Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  

Author Topic: ATTR token and work preferences [SCIENCE!]  (Read 829 times)

taldarus

  • Bay Watcher
  • Awkward Post Master
    • View Profile
ATTR token and work preferences [SCIENCE!]
« on: May 29, 2021, 11:36:04 pm »

Hello Beardlings,

Recently got into a strange mood and dwcided to start tackling my desire to manipulate the dwarven soul to create more specialied workers. Note that I am perfectly aware of SKILL_RATE tokens and their effects.

Scope:
I first set things up with mining in mind.
I then set all but four attr tokens to artificially min/maxed for the dwarven soul.
IE - no 5000 values, just forced the natural range  to be fixed on one point. STR: 2250.2250.2250.2250.2250.2250.2250
I then set one of the learning point skills to the min (toughness)
So I had that min max and three other 'unrelevant' tokens changed.
This was all over tenish generated worlds.

Results:
Mental attributes played a small unspecified role in labor proficiency.
Learned stats appear  to cause a toral of +/- ten percent increase around an unknown fixed point
The fixed point appears to be binary. Mining in each test was always one of two values.
There may have been variance, but it was far from significant (the fifth decimal).
Intuition was tested at several ranges while all learned attr were fixed at max
It reaulted in improvements at 1000 only, and no improvement at 200 and 2000, meaning the ideal number is between 1000 and 2000.

Observations:
Physical attributes provide a consistent improvment as they increase.
Mental attributea provide an improvment in the characters desires. (creativity at 5000 created dwarves that always wanted to create art, but further testing is needed)
All attributes probably contribute to the effectiveness of the labor, but are influenced by an unknown ideal number.
A high focus appeared to be bad for mining, but a low score wasnt necessarily good either.
This holds true to the real world, and may be testing bias.

Conclusions:
The results are inconclusive, further testing will require a very expensive and highly customized analytical tool to measure the exact number instead of just asking Toady. Clearly this is important, and the committee ahould allocate aat least 100,000 dollars to building me a better toy... I mean research platform. (sorry couldnt help myself).

Seriously, ia there a dfhack way to extract the generated values of thousands of dwarves so I can actually figiure this out? Then I would need a really nice computer....

Have !FUN! everyone.
Logged