🚀 go-pugleaf

In comp.sys.ibm.pc.games.strategic Brad Wardell <bwar...@stardock.com> wrote:
> Back a couple of years ago there was a great deal of discussion on Usenet 
> with regards to the GalCiv I AI.  Most people thought it was pretty good but 
> others felt the military tactics were bad.

"Bad" is a bit stronger than I'd put it. It was still better than the
vast majority of strategy games out there, but the vast majority of
strategy games has seriously brain dead AI. I'm glad GalCiv1 at least
made a serious attempt at decent AI, and it was all the more fun
because of it. I'm still playing it.

>  That while the AI provided a 
> challenge, it mostly was able to do so because it was more effective at 
> managing its economy/industry than the typical human and thus could 
> essentially "win" through brute force.

It's not even that it's more effective at running its economy, but
that it gets a big economic bonus. Given time, I will catch up
economically even without wars. Mostly because of the wonders and
trade goods ofcourse, because the AI rarely manages to grab more
than one or two of those.

> In general, my view was that the AI worked pretty well but militarily was 
> disappointed. The main issue was its lack of coordination in projecting 
> military power into a given quadrant.  The result was hat looked like the 
> age-old "Death train" (That's where units uselessly throw themselves at a 
> wall dying with no purpose). There simply wasn't enough analysis on putting 
> units together and using them effectivley.

This is a problem with absolutely every complex strategy game out there.
Their attacks lack focus, which means that in the best cases, they're
decent at defense, but little else. The only exception I can think of
is PG3SE; there, the AI sometimes seems to be mounting a decent counter
offensive, although I think it's usually either prepared by the human
creating the scenario, or the result of some clever scripting.

I think an important part of fixing this is to:
A) make the AI realise that Death Knights without transports aren't
going to win the war, and
B) transports without support from superior firepower are going to
lose the war.

For good combined arms, you need:
1) Some tough ships of the line, and enough of them to fight off the
enemy,
2) Transports to take planets, and
3) Fast raiders who use their speed to jump from behind the protection
of the big ships and wreak some havoc.

The AI should be trying to group his ships in groups with each of these
types present in sufficient numbers (although type 3 isn't really
vital). They don't have to travel to the front in these groups, but
once they get to the enemy planets, something of the appropriate type
has to be nearby, or he's just wasting time.

Also, the AI should realise that his speed 3 Death Knights aren't
going to catch a speed 9 Starfighter, and that adding more Death
Knights isn't going to increase the speed. In my current game, a
handful of Starfighters managed to tie up a gigantic fleet of Arcean
Death Knights while I still had to build my Dreadnoughts with which
I wanted to conquer all of his planets. I think I had a single
Battleship. He caught me with my pants down, and wasted his time
and his fleet on chasing down Starfighters he could never hope to
catch.

> A lot -- A LOT -- of work has been going into the AI of Galactic 
> Civilizations II in order to not just make it "challenging" but to make it 
> play intelligently.  There were significant chunks of the GalCiv I AI that 
> were scripted.  That is, the AI is programmed to do certain things due to 
> play-testing over and over.  This time around, we're trying our best to make 
> the AI build strategies based on analysis and not rely on scripts.

This is really good news. Can't wait to see it. That should help fix
some of the problems I mentioned above.

> The improvements in the AI have led to a lot of changes, probably the 
> biggest one being the elimination of the infamous "The AI knows where the 
> good planets are" issue that GalCiv from the OS/2 days on up to the 2003 
> release of GalCiv.  GalCiv II's AI has to build scouts and look for good 
> planets just like everyone else.

That would be really great. But is the AI smart enough to do that? I
was one of the people who criticised your claim that the game had a
"level palying field", when in fact the AI didn't need scouts at all,
but at the same time, I do understand that smart scouting isn't easy.
And what is the AI going to do when I find a Starfighter early and
start killing off his scouts?

Although I'd love a truly level playing field, an able opponent is
still more important.

> The idea being that when the game's 
> released, the difficulty levels will determine how smartly the AI plays 
> rather than age old "Here's some money (or less of)" way that strategy games 
> have used as a crutch.

Another decision I love. I mean, crutches are better than limping along,
but, to stretch this metaphor a bit further, I would love to see an AI
that can actually stand on its own legs. I've never seen anything like
that.

And again, don't be afraid to make use of the things the AI is good at:
working the numbers, fast processing speed, etc. Make it mean and
nasty and rip my guts out at the toughest levels. It's gonna need it.


mcv.
-- 
"Serenity is a very personal work with political resonance and a
heartfelt message about the human condition and stuff blowing up.
'Cause let's face it, nobody cares about that 'human condition'
stuff... in fact if you notice it, try to keep it to yourself."
                                        -- Joss Whedon on his new film