Testing out CxxC vs Metros for Space Race ability

[Othniel]

I'm going to be playing a couple of games to test out two theories on how to place your cities.

This thread has extensively discussed the merits of CxxC vs CxxxxC city spacing styles. It's a pretty long read and more posts are being added all the time. Be warned too, some of the posts got a little heated!

If you are wondering what mysterious code "CxxC" is, here's a brief description:
The "C" stands for a tile that a city is founded on
The "x" stands for a blank tile adjacent to the city tile.
The more x's there are, the more blank spaces in between city tiles. Thus, "CxxxxC" is a wider spaced style than "CxxC" because there are more blank spots between city tiles.

The CxxxxC style is synonymous with OCP (Optimal City Placement) or Metros. I will use Metros to keep things simple and easy to understand.

Look at this post to see a summary of the debate on CxxC vs Metro as given by Pyrrhos, the creator of that thread. Some other posters disagree with some of his summary, so read on if interested.


Objective:
The objective of my test is to see what placement style is "better" for a Space Race under the most standard conditions I can conceive.

Better will be defined in this test as the higher Firaxis score and / or game date when the Space Ship is launched, the earlier the better.

Test Format:
I will be playing two games on the same exact map. In the first game, I will be using Cxxc style, in the second, the Metro style. I will be using standard game settings, described below.

Space is the Victory Condition of choice because it lasts into the Modern Age. Pro-Metro folks have conceded that the Cxxc is better for the early game, but argue that Metros catch up later in the game. A Space Race gives enough game time to see if this is true.

Mission Statement:
The objective of my test will not please everyone. I know that. Many people, especially those who embrace the Metro style of city placement, do not play Civilization to maximize their score or win a game as quickly as possible. That's fine.

But I want to do some "testing" on how high each style can score for a Space Race. I will attempt to play each style as efficiently as possible while not violating the spirit of each style. I understand that this "test" is not scientific. There are too many uncontrolled variables and regardless of what the outcome is, people will have legitimate reasons to quibble with the "results".

I merely hope to provide some interesting, possibly compelling, in-game examples of the two styles. Going for the same Victory Condition using the same Map over the course of two games should provide some comparison fodder.

I welcome input both before and after the tests start. If you have suggestions on how to setup the test, let me know. If you feel I'm skewing the test toward one style by doing this or that, let me know and we can bicker.

Some people will care about what I'm doing here, many others probably won't care a bit. This will take me a lot of time to play and post, so I'm doing this for those who care, even if it's just a few of you, and for myself of course! I'm personally interested in the outcome.

onto the game setup.....

 

[Othniel]

Game Settings:

I've incorporated suggestions from Aabraxan, TheOverseer714 and timerover51 into the lists below.

Map and Rule Settings
* Standard continents, 70% water, temperate, normal, and 4 billion years.
* No barbarians
* Emperor difficulty
* Preselected AIs, no Science opponents
* Preserve Random Seed OFF
* Culture Flips ON
* SGLs OFF
* All VC's enabled except wonder VC will be turned OFF
* Culture linked start locations OFF

Civilization
MAYA...Portugal!

Portugal was the final choice for my civ, although civ choice ultimately doesn't matter because...all civ traits were removed anyhow. That means Portugal is stripped of its seafaring and expansionistic traits.

Other Rules and Game-Things-I-Intend-To-Do:
*Once I clear my starting continent, I won't go conquering overseas. I think this hurts the CxxC style more because this means less Science Farms, but I'll go with this limitation for now.

*No suicide ships will be used to reach other landmasses. Suicide ships will be allowed.

*The Republic Sling will be utilized and I will remain Republic the whole game

*No limitations on Trading.

*The same Tech path will used, modified by trades of course, with one exception: I will research Sanitation as fast as possible in the Metro game.

*Wonders to be built: Newton's and Copernicus in one city. Also, ToE, the UN, and possibly Hoover Dam. No other great wonders will be built.

*I will use ICS'd Science Farms extensively for my CxxC attempt. I will use CxxxxC placement only for the Metro attempt. However, for any hopelessly corrupt cities in the Metro attempt, I will convert them into larger Science Farms.

*All culture buildings will be built in any decent Metros. Factories will also be utilized extensively.

Overall, I will be playing Builder style with the Metro attempt while the CxxC attempt will feature much more emphasis on units, workers, and settlers. But in both cases, I will try to play ultra-efficiently. I want to see if the Metro style, played efficiently, might compete with Cxxc more than I think it can.

 

[Othniel]

Links to Turnlogs located later in this thread:

Dotmap and Pre-planning

4000 BC - 3000 BC

3000 BC - 2030 BC

2030 BC - 1300 BC

 

[Othniel]

(reserved 2)

 

[Aabraxan]

Looking forward to this!

