C-X-X-C or C-X-X-X-X-C?

[Spoonwood]

ZzarkLinux, I don't think anyone will really argue your point about finishing the game faster in terms of game date. But, did you notice that I started talking about spaceship finish times or *diplomatic* finish times? Did you notice that the original post talked about military advantages? Did you see the emphasis on beakers and science production? Of course C2C will help one get the game over with quicker in terms of finish date (though not necessarily playing time). I've never won a Civ 3 game by conquest or domination, but having used smallpox in Civ 2 AND realizing the production advantages of C2C it almost smacks me in the face that one can almost always have a faster finish date if one goes for the domination condition or massive "map control" as you dubbed it. Sure, "map control" can make a 20k game easier to control... but then again, if I remember correctly, certain types of wars actually halve the cultural rate of your potential 20k city... you certainly don't want that to happen much in a 20k game. Beelining military tradition and going out like a warmonger consits of just ONE way of winning civ 3. And if you check your hall of fame score you can even see that the game puts cultural vs. diplomatic vs. spaceship in different sorts of categories. XOTMs recognize different types of victories differently and give different awards for each of them. So, try to realize more exists to playing civ III than like a warmonger who beelines to military tradition and kills.

"90% of games are "won" before sanitation or Shakespears come into play."

Some 60% of those games probably get played by warmongers. The other 30% come as played at a difficulty (without creative handicaps) well below a player's ability where the player, of course, dominates the game. I don't think anyone disputes your point here really... I DO think you've entirely missed the points of what metro advocates have here tried to argue.

[And of course, apply this at all difficulties for non-spaceship games.]

It certainly won't apply for histographic victories... as even if C2C works out better since you can control the flow of the game early on... you'll still have DOZENS AND DOZENS of later-game "builder" turns if you kill off most of the AIs early. Sure enough, the HoF histographic games I've seen on Sid use a tight spacing... but if you look through the log of say SirPleb's game you'll find him talking about "milking" and unlike a warmonger he starts building all sorts of city improvements after a while. Sure enough, Moonsinger did the same at some point. More cavalry won't help you here. Also, if you haved enabled ONLY the "wonder" condition and play on harder levels the game certainly won't end before you or the AI builds the last wonder and all that matters comes as the histograph. Interestingly enough in COTM48 I didn't even really research techonology or have all that much going until the end of the middle ages. I certainly didn't pass the AIs in the histograph until the late industrial age. Did the game end before then and I lost since I trailed in the histograph??? I feel sure enough I could have won diplomatically also. Would the game have ended before I built the U.N.?

[And here's the best CXXXXC vs. CXXC diagram I could come up with.]

This sort of silly digram indicates why I wanted to talk about "metros" vs. "cities". What thoughtful player spaces their cities like this when terrain suggets a tighter spacing in some spots and a wider spacing in some other spots? Look... the *relevant* discussion focuses on *fewer* metros which DO get hospitals and lots of city improvements and can usually get to at least 15 at minimum, if not MUCH, MUCH larger (like 35 or 40) vs. a bunch of cities that don't get past size 12 in the city spacing debate.

"Also, I remind people that if you aren't winning your (non-spaceship non-variant non-sid) games by the end of the middle ages, you're doing something wrong."

Yes... that's right... if you simply don't beeline to military tradition and go out and conquer you simply don't do things wrong. If you intentionaly have ONE METRO-ready city in the middle ages for Shake's, because you've decided to go for a 20k game... you've done thigns wrong. If you haven't won your 100k game by the time of steam power... you've done things wrong. Forget the fact that culture needs time to add up. If you still have to manage the AI's attitude towards you and/or repair your reputation/get alliances/figure out who will oppose you in the U.N. elections or just keep the peace and race through the industrial techs to secure the U.N. and get your votes for a diplomatic victory you've done something wrong. If you haven't won a demi-god or deity game histographically for a "final wonder built" condition game by the end of the middle ages, you've done something wrong... who cares if that comes out as literally impossible since the wonder condition game doesn't end until all of the Manhattan Project, Cure for Cancer, Longevity, SETI, and The Internet get built (did I miss one?). Who cares that you can more easily build those *heavy* shield wonders in a metro than a city??? Seriously... here's 25 gold... go buy a clue from Xerxes or something.

Meisen,

[There is also no reason why a city or metro can not use scientist specialists.]

Of course. Why didn't I figure this in? I almost always have at least 6 specialists in most of my massively irrigated metros. And it usually works as somewhat easy to get fast growth for a metro once you have a hospital... especially if you irrigate everything, as if you have 4 food per worked square for grassland (irrigation and railroaded), one more square worked... say from growing to size 14 to 15... means that you get two extra food in the production box. An extra plain worked means you have an extra food in the production box for size 15 over size 14. Specialist farms simply can't grow anything close to the 1 to 3 turn growth rate which happens often during the age of hospitals at any point in their life. This doesn't make metros better as sure enough they have other disadvantages as pointed out more pollution, more improvements needed, fewer tiles worked early on (the biggest problem as I see it)... but they have their advantages too.

 

[Pyrrhos]

Out of curiosity, what was your VC in that game? Seems like some serious warring is necessary to get to that city name.

Conquest or spaceship/Monarch. I fought and destroyed the Dutch, Greek, Celts and Inca on my half of the continent first, then the remaining two, Babylon and Sumeria.

Since the game is designed to "challenge", I haven't had to declare a single time. The AI has done it all for me. All you need to do is get sufficiently far ahead and the AI programming means that even polite AI civs will break treaties asap and obey the call to arms against the too successful. At that point, only the Vikings were not at war with me. The rest of the world hates my guts as they are at Industrialization and I'm almost ready to research future techs. That's what happens when you own a full continent and settle it using CXXC except in the core and a few other places where CXXXC or CXXXXC seems better.

