labor recruiter behavior - City Builders Forums

posted 03-11-00 14:14 ET (US)

Having gotten tired of seeing that "poor access to labor" message flashing over my industries, and having seen conflicting opinions about how far that labor recruiter will actually travel, I decided to take a look for myself. A note -- I did my tests at hard difficulty; I had available unemployed workers, and I only tested hunting lodges and firestations. In all cases, I count the road space in which the recruiter spawns as the first space of its walk.

1) Labor recruiters take two kinds of walks -- short and long. Short walks seem to consistently be 27 spaces long (appear in road, then walk 26 more spaces). Long walks vary somewhat by building (37 spaces for hunting lodges, 36 spaces for fire stations). Approximately 1/3 of all walks are long. If a building is adjacent to the end of a road, approximately 1/3 of its recruiter's walks will be be confused -- the recruiter appears, walks a space, turns around, walks a space, turns around again, and walks 24 spaces in the original direction.

2) A trip is successful if it passes over 2 consecutive road squares which each are within 2 squares of housing.

3) A building needs to reach houses by a short (27 space) walk in order to establish reliable labor access. Once established, this access will be broken by three consecutive failed trips in a row. Contact by a long walk is sufficient to maintain access.

4) If a building can only establish access by a long walk, it will have intermittent labor access. It appears that the closer an intermittent building is to distant (more than 27 space) housing, the more time it will spend on, but I'm hazy on the exact mechanics of this. Buildings types appear to have different needs for labor: a hunting lodge placed 37 spaces away from housing will flicker on for an instant after each successful trip, while a firestation at 37 spaces will flicker off and on with some frequency (strangely; the firestation long walker appears to only walk 36 spaces, but a single road square with labor access seems to be enough to get intermittent labor access going, so it effectively has the same outside limit range as a hunting lodge).

5) Buildings which generate sprites (hunters, firemarshalls, etc.) appear to only need the briefest instant of labor access to do so. When those sprites return, however they won't be regenerated until labor access flickers back on.

6) Labor recruiters seem to be random in their path decisions, but they appear to have two biases -- they like to walk towards a city's center, and they like to walk towards their last labor contact (even if it was a long walk contact). Thus, it seems dangerous to place buildings at a range of 28 to 37 spaces from labor contact towards the city center.

7) 3x3 and 4x4 buildings? I haven't checked them.

That's all I've got. Sorry this is so long. Those labor recruiters were driving me crazy.

-- JWorth

posted 03-11-00 15:09 ET (US) 1 / 12

Your post answers many nagging questions

Keep on it and you'll have a gem for all time. Front page news!

SenetEr ...smile as wide as the Nile...
Drop me a Line

posted 03-11-00 16:26 ET (US) 2 / 12

JWorth: your point #2 is precisely why -- in my thread Rules of Thumb -- I make the statement that worker feeder roads (which overlap housing within 2 squares) should overlap no less than 2 squares (i.e.: 2 houses) and that more than 4 squares is overkill (ala SenetEr - which I happen to agree with). I try to get a 4-square overlap to take up the slack in the event of a population/occupancy variance.

Your observations about the total length of the labor recruiters walking radius bares out my own observations that for best results the industry should be within 26-30 squares of the target feeder houses. Grumpus says 25. Grumpus' figure is probably the best 'Rule of Thumb' distance to insure adequate and steady coverage. Hence -- in my revision of my Rule of Thumb list -- I'm going to reccomend a 'best' distance of 25-26 squares. Still -- it's good to know just what the extreme limits are -- 37 squares. Thanks muchly for that piece of data.

[This message has been edited by user maat (edited 03-11-2000).]

posted 03-11-00 20:19 ET (US) 3 / 12

user maat,

Actually, there's an abrupt shift that occurs between 27 and 28 spaces. 27 will give solid coverage; 28 won't (contact achieved only by a long walking labor recruiter seems always to be intermittent to some degree). You could tell people to aim for 25-26, but why not just tell them 27, but no farther? Particularly if you're trying to give people a first look into the underlying mechanics of the game (which I'm still getting into myself), introducing them to the idea of measuring roads in units of 27 and 54 spaces seems like a good idea, since so many walkers (all the service types, as far as I can tell) base their behaviors on these figures (thanks to Nerdicus for explaining this!).

