Building on a mountain

[TimTheEnchanter]

My conclusion would be "In general, don't build on hills or a mountain in SP games because they really don't help your defense when the cities are all empty."

If the AI is attacking your city you probably already screwed up somewhere else.

 

[Ali Ardavan]

Ali - IMO your reasoning has errors.

2) You ignored "present value". A payment of 5+5+...+5 (11 times) is not worth 55, but more like ...

True, but irrelevant. Because we ignored present value throughout the entire discussion on this theory. If you have a city producing 5 shields a turn, it takes 8 turns to create a settler and meanwhile those shields do not produce any return. Same is true of everything one makes in Civ; there is no payout till you are done.

1) You have confused cost with value. For example, an obsolete wonder might cost 200s, but would have no real value (except for score). Likewise, the cost of a settler may not match its value. ... So, I still think the value of a settler's turn is closer to 5s than 2s or even 3s.

You do have a point here. I can use the settler to build a city on a forest and put my citizen on another forest thus getting 4s per turn indefinitely and save the settlers home city one food. So in effect (again using 1 food = 1 shield) a settler turn is worth 5s. Using this new value the investment is 55s and 5% of it is 2.75S which makes the mined hill (2-3 extra shields) barely worthwhile.

So my conclusion, which also matches my intuition, is a pre mined hill city is not worth it during the rapid expansion phase the same way irrigation, mining, and non essential roads are not worth it. Afterwards, it depends whether you are playing for early conquest or early landing. In early conquest where you may not even discover railroad and the time horizon is short it is not worth it. In early landing games where railroad is about half way through the tech tree it is.

 

[Ali Ardavan]

My conclusion would be "In general, don't build on hills or a mountain in SP games because they really don't help your defense when the cities are all empty."
If the AI is attacking your city you probably already screwed up somewhere else.

My defense concern is more about barbarians than AI but applies to both. In particular, my concern is when a ship loaded with 2-movement units shows up. These units can land and attack in the same turn. If you leave cities which are on the coast or 1 tile away defenseless you will lose some during the game this way. I do leave them defenseless early on during rapid expansion, but afterwards I put a unit in all such cities even if it is a diplomat or a caravan.

 

[TimTheEnchanter]

My defense concern is more about barbarians than AI but applies to both. In particular, my concern is when a ship loaded with 2-movement units shows up. These units can land and attack in the same turn. If you leave cities which are on the coast or 1 tile away defenseless you will lose some during the game this way. I do leave them defenseless early on during rapid expansion, but afterwards I put a unit in all such cities even if it is a diplomat or a caravan.

True - I am definitely exaggerating a bit, although I often leave even many coastal cities empty. Prior to republic, many may have units, but that's for the benefit of garrison. Defense is just a side benefit. Wonder cities and the capital usually have somebody nearby, but normal cities often don't get that benefit.

(Actually I'm probably also just too accustomed to OCC games where a dip or camel can successfully defend against a barb, and everybody else loves me.)

But going back to the "ROI" approach, consider the likelihood of a barb ship (1) approaching your city unseen (depends on your sailing patterns/blackness in ocean) (2) having 2-move units on board (3) landing close enough to disembark and attack in one turn and (4) actually carrying through with the attack when they land - which they sometimes don't. Take the likelihood of all those things happening and apply that percentage to the cost of bribing back the city from the barbs and building it back up to its previous size. Also use your calculation of the likelihood of it happening to determine an average timeframe to wait so you can determine the net present value of the investment required to rebuild your lost city if and when it ever happens.

Now compare that to the cost of building defensive units for your coastal cities. Note that these are current investments, so the net present value is 100% of the cost. If you're in republic, add sheild cost for any actual military units. Apply your likeliness of attack calculations to the possibility of losing your defensive unit forcing you to replace it at some point. Add those costs and compare them to the cost above.

I would think the present value of the reactionary-plan is almost negligible compared to the cost of actively defending your cities.

One more thing, back to the original topic: if you're defending your cities with dips, I'm not sure a dip on a hill is going to survive an attack any better than a dip on grass. I'd rather have him located back away from the point of attack to swoop in and buy back the city the next turn.

edit to add a bunch of stuff

 

[Ali Ardavan]

Good points, Tim. And for the most part I agree. But in mid and late game when my cities also have a lot of infrastrucutre recovering from the loss is quite costly.

