26-05-2003 08:14									#1
	Jungle Clearing Efforts with 4 Worker's Support Apolyton buy from Amazon

I put together a sim to test this stuff out and using various techniques the best results for our turns and number of worker's seem to be using 4 worker's in the way I have posted them below.

These order's work whether or not you have to cross river's, since we don't have Engineering yet, this is a concern. Especially considering that we should work the river tiles first.

Here's a key to what I did:

Order's
C= a new order of "Clear Jungle"
R= a new order of "Road"
M= a new order of "Mine"
C-IP= a "Clear Jungle" order in-progress.
MA= an order to "Move" to the next tile to be worked.

Example: 2C-IP 2MA, means that 2 workers are in-progress clearing jungle and 2 workers are to be moved ahead to the next tile to be worked.


Worker status= what tiles the workers are on

JT 4/12= A "Jungle Tile" is being cleared and will have 4 out of 12 turns completed towards that task at the end of this turn.

MT 3/3= Same as above, but for a "Mined Tile"

RT 2/2= Same as above, but for a "Roaded Tile"

code:


            ORDER'S                      WORKER STATUS                   TILE STATUS
Turn:   1   4C                       4 ON TILE 1                   JT1  4/12
Turn:   2   4C-IP                    4 ON TILE 1                   JT1  8/12
Turn:   3   4C-IP                    4 ON TILE 1                   JT1  12/12
Turn:   4   3M  1R                   4 ON TILE 1                   MT1  3/3    RT1 1/2
Turn:   5   1R  3MA                  1 ON TILE 1   3 ON TILE 2     RT1  2/2                   TILE 1 FINISHED
Turn:   6   3C  1MA                  4 ON TILE 2                   JT2  3/12
Turn:   7   3C-IP  1MA               3 ON TILE 2   1 ON TILE 3     JT2  6/12
Turn:   8   1C  3C-IP                3 ON TILE 2   1 ON TILE 3     JT2  9/12   JT3  1/12
Turn:   9   4C-IP                    3 ON TILE 2   1 ON TILE 3     JT2  12/12  JT3  2/12
Turn:  10   3M  1C-IP                3 ON TILE 2   1 ON TILE 3     MT2  3/3    JT3  3/12
Turn:  11   2R  1C-IP  1MA           2 ON TILE 2   2 ON TILE 3     RT2  2/2    JT3  4/12      TILE 2 FINISHED
Turn:  12   1C  1C-IP  2MA           4 ON TILE 3                   JT3  6/12
Turn:  13   1C  2C-IP  1MA           3 ON TILE 3   1 ON TILE 4     JT3  9/12
Turn:  14   1C  3C-IP                3 ON TILE 3   1 0N TILE 4     JT3  12/12  JT4  1/12
Turn:  15   3M  1C-IP                3 ON TILE 3   1 ON TILE 4     MT3  3/3    JT4  2/12
Turn:  16   2R  1C-IP  1MA           2 ON TILE 3   2 ON TILE 4     RT3  2/2    JT4  3/12      TILE 3 FINISHED
Turn:  17   1C  1C-IP  2MA           4 ON TILE 4                   JT4  5/12
Turn:  18   2C  2C-IP                4 ON TILE 4                   JT4  9/12
Turn:  19   4C-IP                    4 ON TILE 4                   JT4  13/12
Turn:  20   3M  1R                   4 ON TILE 4                   MT4  3/3    RT4  1/2
Turn:  21   1R  3MA                  1 ON TILE 4   3 ON TILE 5     RT4  2/2                   TILE 4 FINISHED
Turn:  22   3C  1MA                  4 ON TILE 5                   JT5  3/12
Turn:  23   3C-IP  1MA               3 ON TILE 5   1 ON TILE 6     JT5  6/12
Turn:  24   1C  3C-IP                3 ON TILE 5   1 ON TILE 6     JT5  9/12   JT6  1/12
Turn:  25   4C-IP                    3 ON TILE 5   1 ON TILE 6     JT5  12/12  JT6 2/12
Turn:  26   3M  1C-IP                3 ON TILE 5   1 ON TILE 6     MT5  3/3    JT6  3/12
Turn:  27   2R  1C-IP  1MA           2 ON TILE 5   2 ON TILE 6     RT5  2/2    JT6  4/12      TILE 5 FINISHED
Turn:  28   1C  1C-IP  1MA           4 ON TILE 6                   JT6  6/12
Turn:  29   1C  2C-IP  1MA           3 ON TILE 6   1 ON TILE 7     JT6  9/12
Turn:  30   1C  3C-IP                3 ON TILE 6   1 ON TILE 7     JT6  12/12  JT7  1/12
Turn:  31   3M  1C-IP                3 ON TILE 6   1 ON TILE 7     MT6  3/3    JT7  2/12
Turn:  32   2R  1C-IP  1MA           2 ON TILE 6   2 ON TILE 7     RT6  2/2    JT7  3/12      TILE 6 FINISHED
Turn:  33   1C  1C-IP  2MA           4 ON TILE 7                   JT7  5/12
Turn:  34   2C  2C-IP                4 ON TILE 7                   JT7  9/12
Turn:  35   4C-IP                    4 ON TILE 7                   JT7  13/12
Turn:  36   3M  1R                   4 ON TILE 7                   MT7  3/3     RT7  1/2
Turn:  37   1R  3MA                  1 ON TILE 7   3 0N TILE 8     RT7  2/2                   TILE 7 FINISHED

