| |
Q: One of the tactical elements present here that we haven't
seen in a simulation like this before is rules of engagement.
How do these work?
Troy: ROE is basically a set of rules for the AI to follow when making
decisions. We wanted to deal with the case of the player who
wants to both command a division or squadron, but also work the
systems on his own ship. To do this, we needed a mechanism to
allow the player to communicate his intentions to the AI units
under his command without requiring him to micromanage their
actions.
The great flaw with most real-time strategy games is, for me,
the need for the player to divide his attention between several
different processes (usually resource management and combat.)
This prevents the game from having a lot of depth in any process
and makes play-balancing the whole a real nightmare. We
emphatically didn't want this to happen with DC, so we came up
with the ROE concept.
In a nutshell, ROE allows the player to tell the AI what types
of targets it should prefer, what its operational objectives are
and how much initiative it has in pursuing those objectives. The
player can then point his ships in a given direction (or keep
them in formation with his flagship) and expect the AI to behave
within the guidelines he has set.
Q: Do morale and fatigue factor in the efficiency of my
division?
Troy: This is an issue that's come up, but hasn't really been
addressed in detail yet. Our current thinking is that it should
be based on a system where the player controls the alert status
of his ships. A ship at general quarters will naturally fatigue
over time, but will be able to respond more quickly to threats.
Click to continue
|
|
Q: One of the attractive things about this simulation is that
destroyers perform so many kinds of duty. I saw you make
contact with a submarine and then go to the depth charge
station and configure a spread and then let them go. Tell us
about this weapon system from the technical perspective. What
is modeled, how much control does the player have, and what
aspects can be automated?
Troy: We've gone to a great deal of effort to simulate all of the
weapons systems in detail, including depth charges, torpedoes,
AA machine guns and, of course, the main battery.
With depth charges, we wanted to give the player the ability to
configure a pattern with maximum flexibility. To that end, we've
provided controls to set the number of charges in the pattern,
the interval at which they are dropped and the depth at which
they detonate.
In addition, the number of charges and drop
interval may be set independently for the racks and the throwers
(throwers, or "k-guns" are designed to hurl depth charges to
port and starboard, allowing the ship to drop a much wider
spread of depth charges.) Our goal was to provide an interface to something that was normally handled by a lot of men working in concert according to
their training. This challenge arose repeatedly in the design of
DC--particularly with the weapons systems which were normally
crewed by teams of men.
For example, a single AA mount has a
pointer, a trainer, a mount captain and several ammunition
handlers. Another example is the main battery director, which is
crewed by about six men, with a whole second group of men
manning the computer and the stable element. Naturally, we were
forced to abstract a lot of the functions because a single human
couldn't possibly control all of these aspects in a completely
realistic model. We think we've done a pretty good job of
suggesting realistic functionality without overburdening
players.
Go to Part III.
|
|
