CXXC or OCP - An Investigation

Pyrrhos · Mar 16, 2008

Chamnix said:
I understand that CxxC is going to be better - my issue is with the claim that the distance corruption of OCP is the relevant factor. The main reason that CxxC is better is not lower corruption.

I'll try to post something in a couple hours that demonstrates clearly where I think the difference comes from ...

Then take a look at the 4000BC sampling again:

OCP corruption 7/35 or 20.0%
CXXC corruption 8/42 or 18.6%

OCP waste 9/35 or 25.7%
CXXC waste 8/38 or 21.0%

At 3000BC, the figures were:

OCP corruption 21/71 or 29.6%
CXXC corruption 18/80 or 22.5%

OCP waste 22/73 or 30.1%
CXXC waste 13/66 or 19.7%

Since the number of towns is identical, so is the rank corruption. Rank corruption can therefore be ruled out as the explanation for the higher corruption and waste of the OCP configuration. The only explanation for the higher corruption and waste of the OCP configuration is therefore distance corruption.

The difference in income can be attributed to one factor - the location of each town. As an example, look Nineveh! In OCP it is located on top of the plains + sugar tile, in CXXC it is next to it. That is one source of the higher CXXC income.

chuckiferd · Mar 16, 2008

will there be a CxxxxC test? And why hasn't anyone thought of CxxxC? It would be partial to both CxxxxC and CxxC

also I have some questions about the test

(in theory) OCP and even CxxxxC really shines in the end game why didn't you play it all the way through

I think 18 cities and workers is too much to start off with, it gives the civ with less production power (the one with the ability to make less settlers and workers) an advantage over the one with more production power

EDIT: I have an idea, test a Communist OCP or CxxxxC versus a ____________ type CxxC Communism would work perfect as it eliminates the distance factor of corruption

Tribute · Mar 16, 2008

Exactly, chuckiferd. CXXXC rules all! We already know that OCP became unwieldy after rank corruption was changed in C3C. And that you won't be getting very far in terms of territory if forcing yourself to expand CXXC only. CXXXC allows for you borders to connect without needing cultural buildings.

Now, that test was rather unfair, wasn't it? CXXXXC being forced to make temples.... Doesn't anybody else think that the upkeep was a rather LARGE and GLARING disadvantage? But it is a reality that CXXXXC NEEDS TEMPLES. (By CXXXXC, I do refer to OCP since you'd be foolish to think that leaving unworkable tiles everywhere in your empire could possibly give you an advantage.)

However, neither CXXC nor CXXXC need them. CXXC because the borders automatically touch. And CXXXC because the borders (artificially) extend if two borders of the same civ are one tile apart. CXXXC gets the selection of more tiles than CXXC without needing culture, something that CXXXXC cannot avoid.

Someone really needs to compile a table of advantages for each settling style. And also, I suggest running this test with my CXXXC with diagonal CXXC pattern. *looks at Pyrrhos*

CXXXC with CXXC diagonals looks like:
-----E-------
---------E---
-------------
--E----------
------E------
----------E--
-------------
---E---------
-------E-----
(of course, this is more vertically stretched than it really is. But you can see that there is one set of CXXC cities, all diagonal. The above shows 6 cities (E) around a center city, the hexagonal pattern)

Anyway, nobody really builds in an unnaturally fixed pattern of squares (CXXC), hexagons (CXXXC with CXXC diagonals), or large hexagons (OCP). I just build so that each city has a fair amount of tiles that it can use so long as no tile is left unworkable (unless it's an unnecessary desert tile). Grab what's good when you can. And don't feel constrained to these foolish patterns.

Moving on, I really don't think the test was very fair. Settlers actually have to walk to their respective settling sites, and who knows? Maybe it's rather difficult to expand while doing OCP. (How long would it have taken them to make 18 extra settlers compared to CXXC, anyways?)

chuckiferd · Mar 16, 2008

why is everyone so obsessed w/not leaving unworked tiles in between cities? The concept of it seems so wierd the more you expand outward the more land you earn right off the bat then you can fill in the gaps later or simply wait for culture expansions to fill it in. Culture is not a neccessity, I say run the tests again with EVERYTHING the exact same, meaning OCP doesn't need temples

PS. If a enemy unit builds in between your cities either trade for it, wait for it to forfeit to you, or capture it. The submission to your culture while surrounded is pretty inevitable, or take the shortcut and put that treasury sucking standing army to use.