What also happens is that with cities running to New Chengdu 4, city and production management becomes too much of a bore so I left the game at that point.

 

[Pyrrhos]

Doug! Your post #189, brilliant! Of course I must be doing something wrong if I do not play the game according to others' expectations! If I do, it implies they might not be right about everything and that to many is a scary thought.

Bottom line is everyone can't be right about everything all the time and there a more ways than one to make an omelet.

 

[Chamnix]

Maybe you could explain how these farm exploits are being used in your game and how you defend them from your opponents so what happened in overseer's earlier posted game doesn't happen to your team....for those of us who don't want to spend an hour or so going over your logs and maps.

It helped win the game just like it helps to win single player games – we had so many scientists that we were researching Industrial Age techs in 4 turns with the science slider fairly low. We had so many cities that our unit costs were much lower than anybody else’s.

To defend ourselves we had lots of fast moving attacking units. Our border was mostly jungle so to reach our cities, the enemy would have had to stop in our territory where our fast moving offensive units could clean them up. We used our economic advantage gained by placing cities close together early in the game to build a powerful military to overrun our neighbor by the late Middle Ages so that we ruled our continent.

 

[Othniel]

Sure Othniel... but then someone almost want to argue for metros in the core and specialist farms elsewhere. On an archipelago map on some levels that can work. But, on a continents or a pangea map you'll have trouble since you'll need to take territory by force and doing that with pre-metros works out harder than with a tight spacing.

I don't really understand your answer, sorry.

Let me say something plainly: the manner of comparisons that you and meisen have been drawing between metros and specialist farms are invalid. Farmers don't build farms in the core.

If, as meisen claims, in an un-modded game with the FP already built, you can get a town from 90% corruption to 50% corruption by just building a courthouse and police station, I might consider that a core town myself. But I think that vastly overestimates the corruption reduction from those building. Also, there would need to be consideration of the shield and maintenance costs, among other costs, of getting those buildings up and running. It is not cheap to rush a court and police station in some 90% corrupt town. That's a lot of gold to payback or else the rushing was useless from a money point of view. Even if the corruption reduction is really that much, it still may not payback.

So please, if you wish to continue comparing farms to metros, compare it using areas that are corrupt enough to actually tempt a farmer to farm there. For me personally, that level is 80% or greater corruption with the FP already built.

In general, my answer to any "metro vs other style" questions is...look at the GOTM. We have countless amounts of data, refined under competive pressure, that suggests metros are a suboptimal way of playing for score. Have people played GOTMs and won using metros? Sure. But I think that you will be hardpressed to find someone who has consistently scored high / won awards using that style. This is especially true if you look at the more recent games, say the last 2 years, when people have figured out the cxxc power-game style.

Playing the GOTM/COTM gives a chance to see how effective different styles are in a competitive crucible. You've been doing that Doug, so good for you. I think doing this experientially answers a lot of questions on the metro vs cxxc debate. You might wonder how some people score so dang high and you'll see that cxxc has probably been used.

Not that I think I'll get a lot of takers for this idea but...I'd also be happy to play two games on the same exact map, one where I can use farms and cxxc on one try and one where I use primarily metros. The VC would be space. I highly doubt any "proof" from this test would silence the debate--I don't think any test will do that--but it would be fun for me. There would be no issue with "different skill levels among players" polluting the test, there would just be me. And up to a year ago, I was a hardcore builder type (betcha didn't know that about me ) so I think I could play the metro style pretty honestly.

And no, using cxxc and warmongering is not the only way to play and win at civ. Of course not. But there is so much evidence (GOTM...) that it maximizes score and power in most games across different victory conditions. If you do not care about score, that is fine, and people have said that everywhere across this forum ad naseum. Having fun is a valid reason to play civ.

 

[ZzarkLinux]

@Doug

You make many points, so I can't address them all.
So I'll address 2, and try to be clear.

Point:
Histograph & diplomacy are easier with 4-tile spacing
I'll contend the "diplomatic" and "histograph" games are won better without 4-tile spacing.

I'd assume that for histograph, you take every tile but 1 city, then mass lux slider happiness while building nukes. Oh, and mass culture like in a 80K game (just turn culture victory off).

And diplomatic games it's a toss-up too.
For more, see this link.
You just have to rodeos your opponent so you can easily bribe them when UN comes. That involves war, then bribery using your territory / luxes / resources conquered.

Kinda reminds me of American politics. Steal everything, then offer a fraction of it back as "goodwill" and "charity".

Point:
I think there's a hint of "having a size 14 city is better than having a size 9 city"

This is too broad of an assumption with too many holes.

Example:
Consider building a infantry. I guess they cost 100 shields.
100 shields = 1 turn infantry
50 shields = 2 turn infantry
33.4 shields = 3 turn infantry
25 shields = 4 turn infantry
20 shields = 5 turn infantry
etc...

The city size doesn't matter above. All that matters are the "thresholds" ie making a city go from 50 shields per turn (any size) to 60 shields per turn does nothing. Science is similar, since there are thresholds for techs (cost / 2 cost / 3).

As unit / tech costs become higher and higher, and I contend that city size doesn't scale to keep up with it. You need more-and-more metros to keep up, and that's where corruption would kill 4-tile spacing... Just mass farm strategy

BTW I smell trolls in this thread

Edit: I applied an edit for clarity.

 

[Pyrrhos]

All that matters are the "thresholds" ie making a city go from 50 shields per turn (any size) to 60 shields per turn does nothing.

