timerover51
Deity
In some respects a better question would be, can you win both ways? If so, then either way works for people.
This is why CXXXXC is VASTLY superior to CXXC
- Thread starter Pyrrhos
- Start date
Mursi lives
Warlord
- Joined
- May 27, 2007
- Messages
- 250
You clearly proved the superiority of CxxxxC when the game is modded to start like that: "<leader's name>, Sid gave you that vast territory, plus Divine Knowledge of every tech until Education. Now you must select 15 places in the Promised Land and Sid will make 15 fully developped cities appear there. He will also create your first 180 loyal servants and you will reign over them... "
In another test of yours (thanks for creating them!) you also demonstrated what happens with a normal start: Remember India vs Korea? You compared India and Korea speed when REXing a large island, using loose city placement. Then MAS rerun the test with Korea, using a much tighter city placement and the same number of turns. Did he end up with half-sized territory? Not only did he colonize the whole island, he also improved many more tiles, cleaned every marsh/jungle ones. He even put a foot on next continent....
Any city placement designed to get 12 citizens per city and no unused tile will obviously have ratio [city tiles]/[total tiles] = 1/13. Now, let's consider "strict" grid patterns:
CxxC = 1/9
CxxxC = 1/16
CxxxxC = 1/25
So yes, ZzarkLinux's pattern is somewhere between CxxC and CxxxC.
DJ Bonebraker
a.k.a Laura
TBH, I generally use Zark's pattern (in between CxxC and CxxxC), due to the fact that I always play a no-pollution variant... I.e. I don't build any pollution producing city improvements, nor do I allow cities to become metroes. This means that I can never have more than 12 citizens per city (unless I play mods with modded population settings, such as the Final Fantasy mod or TCW), so I settle in a pattern to ensure 12 C/C. Using Zark's pattern (taking in account such things as coastal, rivers, etc), I was playing a very successful game as the dwarves in the Warhammer Fantasy mod.... the link to which is in my sig... You'll note that while I may have one of the smaller nations, territory wise, I am #2 in population, because that small territory is packed with size-12 cities. Furthermore, I conducted a successful war with the Vampire kingdom of Sylvania (which didn't get added to the story write up, but once I get the FF mod finished, I'll see about finishing that story), which, while they had the same approximate area as me, they had half the cities I did, and consequently, I could spam troll slayers and other military units and I eventually overwhealmed them through sheer productivity.
In games where the population is modded, I use a looser pattern, but then that changes the dynamics again, because, in the case of the FF mod, you can get up to 12 citizens before needing an aqueduct, and 24 before you need a hospital, so looser placement actually pays off sooner....
Here is a pic of my territory in that particular game... I was #2 or #3 in pop, but somewher around #8 or 9 in land area.... Not long after that map screenshot was taken, virtually all the Sylvanian cities in view were, uh, assimilated into the Dwarf Kingdom.
In games where the population is modded, I use a looser pattern, but then that changes the dynamics again, because, in the case of the FF mod, you can get up to 12 citizens before needing an aqueduct, and 24 before you need a hospital, so looser placement actually pays off sooner....
Here is a pic of my territory in that particular game... I was #2 or #3 in pop, but somewher around #8 or 9 in land area.... Not long after that map screenshot was taken, virtually all the Sylvanian cities in view were, uh, assimilated into the Dwarf Kingdom.
In general, certainly. You can win on chieftain either way on any map. Some players can win on deity either way on some maps. But are there maps/levels that a given person will lose if he places cities suboptimally but could have won with better city placement? I'd say so.
i like pies
Chieftain
i know im not in the right thread but can someone tell me how to get a good start on a civilization
.......every time i try i screw it up
.......every time i try i screw it up
Without more specific information, i like pies, it's a little hard to give specific advice. Expansion in the early game is very important. I suggest reading Cracker's Opening Plays as a starting point.
i like pies
Chieftain
can someone here tell me how to get a decent start on a civ ..... i always screw it up
i like pies
Chieftain
thnx a ton dude .......btw if it doesnt help can i amail u the specifics......maybe u can help
DJ Bonebraker
a.k.a Laura
It also depends on what variant you're playing, too: OCC games, the point is moot, since you won't be building any more than one city.... In 3CC and 5CC, very loose placing would be best, since you're only going to be building 3 or 5 cities during the course of the entire game, so having those cities control as much territory (and resources) as possible is critical.
Actually, what I would recommend is that you do is:
- Start a game
- Play 80-100 turns
- Keep a log of what you do
- Start a thread & post the log and save
Here's an interesting question/challenge for anyone who dares try... given an 80% water, and we'll even make it archipelago... what's the hardest level you can beat with only 5 cities and ICS spacing?
DJ Bonebraker
a.k.a Laura
I don't know, but I DO know of at least one deity game that was won with a one city challenge. 
...it should be in the legendary stories archive in the stories and tales section.