 

[Lord Emsworth]

I hope that with CxxC you mean 13 tiles per town, and not the strawman that would restrict you to 9 tiles. (After all, CxxxxC doesn't really mean 25 tiles (4 unusable) per metro either)

ETA:
And about the wonders ... I can see the point in making ToE and UN. I don't really see the point in making the Hoover Dam though. How about, instead of the Hoover Dam, making it a requirement to at least attempt to build both Copernicus's Observatory and Newton's University?

 

[Spoonwood]

"I will build the UN, the Theory of Evolution, and Hoover's Dam (I don't think these wonders will give me a Golden Age ). No other wonders will be built."

Hoover's Dam, by itself, WILL trigger your golden age.

"However, for any hopelessly corrupt cities in the Metro attempt, I will convert them into larger Science Farms."

This means courthouses and eventually police stations, correct?

Lord Emsworth might have a point... even though I always have built Hoover's Dam in my spaceship games, and used factory metros with Hoover's Dam to build spaceship parts, production doesn't get me a faster launch date as long as I pre-build spaceship parts. Copernicus's and Newton's come as more relevant.

As I said elsewhere, I think cxcxxc or cxxcxc also makes for an interesting test... I know... I know.... that's more time and I hear you there... so just a suggestion and I'll stop making it now.

 

[Pyrrhos]

I hope that with CxxC you mean 13 tiles per town, and not the strawman that would restrict you to 9 tiles. (After all, CxxxxC doesn't really mean 25 tiles (4 unusable) per metro either)

ETA:
And about the wonders ... I can see the point in making ToE and UN. I don't really see the point in making the Hoover Dam though. How about, instead of the Hoover Dam, making it a requirement to at least attempt to build both Copernicus's Observatory and Newton's University?

:suicide:

Still trying to swing things your way I see, "damn your eyes Igor!"

First of all, no more than twenty tiles are useable per town no matter how much space you give it.

Second, where did you get the figure of 25 from?

Third, now go grab a pen and paper ffs and draw the respective grid maps for the two configurations and work out MATHEMATICALLY how many usable squares there are per town/city/metro with CXXC and CXXXXC respectively.

Fourth, if you don't understand the reasons behind Othniels choice of wonders, don't draw everyone's attention to the fact.

Fifth, if you are going to object to the way people who actually take the thankless task of performing such an extensive test upon themselves set things up, go make your own tests! If you do, you can draw the conclusions that you are happy with and present them whichever way you want. Thing is, I wouldn't trust your objectivity.

Aaarrrrrgh!

 

[Lord Emsworth]

First of all, no more than twenty tiles are useable per town no matter how much space you give it.

Yes. (21 tiles actually, if you count the CCs as usable)

Second, where did you get the figure of 25 from?

Third, now go grab a pen and paper ffs and draw the respective grid maps for the two configurations and work out MATHEMATICALLY how many usable squares there are per town/city/metro with CXXC and CXXXXC respectively.

Which versions of CXXC and CXXXXC respectively do you want? Here are the strawman versions for each of the patters:

Strawman-CxxC:

XXXXXXXXXXXX
XCXXCXXCXXCX 
XXXXXXXXXXXX
XXXXXXXXXXXX
XCXXCXXCXXCX
XXXXXXXXXXXX
XXXXXXXXXXXX
XCXXCXXCXXCX
XXXXXXXXXXXX

9x12=108 tiles
108 tiles for 12 cities means 9 tiles per city

Strawman-CXXXXC

OXXXOOXXXO
XXXXXXXXXX 
XXCXXXXCXX
XXXXXXXXXX
OXXXOOXXXO
OXXXOOXXXO
XXXXXXXXXX 
XXCXXXXCXX
XXXXXXXXXX
OXXXOOXXXO

10x10=100 tiles
100 tiles for 4 cities means 25 tiles per city (4 tiles per city aren't usable, 16 overall)


If you insist on using a strawman version for CxxC, why not also stick to the strawman version of CxxxxC as well. Or, better yet, use fair representations for both? (I leave it up to you to draw up more fair representations for both patterns. They are both available in the CxxC vs CxxxxC thread.)

Fourth, if you don't understand the reasons behind Othniels choice of wonders, don't draw everyone's attention to the fact.

Othniel's attention would suffice.

 

[Optional]

Othniel, in what type of sequence are you going to play these two games? If you would play 10 turns in one game first, you would unavoidably take that terrain knowledge into the second one. This can make a crucial difference. Later in the game it might not matter that much, if the whole world is known, and each game has gone its own way.
But in the early game, are you going to play each game one turn at a time? That seems a bit tedious, but is probably needed.