Oh really...

First. Going from 50 to 60 shptnet does a lot! It means you can build Mech Inf in two (two) turns with a waste of only ten shields or Modern Armour in the same number of turns without any waste of shields. Your 50 shptnet city requires three (3) turns to build either with a waste of 40 and 30 shields respectively.

Second. If you have three towns/cities/metros producing 40, 50 and 55 shptnet respectively, you set the first one to produce Modern Armour (3 turns) or Artillery (2) with no sh wastage. The second will produce Bombers in 2 turns with no sh wastage and the third Mech Inf in 2 with no sh wastage.

As for the rest of your reply, I gather that you disagree with Doug but you've not managed to explain why.

PS. If you worry about the smell of troll, may I suggest you use a good deodorant/antiperspirant? It blocks the smell and thus won't alert other posters to your presence.

 

[Aabraxan]

Wow! There's been a lot of activity in this thread and lots of excellent points made.

The below-quoted material is a good analysis, Doug, but there are a couple of points I'd like to make.

If you draw a 5x5 grid (that's 4 vertical lines and 4 horizontal lines) and put Xs at the corners you have the amount of territory claimed by an ideal C4C metro (not a C3C metro). You can fit 7 Cs in this grid without any of them touching. This doesn't mean 7 cities per metro though (given all grasslands), since the metro space in any other direction can only have 6 Cs. So, *given all grasslands*, you have something like 6x4+7=31 specialist farms per 5 super-metros. 31 specialist farms will give you 31x9=369 beakers (again... given all grasslands and you can get them all to size 6 with 3 specialists)... while 5 super-metros... the center has an extra commerce square, right?... will give you (20x2+3)x5=40x5+3x5=200+15=215 base commerce. So, if you have 100% research you'll get 215 beakers from those metros. So, 215+214=429 beakers with a library and a university... given no corruption. Corruption might lower science output quite a bit. But... you simply won't have that much grassland... you'll inevitably have some hills and mountains.

I don't get what you mean by "the metro space in any other direction can only have 6 Cs." Are you referring to adjacent "metro spaces?" Obviously having too many hills limits the size of the science farms, as well as the ability to irrigate effectively. On the other hand, if I've got a few hills surrounded by grass, what I actually do is build on the hills. The 1-food hill becomes a 2-food city center and I can irrigate through the city, pre-electricity.

Consequently, I would guess it comes out as something like 3 or 4 specialist farms packed tight as possible per metro. 4 specialist farms will give you about 36 beakers. 5 specialist farms give you 45 beakers. 6 specialist farms 54 beakers. 1 size 18 metro will give you 3+2x17=57 base commerce before corruption... probably around 30-40 after on average. Then 30-40 more for the library and university. So, 60-80 beakers for 100% science. Adding on research labs gives you another 30-40 beakers. If you add on rivers and coastal areas (especially with commercial docks), the metro gets even more beakers/commerce since extra commerce goes towards the base rate. In this way a metro on a river, I would think, almost always works out as more productive in the long term than specialist farms (given that it doesn't have too much corruption). These sorts of comparison doesn't work as fair since you have to consider them both over time. The metro strategy also has the advantage that one can get money out of one's metros merely by changing the science slider, instead of mass-changing specialists. Then again, maybe you can lower your research rate *in general* and get more cash overall for specialist farms.

I agree with Othniel that this is a good analysis. I also agree that corruption puts a fly in the ointment here. Without corruption, the metro that you discussed above . . . well, I wouldn't make it into specialist farms, either.

As Othniel points out:

. . . . Metros ARE superior to farms in low-corruption areas but no one is arguing to the contrary, I think.

I'm certainly not claiming otherwise. Any city that's producing 30-40 (or even 20) commerce after corruption should be fully improved. Now, if it were in an 80-90% corrupt area, then I'd turn it into farmland.

Another item that I'd mention with regard to this analysis is building upkeep. If you've accounted for that in the above analysis, please forgive me for having overlooked it. OTOH, if it's not in there, it needs to be taken into account. In low- to moderate-corruption cities, the improvements are good investments. In farms, they're not.

. . . .
CONCLUSION
You cannot compare "late-game" decisions to "early-game" decisions. You have to examine the "time-till-payback". . . . .

This brings me to another point that I've made before on farming: Growth time. On the one hand, with specialist farms, I can have many more of them growing at once than I could with a few large cities (pre-metros). On the other hand, I use my specialist farms to peel off workers and settlers as I go, which slows down their growth. In Doug's analysis above, all of the towns compared were fully grown. I think specialist farms come out better in this part of the analysis, but I can't say that I've ever run any spreadsheets or other tests to confirm it.

. . . . Sure, "map control" can make a 20k game easier to control... but then again, if I remember correctly, certain types of wars actually halve the cultural rate of your potential 20k city... you certainly don't want that to happen much in a 20k game. . . . .

Mostly off-topic: Are you referring to wartime mobilization?

You know, I've kinda wondered about that. The wider spaced cities do leave tiles unworked till they fill up, but they get to choose the best tiles. So that city for city, the wider spaced ones will probably be producing more than the tightly spaced cities, even though both may still be at the same population. You will only be able to build so many cities before you run out of room and early game, the difference in the number of cities built will be small (due to the time it takes to build settlers). In other words, you trade a slightly larger number of cities for a larger territory and better individual cities. . . . .

This is slightly off-topic and not intended to distract from the merits of the discussion. I just find it interesting.

One of the points that I always find interesting in the city spacing/metros vs. cities debate is the perspectives of the different sides. The proponents of wider city spacing and larger metros frequently point to a city and say "Look what my city can do!," while proponents of tighter city spacing frequently respond with, "It doesn't matter what one city can do! Look at what my empire can do!"

