"Emergent Puzzles to beat the Grind"
ben...@gmail.com
situations, and reward player alertness and creativity. I would love
to get some discussion going on this -- there's a comment section on
the page as well.
http://benhem.com/games/wayfarer/puzzles.htm
In Wayfarer news, you can see some silly eye candy demos (and source)
from the new "ice storm" and "fireball" wands at:
http://benhem.com/fireball and http://benhem.com/icestorm
--Ben
The Aging Minotaur
> http://benhem.com/games/wayfarer/puzzles.htm
Thanks for an interesting read. From my own perspective, I try to
develop my game more as a board game than a "simulationist" RL. So
getting, say, a realistic behaviour for monsters is less important to
me than making monsters that can be combined in the kind of emergent
puzzles you talk about. It's a game where the player is weak (and
can't expect to get perverse stats and health), so hacking down
everything is really not an option. Instead I try to balance by
implementing some features that can be used to your advantage in some
situations, or must be overcome in other ways. For instance, the game
features a kind of giant moth ("night swarmer") that is never hostile
to anything, instead moving completely at random and performing an
attack on whatever comes in its way. Crossing a cave with these beasts
is a trying task, but you can also lead hostile monsters into such a
flock and retreat to a safe distance. There are also groups of
adventurers, who are neutral to the pc, but will attack certain
monster types. OTOH, these bands tend to clog passages the pc needs to
get through, or use area attacks that can harm the player if s/he gets
in the way (or even their comrades, causing in-fighting, which is
amusing in itself, and tends to leave some fair equipment lying around
when the dust has settled). In a game where the player can only move
orthogonally, there are a few monsters that move along the diagonals.
Snakes became an interesting addition: they're only allowed to move
diagonally, wants to attack everything, and retreats from enemies
standing orthogonally adjacent to them. This makes them difficult to
hit, but usually easy to avoid. If you are being chased by some
brutish melee monster, however, a snake pit can turn out to be either
a useful dungeon feature or a deadly trap.
Of course, this means that developing the features is a slow process,
since I want every new beastie to have some new schtick. If ogres and
brigands are basically the same, in gameplay terms, I can't see why
you would need both. At the moment, most of the "emergent puzzles" in
my game involve making one monster attack another, or step on a trap,
but I'm working out some ideas for more interesting behaviour, ie.
predefined or randomly generated cultural traits. Some intelligent
species will have taboos and customs that can be gleaned through
rumors/books or experimentation, or have other actions they like to
perform: digging tunnels or changing the map in other ways, picking up
loot or accepting gifts, turning hostile on beings who pick up stuff
in their home, setting traps, leaving home to explore the level, etc.
Hit points are a commidity (no regeneration, and a default pc starts
out with 8 hp), so it's no good to lose even a single point, which is
to be expected if you want to beat down something like an overworld
citizen. This means that plundering the starting village becomes a
mini-game in itself: people tend to stay in their homes, and don't
like stealing; if a forest is generated, it may contain hollow trees
with treasure, but also ferocious animals.
As always (ranting),
Minotauros
PS.
> In Wayfarer news, you can see some silly eye candy demos (and source)
> from the new "ice storm" and "fireball" wands at:http://benhem.com/fireball and http://benhem.com/icestorm
My browser tells me that your signature is broken, and goes on to
display a black square. (Running Debian stable, Epiphany, and Sun's
plugin that comes bundled in Debian's non-free repositories)
Kusigrosz
> On May 30, 3:41�am, benh...@gmail.com wrote:
>> http://benhem.com/games/wayfarer/puzzles.htm
>> Introduce greater risks, more give-and-take. A powerful spell has
>> a chance to render the caster confused and vulnerable for a few
>> rounds, or a wand has a chance to backfire.
Hmm, I'm not sure if it leads to the kind of gameplay that I like.
In a long game with permadeath, it might end up being used only
in desperate situations (1 turn from death anyway), or when the
worst known consequence is easy to absorb.