ZzarkLinux · Mar 16, 2008

OCP doesn't need temples

What do you mean by this? I may agree / disagree based on what you mean. Please clarify.

a) You mean OCP = settling the best city sites. Going the extra square to get the bonus wheat, etc...
because the wheat is RIGHT NEXT to the city. No culture expansion needed.
- I totally agree. OCP and CXXC are ideals, not strict rules, and things will vary on terrain.

b-1) OCP means you don't need temples to work all 21 squares in city radius.
- This is untrue. The very definition of OCP is that each town can work 21 squares without clash.
- This is easily disproved. When you build your capitol, how many squares can it work on turn 1?? 8 or 21?? Only 8.
- The palace provides culture...

b-2) You are waiting for libraries to expand culture...
- You're waiting too long... CXXC would have an extra town doing work while OCP just sits there, waiting...

the more you expand outward the more land you earn right off the bat

Notice that I play higher difficulty, and I run out of land really really fast.
The effect of 1 extra city getting 6 extra tiles is non-existant... Cause then the AI's knocking at your door.

I'll close with the beautiful thing you said in the other thread:

I would use tight placement, when all available land is used up, then I would go back and build cites in between my older cities

I totally agree. :thumbsup:

Settle the best spots first, then fill in the rest

CXXC or OCP don't fully embrace this concept... You gotta "work-it"

chuckiferd · Mar 16, 2008

What I mean is that people shouldn't fret over leaving unclaimed tiles, they would probaly not be getting worked untill the nation is a metropolis anyways. So according to that you do not need those expensive temple in the test. Removing them would get a much more acurate result

Aabraxan · Mar 16, 2008

chuckiferd said:
will there be a CxxxxC test? And why hasn't anyone thought of CxxxC? It would be partial to both CxxxxC and CxxC . . . .

As others have noted, CxxxC has some of the advantages of both CxxC and CxxxxC. The arguments, both here and in the other thread, both fall prey to one problem: terrain. Both CxxC and CxxxxC sound great in practice, but when you fire up the game, there are rivers, mountains, deserts, dyes, etc. to deal with. You grab the best tiles first. What part of it boils down to though is: If I get to choose between wider and looser spacing, which do I choose?

chuckiferd said:
. . . .(in theory) OCP and even CxxxxC really shines in the end game why didn't you play it all the way through

That's a big part of the aforementioned discussions. You are correct, though, in that CxxxxC's strength is late game. Unfortunately, that means that, when I'm busy settling cities during the expansion phase, I have to leave a lot of tiles open that I could be using if I used a tighter expansion. Those cities without fresh water will only be able to work 7 (6+ city center)tiles, maximum, until aques are constructed. Those with fresh water or an aque will get to work up to 13 (12+ city center)tiles. None of them, however, will get to work more than 13 until I've built hospitals. Further, Sanitation is an optional tech, which I will rarely research. By the time I'm even capable of researching it the core of my empire is already laid out.

chuckiferd said:
EDIT: I have an idea, test a Communist OCP or CxxxxC versus a ____________ type CxxC Communism would work perfect as it eliminates the distance factor of corruption

I can't tell if the blank in this is an expletive that you didn't want to post or if there was some other adjective that was supposed to go there. :lol:

Like Sanitation, though, Communism is an optional tech that I rarely research. There may be some games in which Communism is outstanding, but I don't usually want to deal with a second anarchy.

If you start the test with full-grown metros and communism, I strongly suspect that the Communist CxxxxC empire will be stronger. But that eliminates the progression of the game up until that point.

chuckiferd said:
why is everyone so obsessed w/not leaving unworked tiles in between cities? The concept of it seems so wierd the more you expand outward the more land you earn right off the bat then you can fill in the gaps later or simply wait for culture expansions to fill it in. Culture is not a neccessity, I say run the tests again with EVERYTHING the exact same, meaning OCP doesn't need temples

Culture may not be a necessity, but for anything over size 12, sanitation or Shakespeare's is. That means that until the Industrial Age, no city will grow past size 12. If you're leaving

Why are some of us so obsessed with not leaving unworked tiles? I don't claim to speak for anyone else, but I see unworked tiles as untapped potential. Potential food, shields and gold that could be turned into settlers, military and beakers. Why should I let a perfectly good grassland sit there, unworked, until I can get around to building a hospital?

chuckiferd said:
PS. If a enemy unit builds in between your cities either trade for it, wait for it to forfeit to you, or capture it. The submission to your culture while surrounded is pretty inevitable, or take the shortcut and put that treasury sucking standing army to use.