. . . . Not that I think I'll get a lot of takers for this idea but...I'd also be happy to play two games on the same exact map, one where I can use farms and cxxc on one try and one where I use primarily metros. The VC would be space. I highly doubt any "proof" from this test would silence the debate--I don't think any test will do that--but it would be fun for me. There would be no issue with "different skill levels among players" polluting the test, there would just be me. And up to a year ago, I was a hardcore builder type (betcha didn't know that about me ) so I think I could play the metro style pretty honestly.

I'd be interested in seeing this. I was considering doing something similar. However, I've never been much of a builder, so I'm afraid that I wouldn't do metros justice. One item that occurred to me was this: Why not play from a save in which you've already played to the end? That prevents one "style" from having the advantage of foreknowledge of the map.

. . . .
Point:
I think there's a hint of "having a size 14 city is better than having a size 9 city"

This is too broad of an assumption with too many holes.

Example:
Consider building a infantry. I guess they cost 100 shields.
100 shields = 1 turn infantry
50 shields = 2 turn infantry
33.4 shields = 3 turn infantry
25 shields = 4 turn infantry
20 shields = 5 turn infantry
etc...

The city size doesn't matter above. All that matters are the "thresholds" . . . .

I got x-posted by Pyrrhos while writing all of this. I view these "thresholds" in light of what I'm building. If I'm building cavs (80 shields), then going from 40-50 net shields does me no good. If I'm building bombers (100 shields), then going from 40-50 net shields does me some good. In terms of building an infantry, going from 50-60 net shields does nothing. ZzarkLinux is right about that. But that analysis changes if you're building Mech Infantry, as in Pyrrhos' example. In that case, it certainly does make a difference.

 

[ZzarkLinux]

PRELIMINARY CHECK:
Okay lunch break

TECHNICAL-SUMMARY:

If I'm building cavs (80 shields), then going from 40-50 net shields does me no good. If I'm building bombers (100 shields), then going from 40-50 net shields does me some good. In terms of building an infantry, going from 50-60 net shields does nothing.

So what do you do if you are at one threshold, and you cannot quickly hit the next threshold?
2-Tile spacing: build workers and settlers with extra (useless) population
4-Tile spacing: spend gold and shields to make that extra (currently useless) population happy

Most of the time, efficiency goes to the 2-Tile spacing in the above example.

And this analysis isn't unreasonable. It often happens with 70shield Knights...
You can hit the 3-turn or 2-turn, but never the 1-turn threshold (70shields is alot in that timeframe)

CONCLUSION:
So once a city hits a threshold, and cannot hit the next threshold (shield or science), then you now have excess population which doesn't contribute anything.

This excess population turns into workers and settler spam !!

POST-CONDITIONS:
Now back to progamming

Edit: I applied das edit

 

[Pyrrhos]

No, no, no! I'm afraid it looks as if you're wearing tinted shades here, Zzark:

* Building workers or settlers - and recouping the loss in pop and productivity - is much easier with the large city that drops from 12 to 10 with a full box (=11 next turn) than with your pop 5-er, which drops to three and, I assume since you're against waste, only a half-full box

* If you hit a threshold and feel you waste too much building cavs, there's always something useful to build such as a uni, bank, airport, commercial dock or why not a worker or two. Sometimes it is better to build 118 shield, three-turn cavs than abstaining because you lose 38 shields.

* With a large spacing, each town usually - at least until very late in the game - has a wide choice of tiles to work. This means that if you indeed do hit a threshold and don't want to or can't join workers to get over it, you can exchange that mined hill and mountain for a couple of 3+ commerce tiles. The excess population (as you term it) contributes food and gold now instead of surplus shields

 

[ZzarkLinux]

Pyrrhos I liked Post #196 more than Post #199
BTW post 196 made me laugh so much (in a good way ) I just decided not to reply to it

Your first point above: *Get back that extra population
Dang, better hurry to get that excess population back that doesn't make me get knights faster.

Your second point above: *Switch to build something else
Valid, yes, but not when you're massing knight armies to win most victory conditions

Your third point above: *That extra population gives more food and commerce
What if that commerce doesn't get you X tech any faster??
Then all that extra food gives me .... more excess population

Edit: I applied a edit

 

[Pyrrhos]

Zzark, you're talking about a particular strategy to win in a certain fashion. Given those restrictions, I can see your point of view.

However, as the others have pointed out, whilst it is possible to win every game by treating it as a Middle Age conquest and while people who do so place highly in the HoF for cultural, spaceship etc, it is not the ONLY way to win or to do well! There are disadvantages with your way as well as advantages. There are advantages as well as disadvantages with other strategies.

Now, if everyone agreed that there only was a single way to be successful at civ, how fun would that be and what are the chances that new, and perhaps even more powerful strategies would be investigated/developed?

 

[Othniel]

In my case that is definitely a misrepresentation of what I've written. You even note my comments about 90% corrupt cities below. So you definitely knew I was talking about post 90% corruption. Either you have severe comprehension problems or you are dishonest. Nowhere do I claim anyone uses farm exploits in lower corruption cities. If you are going to respond to what I write, I would appreciate it if you would do so honestly. If you do not understand something I wrote, ask for a clarification. One aab on the board is more than enough....

Honestly? You and Doug wrote a lot of words about metros vs specialist farms and I didn't feel like quoting it all. Shoot me.