 

[Peaster]

... consider the likelihood of a barb ship (1) approaching your city unseen (depends on your sailing patterns/blackness in ocean) (2) having 2-move units on board (3) landing close enough to disembark and attack in one turn and (4) actually carrying through with the attack when they land

This happened to me (with AI instead of barbs) in a recent GOTM when I only had about 4 cities, no dips yet, and no good response; so I quit. But that GOTM was artificial, a scenario really. It was a rare thing, and I agree that leaving cities empty is usually OK.

Ali - AFAIK I never ignored present value in these talks. It is one of the main reasons pre-mining is inefficient; you have to wait so long for the benefits to start. I agree with Tim that hills are usually not good sites in SP, but I assumed that you wanted to build on one anyway, and only debated whether to pre-mine or not.

BTW - Maybe one good use for hill-cities is the distant overseas colony, which some EC players use to pump out crusaders. They are very susceptible to AI sneak attacks when first built, and the city's growth+production is not really an issue, since the units there will be mostly RB'd. Of course, you would not pre-mine these.

 

[starlifter]

(Or my favorite, a gold mountain)

This is an excellent city building site, especially if you can set a miner loose on it before founding. It is single the most valuable resource location in the game.

As far as the SP discussion about mining a city on a hill/mountain, that's normally the only way I'll found such a city. I'll delay until a unit can do the mining.

The reason is because of shield multiples, not shield payback. In the life of a city, if it produces (at some point) say 9 shields, then I'm forced to buy another line of shields in a unit IRB. This is 25 gold. In late game, if the city is say 37 net shields (lets say gross, too, for discussion), then with Factory+Hydro, this means a mined hill would be producing 41 (40 for the first shield, then 41 for the 2nd from the mining, assuming post-RR). This allows me to support a unit, and save 80 gold in a city improvement IPRB (incremental partial rush buy). For example, the bank-stock exchange. Build the bank, don't touch production when complete, let the box fill with 41 (of 120), then buy the bank (now you have 120 shields, for 158 gold), then switch to Stock Exchange... now you save 80 gold and still complete the construction on the next turn, yet waste only one shield ... and (in this "common" example) it is due to the extra shields from the mining before foundation.

So if I found on a mountain or hill, I normally won't found until I can also send a miner, since once founded, you can't later change your mind and decide you need those shields and then mine.

 

[civdood]

mountain make unit powerful like behind city wall. so strong. but colossus give arrow for this? I want it so but think it not. Just imagine suprise of enemy attack when he hit musketman he who have 3+3+3+3+3+3 it total 18 to defend! no but wait! my infant tank he who have 6+6+6+6+6+6 when he sit in city wall on mountain top!! can cruise misslie dent my armor in such position? maybe nuke can win since even 36 can never match attqack 3 time more powerful, but may get lucky?

 

[starlifter]

by Civdood:
can cruise misslie dent my armor in such position? maybe nuke can win since even 36 can never match attqack 3 time more powerful, but may get lucky?

A cruise missile will dent you, but you will live. If your unit is veteran, and you have both SDI and SAM on a river , you will get: 6*2*2 + 0+0 + 6+6 + 3 + 3 = 42. This comes from SDI & SAM (the 6*2*2) and Walls (0+0) and Mountain (6+6) and river (3) and veteran status (3).

But against a nuke, the rules change... and no luck. If a nuke attacks your city, there are 2 outcomes:
1. Your city has SDI (or an SDI is in a city within 3 squares). The nuke is simply destroyed, 100% of time.
2. No SDI protection: Nuke detonates on target, destroying all units of all civilizations within a radius of 1 square from impact (e.g., the 8 squares around impact). You have no chance to defend, you always simply lose all units. If you could defend (but you can't), your unit would defend with: 6*2 + 0+0 + 3 + 3 = 18.

 

[civdood]

Civdood live short tiem on city on mountain, it work well and had big airdrome, it named Lapaz in Sud Amrica, in county of Bolivia maybe 4000m up in air, very tall mountain where air not thick at all. It hard to take city on mountain, very very good defender from invader! Civ 2 is so real to life.