A Theory Contest

[Peaster]

Can we assign meaningful values to different Civ2 resources such as shields, beakers and food? For example, many players feel 1 shield = 2 gold. This has been debated in the "Hurry Production" thread and several others. Various opinions have been offered, but some are vague and there is not much agreement.

I propose a contest. I will create a 4000BC save on a very simple map (grass + a few forests, no huts, no specials, no barbs, no AI). You each get a small city, a settler, monarchy and 100 gold. You have to

A) State your values or your theory of the position as simply as possible, and

B) Act on it.

You have a week to do part A). You can use the saved game below for practice. I will provide a similar save for the real contest. It is a 2.42 save, only 1 civ, with monarchy. I chose Deity this time, but may choose King next time for an even simpler game. For the same reason, there will be no tech advances. You must leave your science bar at zero, and you cannot use any civ2 tricks (but rush buying is OK).

Then I will play "Banker" and sell shields and/or food to the highest bidder before play begins. I will use Cheat Mode to revise the save for each player (assuming there are not too many !). You have to bid consistently with the values you stated in part A.

Then everyone plays to 1000BC. Again, your play must be consistent with your statement. I feel this should be done rather quickly, without massive micro-management, since the contest is about your values more than your playing skill. The winner is the player with the highest Mfg (shields per turn) in 1000BC. Or, we can use some other goal that the players agree on.

I'm interested to see if anyone can find the "right" values, and an optimal strategy, for this simplified version of Civ2.

[Peaster]

The save:

[Grigor]

Is size 1 settler a "civ2 trick"??

[I have personally decided that it is a conscious strategy for higher levels by the game designers on the basis of its being disallowed at chieftan level.]

And what is our goal at 1000BC since we will have no science? How will positions be evaluated?

Apart from my confusion, this sounds like great fun.

[Peaster]

Yes - I consider the size 1 trick to be a trick for this game. I want the game to be pretty simple, so that it may be possible to find the "right" strategy and the "right" values of things for this situation. But if most of the players want such a trick allowed, I'll say OK.

I suggest we evaluate the 1000BC saves by looking at the total production of the player's cities, measured in shields per turn. This is the Manufacturing statistic on the Demo screen. But if the players can agree on some other method, such as population, or gold, or some combination of stats, I'll OK that too.

[rjmatsleepers]

I like the idea. I'm not sure I understand the details yet, but no doubt it will become clear. For example will we be able to bid for shields each turn or only at the begining? Can we only bid for shields, or can we also bid for food?

RJM at Sleeper's

[Peaster]

You can only bid the first turn, mainly because selling you things will be a bit of work for me. With cheat mode, I can adjust your city's shield box, or city size, or create a food caravan to give you half a food box.

I was planning to start you with 100 gold, but maybe 200 would allow a wider variety of starts, and a clearer outcome.

After that, you will choose to place your workers on forests or grassland or ocean depending on how you value shields, food and arrows. You can also include strategic ideas into these decisions (for example, on the last turn all possible workers will be on forests to max your Mfg score).

[Grigor]

OK, so I did two quick trials and have uncovered some of the issues:

(1) My first try using straight ICS got to 57 shields. My second try, aiming to expand toward more forests earlier, got to 67. It seems that the challenging goal is 100.

(2) On the last turn, all workers are placed on extra shield squares. But it doesn't stop there. Without CB, unhappiness is rampant (and uncontrollable by martial law). By using Elvii one can avoid cities in disorder (it takes some serious micromanaging). However, when I maximize shields for the last turn, I dismiss the Elvii, and every city will riot next turn. So I get a theoretical shield output which is actually unusable. This seems to be what Peaster has in mind, although it sems on the face of it not to be a particularly practical result.

(3) Since growth slows to a crawl once the black hats come in, it is not certain that maximum exponential growth (ICS) is the optimum strategy. Creating a lot of settlers and then plopping them all down on the last turn seems to be an effective strategy

[Peaster]

IMO this unhappiness is an undesirable complication. So, I will probably change the settings to Warlord for the real game and/or give you the CB tech and/or HG. Any suggestions on this ?

[Grigor]

It keeps getting more complicated. Another interesting strategy is to build settlers and use join city to get to size 5 - then just Xin up through CB, Alph, and Mysticism. I will try that tonight.

I think both CB and HG would be useful. HG allows celebration which increases arrows.

What exactly is this all supposed to be a test of?

[Peaster]

Yes, please let me know if it gets complicated, because that is what I want to avoid for now.

Kramsib's thread about maximizing production and marginal values might provide some background for this one. Some of it is rather abstract, so I thought a concrete example would help us all understand it better.

For me, the basic question is whether it is practical to devise a value system to improve Civ2 decision making. It seems theoretically possible, but it may be too hard to do, and the answer may depend a lot on the situation.

I am hoping someone can find an answer, at least for this over-simplified game. It may not apply very well to real Civ2, but if this experiment has a positive result, we can try again from a more realistic start.

I prefer to remove all tech from this game, except monarchy and maybe CB, which will be given at the start. I'd also prefer not to allow celebration. The player should have just a few routine decisions to make - such as placing worker tiles, placing cities, production, whether to build roads, etc. So, part B) may be relatively boring, which is why I suggest a 1000BC ending date. But I think getting part A) right will be challenging, even for such a simple game.

[Peaster]

I am thinking seriously about giving you Fundy at the start, to eliminate unhappiness completely. This would make warriors useless (except for disbanding purposes). Any objections?

Here are my first thoughts about solving this position:

1) IMO the optimal growth strategy is to send a few Settlers from Berlin to the major nearby forests (no warriors, no roads). Then, the new cities mutliply in clusters around the forests. With Fundy, I got about 115 s/t in 1000BC this way.