Here's the excerpt of what you wrote that I responded to:

With courthouses and police stations and using the forbidden palace, even 90% corrupt cities should see about 50% usable shields, gold and beakers. So using your figures, with an average of 4 farm exploits per metro, even base 90% metros should be about equal to the exploits in producing beakers once the metros have been filled out and outfitted (and since most of these are usually captures with a decent pop level already, it takes less time to get these metros up and running than razing the landscape and building up farm exploits from scratch).

In this post, Doug made a comparison between specialist farms and metros that were placed in areas only somewhat corrupt (his words). In the same post of yours that I quoted above, you quote Doug yourself, say "yup" to his analysis, then added your blurb that I quoted above.

So, me being stupid or dishonest, assumed you agreed with Doug that comparing specialist farms to metros in marginally corrupt areas was a valid line of reasoning.

My point in response is simple. If the corruption % can be reduced down to 50% by just using a police station and courthouse, I don't consider that a corrupt enough spot to farm either. Pick a more borderline metro, say one that has 75% corruption after all improvements, than compare the results to farming. This is a more valid test. Just make sure to compute all the costs to get things running in a metro (and for a farm). The calcs aren't simple, as I know you know.

It's been a while since I played a stock game, but from memory, I was able to get 90% cities productive to where they could build up their own infrastructure and build units in reasonable, usable times. The reduction % I mentioned was a guess based upon that memory. It might have been only 40%, it might have been 60% reduction of corruption. With metros, there is an excess of citizens that are not needed for food and these can be set to policemen, which further reduces the corruption. Even at only 40%, a fitted out metro can still put out a good deal of gold and shields. The point was is that those cities were productive.

The point on policeman is a good one and should be included in the analysis.

Your "it still may not payback" is also wrong. Payback in civ is not an accounting ledger of what you put into a city throughout the game and what it gives back. Payback is when you build up a city it gives you back what you want it to give back. That means the city becomes a useful cog in your empire's wheel. Be that because it can produce units in reasonable times, concentrate on food and put the excess citizens on science or taxes or anything else.

Actually, payback does have to consider the cost side of things. This is Finance 101. Why do you think the US is outsourcing so many jobs to other countries? Cost, although not the only reason, is a major reason. And why is the US so in debt? Again, complex answer but one reason is we collectively keep on buying expensive stuff that doesn't return enough on the investment to even cover the startup costs.

Say I put 316 gold in to rushing a court (79 * 4 assuming one shield in the box to avoid double penalty), but the court only raises the city commerce from 1 to 3gpt post-maintenance cost. It would take 158 turns (316 rush cost divided by 2gpt increase) to simply payback on a money side. If there are not enough turns left in the game, I just threw money down a black hole.

I know this example ignores any shield saving from waste reduction. It also might be an extreme example because the court might be more effective than this. I just write it as a hypothetical example of why payback needs to consider costs.

It's much easier replacing pop in large cities and metros. And much easier in cities where every tile is not needed to keep the city productive, as frequently is the case with tightly spaced cities. Who can lose a great deal of their usefulness if a couple of pop get subtracted.

BTW, are these accountants allowed to build granaries? I seem to remember some of them, at least, writing that these buildings are unneeded. And they do cost 1 gold upkeep each. That's such a huge, empire busting waste...

Uhh....

Ever heard of settler factories? Us accountant types dig them. They make us figure out the exact sequence of shields and food to be able to pop out settlers in record times, such as a 4-turner. All that number crunching and mental exercise, yummy.

And it is blatantly easier to replace pop in a size 6 city or smaller. The food box is only 20 compared with 40 for the larger cities.

Oh, and thanks for comparing me to Aabraxan. I consider that a compliment. He consistently shows himself to be a patient, courteous debater. That is commendable.

 

[Spoonwood]

meisen,

"So that city for city, the wider spaced ones will probably be producing more than the tightly spaced cities, even though both may still be at the same population."

Sure, but empire wide I can see how more tighter space cities produce more shields than wider spaced cities before hospitals. Medieval and ancient units generally run cheaper overall in terms of shields than more improvements also.

"I'm not so sure the closeness of the cities saves that much travel time for the settlers, either. Another thing constantly said to be an advantage. The time spent making a settler is the same irregardless whether that settler has to travel an extra 1 or 2 tiles to reach its settling spot."

One can use cities as roads for workers and settlers (play smallpox on any civ game and you'll do this A LOT). Settlers sort of function as "super-workers" in a way. A tighter spacing will save travel time... how much on average comes as the question. The time spent making a settler comes as the same, but a tighter spacing has more cities built faster that can conceivably produce settlers.

"Also, the time taken to build those extra settlers to fill in the core's initial wide spacing (if that was used to gain more early territory) could be put to better use building infrastructure and fighting units."

Maybe.

Othniel,

"If, as meisen claims, in an un-modded game with the FP already built, you can get a town from 90% corruption to 50% corruption by just building a courthouse and police station, I might consider that a core town myself."

It sounds to me like you've swapped definitions of "core town" now. Look, I do grant you that we have to speak carefully about these comparisons. We can't talk about our capital's area vs. 5 specialist farms... that doesn't make much sense. We CAN talk about specialist farms vs. cities in the 2nd, 3rd, and/or possibly 4th ring (maybe not the 4th). So we have a clearer definition at play, any city placed within 10 tiles of the we'll call a "core town". We'll say that anything else doesn't lie in our "core". If you don't like this definition, then please propose a clear one and let's have a discussion about that so that neither you NOR US unconsciously switches definitions. Also... if you think that courthouses and police stations can lack that power *in some instances*, then go look at my saves for COTM48 and sell the courthouses and police stations in my towns out near the northeastern English colony on my island (not the one's near the tundra... in the northeast).

