Crafting 3.0 Q&A

Thanks for all of your crafting feedback, folks! In the month since the last entry, I've gone back and refined the 3.0 design, poking at it with your perspectives in mind and with a few new ideas of my own - I am VERY excited about this system, and I'm hoping you all get as much joy out of the exploration as I have out of the design. I'm in the very early stages of coding right now.

I'd prefer to keep my crafting beans primarily unspilled, but I will very briefly answer some of your questions from comments and email since the last post:

1. PK mixed with crafting? Yes.

2. Items transferrable? Yes.

3. Stones/marbles as crafting currency? Not planned, but sun stones will allow for respecs.

4. Specializations? Yes.

5. Forced into crafting? No-ish. Crafted gear won't provide stat boosts as large as in 2.0. However, there needs to be a reason and incentive to craft. "Forced" is partly mindset. You can buy gear from crafters.

6. Obfuscated formulae? Not as obfuscated as 2.0. However, understand that a sense of discovery is important.

7.  Auction house? Not planned, trade channel for now. Someone could always set up a trading forum on compose. =)

8. Existing gear/Ruushi overlap. Yes, definitely some overlap with Ruushi gear (which I think is still a bit overly strong even post-nerf).

9. One size shouldn't fit all? Noted. The shape of the old 2.0 system is due in part to expediency (because of the underlying structure, adding class-specific gear would have massively inflated certain portions of the system). In the past month's revision of the new 3.0 system, I've worked in more distinction.

10. Materials/inventory/bank? This is a good point. I don't have a solid plan for this yet. The 3.0 system is rather different from 2.0, so I'd like to see how it plays out before devoting design/imp to it.

11. High barrier/investment and reliance on others? Noted and agreed. The trick is to leverage that aspect (which makes crafting valuable) without providing impossible barriers to entry for future crafters.

12. Crafting bags? Not planned at this time, easy to add if future behavior indicates this would be desirable.

The questions I haven't addressed above relate to aspects I'd like to keep in my pocket for now. =) There's nothing quite like the kick I get out of seeing people unravel the mystery of a fresh new system!

Something completely different

This may be one of the least essential posts on this blog ever, but I thought that this stuff might be interesting to one or two of the more curious players, so please bear with me. =)

While pulling out some previous design work, I dusted the electrons off of some very old files and found a lot of enlightening/embarrassing notes from a different era.

I had a blast going through them, and I thought I'd share an arbitrary excerpt from one of these old designs. These are some of the notes for an area I was planning called the Auriakan Gauntlet - a questing area extending beneath Aur-Vindi filled with many dozens of "challenge rooms" linked to each other, which players would navigate through in a competition to progress as deep as they could before their timer expired.

It was a horrible design, for many reasons, and I eventually abandoned it. By the way, these notes are over 10 years old(!!) The tunnels still exist, however, their hollow corridors running deep beneath Aur-Vindi, filled only with the echoes of deserted dreams. One day...


<Title>:  Rakelle's Blossom
Spiral-patterned series of one-way rooms as below. Start at s, exit at e. Secret
doors bar progress from each room to the next. Clues to solve each secret are in
descs. Rakelle's spirit in inaccessible room with one-way exits into s and *,
mental tiaraning while player is trying to solve clues.

[*]}[ ]}[*]
 ^       v
[ ] [s] [ ]
 ^   v   v
[*]{[ ] [ ]
         e

<Title>:  Flags
Remember to use remortFlag and make Rakelle an NPC race so she doesn't chase the
player after the first tiaran.

<Title>:  Non-replayability/scripting
Sucks that this is a non-challenge once solved and can simply be scripted.
Possible semi-solutions:
1) Clockwork golem with very long delay (35min?) revolving the entrance access
   between maybe 4 or 6 challenge sequences?
2) Extend idea (1) to provide branching access midchallenge?

Both of the above would require giving Rakelle one-way tiaran access to more rooms
(possible simply by adding n/s/e/w extensions to her home).

<Title>:  Scrap this idea.
This is really a pretty dumb idea. Basically a test of the player's Con and HP
more than anything else.

Scrap.

<Title>:  (sentry statues)
Sentry statues that come to life after a player has spent a certain amount of
time in that room and attack. Very high powered. Purpose is to keep player
moving. Can come in differently timed varieties and used in multiple challenge
rooms.

