A Tale of Two Cities

[notyoueither]

Some of the discussions re health gave me pause. Hmm, I thought, is it really better to settle on the hill?

Well, easy enough to game it out.




The answer is that the hill is better for a very short time. Very soon the larger health limit and better use of food favours the fp city. By 1520BC settling on the start tile is a bit better and it goes from there.

In neither case did I get a religion. The AI goes for Meditation religiously (harr harr).

In the fp scenario the 7th pop (need +1 Ha, like from religion) can work a mined plains hill, giving it 15hpt while the hill city is stuck at 12hpt and the next pop has to work fp to break even.

Surprisingly, the two cities build roughly the same amount to 2000 and 1520BC. The hill city gets an extra skirmisher and one less warrior. These include the starting warrior.

     Plains Hill    Flood Plains
    2000BC 1520BC  2000BC 1520BC
Wa    2      2       3      3
Sc    2      2       2      2
Wo    2      2       2      2
Sk    2      3       1      2
Se  in 7  1+ in 5  in 7  1+ in 5

Tech rates were very similar as well. Got Priesthood from a hut in the fp scenario, and archery from a hut in the hill scenario (speeding up the hill scenario).

fp: Hunting, Mysticism, Meditation, AnHusb, Agriculture, Pottery, Archery by 2000BC with BronzeW and Writing by 1520 with IronW in 6

hill: Hunting, Mysticism, Meditation, AnHusb, Agriculture, Pottery, Archery (free) by 2000BC with BronzeW and Writing by 1520 with IronW in 3

In both cases we see horses and copper before moving the first settler, and iron before the second.

In the longer run, health resources would improve the situation for the hill city, but... we can't chop many trees, and we're farther from those that can be chopped.

OTOH, the hill city is safer and it does kick out some units sooner. It does a warrior and a scout in the time the fp city does a single warrior (if both are set to max growth).

What do you think? Safe short term on the hill or power long term on the flood plains? That's what it boils down to.

[Theseus]

I was just doing a Sim run on the fp myself... separately from nye's observations, note that moving to the fp only really costs one 'good' fp tile and one 'okay' irr plains tile, while gaining 3 forested tiles.

The loss of the city hammer doesn't "feel" all that different.

[Dominae]

I like the long-term site much better.

While it's nice to get an early-game production boost from settling on Plains Hill (site A), the Flood Plains site (site B) is superior if this game goes long, which it should.

Site A:

+Early-game production boost.
+Additional Flood Plains tile to work; eight total Flood Plains accessible.
+More River tiles.
-Lower Health cap.
-Fewer Forests tiles available.
-Poorer tile quality overall (i.e. two Desert Hills, ugh).

Site B:

+Highter Health cap.
+Greater overall tile quality.
+Better sharing potential with second city upriver.
-Fewer accessible Flood Plains tiles.
-Fewer River tiles.
-Takes time to come into its own.

I guess it's pretty close. If our Health resources end up being cattle or Calendar-enabled, a lot of site A's potential goes out the window. So I prefer site B. We had a nice look around, now let's return to the starting spot!

[Blake]

Gosh that is a tough call, +1 hammers vs -2 health... it's an easy call for a single city, but also of consideration is that the city will hold and pump workers/settlers at some point.. possibly for quite a while. +1 hammer is enough to offset -1 health for this purpose.

I suppose the thing to do is calculate how many turns it takes for SiteA to get to -1 health, that number of turns is the number of hammers we gain. Then calculate the point where siteA hits -2 health, that's when SiteB is profitable over SiteA.

If we spend a long time between -1 and -2 health, "holding position" and training settlers, then the sites are fairly equal but with SiteA we have the advantage of an extra unit or whatever.

OTOH if we blow straight through to the point of -2 health then SiteB is clearly and unquestionably better.

edit: It does look like SiteB is probably better, given the extreme ease of growing.

[Theseus]

Arguing for SiteA (hill) again, I don;t think it's a matter of when we reach -1 or -2 health, but rather of of the point at which food / hammer production slows unacceptably.

Running through Sims with the city governor on, I'd poprush when growth would take longer than 6 or so turns, making sure to use 3 pop in the process (Settlers, Barracks, and Library all do so, for instance). There's a lot of food available to push that slowdown back, however.

A calculable point in development as well... just not by me (bedtime ).

ps: Hope that made sense.

[nbarclay]

I think the key issue is probably how early we'll try to grow our empire. If the heuristics I normally use are right, the hill site works best if we start work on our first settler about the time the city reaches its health limit and then keep building additional settlers and workers to help our civ grow in size more quickly. In that scenario, once we reach that point, we get our military production from other cities - especially while waiting for them to grow to their health and happiness limits. If we do that, the health limit doesn't bite us all that hard because cities don't grow when they're building settlers and workers anyhow, and we'll get our additional growth from other cities instead of in our capital.

On the other hand, if we focus a lot on early units so that we don't go into a heavy settler-worker mode until we're past the health limit for the hill site, the lower health level really bites us. Under those conditions, the loss of health is transformed into a loss of production capacity for building settlers and workers (since the food component of settler/worker production is smaller), and into having to work fewer production-oriented tiles to get the same rate of growth.

So I think the question of where to settle is intertwined with the question of how we plan to use the capital.

[notyoueither]

The fp city is a single happy point away from another mined, river, plains hill.

Long term it's a slam dunk winner for whatever we choose to build in the cap.

[Blake]

You're right.
And that happy point will be EASY if we snag buddhism, if we don't then we'll have diverted to another tech and thus be further along the timeline elsewhere...

So at size 7 ... and with another health resource or a farm ... the city can work 3 plain hills mines and bring in a respectable 15 hammers/turn... we can do quite a lot with that.