Tiles will cleared, roaded, and mined on turns 5, 11, 16, 21, 27, 32, 37 
counting from the turn that 4 workers can begin clearing a tile.

Here's some math to further confuse you.

4 worker's at 37 turns equals 148 worker turns.

It takes a minumum of 17 worker turns to clear, road, and mine a jungle tile.

So 17 X 7 tiles cleared= 119 Worker turns.

Since some worker turns will be wasted on movement, we'll subtract those from the 148.

4 Worker's moving to 7 jungle tiles= 28 Movements.

So 148-28=120

120 worker turns used on tasks.

Since 119 is the minimum value needed to do the tasks I will divide it by 120 to get the efficientcy rating.

119/120=99.17%

Conclusion:


The jungle tiles were completed with turn spacing between finished tlies like this:
5, 6, 5, 5, 6, 5, 5

I suspect that pattern would continue.

We get a completed tile every 5.33 turns, not including the first tile.

With 4 worker's this was the best I could do; maybe someone else can do better??

I may run some sims based on 3, 5 and 6 worker stacks; but I think with the preliminary numbers I ran on those, we wont get the efficiency that we would get with 4 working on jungle tiles.

9 worker stacks gave very good number's but I didn't look into it too much, maybe at a later time.

Anyways, I hope this can help ET out a little.

	26-05-2003 13:06									#2
	Support Apolyton, buy Galactic Civilizations

I believe you can't take the movement out of the equation for the purpose of determining efficiency (as the extra movement actually lowers efficiency a lot, compared to "100% efficient" single-worker jobs). Add 1 more turn to the single-worker time (move, clear, road, mine = 1+12+2+3) and do not subtract movement times later on.

(1+12+2+3) x 7 = 126 worker-turns (single-worker jobs)

126 / (4x38) = 85% efficiency (your scenario)

	26-05-2003 13:16									#3
	Support Apolyton

You can if you want to, depends on how you look at movements. When a worker is assigned a task, it is not a movement. A worker gets 1 movement point on regualr tiles, 3 moves on roads; but that does not translates to 3 worker tasks either.

To me, since we are improving "unimproved" tiles, the movement points it cost to get there shouldn't count in the equation. If we were to further improve upon tiles that were roaded, then you'd see that bonus come into play. That's when I wouldn't take away the movement points from the equation.

I look at it like two different situations. One where you must expend a movement point to get to your destination because no roads exist there. The other where roads are already present and you can move a worker to a point AND set him on a task.

It just depends on how you look at it.

Efficiency to me is not having worker's waiting around on other worker's to finish a job. Inefficiency is when you road a tile first , then clear it afterward. To me, it is not inefficient to move a worker onto a tile that was not previously roaded if that tile is to be improved upon by that worker.

	26-05-2003 13:26									#4
	Browse Apolyton AD-FREE

Oh, I agree with clearing first, roading next, of course.

The reason for having to keep the movement part of the equation is that if you are using 2 workers to improve 1 "virgin" jungle tile, you spend 2 worker-turns entering it (unlike with 1 worker, which uses 1 worker-turn only). Entering a new, unimproved jungle tile, is always a 1-turn job for a worker, no matter how you look at it.

Example (clearing 1 jungle tile):

1 worker ... 1x(1t moving, 12t clearing) = 13 worker-turns
2 workers ... 2x (1t moving, 6t clearing) = 14 worker-turns

efficiency: 13 / 14 = 93% (time: 7 / 13 = 54%)

	26-05-2003 13:30									#5
	Support Apolyton, buy Civilization III: Complete

BigFree, compare to time spent with 4 workers each clearing a single tile each, then roading them, then ganging to mine/irrigate.

Effeciency needs to be measured over longer terms, where you ask how much can 8 or 10 workers accomplish in 50 or 100 turns if using plan A or plan B.

Movement 'wasted' to move into the clearing tiles factors into it to a very great extent. Every additional worker moving into an unroaded tile over the first one, is an additional worker turn 'wasted'. With 10 workers, over 50 turns, that can add up significantly.

