I'm Not Sure About
the benefit of 20/20 hindsight, it's easy to identify some things
as having gone "right" and other things as having gone "wrong."
However, some of the most interesting things to consider are the ones
that aren't so easily pinned down. Here are some questions that are
still very much open in my mind. (If any of you have answers, feel free
to share them!):
- Is it better to start the design process with fiction and high-level
gameplay goals or to dive right into game systems? We did a much
better job, I think, of the former than the latter. Too much of our
system stuff had to be rethought relatively late in the project. Only
the conversation and inventory systems are largely untouched from
where we started. We remained true to our high-level goals, but I
can't shake the feeling we could have done a better job early in the
project on the system design front.
iconic/abstract representation of characters, power-ups, player rewards,
tools, objects, and so on better than realistic/specific representation?
In other words, are instantly identifiable floating crosses better
as healing items than a med-bot, something the player may or may not
be able to identify? Is it compelling to wonder if that guy over there
is a good guy or a bad guy? Or is it better to know just by looking
at him, so you can plan accordingly? As in so many things, we went
with a hybrid approach -- nothing as extreme as floating health restorers,
but instantly recognizable good guys, bad guys, rewards, and objects.
it better to stop the action while the player is in interface screens
or to keep the action going à la System Shock 2?
We chose to stop the action because, for us, the tactical decisions
this allows outweighed the artificiality and the loss of immersion.
Was that the right decision? Probably, but there's no way to assess
the road not taken in this case.
you get to name your character or not? A holy war almost broke
out on the Deus Ex team about this. "If you can't name
your character, it's not an RPG," said some. "If we don't
name the character, how do we write and record compelling conversations
and create a cool story?" said others. "Story isn't the
point…" "Yes, it is…" and on and on and on.
We compromised: we gave the player character a code name and back-story
but let the player select his real name, which came into play in various
ways (though never in speech).
it better to worry about graphics and art direction and cinematics
early or late? We did zero flic work until the very end of the
project, and though the game looks good, we left it to the end to
go back and make a pass at consistency (getting the lighting and texturing
just right, and so on). I decided it was more important to get the
gameplay under control than to get the game looking good. We did make
our art direction pass and we did make the game look better -- really
good, in fact, in my totally prejudiced opinion. But it's unclear
to me whether the game could have looked even better if we'd gone
the other way and dealt with art issues sooner rather than later.
The Bottom Line
of the challenge of game development is making the tough decisions along
the way, leading to many difficult junctures when you have to determine
that something that can't be done right in the game shouldn't be done
at all. Notice the complete lack of references to multiplayer action
in this Postmortem. We wanted to provide multiplayer support but didn't
have the time to do the job we knew we needed to do, and so it got cut.
generalize from that point: It's all well and good to have design goals
and an ideal game pictured in your head when you start, but you have
to be open to change and realistic about what can and can't be done
in a reasonable time frame, for a reasonable amount of money, with the
personnel and technology available to you. And if you don't have time
to do something right, cut it and do everything that's left so well
that no one notices the stuff that isn't there.
not saying we did that perfectly on Deus Ex. We certainly didn't
ship a perfect game. But if we hadn't gone into development with the
attitude that we'd do things right or not at all, we would have fallen
far shorter of perfection than we did. How close we did get is something
all of you can decide for yourselves. All I know is we're going to get
closer next time.
Publisher: Eidos Interactive
Number of Full-Time Developers: Approx. 20: 1 of me, 3
programmers, 6 designers, 7 artists, 1 writer, 1 associate producer,
Number of Contractors: Approx. 6: 2 writers, 4 testers
Time: 6 months of preproduction and 28 months of production
Release Date: June 23, 2000
Target Platform: Windows 95/98/NT/2000 plus third-party
Macintosh and Linux ports
Critical Development Hardware: Ranged from dual Pentium
Pro 200s with 8GB hard drives, to Athlon 800s with 9GB fast SCSI,
and everything in between. More than 100 video cards were cycled
through during development.
Software Used: Visual Studio, Lightwave, Lotus Notes
Notable Technologies: Unreal engine and associated tools
such as UnrealEd and ConEdit (our proprietary conversation editor)