quote: Originally posted by Ribannah on 02-14-2001 02:33 PM Maybe you can tell me exactly what your approach is, or if you like, propose a map we will both play and we'll compare our games

Whoops, sorry. Got caught up in the moment there.

I fall in the settle soon camp. Early claiming of land is so important in my games. I play as a peaceful expansionist on small worlds so War4ever's note about city sites being at a premium is particularly true for me. My success is directly proportional to the amount of land I claim in the beginning (either through fast city founding or warriors penning in the AI). My experiences are similar to geofelt's, DaveV's, and Ming's - rapid expansion is best.

Rambling comments follow...

quote: East Street Trader I have not attempted a quantification but my instincts are that the volume of extra arrows, wheatsheaves and shields due to the work of an unsupported settler would exceed the net contribution of one city.

quote: Ribannah - but I can tell you now that such an approach would be worse than either of the extremes chosen in my sample runs.

Ribannah's point about hitting the unhappiness wall before the Hanging Gardens is well taken. This somewhat lessens the advantage of building cities at that point.

The fact that new techs double in cost over time means that soon the "keep the settler" approach makes you behind by one tech at most. However, early Monarchy is quite important (and you can never trade for it).

p.s. I'll have to try the "build a settler from your only city while it's size 1 yet not have it disband" trick (which obviously requires you to keep that "extra" settler at least a little while).

p.p.s. What's the etymology of your name, Ribannah? Like "rah", it sounds like something from the rabbits' language in the book "Watership Down".

quote: <font size=1>Originally posted by Ribannah on 02-14-2001 02:33 PM</font> You can count the size of your empire in any way you want, it doesn't make the expansion exponential.

quote: >An expansion is never exactly exponential: you must always suppose that the exponent varies within the time. By this definition ANY expansion would count as an exponential expansion, because you can always write its formula that way. In other words: the term exponential would give no information whatsoever. I don't think your definition is very practical!

quote: Maybe you can tell me exactly what your approach is, or if you like, propose a map we will both play and we'll compare our games

To choose a map, to publish it, and let play people up to monarchy. Any participant would publish saved games or pictures with a list of all actions (micromanaging of cities included) of all turns with very detailed explanation of decisions. Then we could discuss everything.

Of course, less detailed (and less time consuming) approach would be useful too.

The problem is I didn't start thinking about the best civ algorithm yet (In other words, I have no own approach). I will contact you later.
Of course, I am not afraid to be behind you in that comparison game now...

quote: Originally posted by SlowThinker on 02-14-2001 03:35 PM Will anybody tell me what "LOL" means? I speak seriously for once.

LOL = Laughing Out Loud ( )

quote: I agree. By my definition ALL expansions are exponential, by your definition NONE real expansion is exponential

quote: ... To choose a map, to publish it, and let play people up to monarchy. Any participant would publish saved games or pictures with a list of all actions (micromanaging of cities included) of all turns with very detailed explanation of decisions. Then we could discuss everything.

Exactly. Monarchy may be too early to stop, but we could make that decision when we get to it.

quote: The problem is I didn't start thinking about the best civ algorithm yet (In other words, I have no own approach). I will contact you later.

Slowthinker,
I read the post and it is not technically correct. Please read my explanation on page one of the this thread.

quote: Originally posted by Edward on 02-14-2001 03:16 PM I fall in the settle soon camp. Early claiming of land is so important in my games. I play as a peaceful expansionist on small worlds ....

On small worlds (with 7 civs) I am more inclined to build my second city right away, too.

quote: There is a point that's reached pretty early where building a road or two on your way to found a city saves time for future settlers to get to the borderlands. I can't imagine that a purist stance of never building roads until you've laid down all your cities could beat one that incorporated their prudent use.

Not ALL your cities, indeed! By the time you are researching Monarchy (or Republic) it is usually wise to have settlers improve at least some of the terrain.
If you kept your NONE Settlers, you will gain 3-6 turns during that stage of the game.

quote: What's the etymology of your name, Ribannah? Like "rah", it sounds like something from the rabbits' language in the book "Watership Down".

It's Elf, and translates into something like Herb Gatherer, quite appropriate for a Druid

Ribannah,

quote: Originally posted by Ribannah on 02-14-2001 04:28 PM Exactly. Monarchy may be too early to stop, but we could make that decision when we get to it.

quote:

quote: Originally posted by SlowThinker on 02-14-2001 07:11 PM That's not all: this is all my civ experience: 2 unfinished games of civ1 couple years ago (one to cannons, one to bombers), couple of starts of civ2 (max. to the beginning of A.D.).

You can appoint a champion to do the playing for you, if you wish

A. NONE settler builds 1 road, then builds city asap; produced settlers build 1 or 2 roads, then build new cities.
B. NONE settler keeps improving land; produced settlers don't improve land, but build cities asap.
C. hybrid solution: NONE settler builds roads in 2 or 3 land squares, then builds city; produced settlers do the same.
Except for case B, all roads are build in 'easy' terrain only (grassl., plains, desert, tundra).

Then I compared situations in 2000BC:
A. 4 cities (3 size 2, 1 size 1), 4 warriors, 1 phanlanx, 2 settlers on their ways, monarchy, but no complete connection between cities, 89 gold, 5 techs.
B. 4 cities (2x 2, 2x 1), 2 warriors, 1 NONE settler, despotism, all cities connected, more terrain revealed, better spots (whales) for future cities found, 71 gold, 4 techs.
C. 4 cities (3x 2, 1x 1), 4 war., phalanx in construction, 2 settl., monarchy, 1 road missing (in forest) for complete connection between cities, 93 gold, 5 techs, same revealed terrain as B.
All 3 cases had same income: 5 and research: 10

I was an A player, I think I will evolve toward C.