 

[Optional]

Err...after a bit of further thinking, one turn at a time is probably not good. It's probably best to start with one turn in one game, and from then on always two turns at a time, swapping between games. Then the games alternate in being ahead, and it's not the same game always that's ahead.

Couple of more thoughts:
You might want someone to generate a map for you. You'll want something average. The random map generator can give you pretty weird stuff, even when choosing average settings.
You'll probably want two more or less equal continents, with an even distribution of lux. And you'll want a start position that suits both CxxC and CxxxxC.
And I would strongly recommend, if you're going with the Mayans as your civ and are not planning to build the Javelin Thrower, that the Acher is checked as a build option for the Mayans. That allows you to build a 20 shield attacking unit, which is much more like it would be in an average game.
Another possibility would be checking 'no civ specific traits'. I don't recommend doing that. The game would just lose flavour for the sake of appearing to be more scientifically sound. You're never gonna get it completely scientifically sound, so why sweat it?

I hope you're gonna have fun with this in the first place. If you would like someone to generate a map for you, with an appropiate start position checked, and maybe a few things manipulated in the editor, I'm happy to do that. I brought it up in the first place, so it's only fair that I put my name down. But no problems if you want to do it your own way. It's your experiment. I'm looking forward to following this!

 

[timerover51]

Othniel. I already have a possible map generated that would likely work, with two roughly equal continents. There would be a minimum of 4 different luxuries on each continent with a minimum of two of each type. All strategic resources would be on each continent, again with a minimum of two of each type. If more than two are generated, I would not take any out. Some editing of terrain would be done to avoid a massive block of forest/jungle/mountain terrain skewing the outcome. Probably would be best to have someone review the map before you play it to insure a fair outcome.

I would follow Optional's recommendation and give you the archer instead of the Javelin Thrower. No sense limiting you unduly against the AI.

I will follow the test with interest, but will likely continue to play the way I have been. I am curious as to the results.

 

[AutomatedTeller]

well, he did say that he'd not have a golden age, and hoover's does create a golden age.

I think his point is valid about the tiles worked - I've never played CxxC where I don't have 12 tiles per city.

Generally, I use a CxxC where the towns are offset, just like you would with CxxxxC to get all tiles used with a 1 tile overlap between cities.

 

[Othniel]

Lots of good questions / suggestions have been raised. Thanks folks!

Regarding wonders...

@Pyrrhos,
Thanks for being considerate of my time and efforts into putting this test together. However, I did ask for input and Lord Emsworth's questions are legitimate and helpful.

Thanks also to those who pointed out Hoover's would give a Golden Age. I remembered it was an INDU wonder, forgot it was an AGRI one also. That means Hoover's is out.

FWIW, I generally build Hoover's for Space race games, especially if I feel I need to still do some warring late in the game. I'm a bigger Factory fan than some other CxxC folks, I think. Perhaps my affection for Hoover's and factories is a holdover from my more builder-ish days. Even in my Deity solo games, I sometimes get "lazy" in the late game when I know I've won and start building a lot of unnecessary improvements. Probably hurts the score a bit.

@Lord Emsworth
Why would I build Cops and Newton's? I realize they boost science, but that doesn't seem like enough reason in and of itself for a test like this. Is it to see how well CxxC can build wonders or is there another reason I'm missing?


Regarding city spacing...

Basically, I don't intend to nerf either style by using a strict grid. I use a looser CxxC in my solo games, often with a little CxxxC thrown in, so I intend to do that for this game. Most core cities should get more than 9 tiles. Of course, the same will apply for my Metros. They will be optimally placed, generally, but they will overlap a little so that tiles aren't wasted.


Regarding how I play the turns...

One of the biggest drawbacks to my test format is that I play the same map twice. This means advance knowledge, etc. To mitigate this drawback, I intend to preview the map carefully before I play my first attempt. I'm not actually sure what this will do to my strategy because I've never seen a map ahead of time in any of my other games. It will be interesting I'm sure.

Optional's suggestion that I play the games virtually simultaneously is interesting. However, my personality is not split enough to keep the separate strategies clear, we thinks.

So, my thought was to play one style through from start to finish. That would give me a benchmark in the second game to shoot for. That way I can clearly focus on playing each style to the best of my ability. Is this a bad idea?

Regarding the map itself...

I'd like to see your map, timerover. It'd be nice to have Optional review the map also to give the stamp of approval.

Great suggestion on modding the game to give me archers.

Agreed on editing out any huge clumps of jungle or march. Keep some of those tiles in play, this is a normal obstacle, but a nasty map won't be much fun even if doesn't skew the results (not sure it would). Desert tiles are fine to keep because I'm AGRI.

Thanks to all you who have expressed interest in this test!!

