Slingshots vs. Expansion: a comparison

[gus_smedstad]

Recently I did some investigating into the idea of executing the Civil Service Slingshot earlier than normal by generating a Great Prophet. Generally the limiting factor in a Slingshot is the time to researching Code of Laws, so perhaps a Great Prophet could hurry things along a bit.

After some experimentation, I discovered that Philosophical leaders need only build Stonehenge or a Priest in to get a Great Prophet in time. Non-philosophical leaders need both Stonehenge and and a Priest in order to get a Great Prophet before you can execute a normal Slingshot.

In this particular example game, I'm playing Gandhi on Monarch at Standard speed on a Continents, Standard size map. After successfully getting Civil Service in 1520 BC, I replayed it to execute a regular Slingshot. I discovered that with no nearby commerce bonuses and no food resources, I had to focus rather tightly to finish it before the AI finished the Oracle.

After succeeding with a regular Slingshot in 1240 BC, I continued the original game until 1240 BC for an apples-to-apples comparison, and then replayed the game from the start using a conventional expansion strategy as a control, again until 1240 BC.

Here's the starting starting save file:

[gus_smedstad]

Here's the starting position. Note that the only nearby special is Cattle, and since it's on a Plains tile, the maximum food is 3. Normally I want a 5+ food tile or a gold mine inside my starting radius to execute a Civil Service slingshot. In none of these examples did I manage to generate a Great Scientist before getting Civil Service.

[gus_smedstad]

Here's the timeline for the first game:

4000 BC: Found Delhi. Start researching Polytheism, begin building Stonehenge.

3480 BC: Discover Polytheism and found Hinduism. Start researching Bronze Working.

2880 BC: Discover Bronze Working, begin researching Priesthood.

2800 BC: Finish building Stonehenge, begin building Worker. 49 turns to a Great Prophet.

2600 BC: Discover Priesthood, begin researching Writing. 44 turns to our Great Prophet.

2400 BC: Finish worker, begin building temple.

2280 BC: Worker chops a forest to finish Temple. Begin Oracle. Assign Priest, which kills our growth, but decreases time to Great Prophet to 15 turns. 17 turns to Writing + Meditation, which is all we need for Civil Service, 17 turns to Oracle.

2000 BC: Discover Writing, begin researching Meditation. 8 turns to Great Prophet, 10 turns to Oracle, 10 turns to Meditation.

1800 BC: Remove priest and place worker in Grassland River tile for slight increase in research. 5 turns to Meditation, 6 turns to Great Prophet, 7 turns to Oracle.

1600 BC: Discover Meditation, begin researching the Wheel. Great Prophet in 1 turn, Oracle in 2 turns.

1560 BC: Moses born in Delhi, discovers Code of Laws. Begin researching Hunting.

1520 BC: Oracle completed, discovers Civil Service. Begin Warrior.

1500 BC: Warrior complete, begin Library.

1320 BC: Library complete, begin Settler. Discover Hunting, begin researching Archery.

1240 BC: Empire stats: 16 beakers / turn, 13 hammers / turn. First settler under way, 2 warriors.

[gus_smedstad]

Here's the timeline for a regular Civil Service slingshot:

4000 BC: Found Delhi, begin Warrior, start researching Polytheism.

3680 BC: Finish warrior, start another warrior.

3480 BC: Discover Polytheism, start reseaching Priesthood.

3440 BC: Finish warrior, start warrior. Army is now 3 warriors including starting unit.

3240 BC: Finish warrior, start warrior.

3200 BC: Discover Priesthood, begin researching Writing. Shift production to Oracle. 38 turns to Oracle.

2600 BC: Discover Writing, shift production to Library.

1840 BC: Finish library, hire a scientist. We can't afford 2 scientist because we have only the basic 2 food surplus. Shift 2 forest-grassland workers to river-grassland workers for the marginal extra commerce. 14 turns to Code of Laws, 15 turns to Oracle.

1280 BC: Discover Code of Laws, begin researching Hunting.

1240 BC: Finish Oracle, discover Civil Service. Empire stats: 16 research / turn, 10 hammers / turn. We're missing The Wheel, Hunting, Archery, Meditation, and Bronze Working, compared to the Great Prophet approach. We have 2 more Warriors, but don't have a Worker, and haven't started a Settler.

[gus_smedstad]

For my 3rd replay, I went for conventional expansion. This pays off in a slightly larger army, most of which is tied up defending the extra cities, and slightly higher production. Compared to the Slingshot approaches, there's far less technology. Because we have tundra to the south and desert to the north, there are rather fewer attractive city sites than normal.

4000 BC: Found Delhi, start Settler. Begin researching Polytheism. Beelining for Animal Husbandry may be slightly better for expansion, but we can't get it before we build our first Settler, so getting a religion sounds better.

3680 BC: Discover Polytheism, start researching Agriculture.

3120 BC: Discover Agriculture, start researching Animal Husbandry.

3000 BC: Finish Settler, begin Worker. Now we need to choose second city site. Here are our choices:

[gus_smedstad]

2920 BC: I found Bombay to the northwest.

[gus_smedstad]

2640 BC: Discover Animal Husbandry, begin researching The Wheel.

2400 BC: Finish worker in Delhi, start Warrior.

2360 BC: Discover the Wheel, start Pottery. We need some cottages to fund our expansion.

1960 BC: Discover Pottery, begin researching Writing.

1600 BC: Discover Writing, begin researching Bronze Working.

1520 BC: Found Madras. Unfortunately the most attractive city site is getting somewhat distant from our capital.

[gus_smedstad]

1280 BC: Found Bangalore. This site is awfully low on food, at least until we improve those Sheep.

[gus_smedstad]

When 1240 BC rolls around, we've got a couple of cottages, 4 cities, Warrior garrisons, and a couple of extra warriors fighting off Barbarians. After we balance the budget, we have a net research of 8 beakers / turn, half what either Slingshot gives us. We have a total production of 18 hammers per turn, a bit more than in the first game. We have Agriculture, Animal Husbandry, and Pottery, but we don't have Bronze Working, Meditation, Priesthood, Code of Laws, or of course Civil Service. The difference is about 1400 beakers of research.

