people works best when they're given loose instructions, the boss has to know which works best for his underlings. I didn't work in creative job per se, but I had three bosses in my previous work; second one was very loose one, he gave me wide berth to do my job the best way I saw fit and gave me feedback when it was necessary, the last one was always looking over our shoulders and micromanaging us all the fucking time and "driving from the back seat", which made me go crazy. The first boss was something in between, I worked best under the second boss, but some of us worked best under the last one.

 

[Prime Junta]

By the way, in case somebody's wondering how come I have such strong opinions on this, there's a reason.

Where I work used to be divided into two departments: development and projects. Dev worked on our base products; projects customised them for specific customers. Dev was organised on a "flat" hierarchy, and projects were organised on a traditional top-down project hierarchy.

The projects side was a constant underperforming PITA while problems on the dev side were, by comparison, pretty quickly and easily solved.

Moreover, we finally eventually ended up abolishing the division and a lot of the hierarchy on what used to be the projects side, introducing practices we had found to work well on the dev side. Things immediately started improving.

So I've seen very close up how these two different organisational principles operate on the ground. The top-down hierarchical organisation just performs much worse than the flatter bottom-up organisation in almost every measurable respect.

 

[Grauken]

No matter what's the organizational structure, people in top needs to leave the professionals to do their work. Based on what Chris Avellone has said, this doesn't seem to have been case in Obsidian. Also based on my own experience, some people works best when they have clear instructions what are expected of them and some people works best when they're given loose instructions, the boss has to know which works best for his underlings. I didn't work in creative job per se, but I had three bosses in my previous work; second one was very loose one, he gave me wide berth to do my job the best way I saw fit and gave me feedback when it was necessary, the last one was always looking over our shoulders and micromanaging us all the fucking time and "driving from the back seat", which made me go crazy. The first boss was something in between, I worked best under the second boss, but some of us worked best under the last one.

The problem with adapting to micro-managers is that once you get back to a less-structured environment, its a fucking pain to shift back to working like that

 

[Ninjerk]

By the way, in case somebody's wondering how come I have such strong opinions on this, there's a reason.

Where I work used to be divided into two departments: development and projects. Dev worked on our base products; projects customised them for specific customers. Dev was organised on a "flat" hierarchy, and projects were organised on a traditional top-down project hierarchy.

The projects side was a constant underperforming PITA while problems on the dev side were, by comparison, pretty quickly and easily solved.

Moreover, we finally eventually ended up abolishing the division and a lot of the hierarchy on what used to be the projects side, introducing practices we had found to work well on the dev side. Things immediately started improving.

So I've seen very close up how these two different organisational principles operate on the ground. The top-down hierarchical organisation just performs much worse than the flatter bottom-up organisation in almost every measurable respect.

Yeah, it's not like you to have strong opinions.

 

[Jarpie]

No matter what's the organizational structure, people in top needs to leave the professionals to do their work. Based on what Chris Avellone has said, this doesn't seem to have been case in Obsidian. Also based on my own experience, some people works best when they have clear instructions what are expected of them and some people works best when they're given loose instructions, the boss has to know which works best for his underlings. I didn't work in creative job per se, but I had three bosses in my previous work; second one was very loose one, he gave me wide berth to do my job the best way I saw fit and gave me feedback when it was necessary, the last one was always looking over our shoulders and micromanaging us all the fucking time and "driving from the back seat", which made me go crazy. The first boss was something in between, I worked best under the second boss, but some of us worked best under the last one.

The problem with adapting to micro-managers is that once you get back to a less-structured environment, its a fucking pain to shift back to working like

I solved it by quitting the job, as I'd gotten way too fed up with the work anyway. Probably best decision of my life.

 

[Kyl Von Kull]

Yes, I enjoy that game very much, and am working on something similar for the same reason: I think it's a fun style game, and I wanted to work on something like it. It was a complicated decision to say "yeah, I'm interested." No details I can provide yet, though.

Hmm, are there any other top secret projects you're working on that you could shed a teensy tiny bit of light on via this kind of analogy? Apologies in advance; this question feels inappropriate, like I'm asking a genie for infinite wishes. But, you know, if infinite wishes are on the table...

 

[Mustawd]

Chris Avellone, are you familiar with the premise of Project Indiana? And if so, what are your thoughts on it?

Not trying to get details, just your thoughts on it.

 

[Chris Avellone]

Chris Avellone

I recall an article where you said writing/doing work for a mod was one of the best ways for someone from the outside to break into the industry, have you personally seen any mod-folk hired at Obsidian? Did you guys actually check out the mods the applicants worked for in-depth?

Oh, and what about Twines? Not very familiar with the program, but I'm assuming you spend a few minutes checking out how it's written?

Yes. Jorge "Oscuro" (FNV, DLCs, PoE) was hired for FNV based on his mod work for Oblivion (he rewrote the Obsidian content layer in his spare time, did patch notes, re-released, etc. - all the things you do for a game release, he did on his own - and because he did it for fun on his own time and did a thorough job, we knew he'd make a great developer).

