Brad Wardell's views about technology, politics, religion, world affairs, and all sorts of politically incorrect topics.
Published on January 20, 2011 By Draginol In Elemental Dev Journals

You hear about executive producers sending “notes” down to the director or producer of a movie.  That’s what I’m doing these days. “Maybe we should have robots in this scene instead of a dragon?” I’m leaving the dream. Smile

I assume a lot of people read these journals because they like seeing what happens at a game company that they would normally not get to find out about because most game companies these days are publicly held and thus there’s a strict PR leash on everyone. So today I’m going to outline to you what a medium sized game production looks like in terms or organization.

I’m also going to try to show those of you who are aspiring game developers some of the things to look out for that can lead to disaster.

A Stardock Game Org Chart

First, let’s look at the general Stardock game project org-chart.

image
Stardock Entertainment Model

It looks huge but many of these spots are occupied by 1 person. In fact, some people wear multiple hats in here.  For instance, our art lead also does all the UI. Kael is the Project Manager and also the Design Lead.

The object is to find the right balance between maximizing your “resources” usage (people) and the overall output. That is, sometimes, it’s fine for someone to wear more than one hat – it’s often preferable. The challenge is to find out when they are wearing too many.

Sometimes, you end up having little choice but to stretch people thin. That’s where “crunch” time comes in. It’s when you’re asking people to do far more than they really should be expected to do. Crunch time occurs because someone – usually management – screwed up.

Now, if you look at the titles listed in the chart above, you’ll notice that they don’t sound that gamey.  For marketing, we try to translate our software engineering titles into something that remotely conforms with industry norms such as:

Product Manager = Executive Producer & Lead Designer

Project Manager = Producer & Designer

And so on.

Other Org Charts

Other companies do it differently. At Gas Powered Games, the Lead Designer is on a separate track from the Project Manager. This has some advantages in that the designer doesn’t have to be a software engineer or project manager. With larger teams, you really do need someone with a background in project management to run the show.  Gas Powered Games teams tend to have 30 or more people on it. Hence, finding someone who is both a game designer and someone capable of managing that many resources would be very difficult.

Stardock teams tend to be much smaller. Historically, our game teams are 10 or so people. Elemental: War of Magic  got to be about double that and that’s one of the reasons it got into trouble.

For the first 2.5 years of Elemental’s development the org chart looked like this:

image

I wore two hats. Product Manager and AI. 

My job was oversee the overall product and provide AI.  Unfortunately, during a good chunk of this, I got diverted to help out another project that was struggling for several months. This turned out to have been a fatal decision – I didn’t assign someone else to take over for me while I was gone.

When I finally got back into the thick of things, it was clear that I needed to add more hats since we couldn’t readily bring in people easily (one of the downsides of being in Michigan, I can’t just go up the local Digipen and get people).

By the time War of Magic shipped, it looked like this:

image

With the benefit of hindsight and common sense, that’s too many hats (the orange spots). 

Each new hat has its own unique story behind it (Marketing Manager left the company, The printing deadline got changed by a month and I type faster than anyone else, our Producer’s wife was having a baby and could no longer put in the time necessary, the lead developer was having health problems, the tile creation (part of concept) was far behind, etc.). But in the end, I had assigned myself across the whole project. If you’re a new game developer, you may think you can do everyone else’s job. Maybe you can. But you shouldn’t. If you’re the one who gets to decide the game is “ready for release” you shouldn’t be too involved in the parts that determine whether it actually is ready for release. You can get way too close. This is a very very common problem in small software studios. One we knew about but after a shelf of Editor’s Choice Awards, you start to think the rules of software engineering no longer apply.

This is why, if you’re an aspiring game developer, that you should never forget, even for a second, that making a game is an exercise in software engineering  -- and has to be treated as such.  None of the above hat acquisitions occurred overnight. Only gradually, does one realize that people were taking over more and more spots.  Our Art Lead, for example, had taken over all UI and was involved in a lot of the design (XML) implementation. Our Engine lead and I were doing debugging scenarios, etc.  I cannot overstate just how talented our staff here is.  It’s the thing that comes up all the time when people visit us – you did THIS with this few people? It gets dangerously tempting to forget that such a team has any limits at all.

I’m friends with “Chantz”, the guy Atari sent in to try to rescue Master of Orion 3 and we have talked a lot over the years about this kind of thing. Beware of the combination of the phrase “I can do it” and also being the guy who gets to evaluate whether they can actually do it. The guy who says “I can do it” and the guy who gets to determine whether they really can do it should never be the same guy.  Because sometimes, you really can’t do it no matter how sure you are that you can. Fatigue, overconfidence, external factors can all weigh in to create a fractured evaluation of a given project.

