On Skills and Focus

[miolian]

This is a proposed overhaul for gaining skills in C:DDA.

Current Problems:

1. Slower skill-gain, but without increasing the time spent reading books.
2. Grinding skills by repetition is quite easy.
3. The interesting part of gaining skills in C:DDA, for the most part, involves hunting down books. Combat skills, on the other hand, seem to be working just fine.

Real Life Observations:

1. Alternating between theory and practice of a skill is more effective than doing either exclusively.
2. Getting better at something, especially cognitive stuff, often result from breakthrough moments (in other words, one gains skills in short bursts doing something very hard).
3. Skill gain is motivated by circumstance or inspiration.

Proposals:

1. Grinding skills by repetition quickly becomes ineffective. For example, crafting X of anything should give diminishing returns;

2. Each skill get a theory multiplier and a practice multiplier, which are raised by theory or practice, respectively. The multipliers effectively replace focus- they function similarly, though more nuanced. Both multipliers decay exponentially with a skill-specific half-life. The theory multiplier is also raised by sleep, if it is over some threshold and fall if it is smaller; this incentivizes reading theory, sleeping and then doing what you read about. It also is quite realistic. Like focus, practicing a skill will cause the practice multiplier to drop and thus the delta skill points gained for that skill to diminish and will raise the theory multiplier. Similarly, reading a book will cause the theory multiplier to drop and will raise the practice multiplier.

3a. the player’s character gets random boosts to some skill, possibly triggered by some event. For example: “You witness Bob Ross round house kick a child zombie to the chest and are inspired to improve your Brawling.”

3b. For fabrication/tailoring/mechanics/etc. the player could be inspired to build X. This is a bit more involved than 3a), though now that I think of it, we do have the achievement system one could maybe use for this. I guess what I am getting at is a set of template quests which can be as little as “get materials to build X and then build X” or more complicated. Making such inspiration quests moddable would also be interesting.

Some Game Design thoughts on Focus:
One reason I think a bunch of C:DDA players dislike the focus system (including Vormithrax) is because it is on one hand too broad and on the other mostly negative. At least, it is presented that way to the player, and presentation makes a big difference. I think in the game World of Warcraft they had something similar happen with debuffs, so the game designers re-branded the debuffs as buffs by penalizing everything else (and not telling the players) and everybody was happy.

[Sperg]

There’s a really good PR or something somewhere on the idea of crafting/skills overhaul. I think that’s where the proficiency system was birthed from.

It might be worth contributing your input to that.

TL:DR i think it was how to make levelling crafting without just reading (theory) and without just spamming the same thing each level, as well as overall slowing progression. Same kind of thing.

I feel like in theory it’s as simple as - players spam reading or single items to level because there isn’t a meaningful alternative, certainly not enough meaningful things to craft each level each skill to get you to the next level. Provide a better alternative before nerfing the current options.

Randomising skill progression via random inspirations takes the choice away from the player - you’re halfway through a looting run and get a mechanics inspiration, looks like you’re going back to do some random mechanic stuff. You get an inspiration to improve your Brawling, what if you don’t use Brawling? What if you’ve even learnt another, better martial art?

But there’s also problems with providing enough meaningful content to craft per level:

1 - It would require gameplay taking precedence over realism (a little) - cutting down all the fluff recipes, the random but realistic recipes, the redundant stuff no-one ever crafts.

2 - You’d have to find a way to provide enough things to craft per level per skill to matter, while also slowing down crafting progression overall to make it more meaningful and in-line with the rest of the progression in the game being slowed down.

A basic tier structure would need to be implemented with a pass over all items - weapons, armor, clothing, containers, tools etc and would also need to involve not showering you with all of the above from the start. It’d be a kind of pointless system if you get most of the first 2 tiers of tools from the very first house you loot.

Edit: I believe in that PR or elsewhere the idea was supposed to be for crafting to take a back seat, with the focus being on found loot. You don’t want to just sit at home and go through half the games progression, but it’d be cooler to have half and half - you find half the things you need, while having scavenged material to craft the other half and/or optional/side-grade equipment. Material found in more and more dangerous places the higher the tier. Instead of one over the other, make them need eachother - a symbiotic relationship of awesomeness.

[zantanzuken]

mabye there could be something added as a hidden feature behind the scenes, like… if you go out into the woods and cut down a tree, then saw those logs into wood, you try and craft those planks into something, (say a 2 by sword) and then it gives you a few other recipies for other crafting types revolving around wood

or on a higher scale, say you’re crafting makeshift weapons, you start off with a pipe pistol, and crafting that gives you the ability to make a pipe rifle/pipe shotgun, and crafting those… etc etc, until you can craft a pipe weapon of any variety of ammo type.

disassembly already gives you crafting recipes for making things… but what if, for example… you dissassemble a radio, and it instead gives you the recipe for a speaker, because the speaker is one of the components of the radio (and because you spent the time to learn it, it is consumed in the process, assuming you tore it apart to find out how it works… which can change with the more skill you have.) theoretically you could then craft the item you disassembled by making its component parts first.

all of the above… plus… take away learning skill through books entirely. make books ONLY give recipes… make practical application be how you gain skill and learn how to make things. but thats my opinion.