B is not, from my point of view, a good solution because:
- for many early turns, I had improved terrain that benefits nobody (inefficiency).
- The NONE settler lose time in movement, going from one city spot to the other.
- it has as many cities as case B and C, but the 2 last ones were only 2 or 2 turns old and had not even produced their 1st warr. yet... against barbs that are attracted by improved terrain, this can be dangerous.
- case A and C had 2 more settlers, close to founding 2 more cities.

As to the main debate, your point about immediate usage of the improvements, Dry, seems to me an excellent one.

If the "none" settler makes a road in a trade special square which is currently being worked he immediately adds one trade arrow per turn to the civ's productivity. If he merely irrigates a square which, one day, will be worked the lost production that an extra city would be contributing is almost bound to offset the delayed value of the irrigation.

What I think your case C may illustrate is that the 1 shield support cost per turn saved by hanging on to the "none" settler is not really a very significant part of the overall equation. The factors weighing heaviest in the balance are the immediately usefulness of the land improvements (whether for productivity, defence or other advantage) as compared to the production from one more city.

I note the terrain you used for your test. It is my guess that retaining the "none" settler may be a better option when it has been possible to found in a river rich area. I seem then to find a lot of good, immediately productive, work for him to do. Conceivably that would apply in a hill/mountain special rich area also, although I am less confident of that as you sometimes can't afford to work mined iron/coal early.

In my "trade/sleaze" version of ICS on Large/Giga maps I would always build with the second settler. The key to playing the game this way is to have both Gardens + Colossus in your capital. No settlers are built from the capital as it has too much work to do! If you want both the above wonders in the same city the Gardens must be started early.

So settler #2 really founds all the other cities! Only when I reach about 6 cities do I worry about roads. Usually you can find one city busting with food near wheat/fish which will produce a guy to do the roading. The roads must be on the way by the time trade arrives for easy transport of caravans.

If you play in different ways perhaps it is best keeping Settler #2 busy making terrain improvements.
However, too much detailed theoretical analysis is lost
in reality. (What former British PM Harold Macmillan referred to as "events, dear boy") In the action of the game we all know how much things can change resulting from the early discovery of our neighbours, the persistent arrival of Barbs or that Advanced Tribe discovered in 3750BC!

IMHO more important than using the slide rule over each and every possibilty is to have a series of short, medium and long-term goals. Micromanagement is fine, but sometimes you get so close to the coalface that all you see is coal dust flying in your eyes. Manage the global nature of the game as well!

After playing and enjoying many of the Succession Games it seems to me that too many games just bump along without a real purpose. (this is partly in the nature of succession playing) This is relevant when we come to look at the terrain improvements some players have seen fit to make with precious settlers. Often there is extensive mining in places which would starve if the mines were used. Over enthusiatic irrigation is very common, as well as putting a road through a "useless" square when a trade special is crying out for one!

Whatever you do with Settler #2, make sure you use him wisely!

-------------

SG(2)

In the past i have tried to keep the non settler around as long as possible. which means building terrain improvements that wont be used until much later. I think i will do better building a road or two, then starting 2nd city. Im sure there are strategies into which keeping the extra settler would fit, but i think for a diety beginner the start new city approach is easier, and less likely to tempt me into perfectionism (not that im doing ICS, at least at this point)

BTW, my current game (the one mentioned above as the Persians) is going well. After a long mediocre beginning and midgame, but with peace and steady expansion, i got SOL around 1800, went commie (from monarchy) went fundie when the AI started a war (in 1864) and made supreme by 1885. Its now 1905 and ive just gotten howies, and have a reasonable chance of winning, i think. Winning "ugly" to be sure (fundie and howies) but at this point i'll take a diety win anyway I can, and will try improved starting strats next time.

BTW ive also checked out matthew's beginners guide to diety. Any thoughts?

Lord of the Mark

quote: Dry: - for many early turns, I had improved terrain that benefits nobody (inefficiency). - The NONE settler lose time in movement, going from one city spot to the other.

Yes, that can be a problem for the NONE settler in the beginning. In my early game I keep all my cities very small so I can't do to much useful irrigating. And also the hand built settlers can build roads where needed along their various routes to the borders. The NONE settler must dash this way and that to get to the right place.

Your observation about revealed map area is also important to note. This is why building a bunch of warriors (in your size 1 cities) to explore in all directions is as important as making settlers.

quote: East Street Trader: What I think your case C may illustrate is that the 1 shield support cost per turn saved by hanging on to the "none" settler is not really a very significant part of the overall equation.

Unit support is really crazy in the early game. I think it's the one food the settler eats that's really the savings. In Despotism and Monarchy, units don't cost any shields until you get a number of units equal to your city size (Despotism). In Monarchy you get to support three units shield-free even in size 1 and 2 cities. I think that unless you're a Republic or Democracy - settlers never require shields for support.

In Ribannah's defense, that NONE settler will save you a food every turn (two food when you get an advanced government) throughout the whole game. Not sure what that adds up to, but in analyzing Dry's three outcomes, remember that outcome B, unlike A & C, includes that bonus NONE settler. Someone might want to make hybrid option D where the produced settlers help that sweaty NONE unit by founding some roads themselves (as they do in A & C).

I still fall on the settle early camp. New cities are just so valuable.