>> Allow monsters to interact in interesting ways. The classic examples:
>> the floating eye that can paralyze you long enough for something else to
>> nab you, or the shrieking mushroom that brings the whole neighborhood
>> running. But this could include cooperative monsters that lure the
>> player into ambushes, feign death, steal items and use them against the
>> player, pin the player's arms or legs while another monster attacks,
>> combine magic to cast more powerful spells, block the player's escape,
>> etc.
What about mon-mon interactions that don't directly affect the PC, but
can be observed and learned from?
> Thanks for an interesting read. From my own perspective, I try to
> develop my game more as a board game than a "simulationist" RL. So
> getting, say, a realistic behaviour for monsters is less important to
> me than making monsters that can be combined in the kind of emergent
> puzzles you talk about.
[snip]
> At the moment, most of the "emergent puzzles" in
> my game involve making one monster attack another, or step on a trap,
> but I'm working out some ideas for more interesting behaviour, ie.
> predefined or randomly generated cultural traits.
I think monster-monster interaction is definitely a good idea.
It is not easy to get right though; on a large map it could leave parts
of the level in a 'strange' state. OTOH, if properly balanced, this
could be part of the fun - maybe even worth running the level without
the PC for some turns immediately after it is created, to allow some
interesting situations to develop.
--
Kusi...@AUtorun.itvk.pl To send mail, remove 'AU' from the address
You see here a scroll labeled "Q-e0fHUoKD8"
ben...@gmail.com
Re: non-cooperation: definitely. It could have an ecosystem angle:
I've been planning to include predation and protection relationships
(glormbeasts protect their glormlings, who are the favorite prey of
the much stronger warvelmaws), etc. (Slime molds: the essential
bottom of the food chain?) Running the level to let the "ecosystem"
develop for a while (maybe before closing all the doors?) sounds
awesome. Can't wait to get a system developed enough to try it.
I am already imagining scenarios where a monster triggers a trap that
happens to summon a rival species. Good times ahead!
Re: diagonal snakes -- sounds like it would go a long way in producing
a puzzle or even board-game effect. I've wondered about including
"chess moves," even the "knight's leap" -- maybe a rather slow monster
that presents little threat until it can hit the player from the L-
point? I haven't tried ChessRogue, but it sounds intriguing...
http://chessrogue.sourceforge.net/
Re: risks too high -- maybe high skills with wands or spellcasting
could negate the risks I posited...or maybe it would just have to be
fine-tuned to ensure that they only produced minor setbacks. I agree
that a total wild card is not tempting in a permadeath world.
Minotauros, how might the NPC customs / taboos manifest in behavior?
An aside on emergent NPC customs: due to a "bug" in the way I average
walkable grids' scent values, items that block movement (like
braziers, fountains, and chests" tend to attract my monsters. So if
the player uses infravision they can often see groups of monsters
marching in odd maypole dances around said objects. Freaked me right
out when I first saw it.
Thanks for the input! I've put up a link to this thread from the
article and tried to paraphrase what I saw as key points.
Oh, and about my applets: the offending certificate is for Sun's own
JOGL bindings, and they really need to get on the ball. It is a
harmless .jar, so I recommend trusting it, but I can understand if
you're leery. I tried exporting some non-JOGL versions, but they're
really slow. -- Ben
The Aging Minotaur
> I've wondered about including
> "chess moves," even the "knight's leap" -- maybe a rather slow monster
> that presents little threat until it can hit the player from the L-
> point? I haven't tried ChessRogue, but it sounds intriguing...
I must also admit to never having played ChessRogue, but I agree the
idea is fascinating. A possible monster might be actual "chess
pieces", as featured in Lewis Carroll's "Through the Looking Glass,
and What Alice Found There" (from whence came the Jabberwock of the
original Rogue).
For my own part, I'm considering a critter that moves a bit like the
pawn, ie. basically always in one direction. If you drop a dozen of
them on the edge of an overworld map and start them moving, it might
make for an interesting swarm of locusts, or something like that.
> Minotauros, how might the NPC customs / taboos manifest in behavior?
