Command and Conquer (et al ) AI
Boiman Oren
I was truely disappointed by the level of AI in this game :
1. the sandbags problem : enemy units never attack them , so you can block
them and do very nasty things with those bags ( i.e. creating a chain
a putting a guard tower in the middle of the screen . . .)
2. It is easy to observe that no major attack is planned in real time :
major attacks always come from the same direction , even if you restart
the game. The real-time attacks are of scattered units , thereby not
being real challenge.
3. Tactics - they say its the ability to concentrate your forces in the
right place and the right time. According to this - Computer player
has NO tactics. Computer forces are scattered and not coordinated.
( e.g. U regullarly encounter launchers alone , not behind a battery
of tanks).
4. Offensive - The programmers simply forgot this part , or maybe they
simply chose to ignore it by concentrating on the defensive, cause
this part can be planned by programmers and not in real time.
It seems to me that a lot of hours were spent preparing this game.
The result is impressive , but someone there prefered to develop
the graphic side ( which is gr8 to my oppinion ) and forgot the
algorithmic side.
As a programmer , I wouldn't have stated that the game has any relation
to AI , judging from the poor result( i'm refering only the algorithmic side)
Maybe its the time and place to remind programers that using
a Best-first-search in your computer game doesn't merit it
with the so abused slogan ' AI inside ' ...
Things like this have brought AI bad reputation.
_
\\- -//
(`o-o')
-----------------------------------------------ooO-(_)-Ooo----------------------
Signed : Oren Boiman , Addr. : boy...@math.tau.ac.il , Phone number : 03-5719617
Home Page : http://www.math.tau.ac.il/~boyman
Smart Quot^H^H damn! ^H^H ^Q ^[ .... :w :q :wq :wq! ^d X ^?
exit X Q ^C^C^C!!! ^? ^?! :quitbye CtrlAltDel ~~q :~q logout
save/quit :!QUIT ^[zz ^[ZZZZZ^H man vi ^ ^L ^[c
^# ^E ^X ^I ^T ? help! helpquit ^D ^d man help ^C ^c help exit
?Quit ?q CtrlShftDel "Hey, what does this button d..."
--------------------------------------------------------------------------------
Chris Wheeler
>
> What AI ?!
> The result is impressive , but someone there prefered to develop
> the graphic side ( which is gr8 to my oppinion ) and forgot the
> algorithmic side.
>
You're right to doubt that this represents true AI.
However, if you consider this game in relation to Dune II, to which it
bears a huge amount of similarity, the only areas improved upon are the
graphics and the AI. Yup, the AI is an awful lot better. Name one
instance in Dune II when you had more than one armoured unit come at you
at once. The only units that banded together in that game were infantry.
They used the wait til you have five units, then attack, school of AS
(Arificial Stupidity).
There are basically no strategy games that I know of, where more than one
piece can be moved at once (unlike Chess, Draughts, Othello, Go, etc.),
in which a computer player has a chance, without cheating (Cheating here
is defined by such things as Civilisation's allowing computer players
vastly reduced cost of production, research, etc., Dune II & C&C's
computer player having vastly better position.
Hopefully there will be one day!
--
Chris Wheeler
Idiot
John Christian Lonningdal
>There are basically no strategy games that I know of, where more than one
>piece can be moved at once (unlike Chess, Draughts, Othello, Go, etc.),
>in which a computer player has a chance, without cheating (Cheating here
>is defined by such things as Civilisation's allowing computer players
>vastly reduced cost of production, research, etc., Dune II & C&C's
>computer player having vastly better position.
One problem is that in realtime strategy games you end up with far too
little CPU time to implement a good AI. Although I thought C&C would
have a better AI, it has the same problems all the real time strategy
games have. The sandbag problem is obviously something the developers
never thought of, or I'm quite sure they would have changed the code
to take care of this. Actually I think the problem is that the market
for C&C was so hot and getting angry for all the delays that they just
didn't care about these problems. They have to draw the line
somewhere. If you want to think about every situation in a game you
end up spending too many years of development. The AI suffered from
this unfortunatly...
Regards,
John
------------------------------------------------------------------------------
John Christian Lonningdal - Web: http://www.lis.pitt.edu/~john
jo...@lis.pitt.edu - (412) 521-9386 - 6611 Wilkins Ave, Pittsburgh, PA 15217
------------------------------------------------------------------------------
Get JOBE 1.5 a sprite editor @ http://www.lis.pitt.edu/~john/jobe/jobe.htm
Try out FUSECUTTER! @ http://www.lis.pitt.edu/~john/fusecut/fusecut.htm
Steven Woodcock
: Actually I think the problem is that the market
: for C&C was so hot and getting angry for all the delays that they just
: didn't care about these problems. They have to draw the line
: somewhere. If you want to think about every situation in a game you
: end up spending too many years of development. The AI suffered from
: this unfortunatly...
Allow me to second John on this one (and I know Eric is out there
someplace likely agreeing with this too). Getting the game to market
is a MAJOR factor in AI design and just can't be overlooked. In
some ways it's WORSE than the problem of insufficient CPU cycles;
if my code is inefficient I can do things to speed it up, but a
deadline is immovable....
Steve
+=============================================================================+
| _ |
| Steven Woodcock _____C .._. |
| Senior Software Engineer, Gameware ____/ \___/ |
| Lockheed Martin Information Systems Group <____/\_---\_\ "Ferretman" |
| Phone: 719-597-5413 |
| E-mail: wood...@escmail.orl.mmc.com (Work), swoo...@cris.com (Home) |
| Web: http://www.cris.com/~swoodcoc/wyrdhaven.html (Top level page) |
| http://www.cris.com/~swoodcoc/ai.html (Game AI page) |
| Disclaimer: My opinions in NO way reflect the opinions of |
| the Lockheed Martin Information Systems Group |
+=============================================================================+