[anothersimulacrum]

Overall:

Practice makes perfect

I’ve been pondering how skill advancement works recently, and I’m starting to get dissatisfied with tallying up performing actions to measure practice of skills, particularly combat skills. This isn’t how it works in reality, you pile a huge amount of time into practicing a skill so that you can perform it proficiently when needed. Sometimes this fits (notably crafting, you do just craft stuff over and over to get better in reality), but for shooting, melee, dodging, or most other active skill…

Big picture future stuff:

• Skill gain speed slowed
• Practice actions added
• Books stop teaching skills, instead teach practice actions
• Focus becomes a long term ‘how much can I learn in this period of time’ meter

[Sperg]

Is there any meaningful difference between spamming practice actions and say spamming washboards to get Fabrication from level 1 - 2?

If practice actions are the new way to go, but still unlocked through books - Will there be a difference in gameplay from the current placing top priority on books to then just take home and level skills off of?

[miolian]

Hey!, to thouse who answered,
thanks allot to you Sperg and zantanzuken for your insightful thoughts to the idea,
I’m quite delighted and flattered thereby.

@Sperg
“There’s a really good PR[…]”
Oh really? I haven’t devled too deeply into the pull requests. There are so many
it is overwhelming… For the same reason I am slightly scared to look deeper into
the code than I’ve already done.
[off topic tangent] I do C++ backend development for my dayjob, somewhere in Cologne Germany,
C for the most part that is, web development has been creeping its way into our
facilities, like it or not. I’d like to see if I would be able to pull something
off in C:DDA, like this idea of mine, or at least attempt to do so and
get something working without submitting anything. Maybe it would be interesting
talking to the dev who submitted said PR. Dunno though, C:DDA is so big and there
are so many talented programmers contributing I feel sort of sub-par. 6 weeks ago
I was looking at the code for the nested inventory system, and it is partially
implemented in the way I would do it --not to say it isn’t what any programmer would
do for any job requiring nested container-esque structures–; despite me understand
the gist, it’s role in this humongous code base stays elusive. Also, I’ve only
programmed in smaller teams for smallish companies and never submitted a single
PR to any open source community project till now :(.

“I feel like in theory it’s as simple as[…]”
I do agree with you, but what is “meaningful”? This idea is far from profound, but
to me it is one where the environment presents an obstacle which forces
the player (or person, this applies to real life just as well) to engage not just
in novel behaviour, but apply some sort of judgement, requiring the player to
asses her environment and maximize for success, avoiding said obstacle and ideally
profiting in some hidden, highly case-dependent manner.
Yes, I am not only focusing on realism, but though this is said to be the main
focus in developing C:DDA, it hardly can be the only focus, for the simple reason
that there are hard limits to how real the game can be, us wanting it to be
won’t make it realer. For example, if a system such as the vitamin system,
proposed in version C.X I think, is implemented to the max, as real as it gets,
if other systems around it are not on the same level, the effect as a whole on
the game will actually drag it away from reality and force the player to worshiping
said ultra-realistic simulation. What I am saying is, it should be important for
C:DDA devs to be aware that compromise is very necessary and that design shouldn’there
solely be about maximizing realisim and forgetting about game-play and player
decision-making. That said, this is my opinion, and I am fine with Granade et al
disagreeing. It is mainly their project and I find this project marvelous. And fun,
I might add. Heh.

“[…]You get an inspiration to improve your Brawling, what if you don’t use Brawling?”
This is excellent, actually. At least from a game design perspective. Look at it this way:
you are trying to maximize for X but you get A, B, C, … you ignore A and B because you
don’t care, but C is so good that it forces you to go out into a city and hunt down something.
Because of this you get yourself into some hairy situation because you weren’t quite prepared
at that moment and you nearly evade death. That’d be quite fun I’d imagine.
Ignoring a random inspiration template quest is in itself a player meaningful decision the
player makes, and getting handed a actually meaningful inspiration is twicely as rewarding
and might often force the player into sub-ideal encounters because of the time-limit and
because of the random event. In C:DDA you currently can plan our your life, get X,
level up skill Y, become Z. The predictable pattern at which zombies evolve poses no
real threat to the skilled player (i.e. greater than 50ish hours real-life gameplay time).

