A Tale of Two Cities

[notyoueither]

Damn!

[Blake]

No worries.

[Aginor]

Originally posted by Blake

In my hill trials I always cottage every floodplain and have had no growth problems.

Give it a go sometime with the farms - I find that it's surprising how much of a difference growing a bit faster can make in the extreme early game out of an FP start. Can always blow up the Farms later, too.

I'm thinking +5 culture/turn border expansions from religion to grab whatever's down in the Tundra, long term. That looks like some nasty turf down there.

I'm startled that we ended up the principals in this debate - we usually think alike! Your work in the long-term planning thread is gorgeous, by the way. I think we have a head for Smith's College. Wonder what American CEOs would say to a 2000 year business plan...

[Blake]

Give it a go sometime with the farms - I find that it's surprising how much of a difference growing a bit faster can make in the extreme early game out of an FP start. Can always blow up the Farms later, too.

I did use farms in some trials (and i didn't mean that as a generic statement on all floodplains). The thing about the hill start is that (thanks to the sheep) it has a healthy and permament +5 food surplus, so another +1food is only a 20% increase in growth.

The FP start actually has less food because it sort of needs to work the Mined Sheep.
So at size 4 it has +3 food from the floodplains, -1 from the sheep and the +2 from the city, only +4. At size 3 (working the sheep) it's only +3. These numbers are definitely low enough to benefit from a farm, so that's why I think we should make the mine and then a single farm.
The farm can then be loaned to the 2nd city.

[Blake]

Wonder what American CEOs would say to a 2000 year business plan...

Probably something about losing sight of things that really matter, like the next quarters profits....

[nbarclay]

Originally posted by Blake

The fact is, on the hill the point of break even (+1 hammer -1 food) is initially at 3 pop, that is when we'll probably stall to train a worker. The pasture then raises the break even point to size 4, which would be a good point to train a settler. So it's only after training this worker and settler that the hill site actually starts to lose out, and a health cap of 3 is quite ideal for whipping down from 6.

I know we've already founded our city, but I'd like to continue this discussion a little longer for future reference.

Without a pasture, the hill site actually reaches its maximum food+hammers (without irrigating flood plains instead of cottaging them) at size 2. Any growth beyond that is merely a break-even proposition, so it would make sense to go ahead and start building a worker at size 2.

Even with a pasture, the hill site reaches its maximum food+hammers at size 3. So at size 3, the hill site has one more food+hammers than the flood plains site because of the advantage of settling on a hill. At size 4, the two sites are even in food+hammers because the hill site loses a food to health. At size 5, the flood plains has one more food+hammers than the hill. And with a pasture rather than a mine for the sheep, the flood plains could have two more food+hammers at size six.

So if the cities would oscillate between size three and size six, always growing at the same rate and spending three population on a pop rush practically the instant they reach size six, their sustained food+hammers potential should be virtually identical. That would give the hill starting position, with its extra early hammers, a clear advantage.

However, I seriously doubt that the timing would work out to always grow at the maximum rate and pop rush three pop points the moment the city hits size six. To the extent that it doesn't, the flood plains site can get an advantage in sustained food+hammers by focusing more on growth tiles at smaller sizes and more on production at larger sizes, thereby spending more time at a size where it has a food+hammers advantage than it does at a size where it has a food+hammers disadvantage.

The big question is whether or not that advantage in sustained output can make up for the advantage the hill site offers until the cities grow enough for the flood plains' health advantage to kick in - especially since early scouts would give us an advantage in popping huts and let us found the second city sooner. (On the other hand, earlier scouting would block a religion beeline, increasing the risk of being beaten to our targeted religion, with the costs associated with that.) I don't really know how that balances.

I'll confess that the hill city comes out of this analysis looking a whole lot better than I'd expected it to, and I wish I'd thought to do this kind of analysis earlier, when there was more time to take it into consideration when we were deciding where to settle. In any case, whether we made the ideal choice or not, it's been a very interesting learning experience - and learning is, after all, is one of Gathering Storm's most important goals. Thanks, Blake, for pointing out a strategy that my focus on the advantages of bigger cities blinded me to.

[nbarclay]

Originally posted by Blake

The FP start actually has less food because it sort of needs to work the Mined Sheep.

The only compelling reason I see to mine the sheep is the problem of getting Animal Husbandry by the time we'd want to improve the tile. Mining the sheep instead of building a pasture is a break-even proposition in food+hammers. But if we want a break-even proposition in food+hammers, we could get it by working the plains hill forest instead of a flood plains part of the time, and/or by miniing the desert hill in addition to the plains hill. Using those tiles to trade food for hammers would allow maximum growth at smaller sizes (aside perhaps from a brief detour to get scouts out a little sooner) and a focus on production at larger sizes. That's the best way for the flood plains start to get an overall production advantage over what the hill start provides. So average sustained output ought to be a tiny bit better with a pasture.

The catch is, if we can mine the sheep tile earlier than we could build a pasture, that lets us get a tile with +5 food+hammers earlier. Since our other tiles are effectively +4 (albeit with flood plains each trading one of their potential food for a cottage), having a +5 tile earlier would bring some advantages.

[Theseus]

Originally posted by nbarclay
Even with a pasture, the hill site reaches its maximum food+hammers at size 3. So at size 3, the hill site has one more food+hammers than the flood plains site because of the advantage of settling on a hill. At size 4, the two sites are even in food+hammers because the hill site loses a food to health. At size 5, the flood plains has one more food+hammers than the hill. And with a pasture rather than a mine for the sheep, the flood plains could have two more food+hammers at size six.

So if the cities would oscillate between size three and size six, always growing at the same rate and spending three population on a pop rush practically the instant they reach size six, their sustained food+hammers potential should be virtually identical. That would give the hill starting position, with its extra early hammers, a clear advantage.