You can't trade for cities in C3C. I think you could in one of the early versions, but I believe it's been patched out. Someone correct me if I'm wrong.

Culture flips -- Well, culture isn't my strong suit, but even in the few games where I've tried to build lots of culture, I found that culture flips to me were pretty rare. I usually take the "shortcut" that you mentioned.

chuckiferd said:
What I mean is that people shouldn't fret over leaving unclaimed tiles, they would probaly not be getting worked untill the nation is a metropolis anyways. . . . .

Here's where you're wrong. As a general rule, every tile in my core and semi-core will be worked, unless I just cannot generate enough food to support someone working it. And I will put them to use as fast as I can. I can't understand leaving tiles empty just because I haven't gotten an optional, Industrial Age tech researched yet.

Gosnork · Mar 16, 2008

chuckiferd said:
What I mean is that people shouldn't fret over leaving unclaimed tiles, they would probaly not be getting worked untill the nation is a metropolis anyways. So according to that you do not need those expensive temple in the test. Removing them would get a much more acurate result

Which is why you use CxxC so that those tiles WILL be used. There is NO reason to let them sit there, if it is in your territory, USE IT! Why would you build loosely spaced cities and NOT build culture, instead of building CxxC. There is no reason to place cities CxxxxC and not build culture. You might as well place cities CxxC and get twice as many cities and almost double production.

Chamnix · Mar 17, 2008

Here are the settings I used in generating a random map:

C3C v 1.22
Playable civ – Babylon (just because that is what Pyrrhos used above).
Level – Deity (because the higher levels have higher rank corruption but no difference in distance corruption – also, the temples are “needed” sooner than on a lower level)
Map – Standard, 70% water, continents, average everything else, no barbarians
Rivals – just one, so hopefully we start on different continents, and we won’t meet each other.

Here is the start I got.

Spoiler First 23 turns :

At this point both games are identical. The crappy world so far:

I didn't stick to a strict settling pattern for either. I also micromanaged both roughly equally incompetently. A=tight settling, B=loose settling.

Spoiler Remainder of QSC :

Spoiler Final Stats :

Pyrrhos · Mar 17, 2008

Commendable thorough, Chamnix! I have a few questions though:

First, city placement. Why settle Nineveh where you did? My choice for both "tight" and "loose" would have been the wood tile at the river's mouth between the wheat and the fish (west of Akkad). What a powerful settler pump that would be in either configuration!

- One irrigated wheat grassland
- One mined grassland
- One fish
- one wood/mined hill once the wood's been cut

With this location for Nineveh, Babylon would become a spearman pump, occasionally producing a settler to keep it from growing too large (unhappiness > clowns/slider = nono)

Second, from what I can see and read, you did not chop woods to help the building of granaries & temples (except two near the "tight" Akkad). This would have saved you MANY turns. Furthermore, it's going to hurt "loose" more.

Now, do you have the 4000BC start file?

Chamnix · Mar 17, 2008

Since both starts were played identically through 2850 BC, I thought people might prefer to have the save with the first 23 turns already played for both, but I have attached 4000 BC as well.

As far as why I settled Niveneh where I did, I was playing the map with no foreknowledge - I didn't know the wheat was there until after both Niveneh and Ashur were founded. I settled Niveneh where I did because it was the highest food location I knew about at that point, and food is :king:

.

I didn't do much chopping in the loose settle-pattern just because my workers seemed always to have more important tasks. I gave priority to connecting cities and making sure all worked tiles were improved before chopping.

Pyrrhos · Mar 17, 2008

Thank you for the save!

Don't you send out your first two or three warriors in order to scout out the terrain for just the purpose of discovering where to settle next most advantageously?

Chamnix · Mar 17, 2008

Obviously I don't do a very good job of it :lol:

. I did have 3 warriors out scouting, but none of them went in that direction :cringe:

.