2) A formula like "1 shield = 2 gold" means both sides of the equation are equally effective in pushing this strategy. It might be possible to deduce such a formula either from reasoning or from testing.

3) From reasoning, I figure these are about equally useful:

25 gold = 10 shields (2 NON-warriors) = 15 sheaves (1 food van)

4) I briefly playtested the effect of adding each of these 3 options into the 4000BC save. The 15 food seems a bit better than 10 shields, which is better than 20 gold (maybe because IRB'ing is hard to do). So, I do not have a final answer yet, but feel I am getting closer.

Disagreement is encouraged, of course.

[Grigor]

Fundy seems a good solution, but Chieftan level may be the same.
Also, chieftan does not allow the size 1 city trick.

Part of the reason you do not see good results from an extra 25g is that you already have 100g. From a base of zero, the effect of 25g should be vastly magnified.

Similarly, the food box seems unusually important because you are starting at size 2. With a size 1 city (and a food box value of 10 sheaves) things might (or might not) be more equal.

[Peaster]

Hmmm. I was thinking of increasing the 100g to 200g, to give the bidding stage more importance. But it is hard to spend over 100g efficiently with this start. Maybe the best remedy is to give you BW and Writing, to allow more IRB's. This probably means 1 shield = 2.5 gold, but maybe someone will find a reason it's not that simple.

Another possibility is to give you 2-3 starting cities, which would add some flexibility, but would also make a bigger game (more micro-management required). I suppose we could end the game at 2000BC in that case.

Grigor; I'm unclear on your food comment, especially "(and a food box value of 10 sheaves)" ? I assume you mean it is half full (rather than having a max capacity of 10), so that a food van would be worth only 10 food in that case. I guess that's true, but I doubt if it would afffect a ratio like "10 shields = 15 sheaves" much though.

[Grigor]

Oh, I can spend 200g just fine in this start. Rushing settlers the minute the cities hit a certain point seems to me to be the optimum strategy. That uses up a lot of gold from 2500 to 1500 BC. I guess there might be a question about whether to rushbuy the settler the instant the city reaches size 2 or to wait until the moment when the size foodbox reaches 18 sheaves so as to keep the city producing at size 2 the whole time.

****

You are correct - I meant that a food caravan has a value of 10 sheaves in a size 1 city.

Starting with 3 cities will just push the game further along - as if we start at 3300BC.

Having the 2nd and 3rd rows available for rushbuying will save gold.

All your suggestions are interesting, and they should have an effect on the marginal cost of shields, food, and gold. I am still unsure as to what valuations you are attempting to test in this context. For instance, I see no reasonable alternative strategy to ICS here, so I don't understand what decisions to make differently except how soon and how efficiently to rushbuy the next settler.

***

My basic strategy in either Fundy or Chieftan level would be to build on the no roads checkerboard pattern (cities every 2 squares on the diagonals) and gradually expand towards the forests. My prediction is that the 3 extra turns of city growth and production should lead to a larger civ in 1000 BC than using the forest squares earlier. Less gold might change that.

[Peaster]

Originally posted by Grigor
Oh, I can spend 200g just fine in this start. Rushing settlers the minute the cities hit a certain point seems to me to be the optimum strategy. That uses up a lot of gold from 2500 to 1500 BC. I guess there might be a question about whether to rushbuy the settler the instant the city reaches size 2 or to wait until the moment when the size foodbox reaches 18 sheaves so as to keep the city producing at size 2 the whole time.

Wow! There are so many plausible strategies even in such a simple position! My first thought was that gold should be spent on the first 1-2 cities ASAP. I changed my mind mainly because of the high rush-buy costs when the shield box is mostly empty. I still think we should spend it before about 3000BC.

Since you said you can spend 200g from this start, I will probably not include BW + Writing. This will make the trade-off between gold and shields more interesting.

Starting with 3 cities will just push the game further along - as if we start at 3300BC.

My practice game got a bit boring after 1500AD because there were so many cities. So, I may shorten the game a little. But if this affects your strategy, I will keep the game as is.

All your suggestions are interesting, and they should have an effect on the marginal cost of shields, food, and gold. I am still unsure as to what valuations you are attempting to test in this context.

Well, at the end of this contest, I want to know:

a) Do Civ2 theoreticians have anything useful to say to Civ2 players? (Of course, some of us try to wear both hats).

b) What kind of thinking leads to the best results in this simple game? Pure reasoning, trial and error, experience, or what?

c) Can we learn anything about Civ2 from such a simplified game? (for example, I am surprised that food is so valuable, even with so much grass around, and I expect this value transfers to normal Civ2).

BTW - You do not have to state your theory of the position using formulas like mine. For example, some kind of algorithm for bidding and playing might be even better.

For instance, I see no reasonable alternative strategy to ICS here, so I don't understand what decisions to make differently except how soon and how efficiently to rushbuy the next settler.

***

My basic strategy in either Fundy or Chieftan level would be to build on the no roads checkerboard pattern (cities every 2 squares on the diagonals) and gradually expand towards the forests. My prediction is that the 3 extra turns of city growth and production should lead to a larger civ in 1000 BC than using the forest squares earlier. Less gold might change that.

Again, a surprising difference of opinion! You may be right about this, but I am not so sure. I agree that City A near forests will produce a new settler 3 turns faster than City B with no forests. "A" may have no food, but its first settler will be better placed than B's. So, I am willing to wallk a typical Settler about 5 extra turns to get to forests. I feel this needs testing - or maybe more thought.