<Title>:  (sweepers)
Powerful automatons that "sweep" the Gauntlet of stragglers. Monitor the area,
if any one player is in for too long (3 hours? 5 hours?), the sweepers stalk
that player. (Non-wandering, reside in rooms with one-way exits into the
challenges.)

Alternately, make them powerful but killable (power level geared to the
challenge they're connected to perhaps?), and have them sweep at constant
intervals (every hour?).

<Title>:  Addendum: Idea!
Sweepers could also be constantly doing a stalk on any player carrying a Seal
that corresponds to the next higher challenge... additional cheat-proofing!

<Title>:  Note: Power Level
Do not link power level of the sweeper to the connected challenge - all must be
uniformly unkillable/dangerous. (Otherwise they won't work in preventing the
comatose-in-a-safe-challenge-to-lock-up-Gauntlet abuse.)

<Title>:  sewers
Twisting sewers, use marsh tile and stone. Various slides lead deeper, one-way.
All paths should be valid (except maybe a few with warning signs), but some
tougher.

Low difficulty early challenge.

<Title>:  Clockwork Labyrinth
Populated by golems and other machinery. Maze is "altered" in realtime by golems
who open and close secret doors at set intervals, closing off some passages and
opening up others. Blindly killing everything (thus killing the "operator"
golems) is likely to leave the questor stuck.

<Title>:  Implemented - let's expand
First clockwork maze complete. (Questor has a fixed amount of time to complete
the maze or the challenge is closed.) Simple path-altering done by clockwork
golems to create a real-time changing maze, requiring mapping/deduction or a lot
of luck.

Expand on idea in later challenges. Current clockwork golems only operate if
no PC is in the room. In description, show (but don't explain) method of
programming clockwork golems (example - 4 chest slots, different gemstone in
each slot, the type of gem in slot 1 determines whether golem operates only if
no PC, or only if PC present... etc).

Later golems can be "programmed" for different actions (for example, only open
passage if player present but close in x seconds). This requires the player to
either deduce the "programming" clues or operate on trial and error. (Example -
a passageway that can only be accessed via the clockwork golem in the room which
operates only when a player is present - the player would have to kill the golem
right after it opens the passage or be unable to return after the golem closes
it off behind them. Thus, once the player has deduced "do not kill clockwork
golems, we can throw them for another curve.)

<Title>:  (blood (sharks? mongrels? worms? I suck at names))
Blood sharks (yes, terrible name - blood worms? blood mongrels?) wander passively
under normal conditions but will aggro any PC with half or less life.

<Title>:  Nightmares! (Dream-eaters?)
Area can be filled with skittish rats or some other mob that immediately runs
away when a PC is in the room (and flees if fighting) - must be weak enough that
they won't actually hurt a player. Clues etched on walls. Actual Seal-loading
mob is in a 1-way room connected to the challenge, and will not move unless the
player is sleeping (in which case it will stalk the player).

<Title>:  No Escape
Yes, it needs a new name.

Center of the challenge: a small hill, on which are trapped a band of holy
warriors (paladins, bishops, priestesses). A Diva in the center of the area
chants forth a circle of warding, the only thing keeping the evil at bay.

Completely surrounding the hill are an assortment of evil spirits (banshees,
spectres, and the like). A few paladins also roam the surrounding area,
attacking the spirits. The toughest evil mobs are the devourer spirits,
which roam slowly swallowing large chunks of earth and whatever is on it
(sacrifice all, to keep the area clear of paladin corpses/gold).

The spirits will aggro any good mob, as well as any player, unless that
player is holding the Saint's Locket from an earlier challenge. The spirits will
shy away from any player holding the Locket (move out of the room, or flee if in
battle).

A White Dragoon commander in the center camp holds the Seal. He will trade the
Seal to any player in exchange for the Saint's Locket. (Thus the player will
obtain the Seal, but will have to fight to the challenge portal as the spirits
will now aggro them.)

If the player did not obtain the Locket earlier, the Dragoon commander can be
killed for the Seal, but this will be an extremely difficult battle, even
harder than forging a path through the spirits (who are of a reasonable power
level except for the Devourers, which are huge but sparse and avoidable). One
possibility is to lure the Dragoon commander out of the camp and into a
Devourer, thus having the Devourer tank.

<Title>:  Notes: Dragoon commander
Must be non-wandering, to avoid wandering out of the circle of warding.