"Also, there would need to be consideration of the shield and maintenance costs, among other costs, of getting those buildings up and running."

Perhaps... perhaps not. If you have a tech-lead in the industrial age you can sometimes, at least, make enough gpt and have your science slider at 100% and still have enough cash to cash-rush police stations and courthouses. At least on demi-god and emperor. If you trade your techs right in the industrial ages you don't really ended up needing more cash ever.

[We have countless amounts of data, refined under competive pressure, that suggests metros are a suboptimal way of playing for score.]

I should check this out... but I do believe your statement as it stands really before I've checked. I don't believe the thrust of ZzarkLinux's statements though.

"There would be no issue with "different skill levels among players" polluting the test, there would just be me."

I really appreciate your civility and your qualifications Othniel, and I think your "test" interesting, and I'd enjoy seeing you doing it. The problem comes as that even though we wouldn't have different players at different skill level playing a different map... we would, in effect, having the same player playing at a different skill level, since we would have the same player playing the second game with advanced knowledge of the map. You could *partially* mitigate against this problem by say playing your 1000 B.C.E. turn on the "metro" map and then before finishing that turn, playing the 1000 B.C.E. on the "city" map. Of course... that only works out as a "partial" mitigation since knowledge learned during the turn may affect your decisions. Such a game... since it involves constant saving and loading of a save, sounds like such would decrease the fun factor of civ III myself.

Zzarklinux,

"Histograph & diplomacy are easier with 4-tile spacing"

No... I didn't really claim that. My point lay in that for a histographic game you inevitably end up doing a lot more than just warmongering and knowing builder style can help. The same goes for diplomatic games to some extent. This doesn't concern city spacing... since with your comments I basically left that debate at some point as you seemed, or actually did imply that playing as a "builder" comes as a waste of time.

"I'd assume that for histograph, you take every tile but 1 city, then mass lux slider happiness while building nukes."

Why would you build nukes when there only exists one "enemy" city left? Don't wonders, libraries, universities, and culture in general increase your score more than nukes (I don't actually know myself... but I'd sure think so)?

"Oh, and mass culture like in a 80K game (just turn culture victory off)."

Your suggestion here of "just turn culture victory off" doesn't work for the HoF games like that of SirPleb as I understand it. SirPleb mentioned how he almost LOST (or thought he would lose) his HoF game due to the civ-wide cultural condition.

[You just have to rodeos your opponent so you can easily bribe them when UN comes.]

But this happens in the industrial age. You mean you still have to finagle your opponents after Theory of Gravity? But you said the game "ended" by the end of the Middle Ages. Whoops... contradiction.

[Consider building a infantry. I guess they cost 100 shields.
100 shields = 1 turn infantry
50 shields = 2 turn infantry
33.4 shields = 3 turn infantry
25 shields = 4 turn infantry
20 shields = 5 turn infantry]

When does Replaceable Parts come??? That's right... in the Industrial Age. What's that "rubber" resource doing there before Steam Power? Oh yea... you can't even see it on the map... so no infantry.

"All that matters are the "thresholds" ie making a city go from 50 shields per turn (any size) to 60 shields per turn does nothing."

Let's see... almost everyone I've read here seems to always, at least want to build the Theory of Evolution. It costs 600 shields. For a 50 shield city it will take you 12 turns... for a 60 shield city it will take you 10. What about Hoover Dam at 800 shields? 16 vs. 14 turns. And the U.N.? 1000 shields... 20 vs. 17 turns. But making *a* city go from 50 shields to 60 shields does nothing? Sure, going from 50 shields to 60 shields does nothing for a 100 shield object. But, not everything a city produces comes as 100 shields... and if you want to talk about "thresholds" you have to consider that *thresholds change when you build different things*.

"Science is similar, since there are thresholds for techs (cost / 2 cost / 3)."

This would only hold if every tech costed the same... and they don't... so it doesn't hold.

Aabraxan,

"I don't get what you mean by "the metro space in any other direction can only have 6 Cs." Are you referring to adjacent "metro spaces?""

Yes.

"Mostly off-topic: Are you referring to wartime mobilization?"

Maybe, but I thought it applied when your capital got attacked or lost a troop or something like that... I don't recall when I've seen this happen... but I know I've seen in 20k games I've played before.

"The proponents of wider city spacing and larger metros frequently point to a city and say "Look what my city can do!," while proponents of tighter city spacing frequently respond with, "It doesn't matter what one city can do! Look at what my empire can do!""

Excellent way of putting it. For a 20k game, I think there exists no question that you want *at least one* metro. So, even if the metro-advocates end up "conceding" or "losing", there still exists a "non-variant" use to knowing how to work a metro and what a metro can do.

"Why not play from a save in which you've already played to the end? That prevents one "style" from having the advantage of foreknowledge of the map."

And both games start at 4000 B.C.E. and I'll add that the save comes as the most recent game... We wouldn't our "test player" to only slightly remember the map... we'd want as fully knowledge of it as possible. Or the method described above. Both slightly imperfect... but close.

Zzarklinux,

"2-Tile spacing: build workers and settlers with extra (useless) population
4-Tile spacing: spend gold and shields to make that extra (currently useless) population happy"

No. For C4C you have a temple and/or cathedral, a marketplace, and you trade for/own lots of luxuries. The "extra" population either works another square which might get production, luxury, tax, or science benefits from the improvements, or it might work as a specialist. With C2C you probably have less improvements, so you don't get as many multipler effects.

"So once a city hits a threshold, and cannot hit the next threshold (shield or science), then you now have excess population which doesn't contribute anything."

