No announcement yet.

Quantitative Analysis of Civ Traits

  • Filter
  • Time
  • Show
Clear All
new posts

  • vmxa1
    I want to say that Scientic value looks good, but in reality is much less. This is due to the large percentage of the labs and university. These are not all that commonly built. Labs are likely to be in few cities as they come so late in the game. Universities will be some, but not most cities and in fact even libraries are not in every city. This really blunts its value, although I still like it.
    I see expansonist as one of the weakest, extra value from huts, what does it mean, 50 gold instead 25, so what. Scout is of some use.
    Military is not bad, but can be done with out.

    Religious giving temple discounts is useful.
    Now the best two IMHO is Commercial and Industry.
    Industry is the best to me, as faster worker will pay off in so many ways and forever. Quick roads to get workers/settlers/troops going, extra shields for large cities to speed production.

    Leave a comment:

  • randomturn
    Interesting stuff. I think nato's quantitative evaluation is useful to help understand how good a trait is for building purposes. It's important to remember though that building is only a part of the game, and not the most important part. Grim Legacy is right that the Frendch are a great building civ. I would rate militaristic and religious very high in some strategy areas that cannot be quantified. They are particularly strong at higher levels.

    Leave a comment:

  • Ray K
    Originally posted by Grim Legacy
    Another nice effort. Excellent!

    Now for some bragging on my part: with my infallible Civ-intuition I had already chosen Commercial and Scientific as the best traits -playing the French.
    I was thinking the exact same thing -- except I was playing the Greeks

    Leave a comment:

  • sophist
    french are not scientific + commercial

    French are commercial and industrial.

    Leave a comment:

  • Dis
    interesting indeed

    although it hasn't changed my views . I still love my aztecs.

    but seriously I like military (it is a significant part of my game). Religious is just convienient, not a necessity. I just hate anarchy , and don't feel like f*cking around with it. I tried the romans, but got wiped off the face of the planet with them.

    Leave a comment:

  • Grim Legacy
    Another nice effort. Excellent!

    Now for some bragging on my part: with my infallible Civ-intuition I had already chosen Commercial and Scientific as the best traits -playing the French.

    Leave a comment:

  • Sevorak
    Great post. An excellent effort.

    One thing though. When you calculate Industrious' worker bonus as more workers, and Militaristic's unit bonus as more units, shouldn't you factor in the maintenance cost you save as well?

    Just a thought.


    Leave a comment:

  • nato
    started a topic Quantitative Analysis of Civ Traits

    Quantitative Analysis of Civ Traits

    One of the most important decisions is made before the game even starts: what civ traits to take. This post attempts to quantify the 6 traits and find which are best.

    Two warnings:
    it is very long
    it has to make some assumptions and approximations

    That said, I do think it is worthwhile if you are interested.

    * If you hate long stuff, skip to the bottom where the nice list is. *

    I tried to find the total value of each trait. Generally that value comes down to gold or shields.

    You can buy shields at a rate of 1 shield for 4 gold, or you can produce gold at a rate of 1 gold for 4 shields (after economics). Therefore, I am counting 1 gold as of equal worth to 1 shield. When added together I will call these gold/shields.

    Tech advances cost beakers, which is essentially gold. I have used the info found by regoarrarr and David Weldon to calculate the gold cost of advances. I have used 24 as the base number.

    The conditions of the game are very important. I am guessing the game is 400 turns long on Regent / Monarch. I am assuming a large continents map. I am assuming you eventually get 25 good, low-corruption cities. When it matters, I assume this averages 12 cities of size 6 during the first 100 turns, 12 cities of size 12 and 12 more cities of size 6 the second 100 turns, and 25 cities of size 12 the last 200 turns.

    Finally, it is very important that things that come later are less valuable than things that come sooner. Things that come late are discounted in value. How much they are discounted counts on your play style! If you are a rusher, late is useless. If you plan to win late, late coming advantages are discounted less.

    Without knowing Civ3's interest rate, I have made no accounting for time. It is very important to remember some of these shields and gold are coming much later, and so are much less valuable.

    Ok now the good part:

    1. Commercial

    a) Corruption: No one seems to know how much corruption is reduced. It seems the effect is rather low, maybe 5%. However, in Civ there are no percents, there is either a whole coin / shield saved or not. My guess is corruption is reduced by about 1 coin per city. If you have 12 cities first 100 turns, and 25 last 300 turns, this is (12 gold * 100 turns) + (25 gold * 300 turns) = 8700 gold.

    b) Extra Commerce: You get 1 extra gold for size 6 cities, 2 for size 12. Using what I described above, this is (12 * 100) + ((12 * 100) + (12 * 200)) + (25 * 200) = 9800 gold.

    A lot of the benefit of both bonuses comes late game, so keep that discount in mind.

    So the total value of Commercial is 8700 + 9800 = 18,500 gold.

    2. Industrious

    a) Extra Shields: This is identical to the extra commerce in commercial above. Therefore it comes out to 9800 shields. However, extra shields are often irrelevant. Sometimes a city producing one or two less shields will take exact same time to build something. Therefore some of the bonus is lost. It is impossible to know how much. On a guess, I'll reduce the value to 8500 shields. Further, a lot of these shields come very late.

    b) Faster Workers: Having double speed workers is somewhat the same as getting a free worker when you build one. What is the value of 1 worker? It costs 10 shields and one population point. City growth would replace the population point; say growth averages 10 turns. Ignoring food, say that pop point would make 1 gold and 1 shield each turn during those 10. Therefore the total cost of a worker is 10 gold and 20 shields. If you make 40 workers in a game, then double speed workers is worth 40 * ( 10 gold + 20 shields) = 400 gold + 800 shields, or 1200 gold/shields.