You should be warned that the 37 space limit isn't a practical figure for most building types. An architect that flickers on (and sends out a walker) every couple of months is probably still a useful building. A hunting lodge is less so, and a copper mine under these conditions would be virtually useless. My guess is that the long walker is principally designed for maintaining labor contact through networks of road intersections (which I try never to make available, anyway).

posted 03-12-00 09:45 ET (US) 4 / 12

In a recent scenario, I observed the following on a simple strip industry area involving 3 papyrus makers, a SY and a reed gatherer. The top configuration gave me really intermittent labor for the industries. I added two road sqaures to keep the recruiters near the house longer. The lower layout gave perfect labor coverage that never 'blinked'.

Legend

...smile as wide as the Nile...
Drop me a Line

posted 03-12-00 13:28 ET (US) 5 / 12

Wonderful

ngel

ahakemhotep

Eyrie, Pharaoh Heaven, Caesar 3 Heaven

Have you consulted the Pharaoh FAQs today?

------------------
Homage to thee, Osiris, Lord of Eternity, King of the Gods, whose names are manifold, whose forms are holy, thou being of hidden form in the temples, whose Ka is holy."
-- Book of the Dead (1240 BC)

posted 03-12-00 18:22 ET (US) 6 / 12

That's interesting SenetEr, cause by my count the recruiter would be within 2 squares of the housing for 5 squares walk time while the bottom one would increase to 9 squares walk time. Have you tried the same layout with a 2x2 house? I seem to have much better time with the labor recruiter even with very small overlaps of a 2x2 house. I have used a single 2x2 house on a loop like the following in a number of places without problems. Of course this would give 6 squares of walk time near the house.

Legend

posted 03-12-00 20:12 ET (US) 7 / 12

Hmm, that is interesting. I find that a layout like this, which has even fewer road spaces with housing contact, also works just fine:

rhh cc
rhh ef cc
rbrrrrrrrrrrrrrrrrrrrrrrrrr

I'm working on upgrading these graphics. The glyph generator and my browser didn't like their first meeting.

I guess the recruiter just prefers 2x2 access.

posted 03-12-00 20:15 ET (US) 8 / 12

Oops, I just checked my last post. Those c's should be a mine down at the end of the blocked-off road.

Oops again. I may have messed up the posting order, trying to recreate my message with tabs. I won't try that again.

-- JWorth

posted 03-12-00 20:57 ET (US) 9 / 12

SenetEr and Nerdicus,

It occurs to me that these experiences could be explained if every road has a labor value, set at the amount of housing within two spaces. Under this system, SenetEr's arrangements would have values of 5 and 9, Nerdicus' would have a value of 10, and mine would also have a value of 10 (for a trip that terminated in the dead end). I'll give this a look tonight.

posted 03-13-00 08:59 ET (US) 10 / 12

JWorth, that is a very interesting idea!

So, in my case the road squares near my house would have a labor value of 0,0,1,2,2,2,2,1,0,0,... which would be added to some kind of recruiter's labor found variable when he stepped on each square?

I really like your algorithm and it would be significantly less computation for the game than having every recruiter look in a range of 2 around itself for each square moved. The road's labor value would only change when nearby houses were added/deleted or lost all residents. This may be a bit interesting to try and prove...but good luck.

posted 03-13-00 09:51 ET (US) 11 / 12

I never really though that they might assign values to the road. They probably just assign it to the square similar to desirability and if a road happens to be on it, it can be used by the recruiter. Sorry, if I am stating something rally obvious to everyone. It's out of my area of expertise.

SenetEr ...smile as wide as the Nile...
Drop me a Line

posted 03-13-00 10:20 ET (US) 12 / 12

From my experiences, this construction:

Legend

is worse than this one

But this construction

is equivalent to this construction

Also, this one

is much worse than this one

Seems obvious in retrospect, but that idea SenetEr had about calculating a desirability equivalent in terms of labor access may have been right on. It would work just like any of the other walkers, but the "column" for worker access would be on the worker-seeker's building and would improve as the walker went through good employee areas.

[This message has been edited by Angel Thorfinn (edited 03-13-2000).]