I think you really need to play a 20k game where you have a prebuild or multiple small prebuilds going in the high or early middle ages and your "thresholds" change as you switch from the palace to Sun Tzu's, or from Sun Tzu's to Knights Templar, or from a university to Copernicus's Observatory, or from Copernicus's Observatory to the Sistine Chapel... and where you can't add in workers... and where you possibly forrest cities land after engineering. And re-read this "With a large spacing, each town usually - at least until very late in the game - has a wide choice of tiles to work. This means that if you indeed do hit a threshold and don't want to or can't join workers to get over it, you can exchange that mined hill and mountain for a couple of 3+ commerce tiles. The excess population (as you term it) contributes food and gold now instead of surplus shields"

"What if that commerce doesn't get you X tech any faster??"

Then change your science specialists to tax collectors.

"Then all that extra food gives me .... more excess population"

Excess population... talk about a contradiciton in terms.

 

[Spoonwood]

Othniel,

[You and Doug wrote a lot of words about metros vs specialist farms and I didn't feel like quoting it all. Shoot me.]

I still need some cash to upgrade that lone pikeman to a rifleman (I've spent my cash rushing courthouses, police stations, univerisities, hospitals, factories, etc.)... or I'll have to play a few dozen turns until I can trade for replaceparts and upgrade a warrior/medieval infantry to a guerrila. I certainly don't have time to train a new military unit... at least not for a while. I didn't have potassium nitrate and yeah... KNO3 didn't seem to serve my empire's plans well. And... well... actually re-checking things it looks like I don't have a barracks yet... so upgrading might have to wait longer too. So, for now... no can do... can I take a rain check?

"If the corruption % can be reduced down to 50% by just using a police station and courthouse, I don't consider that a corrupt enough spot to farm either."

Pay attention to my next post and follow closely please and tell me if you would have built specialist farms in a certain spot (given that you didn't have raging barbarians around even), o.k.?

[Say I put 316 gold in to rushing a court (79 * 4 assuming one shield in the box to avoid double penalty), but the court only raises the city commerce from 1 to 3gpt post-maintenance cost. It would take 158 turns (316 rush cost divided by 2gpt increase) to simply payback on a money side.]

Trade tech wisely and you'll find it difficult to actually lose money at 100% science in the industrial era ever.

 

[Spoonwood]

In terms of the "corrupt areas" debate go here http://forums.civfanatics.com/showthread.php?t=274997 Open the save entitled "victory". You can evaluate MANY cities on this map for corruption and the effects of courthouses and police stations, in my opinion, for this discussion. I selected Vidin and Iconium in the eastern part of the map. I only looked at shield output as this seemed easiest to count, and I didn't want specialists to mess up my numbers... and this indicates how I usually evaluate corruption initally at least.

If you count with me (unless I mis-counted) Iconium lies 12 spaces away from my capital Constantinople if we allow diagonal travel and 15 spaces without diagonal travel (I don't know how diagonals get counted in the game for corruption purposes). Vidin lies 15 spaces away with the diagonal and 17 spaces away with the diagonal. You'll also notice if you check, that my Forbidden Palace actually lies farther away than my capital Constantinople (I don't know how much this matters). Iconium lies in the 4th ring and Vidin in the 5th ring... or should I class them by rings differently?

Anyways, currently with the courthouse and police station you can or will see that Iconium produces 5 corrupt and 6 uncorrupt shields. If you sell the courthouse and police station, you'll see that it produces 8 corrupt and 3 uncorrupt shields. Vidin produces 7 corrupt and 7 uncorrupt shields. Without the police station and courthouse it produces 12 corrupt and 2 corrupt shields. Now that we have some data... and possibly more data given their science and tax outputs... would a specialist farmer have built specialist farms or cities in these locations? I'd guess specialist farms... although at least once they had the courthouse and police station it appears they outproduce multiple specialist farms in terms of commerce and shields (maybe this doesn't matter though empire-wide admittedly). Maybe one should build specialist farms around cities like this and then disband them once you have a police station and courthouse in it. I certainly wouldn't class those cities as core towns and I can't see anyone doing so... but maybe I speak prematurely.

 

[TheOverseer714]

Not farm territory yet. I would that semi-core area, about 50-60% corrupt. My rule of thumb is if there's only 1 uncorrupt shield(90% corrupt), even with a cop, it's time to farm it. Depending on map size, that can be 30+ tiles from the capital.

 

[Chamnix]

For corruption purposes, count 1.5 along N, E, S, W and 1 along NW, NE, SE, SW. Truncate the result. Vidim is distance 16; Iconium is distance 13. Vidim is 90% corrupt with no improvements decreasing to 48% with courthouse and police station, Iconium is 74% brought down to 40%. I’ll be the first to admit both of these are in farmland for me.

Iconium – fully improved as it is with no excess happiness (change entertainers to scientists wherever possible) and with 5 more citizens (it has 11 extra food now so add 5 workers to be scientist specialists) provides 99 bpt and 6 spt. It has the following improvements:

Temple – 60 shields + 1 gpt
Marketplace – 100 shields + 1 gpt
Library – 40 shields + 1 gpt (Byzantines are scientific)
Courthouse – 80 shields + 1 gpt
Cathedral – 160 shields + 2 gpt
University – 100 shields + 2 gpt
Colosseum – 120 shields + 2 gpt
Hydro Plant – free from Hoover Dam
Hospital – 160 shields + 2 gpt
Police Station – 160 shields + 2 gpt

Total – 980 shields + 14 gpt.