	26-05-2003 14:07									#6
	Lose 30 kilos (of popups)

The thing in what you are saying nye, is that you'd have to have a seperate work force working at the same exact pace as the other so as to not have any wasted moves. This would also be neccessary for the tiles to be used as soon as possible by the cities. I tried balancing the amount of workers used to the amount of cleared and ready for use tiles to the number of tiles that our cities can roughly use during a certain time period. There's no use in building too many worker's to clear tiles way ahead of time before they would be used. Tiles to be improved should be used soon after they are completed, esle we have to many workers. Imagine if we could keep pace with the number of tiles used by our Civ with 10 workers, but we had 16 workers instead. Those 60 shields could have went towards a Temple or a Barracks in one or two cities or something else that may be needed like a Galley. Not to say we have too many now, but just keep in mind how many will actually be useful in the here and now and the near future.



If I had a seperate group to gang up on certain tiles, after they had been roaded, the worker's may or may not lose "turns" getting there depending on the situation.

I though about having workers do the clear jungle then road thing, followed by another set of workers who'd use the road network to travel to the task site to either mine or irrigate. But since the tiles we want to work first are river tiles, unless you go completely on one side of the river before crossing over, you'd lose movement points crossing the river. So it became a moot point to try to account for that since I don't think we should only clear on one side of the river. If a tile has been roaded you save one movement point for every unit in a stack you are following it up with that is going to mine or irrigate that tile.

Getting the efficiency up is nice but not neccessarily at the cost of not getting the tiles improved in a timely fashion; for example: there are cities waiting to use those tiles.

So balancing the time in which we need tiles to be ready with the number of turns to get the job job with X number of workers (efficiency ratings go down with more workers) is important.

	26-05-2003 14:18									#7
	Support Apolyton, buy Civilization 2

Try using two to clear as a pair. Move. Clear. Road. 8 turns. Now do another one. Move. Clear. Road. 8 turns. Now move one worker back to tile 1 and mine, while the first mines tile 2. 3 turns to mine both tiles. 19 turns for two cleared, mined tiles. 2 worker turns wasted.

Ideally, we are clearing ahead of cities, unless we want them working jungle to start. Pairs are just about the best you are going to get in preserving long term efficiency while accelerating sinlge cleared tiles for new cities.

	27-05-2003 01:57									#8
	Help yourself to an AD-FREE life

I suggest using teams of 2-3 to clear the important tiles that cities will be using ASAP (as discussed primarily by Panzer32 and I).

With multiples teams like these working within 2 tiles of each other, there should never be any additional "Worker waste", since teams can help each other out when the Clear Jungle is done.

All other Jungle tiles can wait. It is most efficient to put just one Worker to clear these. The same Worker then Roads the tile. After that Mines or Irrigation can be rushed in certain areas, depending on the requirements of the cities.

This last past requires a ton of micromanagement; you need to know exactly which tiles the cities need 20 turns ahead, preparations must be made to improve these, and extra Workers need to be assigned to the remaining tiles, all the while being as efficient as possible with respect to Worker movement. If anyone has worked at doing this efficiently using the scenarios, they'll understand why I resigned as MoI...

I'm certain we're not going to do it as efficiently as possible. It just requires too much planning and micromanagement. I'm just hoping we do not just blindly double- and triple-up our Workers, when the improved tiles will not be needed for some time.


Dominae

	27-05-2003 02:04									#9
	Help yourself to an AD-FREE life

Another thing is how fast we need some of the jungle cleared. In some areas, very quickly, in others, we can take our time. The things is, that there is a trade off between speed and efficiency. We can be as efficient as possible and take a long time to do it in, or we can rip though it as fast as possible and be less efficient.

MoI E_T

	27-05-2003 02:08									#10
	Support Apolyton, buy Civilization III: Complete

quote: Originally posted by E_T Another thing is how fast we need some of the jungle cleared. In some areas, very quickly, in others, we can take our time. The things is, that there is a trade off between speed and efficiency. We can be as efficient as possible and take a long time to do it in, or we can rip though it as fast as possible and be less efficient. MoI E_T

I agree with ET 100%. That was the point I was trying to make somewhere up above, but I used too many words. You've sumed it up nicely ET!

	27-05-2003 02:37									#11
	Support Apolyton, buy Civilization: The Boardgame

E_T's comment is also exactly what I just said, except that I offered way of putting this idea into action.

It's nice to see that we're all on the same page on this one!

The main thing to consider is this: we need to plan for Jungle-clearing now. We do not want to be forced into a situation where we'll need to put 6 Workers on one Jungle tile because some city suddenly has not useful tiles to work. If we plan ahead adequately, we should be able to clear the jungle a lot more efficiently, since we'll know what we will and will not need in terms of improved tiles.


Dominae