“1 - It would require gameplay taking precedence over realism (a little) - cutting down all
the fluff recipes, the random but realistic recipes, the redundant stuff no-one ever crafts.”
I disagree here. Cutting down on the fluff recepies increases realism. Becoming a master taylor
isn’t done by crafting n1 of x, n2 of y and n3 of z. Just doesn’t work like that. Forcing the
player out into the world because he was handed some wierd inspiration to sew a bejeweled
kevlar kilt and then actually doing so, in my opinion is much more like real life. Yes,
repetition and practice has been a big part of acquiring skills in my life (and so was it for
all RPGs I’ve played in my life), but key moments have always decided the course of where
said skill would develop towards and to which extent they would develop.

“2 - You’d have to find a way to provide enough things to craft per level per skill to matter[…]”
This already is the case. The system I propose would however slow down how players
per level level up. For example if you are level 0 and want to progress to level 1 in fab,
you won’t be able to rip up a few sheets and craft 200 (or whatever) makeshift bandages, a
better strategy would be to craft a variety of level 0 ish fab craftables from the materials
you are able to gather during your first run to your first basement. No more: “I grab 6 sheets
and select 20x makeshift bandages” any more, but: “OK, should I consider binging materials x,y,z
to the basement, will they be enough? Or should I rather go with a,b,c…” The player is faced
with a circumstantial dillema in this case and not a static “6 sheets to mastery” one,
and this is both realistic and funner, in my opinion.

“A basic tier structure would need to be implemented with a pass over all items[…]”
I agree. But it already is, the proposed system wouldn’t need to look too deeply into the
specific items too much. Though admittedly this is but a hunch. My hunches are good often tho.

" It’d be a kind of pointless system if you get most of the first 2 tiers of
tools from the very first house you loot."
A pointless (not pointless but without points I take, hahah) system would be ideal but I
wouldn’t exactly know how to go about implementing it. Maybe a system with hidden variables
as a few are in C:DDA and maybe reality itself. I joke.

I need to sleep, though I’ll be back tomorrow if you guys are,

to space,

mio

[miolian]

maybe there could be something added as a hidden feature behind the scenes, like… if you go out into the woods and cut down a tree, then saw those logs into wood, you try and craft those planks into some thing, (say a 2 by sword) and then it gives you a few other recipes for other crafting types revolving around wood

Before I say anything else, the use of the cdots symbol »…« is pure win. But yea, that is a good idea. If the skill point multiplier could be increased by doing related tasks in sequence (and decaying the multiplier between sequence elements proportional to time difference), that’d be pretty cool, albeit a bit more involved than what I had in mind. But this is a good idea, really. It could be designed in a way as to guide the player into specific tech trees, that is the classic build x,y,z scheme, by rewarding the player with a higher skill-point multiplier if doing x,y,z in sequence and in a timely manner, but not forcing the player into any said tech tree. One game design principal I swear by is to give the player more than one path to a given solution, and in C:DDA which strives to be ultra-realistic, this principal is in most cases a must.

“or on a higher scale, say you’re crafting makeshift weapons […] disassembly already gives you crafting recipes for making things […]”

Isn’t that, sort of, already the case? Ah, no, I understand now what you mean. This is sort of the tech tree approach that one finds in many games. What you describe sort of run paralell to my idea, that is, they are independent of each other. If I understand C:DDA’s approach correctly (sorry, I’ve only played about 60 to 100 hours of the game over the last year, so I sort of still am noob. That said, the game is a gem among rogue likes), crafting objects is bound to the level of the skill and recipes found in books. I also dislike this, and I think the requirement to craft X should be bound more to than fab level (for example) and a recipe in a book. Implementing a more nuanced system where one can add constraints would also be cool, though I am not sure it would do too much game-play wise. You’d still end up with these static tech trees where the player must craft x,y,z to become super powerful. In real life, I very much cannot do this in general. And by that I mean you cannot say to most anybody:
“read »B. Stroustrup: Programming – Principles and Practice Using C++« and then read y and z and you can work at Cern”. This is obviously because things are so circumstantial, dependent on the persons interests, skills and environment. Such a suggestion can be beneficial as a guide line, but it isn’t this static “do this and succeed” kinda deal, and tech trees and the way most players play the game momentarily this sort of is the case though. Then again I doubt my idea, if implemented would help that much either… oh I don’t know, I need to think about it more.

“all of the above… plus… take away learning skill through books entirely. make books ONLY give recipes… make practical application be how you gain skill and learn how to make things. but thats my opinion.”

I disagree, though not strongly. From a game-play perspective this sort of makes sense. This would make the system less realistic though. I think the undulation between theory/practice provides for interesting game-play, a realistic model (at least more realistic to what we have now) and a greater level of complexity (by which I mean: the space of nuanced decisions a player could make is “larger”, I guess the space is discrete so you could count, but eh).