...it should be in the legendary stories archive in the stories and tales section.the main problem that i see with this particular demonstration is how you are claiming that cxxxxc is more productive per city, this is a given. the idea behind cxxc is to be more productive per tile. in the early stages while cities are still growing neither way uses all of the available land.
the experiment itself is biased, that is why cxxxxc comes out on top. you are testing it in a way that far better suits cxxxxc over cxxc
the experiment itself is biased, that is why cxxxxc comes out on top. you are testing it in a way that far better suits cxxxxc over cxxc
The claim of such a test as "biased" comes out weak when it worked out that CxxxC came out better than CxxxxC according to the test.
eldar
ChiefTank
I started such a game on Emperor - we cleared the continent. It came unstuck when we found ourselves too far away from the other continent to mount a successful invasion. Plus our city spacing was a rather loose ICS - strung out along a coast rather than packed in as a single "core".
Still, it can be done. The fact is that even with ICS spacing, the human player will make better use of the tiles available than the AI. At Emperor, the production bonus they have over the human is soon swallowed up by its poor land management.
Pyrrhos
Vae Victis
- Joined
- Mar 15, 2007
- Messages
- 712
In science, you try to remove as many variables as possible so that you test only the one and can be certain that the results are down to this variable.
In the first one, this, it only the use of tiles as it releates to productivity - pop, gold and beakers are measurements of productivity too!
As the CXXXC pattern provided by ZzarkLinux makes use of all tiles, reaches pop 12 without having to suboptimise tiles (i.e. irrigating BG and mining G or even irrigating both), produces only a few net shields less and makes more gold when researching the same effective number of beakers (i.e. Navigation in eight turns) it is clearly the best pre-sanitation pattern without need for further testing.
The next step would be to see at which point as measured by how many cities more are needed for CXXC to equal the output of CXXXXC. This is much harder to do, as you would have to assign a viable exchange rate between superior science and superior net shield production or make some CXXXXC-cities into a variant of science farms nobody uses until you have achieved a balance between science and shields that satisfies the question asked.
Once a ratio between CXXC and CXXXXC has been established comes the question of feasibility - does CXXC really produce 1.xx times the amount of settlers in a given time?
Now comes the questions about how they do in real game play and it is time to test in a "real situation" where the AI plays in exactly the same manner in order to answer the questions of "How large rex-territories do CXXC and CXXXXC gain respectively?" and "Over a substantial part of the course of the game (including the rex phaze), played in exactly the same way, does one pattern confer any advantages? What are those advantages and are they quantifiable in such a way that it is possible to say that one is clearly superior of the other with any degree of certainty?"
Where many people go wrong in this debate is that they assume that they already know the answers to the questions asked by the last two tests.
In the first one, this, it only the use of tiles as it releates to productivity - pop, gold and beakers are measurements of productivity too!
As the CXXXC pattern provided by ZzarkLinux makes use of all tiles, reaches pop 12 without having to suboptimise tiles (i.e. irrigating BG and mining G or even irrigating both), produces only a few net shields less and makes more gold when researching the same effective number of beakers (i.e. Navigation in eight turns) it is clearly the best pre-sanitation pattern without need for further testing.
The next step would be to see at which point as measured by how many cities more are needed for CXXC to equal the output of CXXXXC. This is much harder to do, as you would have to assign a viable exchange rate between superior science and superior net shield production or make some CXXXXC-cities into a variant of science farms nobody uses until you have achieved a balance between science and shields that satisfies the question asked.
Once a ratio between CXXC and CXXXXC has been established comes the question of feasibility - does CXXC really produce 1.xx times the amount of settlers in a given time?
Now comes the questions about how they do in real game play and it is time to test in a "real situation" where the AI plays in exactly the same manner in order to answer the questions of "How large rex-territories do CXXC and CXXXXC gain respectively?" and "Over a substantial part of the course of the game (including the rex phaze), played in exactly the same way, does one pattern confer any advantages? What are those advantages and are they quantifiable in such a way that it is possible to say that one is clearly superior of the other with any degree of certainty?"
Where many people go wrong in this debate is that they assume that they already know the answers to the questions asked by the last two tests.
If you check Moonsinger's and SirPleb's spacing in their hall of fame write-ups, you'll of course see a mixture of spacings. Still, I think you'll see that the CxxxC pattern fits their spacing best.
Moonsinger's
http://www.civfanatics.net/uploads7/ms_sid5_1000bc_21towns.jpg
SirPleb's
http://forums.civfanatics.com/showthread.php?t=81424
Moonsinger's
http://www.civfanatics.net/uploads7/ms_sid5_1000bc_21towns.jpg
SirPleb's
http://forums.civfanatics.com/showthread.php?t=81424
The question is why did they space that way? Was it to be the most effective in beating the AI? I doubt it.
IMO it is because that is the most effective way of increasing your score quickly, as they were not trying to beat the AI as much as trying to beat the score that anyone else had posted. They are both elite players and could beat the AI with any city spacing that you'd care to mention.
Sir Pleb gives the reason in the write up, it is because of the very low OCN for a huge map at sid level: -
Sir Pleb said:
Similar threads