- Gus

[Alkis2]

That's a great post. Thanks man.

[The_Paladin]

Indeed! Great post. Very informative.

I'll likely give this a try myself as I have a slightly different build order but yours sounds like it has some nice benefits.

[Alkis2]

I think however that your expansion variation is far too primitive. Expansion as I see it, doesn't mean you don't take advantage of your leader's special abilities. In your particular scenario you are Ghandi, who is Industrious, why not build the Stonehenge then?

Building the Stonehenge is a really really big move. I deliberately use the term big move (Go terminology) rather than good move (Chess terminology) because Civ4 has more things in common with Go rather than with chess.

So my idea was to use your big moves and combine them with an expansion strategy. Also, I wanted to build Stonehenge in my second city so that I don't "corrupt" my Capital with priest points.

With that in mind I played your save up to 1240 BC. I have 4 cities same as your expansion variation, but I also have Stonehenge. Prophet will appear in 29 turns. I have 2 barracks one axeman with two promotions and another two axeman on the way. I have library in my capital and two scientists (I have to be carefull though because I want that Prophet before the Scientist).

I also have 2 workers a scout and 3 warriors. My research is 15/turn and my hammers 23/turn. I have a long way to CS, -Polytheism, Priesthood, Meditation, CoL- but after that, I will use the Prophet to get CS. Compared with your first variation I am a lot behind in tech but you also are a lot behind in development.

If you see the save you will observe some mistakes, I missed the fish on the right for example.

Another thing to note is that your starting location is horrible. Desert, tundra, Ice, Peaks what else to ask

[gus_smedstad]

Originally posted by Alkis2
I think however that your expansion variation is far too primitive.

Probably true. As I was placing settler #2 and #3 I found I didn't really want to place them because Madras was too far away and Bangalore was going to be very poor until I had the Sheep hooked up.

In your particular scenario you are Ghandi, who is Industrious, why not build the Stonehenge then?

Because I never build Stonehenge. The attribute is middling at best, though the Great Prophet points are valuable, and it comes at a time when you really want to build other stuff and have very little production to spare.

On the other hand, I like the Oracle even if I'm not trying for Civil Service. I just didn't think it was enough of a contrast.

In all fairness, this start is pretty bad for just about anything. The terrain to the north and south is terrible, and the resources are badly placed. I did another example game using Arabia, but elected not to post it because the starting terrain was too good. I ended up with 39 beakers / turn in 1480 BC, a bit over twice what I managed with India.

Also, I wanted to build Stonehenge in my second city so that I don't "corrupt" my Capital with priest points.

I think priest points are OK. Really, I like Engineer, Scientist, Priest, and Merchant points nearly equally, though I still think Great Artists are awful. Priest Points can be pretty nice if you found 2 religions, as I did in the first two examples. Priests are pretty good specialists as well. They're not as good for your economy as Scientist specialists, but the extra hammer vs. a Scientist can be pretty nice.

My research is 15/turn and my hammers 23/turn.

I was really, really curious how you did that with this start. In the save game, you have no cottages and no financial resources. What you do have is 2 scientists in the capital with a library and the food to feed them (+9 research). It seems that the critical thing is that you moved Delhi one square west, which puts the Pigs in range.

This is a very, very good move. My main quibble with it is that how would you know that the pigs were there without the foresight of having seen earlier versions of the map? The initial image doesn't show the pigs, and there's no way to move your Warrior to expose them.

- Gus

[Arrian]

Presented with that start, I'd have seriously considered moving the settler 4-7, to the plains hill-on-lake (visible from the start). The other thing I probably would've done is gone for animal husbandry early to get pastures up early.

-Arrian

[gus_smedstad]

Originally posted by Arrian
Presented with that start, I'd have seriously considered moving the settler 4-7, to the plains hill-on-lake (visible from the start).

I wouldn't.

By and large I trust the initial placement of the map generator in Civ 4, compared to earlier games. Almost always there are a couple of good things hidden under the fog that are inside your 2-square radius. Of course, in this case I was dead wrong.

More significantly, moving East by North West loses a couple of grasslands, and there's not much food visible. Further, we can see a desert tile to the north, definitely inside the city radius from the 4/7 plains hill-on-lake.

With full foreknowledge, I agree it's a better initial city site. You get the +1 hammer, you get the Pigs, you get the wine eventually, and we don't use up a river tile. I just don't think you have enough information at game start to know how superior just about any spot to the west is to the starting point.

The other thing I probably would've done is gone for animal husbandry early to get pastures up early.

I'm a big fan of Animal Husbandry as well, but for the Slingshot starts it's an expensive tech. I actually did go that route when experimenting with the conventional CS slingshot, and I ended up losing the Oracle to the AI because I was delaying it while waiting for Code of Laws.

For the expansion start, I did get Animal Husbandry while building my first Settler, and building the Cattle pasture was the second thing I did. As Ghandi it takes some 24 turns to get Animal Husbandry, so making the Cattle pasture a priority means building a warrior or two and a worker, delaying the first Settler by about 15 turns.

On the other hand, if you've moved west and both pigs and cows are available, it makes sense to develop both before building a Settler.

- Gus