 

[Spoonwood]

Othniel,

"I realize they boost science, but that doesn't seem like enough reason in and of itself for a test like this. Is it to see how well CxxC can build wonders or is there another reason I'm missing?"

No, it isn't to see how well cxxc can build wonders, and yes you've missed something. Remember that Copernicus's, Newton's, and SETI (you should build or try to build all three... not just too of them), add 100% research to the base rate. The base rate gets determined by the commerce you have going towards science from the tiles worked in the city. For a cxxc city you'll have 12 or so tiles worked, while for a full metro you'll have many more tiles worked. Consequently, Copernicus's, Newton's, and SETI all produce more beakers from a metro than a cxxc city, once you have the hospital of course. Also, The Colossus (O.K.... you probably won't build that in either... then again you might capture it) also has a greater effect for a metro than a cxxc city, just not in quite the same way. The Colossus adds to total commerce output of the metro more than the cxxc city. So, for any metros housing these improvements *for that city or those cities* those wonders work out better than for a cxxc city. The more interesting question comes as do these benefits come out as significant empire-wide?

Also, I'll ask that you post a screen-shot of your empires before you start taking any enemy territory. I'd like to know have some basis of information as to how cxxc vs. metro spacing works for players who simply don't want to go to war. So, that would give me at least something.

 

[TheOverseer714]

If you play the CxxC first, the OCP game should have the advantage by being second. Either the Maya need tweaked, or you could choose a different tribe that Hoover Dam doesn't generate a GA. It is a favorite of OCP players, so we wouldn't to deprive that game of such a useful wonder. IIRC, there is a check box that keeps GAs from occurring, maybe our modder friends can confirm or disprove that.

 

[Pyrrhos]

Easy!

Just uncheck all traits from the wonders in question using the editor (custom rules +> edit rules +> wonders and improvements)

 

[Mursi lives]

100% research to the base rate. The base rate gets determined by the commerce you have going towards science from the tiles worked in the city. For a cxxc city you'll have 12 or so tiles worked, while for a full metro you'll have many more tiles worked. Consequently, Copernicus's, Newton's, and SETI all produce more beakers from a metro than a cxxc city, once you have the hospital of course.

Right. avoiding science wonders, especially Newton's, may help CxxC. The Colossus,OTOH doesnt mean anything since it is usually captured in 90% corrupt area

Easy!

Just uncheck all traits from the wonders in question using the editor (custom rules +> edit rules +> wonders and improvements)

Wouldnt that also cancel AI GAs? No GA for AI means easier and faster conquest of your own continent. May somewhat help metros.

Maybe Mayas are not the best civ for that test. How about Portugal? None of the GW we want to build is seafaring and the UU wont cause any problem. It may even help you to get decently timed contacts since you wont do any suicide run.

 

[Spoonwood]

"Right. avoiding science wonders, especially Newton's, may help CxxC."

*scratches head and makes that sound an impi makes*. First off, for a spaceship victory, avoiding science wonders will NOT help cxxc to research faster... unless you can take as much territory for science farms from the units you trained instead of building say Newton's university. Newton's university, no doubt, gets built in a rather uncorrupt town . If we have our research at 100% (and you probably can on emperor from the industrial age on if you trade right), for a size 12 city you have about 26 base commerce. This means Newton's gives you about 26 more beakers. You have to have 3 specialist farms at size 6 to equal this. Newton's and Copernicus's cost 400 shields, which equals about 6 knights or 5 cavalry. Will 5 more cavarly yield 3 more specialist farms? Not alone, because you need three more settlers which cost a total of 90 more shields and 6 population points. So, I don't see how one can say that avoiding building Newton's and Copernicus's helps cxxc. I said that Newton's and Copernicus's benefit cxxc less science-wise than they do for a metro-spacing.

"The Colossus,OTOH doesnt mean anything since it is usually captured in 90% corrupt area"

Given a decent starting position, one can build the Colossus in one's second town on Emperor... the only thing comes as that building it might not work as compatabile with building some warriors/scouts then a granary and THEN FINALLY settlers in the capital. When I've built the Colossus on Emperor I've usually built maybe a granary, then one or two settlers, and then the Colossus.

"How about Portugal?"

A seafaring civ gets commercial benefits from coastal cities in the center square. The more coastal cities you have, the more commercial benefits you get from the sea-faring trait... so, in terms of science, seafaring favors cxxc over metro spacing.

 

[Pyrrhos]

Wouldnt that also cancel AI GAs? No GA for AI means easier and faster conquest of your own continent. May somewhat help metros.

Nope, you only edit the Hoover Dam to be trait neutral. Since a) he plans to get it, that won't affect the AI civs anyway, and b) The AI civs will most likely have had their GA much earlier than that through their UUs.