[ August 01, 2007: Message edited by: Mechanoid ]
quote:
Originally posted by chaswick:
<STRONG>or maybe some skills result in much faster increases than others?</STRONG>
I think this is the case, at least in a few areas. Seige operation, for example. Loading a catapult seems to be about the longest job there is for unskilled dwarves, but legendary seige operators do it basically instantly.
Also, to add to Mechanoid's comments, for workshop tasks clutter plays a huge role. In an uncluttered workshop a legendary dwarf can do jobs basically as quickly as they can haul in materials, since it takes them a fraction of a second to actually do the task. Of course, if you have a legendary dwarf running a workshop it's going to get cluttered very quickly, and then it slows down a bit...
[ August 01, 2007: Message edited by: RPB ]
Each began a separate one tile wide mineshaft, dug for about 10 or so tiles. I created doors and locked each 'miner' in a one square room. I set them to continue mining their shafts. They commenced at the same time.
After one minute I observed that all had ended at close to a perfect time (the square they were working on was fully yellow for all). I also observed that no ore or gems were encountered in any of the shafts. Results:
The 'dabbler' mined 7 squares.
The 'novice' mined 9. (1.28 x)
The 'proficient' mined 15. (2.14x)
I did the same thing with three 'masons'. Had a fully stocked 3x3 pile next to the workshop to reduce hauling times for each. Made sure workshop was uncluttered. Results for that test were that the dabbler and novice produced at about the same rate. Proficient was a little short of half ( around 20 secs vs around 35-45 secs) as long as the other two.
Conclusions:
I believe that the classified skill level ('dabbler', 'novice', etc) is what determines speed, not the raw exp. So, I contend a 499 point dabbler is no more efficient than a 1 point dabbler. I believe the times ( in seconds) used in the study are not exact because CPU utilization varies over time. Within a 1 minute window it does not seem to change, but as events unfold in the game or on the computer, comparing raw timing data from one period to another is error prone (hence the range for masonry). I do believe we can say that for mining and masonry a proficient dwarf is about twice as effective as a dabbler. Note: I used a dabbler and not an unnamed dwarf to minimize variables. All peasants did the skill at least once before the study.
Thoughts / comments / other skill comparisons?
Compare a paysant to a Legendary miner... It doesn't matter much if the skill indeed reduce mining time, the Legendary Dwarf will simply zip past the paysant in whatever they do as long as he have one or more points added up in Agility.
There is a base speed for the game, and a creature speed figured from the SPEED tag and the Agility of the creature. The base speed is divided by the creature speed.
Next the weight of objects in the job and the Strength of the creature is considered. This will further reduce the creatures speed while doing that task.
Finally the job has a time count that is modfied by the skill level of the creature doing the job.
Once the creature has been at the job for the jobs time count it is done, and the way that is counted is adding the creatures speed for that job up each time until the total is greater or equal to the required time.
This is why an Ultra-Mighty, Perfectly Agile, Skilled operator can load a catapult just as fast as a Legendary operator without the stat gains.
Also skill levels above Legendary keep going, by they aren't named anything else. Stat gains above Perfect, Ultra, and Super are also there. All of these seem to be done with a hyperbolic progression, so the low and high ends make small differences in terms of what we can see. The middle makes the most rapid improvements.
-Albert
quote:
Originally posted by RPB:
<STRONG>
I think this is the case, at least in a few areas. Seige operation, for example. Loading a catapult seems to be about the longest job there is for unskilled dwarves, but legendary seige operators do it basically instantly.
</STRONG>
Strength is a major factor in this. Dwarves hold the stone while they're loading a catapult, and they must be Very Strong to hold a stone without encumbrance. With no extra strength, they work very slowly.
I realized I had no way of evaluating whether I was gaining or losing by choosing a 'competent' skill or a 'proficient' skill and paying extra. I am sure Toady has considered those in his calcs of point cost. I am interested in what I gain by choosing 'Skilled Miner' over 'Miner', e.g.
Regarding the theory about hyperbolic increases, it DID seem like there was a tipping point where efficiency really starts to improve and jobs get done faster. This could be utilized in designing starting builds by identifying certain jobs in that build where their efficiency isn't needed except at certain points in the game. "If I will need that grower to start producing immediately, what skill level is the lowest I can get away with?" e.g.). I am looking for a more concrete way to identify that tipping point (and also just looking to kill time :) ).
As an offshoot, has anyone categorized jobs by more strenght intensive / more agile / long base times / etc to breakdown the output of various workshops or how many can be done in a (dwarven) day? I myself didn't realize how slow making a door was at first.
Thanks all.