My problem with skill decay.

[Greep]

Personally the reason I don't like high decay is that it forces your colonists to have exactly one speciality, maaaaaaybe 2 if you're short on colonists.  That doesn't feel so realistic, many people have expertise in many areas.

If colonists with level 20 being rare is the goal, it would be better for the skill chart to be adapted accordingly instead.  E.g. going from level 19 to 20 takes 25,000 points or so, without decay you'd use maybe 100,000. Or 80,000 if using a really tiny decay like 10 times lower.

[sadpickle]

Personally the reason I don't like high decay is that it forces your colonists to have exactly one speciality, maaaaaaybe 2 if you're short on colonists.  That doesn't feel so realistic, many people have expertise in many areas.

Yes, but only the most gifted individuals achieve high levels of proficiency or mastery in multiple disciplines. We have traits to represent gifted individuals (fast learning and now great memory), but even then it takes incredible effort. And in the real world, if it's an evolving discipline it means staying on top of the latest developments. Things like tech work and medicine are examples. Because of the amount of innovation occurring on an annual basis, it can be a time-consuming challenge to remain on the cutting edge.

I think skill decay is a good feature, but it's fairly aggressive. I wouldn't mind seeing it scaled back, especially at lower difficulties. I also wouldn't mind seeing it scale with time spent not performing a task; ie, if your builder isn't working on construction things daily, he'll start losing skill, but not if he's staying busy.

[Klomster]

Having a skill at 10 gives good results on any taks they perform. Have you ever done something you haven't in years? It takes time to reaquire some, if not all, lost skill, which brings me to my point:

Yes, several times.
It usually takes between a couple minutes to a day depending on what before i'm back to trim.

My problem isn't skill decay over several years.
It's that it takes SECONDS for pawns to begin forgetting what they learned.

As in an example above, someone can forget what they learned that morning while sleeping the same night.
The what?

But anyway, i knew people and Tynan would like skill decay to be in, but it's far to aggressive at the moment.
Since it's impossible to get some skills high.
While f'ex construction easily go into the double digits, and fast.

[sadpickle]

Since it's impossible to get some skills high.
While f'ex construction easily go into the double digits, and fast.

I think this is evidence that it's unbalanced. I too usually wind up with at least one if not more pawns at level 20 in construction, but it's the only skill that I max out with regularity. Crafting/Art is probably possible if you keep them grinding things out. Combat skills are impossible to max.

[bobucles]

Skill decay does add to the game. It's silly to have colonists build up to 20/20 in everything, but it doesn't really matter anyway. Pawns only have enough time to work one major job and maybe 2-3 minor tasks. When you give them dedicated jobs it's easy to reach 18+ skill in their profession, regardless of their traits.

Granted there is some imbalance between skill rates. Construction skill goes up CRAZY fast especially when working with the extended stone build times. Carpenters are always the first to hit level 20. Social skill seems hard to keep up past low teens, doctors are nearly impossible to train outside of war crimes, and cooks are painfully slow to learn. Shooting and melee are obviously the most difficult to train since you can't learn them outside of subjecting your pawn to unlimited scars, infections and chances to die. There needs to be more ways to learn skills without killing your pawns.

My biggest annoyances are that non passion grows too slowly and some traits hit the XP growth cap too quickly. Oh a carpenter is super passionate and a fast learner, that means you get 5000XP per day instead of 4000XP. Why even bother. Passion should play a role in XP skill caps and maybe decay rates. For example:

- Non passion learns at 80% but burns out early so he caps at 2000XP per day. Skill decay starts at lvl 7 so each level decays 3 tiers faster.
- Passionate keeps on learning up to 4000XP, with vanilla skill decay.
- Extreme learns at 120-150% and caps at 6000XP, and skill decay starts at level 13 so each level decays 3 tiers slower.
- Fast learner gives 1.3-1.5x learning rate and 50% more XP cap.

Something like that. This way all pawns can learn skills up to a decent level, but only passionate pawns can become experts.

[Klomster]

Dunno.

There's lots of people who do jobs they don't like but get darn good at it IRL.
It would be odd that only people who really love something gets good at it. Sure, they will learn more swiftly and have fun doing it, but often they don't get unnaturally better, it has more to do with traits.

Even with no skill decay it would take such an astronomical time to reach so many 20s the entire colony will have either been annihilated, they left or have lived long enough there's an entirely new generation of colonists.
The need for a passion to learn at a decent speed is already reasonable and learning non passions take a good long time.
Only one that's easy to get 20 with having no skill decay is construction.

The easiest fix would probably set in a start time before the decay.
A day, a week, a month, something. Since the whole "I learned how to make better surgeries today, but since i went to sleep i forgot how to do it." is rather silly and having a simple start timer on decay would do a lot.

[5thHorseman]

Random thought: What if it worked exactly like it does right now, but the game also remembers the maximum skill your pawn was at and THAT is the skill level of the pawn? So your surgeon won't unlearn how to do surgeries but doing one won't increase his actual level if his decayed level is too low.