Iconium takes up 16 tiles total. Specialist farm exploits generally have 2 worked tiles per city tile, so say 5 cities and 10 worked tiles. Each of the farms provides 10 bpt (3 scientists plus 1 uncorrupted gold) and 1 spt giving a total of 50 bpt and 5 spt. To get 5 farms, you need an additional 4 settlers beyond what the metro needs. Each settler costs 30 shields and 2 citizens. A lost citizen costs 3 beakers per turn (the lost scientist specialist) until the city grows back. Each citizen should be gained back in 3 turns after the settler is built, so each settler costs 30 shields plus 27 beakers (6 bpt for 3 turns and 3 bpt for 3 turns).

Totals – to make Iconium a metro costs 980 shields and provides a net after expenses of 85 bpt and 6 spt.
To make Iconium’s territory full farmland costs 120 shields and 108 beakers and provides a net of 50 bpt and 5 spt.

If you use the standard conversion of 4 gold to one shield, the metro costs an additional 3332 gold for a gain of an additional 35 bpt and 1 spt.

What am I missing?

 

[Othniel]

What am I missing?

Time.

Time to rush enough buildings to get the metro up and running.
Time to get the farms fully functional--you included the beaker cost for the time it took other farms to produce settlers, but not the time it takes for the farm to get to size 5.
Time until the increased investment in the metro's buildings actually return a profit.

IMO, the first two things only make only a thimble full of difference though. The third reason, time until payback, is an obvious huge strike against building a metro in this location, at least with that many city improvements.

Calculating the time effects for the first two are also quite difficult with many factors.

Chamnix's excellent analysis demonstrates what I meant when I said this:

If, as meisen claims, in an un-modded game with the FP already built, you can get a town from 90% corruption to 50% corruption by just building a courthouse and police station, I might consider that a core town myself. But I think that vastly overestimates the corruption reduction from those building. Also, there would need to be consideration of the shield and maintenance costs, among other costs, of getting those buildings up and running. It is not cheap to rush a court and police station in some 90% corrupt town. That's a lot of gold to payback or else the rushing was useless from a money point of view. Even if the corruption reduction is really that much, it still may not payback.

(bold added)

I was wrong on how much the corruption was reduced by police stations and courthouses. Their effect was greater than I imagined. Farms, however, look to be the clearly better investment unless your game is going to last 1000s of turns, and most games that I know of don't last that long.

For the record, I would have definitely farmed the 90% corruption farm and probably the 75% corruption one (this is without looking at the save, no civ access right now). I wouldn't have ever figured how much the corruption could be reduced on the 90% city because it would have been farmed long before police stations become available.

It sounds to me like you've swapped definitions of "core town" now.

You're right. To simplify the terms, I went with just two distinctions: "core" and "farm" areas. Core cities, for me, are cxxc spaced, get some level of improvements and build stuff. Farms are cxc spaced, get no improvements, and build workers, settlers, or wealth slowly (some people build other stuff with the 1spt such as artillery). I've been throwing out 80% base corruption as my personal level of when a city goes from "core" to "farm". After seeing Chamnix's numbers, I might start farming at the 70% corruption level.

It's actually an interesting question: what to do with towns that are in the 50-70% corruption range before any city improvements are added. I assume an FP has been built and the Republic government is being used for this question.

In the past, I would likely build workers or settlers from these towns for the first bit of the game. At some point, I would probably start a courthouse build there unless the terrain is bad. With the court, this town would become part of a "semi-core" group of cities. This semi-core would perhaps get a rax and build units indefinitely. I might consider some basic improvements like a market or library if I think they would payoff. The cities closer to the 50% pre-courthouse corruption level would probably be pretty productive and get more improvements than this, such as Universities. But the cities closer to 70% would be unlikely to be worth expensive improvements like the Uni, IMO.

So, I throw out that thought as a general question. What to do with the semi-core? I'm honestly not sure what the optimal (for score) way to handle these areas. I prefer that we don't include Communism in any response because that is a whole different ball of wax. I'm not trying to stifle debate on Communism , just do it elsewhere. Think Republic if anyone wants to respond.

I think your "test" interesting, and I'd enjoy seeing you doing it. The problem comes as that even though we wouldn't have different players at different skill level playing a different map... we would, in effect, having the same player playing at a different skill level, since we would have the same player playing the second game with advanced knowledge of the map. You could *partially* mitigate against this problem by say playing your 1000 B.C.E. turn on the "metro" map and then before finishing that turn, playing the 1000 B.C.E. on the "city" map. Of course... that only works out as a "partial" mitigation since knowledge learned during the turn may affect your decisions. Such a game... since it involves constant saving and loading of a save, sounds like such would decrease the fun factor of civ III myself.

Thanks for the feedback on the test idea. I like the idea of doing it because it might give some decent anecdotal evidence on the CxxC vs Metro debate, plus some stuff on the power (or not ) of specialist farms.

It is a drawback that I (or someone) would be playing the same map twice. Aabraxan's suggestion that I play a map I've already seen to begin with is quality solution to this drawback. Not perfect, but the best I can think of.

So, if I was to do this test, what "rules" would I play by for each? Definitely I would go for Space as the VC since that means Modern Age play. But what about other things. Can I build farms and metros on the same map, or reserve farming for my CxxC try? I'll think about this rules questions further, but shout out anyone if you want to see a "test" and have ideas.

EDIT: x-post with meisen

 

[TheOverseer714]

I say no metros at all on the farmer game, don't even research Sanitation. Other factors might be same territory, i.e. take a continent only, leave the rest for your AI opponents. Also, no suicide ships, no scientific civs, no barbs(no goody huts), and no overseas contact until magnetism. Standard number of opponents, pre chosen. A neutral party could provide the save. I would say launch date and score be the primary criteria, with population being secondary.