If it’s a team of a half dozen guys, you can wear a lot of hats. But as the scope of the project gets bigger, you have to be careful because those hats start to get really really big. Smile

Process Matters

This is why delaying a project is pointless if the process is broken. If the process is broken, the result will be broken. Time isn’t the issue in such a case.  This was one of the principle reorg challenges we went through this Fall. Hiring very talented people to fill in some of these key roles.  The biggest piece of advice I can give to game start-ups is this, make sure you separate the business decision makers from the development decision makers. 

I hope you guys find this interesting and useful.


Comments (Page 1)
on Jan 20, 2011

BTW, Frogboy == Draginol. Just a different account.

on Jan 20, 2011

Where am I on this chart?  I should be on there somewhere... #1 awesome poster, or maybe Stardock Mascot..Toby's little secret...something like that.

on Jan 20, 2011

Awesome!

Thanks for providing some really amazing incite!

on Jan 20, 2011

Thats all fine and dandy. But, it doesn`t change a thing. Brad, did you even play the game before it was release ?

 

Freebird out.  

on Jan 20, 2011

That's interesting Brad. The challenges you describe are very similar to those encountered in any software development shop that ramps up too quickly with the structure in place. My company went through the same thing about 4 years and put out a buggy version too early (corporate timelines...) that took 3 years to clean up. Lot of restructuring and people changes (some voluntary, some definitely not) was the result.

on Jan 20, 2011

Xia, you are on there.  Lower left corner, under asset support.

 

Thanks Frogboy, although it confused me until you posted that you and draginol are the same person.  Did you have some many hats that you developed a split personality.  I think it might be time for you to talk with someone.

 

I nominate Xia.  He is very understanding and compassionate.

 

on Jan 20, 2011

Freebird1956
Thats all fine and dandy. But, it doesn`t change a thing. Brad, did you even play the game before it was release ?

 

Freebird out.  

Every day for hours on end.  

 

on Jan 20, 2011

Frogboy

Quoting Freebird1956, reply 4Thats all fine and dandy. But, it doesn`t change a thing. Brad, did you even play the game before it was release ?

 

Freebird out.  

Every day for hours on end.  

 

 

I think that was part of the point behind this post - He helped build so many bits and pieces that it was hard for him to see whether or not the game was ready for release. And I understand his perspective. When I was younger I used to write - A lot. I thought it was great, no one else really did. Now that I've distanced myself I re-read some of it and think "Wow, I suck." But when you're attached you don't get that perspective.

Now that he's in his proper role, and after a few months, he can go back to seeing what the game really was to the rest of us. And he's learned to avoid taking on too much when he's the one making the decisions.

on Jan 20, 2011

What's a mock review?

on Jan 20, 2011

Please add a box for Lord Xia (and the rest of the community).  You can label it "Fargin Iceholes".

on Jan 20, 2011

Thanks for sharing, Brad!

on Jan 20, 2011

Nice illustrations of the creeping of the hats.  Personally I find this more dangerous than the more infamous scope creep.  

 

 Are you trying out some of your slides for the conference talk on us?  Or just the illustrations?

on Jan 20, 2011

I chuckle at the assumption that Brad didn't play the game before release.  I think the deeper people are into any project, the more they tend to overlook smaller details because of such a broad, but deep amount of knowledge. 

I had a friend once that was in a stoner band with a few of his other freinds.  They practiced, they wrote their own music, they went to the recording studio and put it on a CD...they worked their arses off!!

In fact, they lived it. How can you tell them that their music just sucked?  Singer way off tune, lack of any compelling melody, yet not having any harmonic depth either.  Just...awful....

I posit this:  if they were to have taken 2 years off, get out of the process of the band, the music, the bias toward self...they would listen to it in 2 years and would have immediately known that this isn't how they remembered it.

Anyway, the best person to test the game would probably be the people on the business side, assuming they are gamers =D

on Jan 20, 2011

Frogboy you really don't need to respond to posts like the one from Freebird1956. It was a childish post.

P.S. Thanks for the frequent insight into the gaming industry.

on Jan 20, 2011

If you check Freebird's history he has a knack for misunderstandings or blatant no-use posts. There's a thread where he complains about the game losing his interest back in 1.09 because of all the crashes, but he doesn't offer to upload his crash data. There's several threads where he says he's leaving the forums, and he doesn't.

He can't admit to himself that he's still hooked and still wanting to see what becomes of Elemental. So I respond to him as if he truly means what he puts, because I respect the fact that he's still here in spite of what he considers problems with the game.

Yes, pot shots like what he took at Brad in a post that Brad put together for our benefit, and not really in relation to Elemental in itself, don't serve much purpose. But if he's lashing out he still cares.

Meaning that he's one more valuable member of an awesome community.

Lord Xia, I still want your babies. >.>

Meta
Views
» 14234
Comments
» 36
Sponsored Links