Camping prevention: load him in his tent, with an Ember. Tent must have a secret
door to seal it. Once there are no players in the Gauntlet, destroy the Ember
and open up the tent.

If a player is in his tent while he has the Ember, request that the player leave.
If s/he doesn't, attack. Probably need to include a hunting dog or other companion,
which joins a fighting mob if MOB GOOD, made very powerful to discourage powerful
players from camping in the tent.

<Title>:  Companion
The companion can be asleep, and set to wake and join the fighting mob if
MOB GOOD and MOB HAS_ITEM (Ember). (Need the ember check, otherwise killing the
Dragoon for the Seal will not be a viable strategy without a ranged attack to
lead him out.)

<Title>:  Notes: Locket protection - abuse prevention
Possible minor abuse: obtaining the Locket to ward off the spirits, but instead
of trading the Locket to the Dragoon commander, lure him out of the protective
circle until a Devourer tanks him. Little to no risk to player, as the spirits
will fear the Locket and s/he will not need to tank the Dragoon commander.

Solution: make sure the spirits do a check on RANDOM PC ROOM for HAS ITEM
(Saint's Locket) frequently, even if fighting, and flee if check is passed.
If the player is attacking the Dragoon, this should result in the player
fighting the Dragoon.

Dragoon should be a moderate-high damage mob with good hp. Devourers should be
a low-moderate damage mob with extremely high hp.

<Title>:  Notes: circle of warding
Circle can be simulated via RACE flags. Allow all player-able races into the
circle, perhaps allow for non-playerable races also, but deny access to spirits.

<Title>:  "buttons and levers" substitute
An idea to enforce solving a puzzle in a semi-linear fashion, without needing
button and lever-type interaction.

The puzzle must require the player to "double back" to previously visited areas.
Section A leads into section B via a one-way passage. B leads back into A via
another one-way passage; however, this passage has one or more doors on the
A-section side. Each door is nonsecret (although possibly locked) from the
A-side, but is a secret door (no clues, not intended to be solved) on the
B-side.

Thus, while the player can move from A to B at any time, they will not be
able to return to A unless they pre-open the doors from the A-side. Note
that repop may get in the way of this, so best implemented in later
challenges (when the player will likely have already seen at least one
Gauntlet repop and thus "should" be keeping track of time).

<Title>:  Hunting Competition
Area loads with many, many copies of a tiny mob, one of which loads the Seal.
Player begins in a sealed room with some kind of messed-up hermit. Hermit invites
player to participate in a game - a race to find the Seal. Hermit then opens up
the room and runs off, slaughtering and looting mobs. Player can do likewise, but
placed liberally in the area are larger aggro mobs - very low damage but high hp,
designed just to slow the unwary player down.

If Hermit finds Seal first, only recourse is to wait for repop or try again later.

<Title>:  amendment
Game should actually be for some item other than the Seal, which the Hermit will
trade the Seal for. Hermit can do a pre-game check to make sure the player does
not have extra copies of the item (use same abuse prevention measures as with
the Keepers). This prevents Seal camping.

<Title>:  How am I going to do this :P
Heh, is this even going to be possible? Thoughts...
Hermit must be wandering, otherwise "returning home" will mess up his hunting.
However, he also must be able to somehow return to the starting room and stay
there, to introduce the game to each questor.

Perhaps every room in the area can have a secret down trapdoor which leads into
a mortal-only room, which in turn leads back to the "entrance" room. He can
periodically open the trapdoor, go down, close the trapdoor, and go down again
to return to the entrance, then rest. This should work fine UNLESS he manages
to wander in the middle of the return process, which will throw everything off.
If he does this every so often, though, it should set him eventually.

We don't want him doing this while hunting, though. While hunting (before each
kill?) he should do some checks to see if he should stop hunting and revert to
reset mode. The following should end the hunt: ME HAS_ITEM, MOB HAS_ITEM (for
RANDOM PC AREA), or INVALID_TARGET RANDOM_PC_AREA.

Need to test this locally.

<Title>:  I did it! And it sucks.
Managed to write the AI. Mobs operate perfectly once the challenge is properly
built... but damn it's a boring challenge. :P

Saving AI for possible future use, removing challenge.

