[SPOILERS] Bacchus and India's many gods

[Bacchus]

Here is my understanding of great people system:

There is a long list of great people, split by era and highly unevenly distributed across era and categories. The list starts with a GP in each category available for recruiting, drawn randomly from the set of GPs of the earliest era pertinent to that category -- e.g. the earliest engineer is medieval. Once a GP is recruited next GP is drawn in that category from an era equal to some calculated average era across all players. I don't know, and I'm not sure anyone outside Firaxis knows, how exactly the state is procced to the next era. There is no way to see who the next GP is, albeit you can at least guess at the bucket. Only admirals, generals and prophets have identical generic uses, all the others come with unique, although sometimes similar abilities.

The cost of GP recruitment is determined by era and nothing else. I don't have a schedule of costs handy, but it's not scaled by game speed or map size. 60 for Classical, 120 for Medieval (used to be 90 pre-patch, I think).

GP points are earned primarily through districts, district buildings, wildcard policies and district projects. Lacking points can be made up by faith and gold. There are equations out there on the interwebs, these are not obvious, I.e. it's not just X gold/faith per 1 lacking GPP. Significant boosts to GPP generation come from Divine Spark pantheon, Oracle wonder and some other minor sources.

GPs can have charges -- I'm not sure whether this information is displayed anywhere consistently, I tend to search the web to confirm. The actual GP screen just tells you what they so, but not how much, which can make a GP impossible to appraise.

What else? GPs can't be killed, so can be cheesed as superscouts, especially as they have four move. You can skip a GP that you accumulated enough points to recruit, and get a discount on the next one, but as mentioned above its a bit of a cat in a bag.

[Bacchus]

The number and age of the great people is fixed for a given number of players

I don't think this is the case, but it's also really hard to test.

Actually, there are whole lot of scenarios that it would be good to test as a community. Pretty difficult to organize this though.

[Bacchus]

If we go for Wet, I may have to switch my choice to China, the payoffs to extra Builder labour become big, whilst stepwells become relatively pointless.

[Bacchus]

Eh, looks like China it is. To make it more fun, it will be Four-Charge China, in honor of the base Builder action allowance. I will constrain my wonder construction this game to just 4 wonders. The two obvious ones are Pyramids and Oracle, the other two will probably be chosen from among: Petra, Forbidden City, Colosseum, Ruhr Valley, Oxford University, Great Zimbabwe, Terracota Army. Not sure we will ever get to the later-game ones to hold construction for them.

[TheArchduke]

Eh, looks like China it is. To make it more fun, it will be Four-Charge China, in honor of the base Builder action allowance. I will constrain my wonder construction this game to just 4 wonders. The two obvious ones are Pyramids and Oracle, the other two will probably be chosen from among: Petra, Forbidden City, Colosseum, Ruhr Valley, Oxford University, Great Zimbabwe, Terracota Army. Not sure we will ever get to the later-game ones to hold construction for them.

Shame, I was looking forward to Varu action.

[gunnerxtr]

They are back to a standard rainfall map setting so hopefully that swings him back to India?

[Bacchus]

They are back to a standard rainfall map setting so hopefully that swings him back to India?

Yep.

[Krill]

What is OLORAM?

Also, how do you value tile improvements given constant cost increases to builders, and all the different adjustments to the number of charges, costs and abilities of the builders?

[Bacchus]

What is OLORAM?

Also, how do you value tile improvements given constant cost increases to builders, and all the different adjustments to the number of charges, costs and abilities of the builders?

OLORAM is the abbreviation for "Our Lady of the Reeds and Marshes", one of the pantheons you can pick after you accumulate 25 faith, which normally happens via the God-King economic policy around T35. All pantheons have global effects for your empire the instant you pick them (you don't "spread" a pantheon), and this one gives +1 cog to all tiles which are floodplains, marsh or oasis. Normally these tiles are just 3 food, and food is relatively useless, but with OLORAM they became strong. A 4 foodcog yield tile is what you normally expect post-improvement, for example grass hills are 2-2 with a mine. You also get these 4-yield tiles in case of a forested hill, for which reason forested hills are much in value, they just give you what you would normally need a builder for.

Improvement valuation is very dynamic, there is no spreadsheet approach here of a sort that could help in Civ4. Most importantly, improvements frequently give Eurekas, so beakers and culture. Some gains are indirect, for example for Civil Service civic you need to grow a city to 10, something that we will likely be doing through stepwells, as the greatest barrier to that growth is housing. It's all made easier by the Builder labour generally being scarce, you normally have very pressing things you clearly need a builder for and valuation is by weighing these against other very pressing things. Especially because of the policy dynamism the decision-making is very tactical temporally, but empire-spanning geographically -- once the need is big enough, switch into Builder-oriented policies, at which point you might as well create some Builder labour for marginal needs.

[Bacchus]

On re-reading the above, I realise that "valuation is very dynamic" sounds pretty close to "I don't have a clue", so let's look at an example. People tend to build at least one builder pretty early on. Why? Because Craftsmanship civic is boosted by having three improved tiles, the boost is worth 20 clefs, so in addition to the improvement gains you get that. How much culture is 20 clefs? Well, the early culture-producing building, the monument, gives 2 clefs per turn, so that's like having 10 turns of the monument. Monument takes 60 cogs to build. I'm not sure what the discount rate is on culture, but I would say present value of all the culture the monument would produce is somewhere between 40 and 60 clefs. By this comparison, the Builder provides a hands-down better return on investment in culture terms than the monument, so you would never build a monument before the builder.

That kind of analysis takes care of builder vs infrastructure dilemma, but choosing between builder and other units is much harder. Except where it isn't. For example -- Builder or Trader? If you are about to start researching Currency, almost definitely Trader. If Trader procs an envoy to a currently uninfluenced City-state -- probably also Trader. In some, very rare cases, you would actually get to compare Trader to Builder on their own merits, i.e. the extra yields, rather then incidental bonuses, and here Builders tend to win out in the early game on food and cogs (especially after Apprenticeship), but lose out on gold, science and culture.

Hopefully this outlines my thinking a little better. Then there are chops and harvesting, which I will discuss a little later.