[Greep]

Well if the idea is to make sure colonists don't end up with 19-20 in every or multiple skills, that would fail:  you could just reach 20 then move onto the next.

[5thHorseman]

Personally I'm a bit tired of hearing "Someone could invest hours of of micromanagement into this to game it so we're going to mess it up for the rest of you." I'm trying to meet them halfway.

[Brainsample]

There's a fancy way of doing this, but it requires an extra field for each skill.
It's fairly simple:

- The first layer is the main skill level. It grows when the skill is used, there is no decay.
- The second layer decays when the skill isn't used, and grows towards the first layer when the skill is used.

So if a pawn has been using other skills for some period, you'll need a bit of time to recover the original level.

I'm not sure if it's worth the effort, but it does seem to be intuitive.

[erdrik]

My first thought is to tweek the decay delay. Especially for cooking. I agree they really shouldn't be forgetting how to cook seconds after leaving the stove.
But that really doesn't solve the issue with maintaining what Tynnan wants in preventing lvl 20 spam.

So, I have had an additional idea Ive had for other skill developing game types, but may be a bit complex...

Results based skill advancement:

• Any skill below 5 will grow while the pawn does the task, AND gain a single shot bonus growth after the task is complete if the task fails, succeeds, or succeeds with poor results.
• Any skill between 6 and 10 will grow while the pawn does the task, AND gain a single shot bonus growth after the task is complete if the task fails, or succeeds with poor results.
• Any skill between 11 and 15 will grow SLOWLY while the pawn does the task, AND gain a single shot bonus growth after the task is complete if the task FAILED.
• Any skill between 16 and 20 will NEVER grow while the pawn does the task, AND ONLY gain a single shot bonus growth after the task is complete if the task FAILED.

This will of course require rebalancing the growth rates to account for the single shot bonus growth that comes at task completion. And obviously wouldn't work for skills where failure is a REGULAR occurrence even at 15+ levels.
(I don't recall ever having a 15+ Warden, but recruitment is an example of a task that fails alot before succeeding even at higher levels in my experience)

The idea is that when the pawn is still a noob at a skill, they are learning the skill and knowledge from EVERYTHING they do. But as the pawn becomes better at it, advancement transitions from being about learning the basics, to practice, then honing the craft to perfection by catch what few mistake are left to be made.

[bobucles]

Skill decay isn't bad, but some of the skill growth rates are definitely not balanced. Cooking skill is way too slow and construction skill is way too fast. The fault for the latter can be blamed on non passion being too crippling because Tynan wants constructors to be viable regardless of learning rate. The current 33% rate for non passion means it takes 3x longer to learn things. So in order for a non passion to learn construction and not completely break your starting game, the default XP rate needs to be ridiculous or they can't learn at all. Buff non passion and nerf construction XP at the same time.

Non passion shouldn't make a pawn full on retarded. Lacking passion simply means the pawn isn't interested in becoming good at a skill. XP caps can stifle skill growth, and decay rates can be modified to hit non passion harder. That way a non passion can still reach a viable 5-8 skill level without pulling your hair out. The learning rate shouldn't be any worse than 50%. Taking twice the time to learn PLUS the a skill decay that goes down to 7-8 means they'll probably never hit more than 13 skill at the absolute max. That's not unreasonable for making a pawn live and breathe something they don't like.

Currently a lack of passion is a stone's throw away from being completely incapable. It hits too hard IMO.

[Klomster]

Random thought: What if it worked exactly like it does right now, but the game also remembers the maximum skill your pawn was at and THAT is the skill level of the pawn? So your surgeon won't unlearn how to do surgeries but doing one won't increase his actual level if his decayed level is too low.

The idea isn't bad but doesn't really solve the problems.
Sure, it makes it so that one cannot randomly drop learnt levels, but the whole forgetting things is sort of still there.
And the problem of theoretically having 20 in all skills still persist.

But isn't a bad idea.

Personally i think giving a skill decay start limit is the main deal. Doesn't matter how long it is (Although an ingame hour minimum, preferably a day or week).
That way, one cannot "just" raise up all the skills to 20, which would still take a silly amount of time.

[Brainsample]

...

Another thing to consider would be multiple skills; It should still be hard to have high skills in everything. I'm sure you could make up some formula, so that for a single skill, you could get to level 20, but for a handful of skills you could only get to around level 15, and for a lot of skills you'd be stuck around 10 or lower.

decay = c * (s(1)^2 + s(2)^2 + ... + s(n)^2 - s(i)^2)

Where c is a constant, s() is the list of skills, and i is the index of the skill under decay.

In words: Decay is proportional to the sum of the squares of all other skills.

Effect: If you have only one high skill, there's (virtually) no decay.
If you have a few high skills, there's decay.
You can't get many high skills, there would be too much decay.

It would probably take a lot of tweaking to implement this in a balanced way.

[Third_Of_Five]

To be honest I think skill decay on any particular skill should slow down as the skill level gets higher. Other than that I think skill decay is fine. It encourages specialization.