    So the total value of Industrious is 8500 + 1200 = 9700 gold/shields.

    3. Religious

    a) 1 Turn Anarchy: This saves you the production otherwise lost during anarchy. Non-religious anarchy seems to take 4,6, or 8 turns, so average 6. Religious anarchy takes 1 turn, so each government change saves 5 turns of production. I am assuming 2 government changes, 1 early and 1 later. I looked at a 25 city game and added up the commerce and shields made at an early time and a late time. Early (which wasn't 25 cities yet) was about 30 commerce and 30 shields each turn. For 5 turns that adds up to 150 gold + 150 shields. Later was about 300 gold + 275 shields. For 5 turns that adds up to 1500 gold + 1375 shields. Therefore the anarchy bonus adds up to 1650 gold + 1525 shields, or 3175 gold/shields. More or less revolutions will change this value, as will size of your empire.

    b) Half Price Temples and Cathedrals: Ah so straightforward.
    25 Temples saving 30 shields each = 750 shields
    25 Cathedrals saving 70 shields each = 1750 shields.
    Total bonus = 2500 shields. Due to time discount, those temple savings may well equal the cathedral ones.

    So the total Religious bonus is 1650 gold + 4025 shields, or 5675 gold/shields.

    4. Scientific

    a) Three Free Advances: Advances cost a number of beakers = tech cost * base number. I am using base = 24 and picking the lower tech cost advances, since that seems to be what the computer gives you. Also I am ignoring the effect of other civs having the advance. Therefore the value is (24 * 26) + (24 * 84) + (24 * 180) = 624 + 2016 + 4320 = 6960 gold.

    b) Half Price Libraries, Universities, and Research Labs:
    25 Libraries * 40 = 1000 shields
    25 Unis * 80 = 2000 shields
    25 Labs * 80 = 2000 shields
    Adds up to 5000 shields.

    Therefore the total value of Scientific is 6960 gold + 5000 shields, or 11960 gold/shields.

    5. Militaristic

    a) Higher Chance of Promotions: This has two effects. First is that your units have more hit points. Having an elite spearman instead of a veteran one is somewhat like having 20 % more spearmen, since you go from 4 to 5 hit points. This is therefore somewhat like having 25 % more of the shield cost of your promoted units. Say during each of the 4 ages you get 10 such promotions, and the shield costs of the units for each age averaged 20, 40, 60, and 100 for each age. Therefore the value is (10 units * 0.25 increase * 20 shields) + (10 * 0.25 * 40) + (10 * 0.25 * 60) + (10 * 0.25 * 100) = 50 + 100 + 150 + 250 = 550 shields.

    The second effect is getting more leaders. Say militaristic civs get an average of 2 more leaders than normal civs. These will be used to construct wonders, for say a value of 600 + 800 = 1400 shields. Therefore the total value of higher probability promotions is 550 + 1400 = 1950 shields. Obviously if you fight more the value goes up.

    b) Half Price Military Buildings: Normally not all 25 cities get all these improvements, so some are for less than 25.
    25 Barracks * 20 = 500 shields
    10 Walls * 10 = 100 shields
    20 Harbors * 40 = 800 shields
    20 Coastal Fortresses * 40 = 800 shields
    25 SAM Batteries * 40 = 1000 shields
    10 Airports * 80 = 800 shields
    Total value is 4000 shields.

    Therefore the total value of militaristic is 5950 shields.

    6. Expansionistic

    a) Free Scout, Can Build More: The free Scout is valued at 10 shields. Early on, building Scouts instead of Warriors gives about double the exploring power. Therefore each Scout built is worth the price of an additional Warrior, or an extra 10 shields. Say 4 extra Scouts are built, this is worth 40 shields. Total value of Scouts is then 50 shields.
    It is important to note the intagible value of knowing the map earlier.

    b) Better Huts: This is the most random factor of all the traits. My worst expansionistic game gave me 1 Warrior and 1 tech advance. My best gave me 3 Warriors, 25 gold, 1 map of region, and 10 tech advances. What is a good amount for an average take? My arbitrary guess is 1 Warrior, 25 gold, and 4 tech advances. The Warrior is 10 shields and 25 gold is, indeed, 25 gold. The tech advances will all be low end ones. They will be worth perhaps (24 * 4) + (24 * 4) + (24 * 5) + (24 * 6) = 96 + 96 + 120 + 144 = 456 gold.

    Therefore the total value of expansionistic is 50 shields + 456 gold, or 506 gold/shields. Keep in mind this is all right at the start, so there is practically zero discount due to time.

    Total Values in Gold/Shields

    1. Commercial = 18,500
    2. Scientific = 11,960
    3. Industrious = 9,700
    4. Militaristic = 5,950
    5. Religious = 5,675
    6. Expansionistic = 506

    A lot of assumptions and outright guesses were made to produce this list, but they are all pretty reasonable assumptions. They must be kept in mind though; I am in no way pretending these values are at all definitive.

    Also, once again remember, a lot of these bonuses, especially Commercial coins and Industrious shields, are coming much later; some of them too late to affect the game at all.

    This said, if these numbers are anywhere near accurate, they are useful in trying to decide which traits are best.

    Thanks for reading!