[MUD-Dev] RTS aspects in MUDs (LONG)

John Robert Arras johna at wam.umd.edu
Sat Dec 1 15:17:13 New Zealand Daylight Time 2001

On Fri, 30 Nov 2001 21:30:17 -0800 (PST) lhulbert at hotmail.com wrote:
> On Fri, 30 Nov 2001 15:53:13 -0800 (PST)
> Paul Schwanz <paul.schwanz at east.sun.com> wrote:

>> I've often said that MMOPRGs suffer particularly in the area of
>> goal management.  Since this is an area where an RTS seems to
>> shine, it seems natural to me to try to bring the two concepts
>> together into one virtual world.
> So how do we cross-breed RTS and MUD (and MMORPG ;))?  MURTOSG??


> Let's look at some pro's and cons.

<SNIP Pro's which I agree with.>

>   Cons:

>      Pathfinding. 
Possible with bfs and some modifications.

>      AI.  AI in RTS is also non-trivial.

I agree. Start simple and build up the AI one piece at a time.  I
never said that it would be easy.

>      Updates.  Making changes to a world without breaking a
>      running AI and it's current strategy will require a careful
>      balancing act.

I don't see this as a problem. The AI observes the world and tries
to compensate when things go against what its goals are.

>      Unit Building.  This RTS staple just doens't fit well with
>      standard MUD functionality.

On some level yes. But, you don't have to use the exact analogue of
an RTS. You just need units created in certain places based on
certain needs.

> The biggest hurdle I foresee in RTS/MUD programming is dealing
> with people. RTS AI programmers are used to their units always
> following orders.  In a MUD, it simply wont happen.  If you give
> an order to 10 soldiers, and 5 of them are PCs, then only 5 will
> do what you tell them to.  The other 5 might do what you tell
> them, might ignore you completely, or might try to sabotage the
> whole plan, since they really want the other team to win.  The
> only two solutions I can see are either A: some sort of
> loyalty/effectiveness tracking so that the AIs know which players
> are most likely to listen to them, or B) giving out the same
> orders over and over until someone listens.  And B) sounds way too
> much like the current system to me.

This is true, but it's not what I'm aiming for. I agree that a
straight RTS AI probably won't work. I also have little faith in
games where the hierarchies are led by players.

  1. A few kingdoms give general direction to societies and keep
  track of how they fare relative to other kingdoms.  They update

  2. Many local populations (societies) that generate units locally
  and create warriors and workers to gather raw materials and defend
  their location. They update every few minutes.

  3. Individual units have specified jobs and go places and do
  things and when they're done with a job, they check the society
  state to see what to do next. They can also work together to fight
  as a team locally. They get updated every few seconds.

In this setup, players don't get to order things around ike in an
RTS. They are a part of a team. When they do good stuff for the
team, they get points (like quest points), and they can alter things
by spending the points.

Some ideas might be: They can "hire" a guard for an hour by spending
10 points. The guard acts like a pet for that period of time.  (They
might even hire other players and reward them with points...)

Thus, they can hire out 100 guards for a couple of hours if they
have enough points to spend, and if the society is willing to part
with the warriors.

They can spend points to encourage the society to make more
soldiers.  That will up the percentages for making soldiers, but it
will take time for these things to come into play. The points spent
at this level are much more than the costs of working with
individual soldiers.

They can spend money and points to alter the kingdom AI. This should
be so expensive that it would take many players to alter the
direction the kingdom takes. For example, asking the kingdom to
declare war on someone, or to make peace with them, or to order the
local populations to beef up security or production.

However, the filtering of the information down to the lowest level
units would be relatively slow and the players would not ever have
direct control of the kingdoms or anything. It's possible that if
you collected a certain number of points you could place a few units
under your direct control for free, but I can't imagine player-run
kingdoms and things like that, since they're just too unstable.

So, the story moves along and the players can affect it locally to a
large degree, but as the size of the entity being manipulated goes
up, the less direct action they can take to alter it. Also, the more
they have to work with others to make big changes.

I think this idea fixes one of the huge problems with putting an RTS
into a MUD. What happens when the player logs out? In an RTS you are
constantly clicking stuff and moving things around, and if your AI
player were so darn good at this, why would you even need to be
around to play? By letting players affect an AI that's always on,
instead of having make the decisions themselves, you end up with
less of a disparity between when players are online and offline. How
could an RTS MUD with kingdoms led by players survive for a week if
the "King" and his "High Council" decide to go to Cancun for spring


MUD-Dev mailing list
MUD-Dev at kanga.nu

More information about the MUD-Dev mailing list