I do know there's designers at Obsidian who also reach out to modders, and we used modders' work in the past (paying or giving credit) since they've done exceptional work (the A.I. system for NWN2, for example, was provided by a modder because it was very well done).

Twines: I try to play through the whole Twine, but usually 15-20 minutes is long enough to get a feel for the person's aesthetic, word choice, response options, plot pacing, dialogue, etc. I do often ask (or make) a text export of the Twine if I can to check for spelling errors, but I don't make that the sole basis for approval/denying a candidate. In design tests, we do try to establish a limit on dialogue and page counts, as it's often very easy to see a person's writing in a few pages vs. 20 or 30 pages, and there's no need for the applicant to kill themselves doing the test. I also allow applicants with experience and game content they've worked on previously to submit that same game content (ex: a video of one of their dialogues for the dialogue section) in lieu of making something from scratch.

 

[Chris Avellone]

Chris Avellone, are you familiar with the premise of Project Indiana? And if so, what are your thoughts on it?

Not trying to get details, just your thoughts on it.

Not really, but I have faith in Tim, Leonard, Charlie, Tyson, and the rest of the team to deliver something cool.

 

[Alienman]

This thread is a new fall out bible in the making

 

[Grauken]

I think its more like the Necronomicon

 

[Quillon]

20 mins and no one posted anything in here, I can breathe!!!

Damn, what have I done...

 

[Blaine]

Damn, what have I done...

You've just shitposted.

 

[Blaine]

Here's an example of a quality post for future reference:

 

[Riddler]

IMO you need someone in charge of projects and hierarchy to keep track of who's doing what and how they are progressing. If someone's not pulling their weight, for whatever reason, there needs to be process to sort out any problems, I know it's probably problematic in creative endeavors as creative fluids might not flow on schedule, but how else can you keep the project running on time?

In my experience, clear hierarchy is necessary to keep it clear who is responsible for what and who reports to whom, of course it doesn't always work like it's planned as there might be personality clashes and whatnot.

As I said, some hierarchy is probably inevitable. The question is how high is the structure, and how do people in it conceive of their responsibilities. Of course situations where someone isn't working out -- not pulling their weight, causing social problems, whatever -- need to be sorted out, and it's maybe a little bit utopian to expect a comrades' court to take care of it.

In my opinion a software house should be organised a bit like a carpentry workshop. Making software -- and this applies to a game just as well as any other software -- is a craft after all. You've got your master craftsman who understands the whole process. You've got your journeymen who are at least competent, possibly extremely competent in one or more specific areas as well as having at least a partial understanding of the big picture. And you've got your apprentices who range from people who are just learning the craft to a majority who's capable of working productively under supervision, to a few that are already very good at some limited areas. The journeymen mentor the apprentices and the master participates in the work, teaches the most promising journeymen personally, and breaks any ties that emerge in the work.

And of course the master and journeymen agree about who takes charge of producing whichever order is being worked on, and how the journeymen and apprentices are assigned to each of the orders. But you don't really need an Assistant Sub-Manager Of Fishtail Joins with a team of Junior Fishtail Join Engineers working in his team, just somebody who's able to make a fishtail join, and some mechanism to make him available where fishtail joins are needed.

Of course the height of the organizational hierarchy needs to be closely managed so that people can operate without constantly seeking approval and operations devolve into inter-manager politics and constant need for more communication rather than being productive.

One common issue that I often see here is that many people, in particular within SW dev, have had bad experiences with hierarchies and thus tries to have as flat organisations as possible and often to pretty great success. The problem is that the case studies of when this really works well are based on two facts:

• The need of very competent, driven and self-motivated employees, something that is a pretty limited resource when everyone are screaming for more developers.
• The cases studies that the theories of flat organisations are based on are relatively small teams and companies (many cases start ups).

Really flat organisations with a lot of "cross pollination" do not work well when organizations scale up, both due to the problem of finding competent enough employees and the issue of the model not scaling well.

As a management consultant I see the lack of hierarchies as a much more common failure mode than hierarchies. This is not due to flat organisations being bad but because they are overused and not adapted to the realities of organisations that do need hierarchies.

Saying one or the other is better or worse is naïve, both are needed to varying degrees in different types of organisations.

In the end however it is rarely the organizational map that is the issue, it is the managers. Being a manager is very difficult and stressful and the method for selecting people for the role is often very flawed and even random (also promotion to incompetence) which leads to bad outcomes regardless of systems.

Most of the time when we come in to help an organisation straighten out their practices the organisation we propose is often based on (when we are doing a thoughrough job) eliminating or circumventing destructive managers or practices rather than implementing some "master organizational plan".

In the end, If the rot is at the very top (which it often is in startups that have been moderately successful) it's often difficult to do anything really productive.

Edit: organizational theories are kind of like diets. People have trouble taking a measured approach and end up with something too extreme which they can't maintain.