Most of my games, I use a very low military start for a very quick expansion - I typically only build 1 warrior (sometimes 2 warriors) before starting my first settler or a granary prebuild, and those first warriors are more concerned with finding contacts and getting a feel for the overall layout than the immediate surroundings. Because I believe strongly in fairly tight spacing and sharing tiles, my second settler will usually settle just outside my capital's expanded cultural border, and it will be using many of the same tiles my capital will be using, so I usually have a good idea where my second town should go without a lot of exploration. The second town often builds a warrior first, and he will check out the immediate surroundings.

Because of the low-food start (and because I was making an effort to play a more "standard" game), I ended up with more warriors, and I guess I had no idea how to use them all effectively :crazyeye:

. I find low military/exploration on average will get me a much faster start than exploring fully. This map happened to be one where my strategy cost me :dunno:

.

Pyrrhos · Mar 17, 2008

Err...

The save you posted looks a tad different...

MAS · Mar 17, 2008

Uhm, Chamnix, your BC4000 save appears to be a different game... I mean, I get a completely different start.

EDIT: beaten to it by Pyrrhos...

Chamnix · Mar 17, 2008

Sorry about that - I'm not sure where that came from :blush:

. I won't be able to correct it for at least another 6-7 hours though.

Tribute · Mar 17, 2008

@ Chamnix. Good work!

Okay, so we all know that CXXC gives better expansion, better military, and better gold/research in the early game. I mean I hope it would seeing as how in the early game, there's no way you can run out of 'good' tiles to use. And if you do, the effects are minimal. Then there's also lower distance corruption.

Great! But that's the early game. Some people have said that the early game dictates what happens in the future so much so as to outweigh the eventual benefits of CXXXC. Let's see that.

But first, how about actually playing a game where you can include diplomacy? I know it can add so many other random factors (and would totally suck since you'd have to check diplomacy so much,) but what about the greater probability of gaining contacts if expanding CXXXXC? All benefits should really be tested.

I'd suggest getting people to sign up for a couple of saves and having each person play out their game according to both styles turn by turn. That way, foreknowledge of the map and differing playing styles (if having different people do each trial) all cancel out. Then we can eliminate luck and chance by looking at the general CXXC and CXXXXC characteristics. If we can go all the way through the middle ages to victory then that would be great! Although, as people have mentioned, GOTM games are good for this....

Chamnix · Mar 18, 2008

Posted correct 4000 BC save...

Snarkhunter · Mar 18, 2008

Tribute said:
@ Chamnix. Good work!

But first, how about actually playing a game where you can include diplomacy? I know it can add so many other random factors (and would totally suck since you'd have to check diplomacy so much,) but what about the greater probability of gaining contacts if expanding CXXXXC? All benefits should really be tested.

It isn't clear to me that the probability of early contact is greater with CXXXXC, if for no other reason than I found a city--and get an extra tile's range--3 turns sooner under CXXC, assuming unroaded territory beyond my old boundary. But much more important is the fact that virtually none of my contacts ever come from strangers bumping into my empire; they all come from the warriors/curraghs I send walkabout through the land. YMMV.

The central problem in this discussion is that it is not clear how to weigh the precise value of a tempo theoretically. In chess, gaining a tempo is precious; it is easily lost again, so it is overwhelmingly important to find a way to translate it into a permanent advantage. In go, this happens more or less automatically, unless one plays out-and-out foolishly, because go is an additive game. CIII seems more go-like in that respect, but it is possible to waste tempi. The overall point is that using tempi wisely tends to work like a savings account: soonest begun, soonest getting to a ripe retirement fund--and playing catch-up is exorbitantly expensive.

kk

Pyrrhos · Mar 18, 2008

I think that a few are confusing the test with a real game. With the way the test was set up, the prohibition on exploration combined with the larger territory meant that OCP was more likely to get the early contact. In a real game, I agree with you that most contacts are the result of my own units exploring the map. Either they encounter AI units or find their territory, that is my experience. Occasionally, an AI scout or warrior "knocking at my door" is the first I see of another civ, most likely a far-away one.

CXXC or OCP - An Investigation

Vae Victis

Chieftain

Not Sarcastic

Chieftain

Engineering Programmer

Chieftain

Mid-level Micromanager

Warlord

Chasing Time

Attachments

Vae Victis

Chasing Time

Attachments

Vae Victis

Chasing Time

Vae Victis

Deity

Chasing Time

Not Sarcastic

Chasing Time

Prince

Vae Victis

Similar threads