This is still on the idea stage, although it's one of the next things
I'm starting to write. The infrastructure is basically there, but only
tested on much simpler data objects (eg. centaurs are generated
individually from a set of "horse parts" and "human parts", so you can
get the typical "human torso + 4 horse legs", but also stuff like
"horse head and front legs + 2 human legs", which moves a bit slower,
but has a more powerful attack than the classical centaur).
The idea now is to generate random cultures for two species at the
beginning of the game. One is loosely based on Swift's lilliputs, and
one on Breughel's dancing skeletons. The customs will be "molecular"
kits that can be patched onto the data objects used to create these
critters and their environs. A lot will be cosmetic, or have only
minor effects. Details pertaining to archetecture and inventory lists
is one example. In one game, a species might live in small, connected
rooms and favor books for treasure, in another they may prefer to
build houses in larger caves and keep mummies in crypts[1].
Other quirks will set or rearrange switches in the state machine of
species/individuals. Here you've got local and global taboos ("don't
eat in the temple", "don't harm <species X>"), paired up with
reprimands (dislike, attack, give penal quest, verbally abuse, etc.)
You can also get stuff that's almost purely for show (for instance a
quirk to make a species musical, adding instruments to their treasure
lists, and making them susceptible to passing the time by playing when
they're at ease).
The possibilies are many, of course. You can have herders (keeping
flocks of animals, either behind fences or as pets in the wild),
hunters (wanting to kill and collect the corpse of certain species),
farmers (have gardens), etc. I'm aiming for a nonsensical setting, so
the plan is to combine cultural quirks pretty much at random and hope
for interesting effects. The game play value I hope to gain, is to
make exploration of the game world fresh from game to game. In lots of
RLs, you pretty soon learn how the setting is put together, eg. that
you have "kobolds" and "kobold archers". Instead, I let some of the
moving parts vary from game to game. For instance, a species might
always contain a subset of <archers>, but they may be rock throwers in
one game and bombadiers in the next.
Again, the problem is the painstaking process of writing and testing
all the data structures. Even if a garden can be defined in less than
ten lines of a data file (if you already have data in place to
generate flower beds, fountains, etc.), the hard part is coming up
with good and plausible ideas, and writing the pieces of hopefully
nice prose to print when the player interacts with a certain feature
or critter.
> (Slime molds: the essential bottom of the food chain?)
Of course!
> [Monsters] marching in odd maypole dances around said objects.
How wonderful!
As always,
Minotauros
[1] I particularily want the mummies as a possible/probable quirk for
the lilliputian species, since I could then add tiny mummies as edible
artifacts. I'm thinking of Futurama, of course: "This is for you, Fry.
Zevulon the Great. He's teriyaki style."
ben...@gmail.com
<spam.minotaur.s...@googlemail.com> wrote:
> The idea now is to generate random cultures for two species at the
> beginning of the game. One is loosely based on Swift's lilliputs, and
> one on Breughel's dancing skeletons. The customs will be "molecular"
> kits that can be patched onto the data objects used to create these
> critters and their environs.
Sounds great. Do you have anything playable yet, or a progress page?
I'd check in on that for sure. --Ben
The Aging Minotaur
> On May 30, 6:47 pm, The Aging Minotaur
>
> <spam.minotaur.s...@googlemail.com> wrote:
>
> Sounds great. Do you have anything playable yet, or a progress page?
> I'd check in on that for sure. --Ben
The game is still lacking in so many ways! And development is going
very slow, squeezed in between my job and my very young son, that I
like to spend time with. But, hey, since you asked, I put out a
snapshot of the code, as is, no questions asked. I won't start listing
bugs and rough edges. Suffice to say that it's pre-alpha. There's no
win condition, but you can crawl around in a dungeon and whack
baddies. The game needs python and pygame to run. Get the sources
here: http://tekst.fix.no/hagerup/squirm.zip And a random screen
shot:
http://img413.imageshack.us/img413/7326/screenshotsquirm4.png
I'll leave the source lying around for a few weeks, for those who
might be interested. All and any comments are extremely welcome, of
course!
As always,
Minotauros