_Command & Conquer 3_ paradox
Simon Slavin
Itmight apply to other strategic games.
It occurs in skirmish games -- player vs. computer. _C&C3_ allows you
to set up up to eight 'teams' in these games. The human plays one,
and the computer plays the rest. It is very flexible in picking
sides: you can, for instance, set up that the human and two other
computer teams team up against three computer teams, and the other two
computer sides firm a third side. Teams on the same side may not
intelligently co-operate butthey won't attack each-other.
In circumstances like this there is normally territorial advantage to
clumping all the teams on your side into a group so that the ones on
the exposed sides of the group get attacked and the ones in the middle
canconcentrate on producing units and less on defence.
Except that in _C&C3_ this doesn't work. Because the computer teams
on the human's side are stupid. If you produce units they expect you
to use those units to protect them. So what you end up doing is
spreading your units over a bigger area. Naturally this thins out
your troops to the point where they can't protect anything. At this
point the enemy builds up a big army, and steamrollers into one of the
teams on your side. Once they have destroyed this team, you are left
with a huge enemy force on your borders, with no units left to defend
you. This makes you an easyvictim.
On the other hand you could place yourself on the outside of the
group, concentrate on building defences and not units, and leave the
computer teams on your side to produce units. That doesn't work
either. The computer units on your side do not move to your
territory. They remain in their own territory until they are attacked
which, of course, you won'tallow to happen until you're destroyed.
I've played about 100 skirmish games now and it always works out the
same way: if you put other teams on your side, they get killed and the
enemy has a big army on your border. On the other side if you make
the other teams an independent side, it fights the other computer
sides more efficiently, and you can wait until one of them dies then
use the troopsyou've built up to pick off the winner.
I wonder if there's a word for this strategic paradox.
--
I'm trying a new usenet client for Mac, Nemo OS X.
You can download it at http://www.malcom-mac.com/nemo
Thomas A. Russ
> I have discovered a stupid contra-intuitive characteristic of _C&C3_.
> Itmight apply to other strategic games.
>
> ... [snip]
>
> I wonder if there's a word for this strategic paradox.
Well, I don't know if there's an existing term, but how about:
"With friends like these, who needs enemies?"
or in abbreviated form:
"With friends like these..."
--
Thomas A. Russ, USC/Information Sciences Institute
Joseph Nebus
>I have discovered a stupid contra-intuitive characteristic of _C&C3_.
>Itmight apply to other strategic games.
[ .. ]
>Except that in _C&C3_ this doesn't work. Because the computer teams
>on the human's side are stupid. If you produce units they expect you
>to use those units to protect them. So what you end up doing is
>spreading your units over a bigger area. Naturally this thins out
>your troops to the point where they can't protect anything.
I haven't played this particular game but have seen similar kinds
of developments on other games and it's agonizing every time. My hunch
is that if it isn't the complete implosion of artificial intelligences
then maybe it's some kind of moderately dopey idea of preserving game
balance given that the human can get to exploiting game logics. That it
goes horribly wrong anyway is ... I suppose normal overcompensation.
My frustration is in grand stategy games when I can't get my
allies to press the enemy on a second front while I work at the first,
or at least to fill in gaps in my front. (Admittedly, I have a bit of
a tendency to outrun myself so they may not have the chance to keep up.)
I've got a dim memory of an old grand-strategy Europe war game
in which at some point the Mongol Hordes would come from off-map to the
east, and their strength was to be roughly equal to whatever the army
of the easternmost-mapped nation was. So it was to one's interest to
have a pathetic nobody off on the east, whatever the other strategic
sensibilities of the game would be.
>I wonder if there's a word for this strategic paradox.
I don't know one, specifically. I remember ... was it in a
Liddel-Hart book? ... the claim that the weaker ally of the pair gets
disproportionate strength in the relationship because, after all, they
can collapse without dishonoring the alliance. Maybe that gets us
towards a term that works.