<Title>:  Balancing Act
Series of ropes/fallen trees/bridges/planks/some sort of walkway, suspended above
a pit. At least two mob types roaming the walkways - one fairly innocuous (maybe
aggro but not difficult), second sparser by far but much tougher and stalking.
Second mob must be tough enough that players want to avoid it, thus being forced
to traverse the walkways quickly. The "walls" of each walkway are one-way exits
into the pit below, occupied by nasty mobs and with no escape (must recall) -
if the player missteps off the walkway, they've ruined their chances of moving
on. Also, if the tough stalking mob catches them, they must fight to the death
or risk fleeing/vanishing into the pit (unless they have retreat).

<Title>:  Ups and Downs... sneaky
Hehe, this could be pretty neat.

One-way exits are obvious (on the outdoor tileset) due to the blackspace and the
lack of a wall - but only for north/south/east/west exits. Up/down exits that are
one-way don't look any different.

This can be used to good effect if the player "assumes" that an up exit is simply
the other end of a down exit they just travelled (or vice versa). In reality, the
up exit does not lead back to the room they just came from, but another section
entirely. Unless the player has a reason to go back up (or is exploring very
thoroughly), they may walk right past without noticing.

This works especially well if the corridor the player just came from is trapped,
thus their instinct would be to avoid going back up after running down to escape.

<Title>:  Completed Challenges
These challenges are structurally complete and ready for mob placement, requiring
only room descriptions, key numbers, and final inter-challenge linking.
               
<Title>:  Gauntlet subrooms
36516 : Cross of Pain .
36524 : Halls of Ice .
36546 : Inferno .
36651 : Burial Grounds .
36655 : Slow Roast .
36704 : Sewers .
36844 : Sovereign Eternal .
36919 : Faith

The plots they are hatching

Alright! Let's get this started, yeah?

The Great Player Wipe of 2014 devoured bits, bags, stats, mats - the accumulated detritus of a thousand journeys. As the universe rebooted itself, some things were carried forward

You'll never pwipe Christof

Some things took seed and regrew differently

Break it up, clans

And some things were lost forever

Crafting.


Version 2.0 of crafting (which was disabled after the wipe - version 1.0 was a well-received trial run with a very different item set many, many years ago) was balanced around a different era. With the universal reboot came the perfect opportunity to revisit this game-defining system and improve it.

I've spent the past several months doing design work, starting from scratch conceptually to create the crafting system that I'd like to see today. The first two crafting systems were designed from a mechanical viewpoint - that is, I took a look at the tools I had and thought, "how can I use these tools to put together a cool crafting system?" For crafting 3.0, I'm trying a slightly different approach.

This time around, I have a long history of player behavior and feedback to work from. I can now take advantage of the numerous AI and system improvements which have been implemented over the years. I have the basic crafting infrastructure as a starting platform, and much more design experience than when I first unleashed the salvage gnomes onto Aur-Vindi.

Will 3.0 be any good? Who knows, it's a coin-toss. It's exciting as hell to try, though!

I began the design process in August by drafting a statement of purpose. Without any specific mechanics in mind, I looked back at Nodeka's crafting history, revisited various crafting systems throughout other games, daydreamed a little (a lot), and came up with a set of goals for 3.0. I won't be dumping my whole set of design notes here - allure relies on mystery! - but a few of the goals and lessons I ended up with were:

+ Experimentation is key: modular construction of items via diversity of inputs (decisions)

+ Trash: things should be crafted because they are needed; avoid "skill-up trash"

+ The grind: I prefer the focus to be on "skilling up" the item, rather than grinding a player skill

+ Stat inflation: no

+ Lifetimes: too many negative game implications; shift focus to permanent gear

+ The crafter should matter: investment + decisions = unique output, potential for economy

These goals (and a number of others) guided the design process. Coming up with a good solution is not as easy as writing up some rules and following them - there are limitations on scope, coding considerations, contradictory goals - but bridging that give and take with your singular overpowered tool (your imagination) is the fun part of making a game system stew.

Blah blah blah my MotD is still empty

I've decided to post this Sixth Sense entry now because the end of the design process is in sight. There were several false starts and do-overs, but I think I have the outline of a system which fits the framework I was aiming for, and which is also possible for me to implement. It will take another week or so to finalize some details, and then I can start the actual implementation (scripting, resource creation, and coding).

In the meantime, while I'm doing the wrap-up work, I wanted to open up the boards to chat about what you think about crafting. I think most of us have encountered crafting systems throughout our gaming histories. Which systems did you like the most, and what did you like about them? What would your vision be for bringing the salvage gnomes out of unemployment?

Let me know in the comments, and thanks for playing!