[nbarclay]

Originally posted by notyoueither
The fp city is a single happy point away from another mined, river, plains hill.

Long term it's a slam dunk winner for whatever we choose to build in the cap.

That's looking at the capital in isolation. If we would leverage an extra hammer per turn in the early game into quicker expansion, the long-term advantage to the capital of building on a flood plain would have to be balanced against the long-term advantage we would get from expanding a little more quickly. That's a more complex tradeoff, but it only really comes into play if we would consider building our first settler earlier than you did in your comparison.

I'd like to tinker with the test scenario a little before we decide. I'd appreciate it if you could point me to it if it's already avialable somewhere, make it available for download (perhaps in the Smith's thread), or e-mail it to me. (If you don't still have my address from earlier, it's the same username I use here at hiwaay dot net.)

[notyoueither]

More looks would be great.

It's in the Forum Guide thread. Second post.

I put it there as I intend to keep it up to date.

[Blake]

In the mock map I guess the components are NOT the civs of our actual opponents? I assume this since there should be no way for them to be founding Buddhism that early... and it'll also change tech costs slightly, possibly.

edit: nm alexander just greeted himself to me. I guess the mockup map will get the proper leaders sooner or later especially once we start meeting them in the DG.

[nbarclay]

I ran into an interesting glitch trying to load the save. For some reason, the filter looking for a .Civ4SavedGame extension was case-sensitive, so the game didn't show up on the list of games that could be loaded until I changed the extension from all lowercase to the appropriate mixed case. That wouldn't be too surprising on a Unix-type system, but I'm surprised to see it on Windows.

[Blake]

Case studies on possible research and build paths would be interesting.

I'm trying out a whippign strategy which works like this:

Grow straight to size 8.
Whip a settler for 4 population, the overflow completes a skirmisher.
Whip a granary for 2 pop.

At this point we have a 10 turn settler+skirmisher pump, with a little tweaking we can whip for 4 population every 10 turns, this is a rate of ~15.5 hammers per turn (12.5 from whip, 3 from tiles) which is nicer than working the 3 mines (15 hammers/turn), except without needing to actually work or build the mines, our workers efforts can instead go towards cottages.
Since the whip rate is limited basically by the anger cooldown I can't see an advantages of having farms.

At some point we could whip in a temple (they cost 40 hammers for us, so can be 2-pop whipped with 20 carryover and another skirmisher - there might be overflow mangling tho :edit: there is, the overflow is only ~11).

The research is tight, but when we grow very large to size 6 on floodplains we get +50% commerce over what we had at size 1 and this does enable us to get our 5 big techs (Archery, BW, AH, Pottery, Meditation) at about the rate which we can do things with them.

Also when figuring out the exact mechanisms of getting 4-pop settler whips I noticed something I hadn't before - Angry people don't eat while settlers/workers are being trained! I think I'd suspected this before but definitely never made careful observation - not sure if this is useful or not tho since the excess pop can just be whipped into insta-settler.


Anyway...
For the "10-turn whip cycle" this is the state of the play at 1000 BC. Note I've researched Mathematics so CoL a bit before 1000BC would be quite doable. Maybe CoL would be a gimmie for us... noting it can't really replace the Buddhism gambit, since we do need priesthood, and we may as well be researching in that direction while waiting to grow and train our worker.
(Don't mind the Islam, I world-edited it in at the point where we would found Buddhism.)


Would 1000BC be a good year for comparison of the broad overall strength of different approaches?

[Arrian]

Interesting, Blake.

Hrm. I was pretty much sold on the plains hill (I *really* like the early turn advantage you can get out of the production boost), but I have to admit that Blake's "pump" idea looks intriguing, and furthermore even without that there is the health issue...

-Arrian

[Blake]

The main advantage of the plains hill is it basically gives us a couple of extra early scouts. Any other advantages of it are quickly eaten up by the lower health. So what price do we put on a couple of scouts running around our neighbours? It's one of those intangiables...

Economically there is no question that the FP site is the slightly stronger. But it may be that the presence of our scouts will be enough to change us from a "soft" target to a "hard" target. I also wonder what the others would think when our city hits size 8 in the top 5 cities...

There's no question our capital is going to be an appealing target, it's a droolworthy city site and will be enjoying expert management making it strong. If we let it get too economically strong it might become one of those "Something stupid" things that justifies early warmonger activity. (know the concept? That even in a friendly game your allowed to kill or maim people who seem to be relying entirely on the unspoken agreement to keep their assets safe from being plundered)

So lets see:
Smith's College: +2 health gives great economic advantages! Do we really want to see those ugly green faces in our star city for the rest of the game?

Spartan Academy: No-one ****s with skirmishers in a capital on a hill. But maybe we would rather watch our capital burn?


I did do some more test runs, about 4-5 test more, on the hill, off the hill. With religion, without religion. Pop-rush, no pop-rush. Really, it was pretty close all around, usually could have 3 cities founded. One conclusion I've drawn is going for Buddhism early really costs us nothing. We have great research by growing on the floodplains. Having religion is great, it's +1 free happy, and another +1 easy happy. This happy is useful for pop-rushing. But there are other options like Monarchy which we have plenty of commerce to get. But given that we need priesthood regardless (for CoL/Monarchy) we may as well be starting down that chain when we have nothing better to do - at the start. So it costs us little to try, and is far from game breaking if it fails. A good gamble.

Note: One other thing... if we do found on the plains, we probably may as well go Meditation THEN hunting, since it takes 15 turns to complete the first warrior. On the hill it only takes 8 turns and then we can start the scout.