--
Joseph Nebus
------------------------------------------------------------------------------
Thomas A. Russ
> I've got a dim memory of an old grand-strategy Europe war game
> in which at some point the Mongol Hordes would come from off-map to the
> east, and their strength was to be roughly equal to whatever the army
> of the easternmost-mapped nation was. So it was to one's interest to
> have a pathetic nobody off on the east, whatever the other strategic
> sensibilities of the game would be.
Not a computer game, but decades ago I played a solitaire board wargame,
SPI's "The Fall of Rome", which had one scenario when the Huns would
come storming in from the east. And once, due to an unlikely series of
die rolls, it turned out that the local barbarians ended up stronger
than the invading Huns and destroyed them upon their arrival.
Of course, then the emboldened local barbarians turned their attention
toward the Roman Empire. So it all ended up as a big good news, bad
news joke.
Simon Slavin
Slavin<sla...@hearsay.demon.co.uk> wrote:
> I've played about 100 skirmish games now and it always works out
> the same way: if you put other teams on your side, they get killed
> and the enemy has a big army on your border. On the other side if you
> make the other teams an independent side, it fights the other
> computer sides more efficiently, and you can wait until one of them
> dies then use the troopsyou've built up to pick off the winner.
I worked it out.
The AI has a set of conditions for 'attack' and a set of conditions
for 'sacrifice various resources to defend aggressively'. When you,
the human, team up with an AI side, your natural impulse to rush to
the side of your ally in defence prevent it from losing badly enough
to switch to aggressive defence mode. So it never does. On the other
hand, the AI will eventually switch to attack mode ... long after I
think it should. If I play a defensive game for long enough --
securing my territory but never attacking anything -- and if my ally
lasts long enough it will eventually build upa lot of troops and then
eventually switch into attack mode.
For some reason, it does this less efficiently than my enemies: my
enemies will attack me quite viciously long before my allies will
attack my enemies. This may be a bug, or it may be an intentional
move by the designers who want to make sure that the human player
plays a big part in the battle rather than being fenced in a corner
while the computer AIsfight it out.
Anyway, it's still annoying, but at least I understand it better now.
Simon.
--
I'm using an evaluation license of nemo since 56 days.
You should really try it!
http://www.malcom-mac.com/nemo
Michael Emrys
> This may be a bug, or it may be an intentional move by the designers
> who want to make sure that the human player plays a big part in the
> battle rather than being fenced in a corner while the computer AIsfight
> it out.
It might also be an attempt by the designers to balance the game, since
a human is usually a stronger player than the AI, a human player plus
an equivalent AI player would be a stronger team than a side with only
an AI player.
Michael
Joseph Nebus
Yeah, I think I can see it making sense by either interpretation.
It would be tempting particularly to ally with a stronger AI and then get
into a grand war trusting that the ally will do most of the fighting, and
that has some undesirable effects on gameplay.
--
Joseph Nebus
------------------------------------------------------------------------------
Michael Emrys
> Michael Emrys <em...@olypen.com> writes:
>
>> On 2009-02-22 20:09:44 -0800, Simon Slavin <sla...@hearsay.demon.co.uk> said:
>
>>> This may be a bug, or it may be an intentional move by the designers
>>> who want to make sure that the human player plays a big part in the
>>> battle rather than being fenced in a corner while the computer AIsfight
>>> it out.
>
>> It might also be an attempt by the designers to balance the game, since
>> a human is usually a stronger player than the AI, a human player plus
>> an equivalent AI player would be a stronger team than a side with only
>> an AI player.
>
> Yeah, I think I can see it making sense by either interpretation.
> It would be tempting particularly to ally with a stronger AI and then get
> into a grand war trusting that the ally will do most of the fighting, and
> that has some undesirable effects on gameplay.
It wouldn't even be necessary for the ally to have stronger AI. Just a
human player plus an allied AI that's equal to the enemy's could be
unbeatable. That's why I wrote 'equivalent'.
Michael