Ah, but doesn;t the advantage go further to the hill site by virtue of just one irr fp?

Excellent reduction of the entire argument, btw.

[Blake]

Without a pasture, the hill site actually reaches its maximum food+hammers (without irrigating flood plains instead of cottaging them) at size 2. Any growth beyond that is merely a break-even proposition, so it would make sense to go ahead and start building a worker at size 2.

Well growing to size 3 does give +1 commerce, so it's slightly better than break-even (this commerce does add up too). The main reason though is with the religious tech there's nothing for a size-2 worker to actually do, given that the hill site doesn't need mines. Growing to 3 can also be thought of as a head start on growth for when the pasture gets hooked up.

The only compelling reason I see to mine the sheep is the problem of getting Animal Husbandry by the time we'd want to improve the tile.

But given how fast the FP city can grow there is considerable oppurtunity cost in delaying the BW+Pottery techs. So I would say there is compelling reason to not get AH.
I quite often decide to not research an entire tech for a single resource, even something good like cows, as long as there is reasonable alternative employment for the workers. So given free worker time and free science, yeah, that pasture makes sense , but alas many thing vie for the attention of our scientists and workers.

The catch is, if we can mine the sheep tile earlier than we could build a pasture, that lets us get a tile with +5 food+hammers earlier. Since our other tiles are effectively +4 (albeit with flood plains each trading one of their potential food for a cottage), having a +5 tile earlier would bring some advantages.

Exactly. The sheep is a really good tile mined.

The main reason why a mine is needed is because while whipping is nice, in this case it's more limited by happiness than growth (and the fact it's only really possible to whip in 1 or 2 things every 10 turns). So we need another way to convert some of the surplus food into hammers, and a mine is the solution. Then it's simply a matter of putting the mine on the best tile, which happens to be the sheep.
The mine will eventually get replaced of course, but it should stand for about 25 turns, which more than pays for replacing it.

Also a pretty simple comparison:

FP Cottage 3-0-3
+
Sheep Mine 1-4-0
= 4-4-3(+development)

Grassland Hill Mine 1-3-1
+
Pasture Sheep 3-2-1
= 4-5-2

I'd be inclined to take the +1 commerce and cottage development over the +1 hammer.
Of course once the health becomes a factor the pasture gets considerably upgraded.

[vmxa1]

I am getting a headache trying to keep up with it all. It looks very close, but the one thing that could tip the scales is the defense from the hill.

If you need to count votes now, then I would side with Nathan as I have absolute faith in his logic. If the vote will not be done yet, then lets wait for more comformation and possible agreement.

[nbarclay]

Vmxa1, we've already built our capital on the flood plains, so the continued discussion of how the city sites compare is purely a learning experience that could potentially be useful in future games.

Please don't put absolute faith in my logic. My skills in applying logic are strong, but sometimes I discover blind spots where I've been basing my logic on incorrect assumptions, and Civ IV is new enough that I still have significant blind spots. When logic is applied based on false assumptions, it can produce perfectly logical conclusions that happen to be dead wrong. Garbage in, garbage out, as the saying goes.

The hill start is actually a very good example of a blind spot in my logic. Once Blake pushed me to do a more detailed analysis, I realized that a pop rushing strattegy operates under rather different parameters from a conventional building strategy, and I hadn't been making adequate allowance for the difference. Now that I've discovered my blind spot and done a more detailed analysis to adjust for it, I honestly don't know whether we made the right choice or whether building our capital on the hill would have been better; I'd have to do some very careful testing to clarify the picture.

[nbarclay]

Blake, I'm inclined to agree that mining the sheep is better than waiting to build a pasture. But comparing building a mine on the sheep in a flood plains start with building a pasture on it in a hill start is something of an apples-and-oranges comparison. The ability to build a mine instead of a pasture, and thereby to delay researching Animal Husbandry and turn the sheep into a +5 tile sooner, is a something with added value that a flood plains start allows us to do, not something that takes away value that a flood plains start forces us to do. Or if we don't view it as adding value, we don't have to do it.

[Theseus]

Ummm... and not see horses.

[nbarclay]

Originally posted by Theseus
Ummm... and not see horses.

But see bronze sooner. It's really annoying sometimes, not being able to have everything at once, isn't it?

[Blake]

Blake, I'm inclined to agree that mining the sheep is better than waiting to build a pasture. But comparing building a mine on the sheep in a flood plains start with building a pasture on it in a hill start is something of an apples-and-oranges comparison. The ability to build a mine instead of a pasture, and thereby to delay researching Animal Husbandry and turn the sheep into a +5 tile sooner, is a something with added value that a flood plains start allows us to do, not something that takes away value that a flood plains start forces us to do. Or if we don't view it as adding value, we don't have to do it.

Remember that the FP start is already lowered value, it doesn't get +1 hammer/turn. In my view the usage of the sheep is something that largely equalized the two starts, the hill starts out stronger but requires the sheep to be pasture'd early on in order to maintain growth. The FP start doesn't need the pasture, allowing investment in other techs; but it does need those other techs earlier in order to make use of the higher health cap.

One thing about the FP start is since it doesn't need AH it can also afford to research Archery earlier which is a good deal.

Altough I ultimately favored the hill start, the FP site is still very nice in many ways, the main one being that freedom of not needing to research AH (it's not a cheap tech).
With an almost assured religion and skirmishers it will be a very tough nut to crack even without the hill.

But see bronze sooner. It's really annoying sometimes, not being able to have everything at once, isn't it?

Yeah, and for some reason I have a good feeling about copper with this start. Heh, in one trial I popped a copper on the grassland hill .