Re: [glob2-devel] Discussion moved from the forum

[Andrew Sayers]

For those of you just joining us, we've been discussing buildings and
strategies around them on the forum at:
           http://epfl.ysagoon.com/phpbb/viewtopic.php?t=378

It seemed more appropriate to put this on here, since we're getting into
quite deep issues of gameplay now.

> > <about fruiting>
> > 
> > > [instant Inn View is] Unfair. The primary reason for this is that
> > > it can basically force all opponents to take drastic action within
> > > 3 minutes or lose.
> > 
> > How is that a problem?  Any decent player will guard fruit trees in
> > case this happens.  Even if they had chosen not to use them (say, by
> > putting a forbidden area around them), all they have to do is make
> > the trees available and their globs will do the rest.
> 
> Ah, but that's the same in every game. "Any decent player will do
> this" is the same as "Every player should do this in every game". If
> there's a static action that should be done in every game, then that's
> a boring task. A boring task is often OK (i.e. building the first few
> Inns) but not only it is boring, it is also a very constant task (you
> don't choose where to build the trees) and the punishment is too great
> for not doing it.
> 
> Basically, either you don't guard trees, and you get screwed, or you
> do guard trees, and you don't. So you do guard trees. Everyone does in
> every game. It's just an extra botheration, not a strategic device.
> 

No, not at all.  If/when you happen across a fruit tree in a game (some
maps have fruit, some don't), you have to make a decision about how to
handle it.  If it's out of the way, you can choose to ignore it and
nothing much will happen.  If it's in your way, you can put a forbidden
area there and they'll ignore it.  If you want to pursue a fruit
strategy (or defend against one), you can start collecting.  If you want
to make a trap for other teams, you could put towers near it, etc.

> > <about seige towers, unless you can think of something else to call
> > them>
> > 
> > > Ignore them. They would not cause traffic problems, because the
> > > forbidden areas would not be traversed by anyone else.
> > 
> > What if you have one siege tower going towards a battle and another
> > coming back, and they both meet in a one-square-wide alleyway?  If
> > you have other globs walking past the alleyway, 18 warriors will be
> > completely hemmed in.
> 
> That doesn't happen because siege towers don't come back from battles.
> 

So what happens to a siege tower when it gets to a battle?  It
disappears?  It hangs around there?

> > Ah!  So how about saying that a war flag (or any building or area,
> > for that matter) which is over-subscribed should prefer better fed
> > warriors over hungrier ones?  Also, there is a general issue here
> > that the current system for getting hungry is too static - workers
> > that never stray more than a few paces from an inn should keep
> > working until they only have 1 or 2 percent food left, while
> > warriors should automatically eat before setting off on a long
> > journey.  There is a discussion somewhere in the suggestion forum
> > about having a movable inn, which would solve the problem too.
> > 
> > Now you mention it, I have found my warriors peeling off in long
> > battles.  Again, I'm tempted to say this is a good thing because it
> > makes it harder to wipe out an enemy in one fell swoop.  That's not
> > much fun for an attacker, but it means you as a defender might get a
> > second chance.
> 
> It doesn't only mean you get a second chance. It also means that the
> defender might destroy half the attacking army by localized
> superiority, and then march to their territory and decimate the rest
> (if towers are not involved.)

True.  But surely a poorly managed attack *should* be disastrous?
That's part of the game's strategy.

> In general, it seems that hungry units should act like this: 1st
> priority: Go back to the inn if it cannot get back to an inn without
> starting (or coming very close to starting) to starve.  2nd priority:
> If it can get back to an inn in a reasonable amount of time, and is
> busy, it continues its task.  3rd priority: Go to the inn.

That sounds like a sensible scheme to me.

> 
> Also, instead of the siege tower, it does make sense to have a mobile
> inn. Here's my take on the situation:
> 
> War Wagon building. It's like an mobile, armored Inn which also acts
> like a defense tower. It takes 8 wood and 2 stone to build, and stone
> to fire. It moves when there's a War Flag, towards the nearest War
> Flag. Only warriors use it as an inn, and it can't hold fruit. Its
> capacity is 4-6-9, its armor is 8-12-18, and its hitpoints are
> 200-350-500. It does not have extra stone slots, as the defense tower
> does, and it moves at half the speed of a level 1 warrior.

It sounds like your model of combat in Glob2 is that it should be quite
fast and adaptable - that you should be able to push your attack forward
and all around until your globs actually run out of steam.  Is that a
fair assessment?

I'm not sure whether I like that idea yet, but one thing is that model
tends to favour the attacker over the defender, and is likely to make
combat quite quick and decisive.  On the one hand, that's good because
you don't get the sort of ragged end to a game you see in many games
(including Glob2 right now), where one side is clearly going to win, but
it takes them half an hour or more to get there.  On the other hand, I
would expect it to be quite depressing when you're doing really well one
minute, then your base is suddenly overwhelmed and you're dead 5 minutes
later.

> > Personally, my strategy against the computer for this is to beat
> > them back until I can reach their resources, then set up inns and
> > hospitals in their territory, to mount a final attack from.
> 
> The AI is pretty weak. I generally do a tower-rush (getting a ton of
> level 2 towers around your settlement before the first enemy attack =
> ownage :D) and then win however's easiest. That's usually a wheat
> genocide (not applicable in the more current version) or converting
> the enemies.

        - Andrew