premiere. Beyond the licensing advantages, there
was an additional consideration driving us toward
completing the project as originally scheduled: both
the programming and art teams had been staffed
up to take on the larger-than-expected workload of
producing a modern “next-gen” title. The result was
that the project was already over budget, and even
if it arrived on schedule, it was set to be the most
expensive game Play Mechanix had ever produced.
Couple that with the global economic downturn
that started in 2008, and there was considerable
financial pressure to complete the project as quickly
as possible.
Still, for the team at Play Mechanix, the decision
was an easy one. The goal of meeting the movie’s
release was abandoned, and the entire schedule was
extended by six months. The rationale was simple:
without a great game, having a great license and great
marketing wouldn’t mean anything. We also knew
from years of industry experience that, although
the pain of missing the movie’s release and cost of
another six months development could be overcome,
the stink of releasing a turd would linger forever.
In the end, the decision proved more than
justified. This was true even though the game ended
up taking a full 22 months to complete and cost
over $4 million to develop, more than double what
we’d originally planned to spend. The extra time
was used to redefine and re-tune our gameplay. We
quickened the pace, upped the cinematic elements,
and added more variety to the enemy types and
dispatch scenarios. We also took time to overhaul
the game’s visuals, bringing everything—lighting,
character models, environments, particles, and so
forth—up to a higher standard. Finally, the extra
time allowed the new gun and cabinet designs to
be completed.
If we hadn’t taken the extra time to get it right,
we’d have been releasing a mediocre title in a weak
economic climate. Instead we were able to release a
title so strong that it re-ignited the market for arcade
action gun games. It also opened up overseas sales
opportunities for us in markets like Japan, China, and
Russia that we hadn’t even thought possible before.
commands include things like "Travel along this path,"
"Begin attack cycle," and "Wait for trigger." Only simple
flow control (branching and looping) is implemented,
and no generic variable interface is provided. The goal
is for the game objects' logic and AI to be written in
C/C++, while the scripting language provides the
gameplay and dispatch patterns for those objects to
execute. Again, the motivation for this was to keep
the dispatch instruction set as simple as possible so
that even someone with no programming background
could use it. Keeping with this thought, the trigger-and-event system is similarly basic, providing little
more than a method for scripts to set and respond
to named global triggers. For example, the script
controlling one T800 actor might wait for the “T800_ 1_
Attack” trigger to begin attacking the player and then
signal another T800 to begin attacking by setting the
“T800_ 2_Attack” trigger when it’s done. Finally, the GUI
front end provides a point-and-click way to author
scripts and assign them to game objects. The GUI
runs in-engine, which allows scripts and scenes to be
quickly tested and iterated upon.
another. At Play Mechanix, our rule of thumb is that
you have about 10 to 20 seconds to grab a player’s
attention and start showing them a good time. After
that, the eyes start to glaze over and the player is
lost to you. They might play their dollar out, but
they’ll never put another coin in again.
This is a challenge we ran smack into when
developing the gameplay for TERMINATOR. From the
outset, we knew we wanted the game to follow in
the grand “red bullet” tradition of arcade action gun
games past, such as TIME CRISIS and ACTION HERO.
In this style of game, the screen is full of bad guys
to shoot, but only the specially highlighted enemy
shoots the magic red bullet which actually hurts the
player. Everyone else fires white bullets, which shoot
harmlessly past the player. Core gameplay consists
of picking the highlighted enemy out of the crowd and
shooting him before he can fire his red bullet.
On top of this, we had the idea that the
Terminator robot in the game should feel like the
same unstoppable killing machine people are
familiar with from the movie franchise. Instead of
4 GUI SCRIPTING TOOL.
/// Going into the project we knew we’d need a new
scripting tool. All of our existing tools had been
created ad-hoc in spare programmer cycles, and were
either too narrowly focused or too onerous to use
(or both) to be useful for TERMINATOR. We considered
using a high-level scripting language such as LUA
or Python, but we didn’t want to limit the scripting
process to only programmers. Instead, we wanted
a tool that would set the bar for creating dispatch as
low as possible, so that anyone at the company could
contribute to the creative process, even those with
little or no programming background.
To that end, we developed an in-house system
which combines a basic scripting language and a
simple trigger-and-event system with a GUI front
end. The scripting language is easily extensible,
but deliberately somewhat simplistic. Typical
Ultimately, this combination proved very potent.
The dispatch team quickly learned to combine the
scripting language's simplistic commands and
triggers to produce much more complex results.
Moreover, by keeping the scripting tool simple and
easy to use we opened up the design process,
allowing anyone at the company to contribute
dispatch ideas and scenarios. The net result was the
dynamic, layered scenes and varied dispatch that
give TERMINATOR its cinematic and visceral punch.
5 TUNING THE “RED BULLET.”
/// One of the toughest parts of arcade game design
is that you have an excruciatingly short time in
which to hook a player. Buyer’s remorse might keep
a player spelunking away at a $60 home title for
hours to get to the good stuff, but a player who has
only invested $1 into an arcade title demands some
serious immediate gratification before they’ll put in
scoring one-hit kills, we figured the player could
have fun slowly shooting the Terminator apart bit
by bit, first shooting off an arm, then a leg, then
maybe a section of his chest, and so on.
WWW.GDMAG.COM 23
