This is NOT Pangaea

[xxFlukexx]

So I start a game of Civ4 in a good mood, lalala, dreaming of happy strategy gaming goodness.

Instead I find this incredibly hostile world, my brothers in faith apparently hate me(even though the only minuses we have is - close borders, I have more + than - with everyone) I war with Montezuma in the B.C.s, raze one of his cities and before my units can turn around, Alex attacks then Tokugawa joins in.




So I decide to give up and hit the World Builder.



What the heck is this, this is NOT Pangaea at all, it's some freaky map that I do not want to play on. I guess I should've been tipped off when I made contact with 6 nations by 3600 B.C. But oh well, maybe it was user error, let's check the settings:



Nope, Emperor, Pangaea, medium sea level, normal speed, all defaults on, just how I like it. I do hit pressed shoreline because I *think* that it gives a more solid line instead of small 1-2 tile islands breaking off from the main Pangaean Island which I despise.

I have had this happen before, but not to this extent, that is a huge landmass that should just plain not be possible when selecting Pangaea. I'm starting to wonder if I should open the WorldBuilder(and I guess never be in my own HOF which is disappointing) at the start of every new game just to make sure I don't get some ridiculous map.

I remember the random world generator in Civ2 and Civ3 being pretty good, or is my memeory wrong, was the random world builder always junk? Is there anyway to fix it?

[Sirian]

Hi. I'm the designer and programmer for the Pangaea map script.

I'm sorry that you had this problem. It's the first instance of pangaea failure that I've heard of since release. You may be interested to know that the Civ3 pangaea generation produced more than one large continent more than a third of the time, so Civ4's pangaea is a definite improvement in that department.

I fired up Civ4 and tested Pangaea on standard with Pressed shoreline. I've attached a picture of the result I got, which is typical for that setting. (Nothing wrong with it.)


Knowing the code, the anomaly you show should not be anywhere near remotely possible on the Pressed setting. It might be possible on the Solid shoreline setting, though, if you got extremely unlucky.


I can think of a few things that might possibly cause your anomaly. I'll list them in order.

1. You chose Solid shoreline instead of Pressed, and drew the worst anomaly on it I've ever seen (extreme luck of the draw!)

2. Any modifications you may have made to the XML.

3. A new bug introduced in the 1.52 patch that interferes with the "Best Strip" recentering process for the core fractals.


Item #3 should show up readily, though, in running tests on the script at the same settings. (So I don't think that is the explanation.) If you haven't made any edits to the XML, then Item #1 looks like the most likely culprit.

Solid pangaeas regularly have a small piece or two broken off the main landmass, but I've never seen any that large. AFAIK, nobody inside Firaxis is doing anything new on maps in the patches.


If ANYBODY runs in to any more games like this from Pangaea, please post about it, including the settings that produced the anomaly and whether or not you have modified anything with the scripts or XML settings.

If the error rate is significant, then I'll see about adding more error handling to prevent anomalies. Or if there is a new bug, I would want to track it down. However, if the error rate is below 1% (less than one in a hundred games) then you will have to live with it. Because 1% is a heck of a lot better than the 35-40% failure rate evidenced in Civ3.

So far, this is the first anomaly I've ever seen of this type on this script. So that's still in the "insignificant" error rate category.


- Sirian

[Sirian]

Re: This is NOT Pangaea

Originally posted by xxFlukexx
I do hit pressed shoreline because I *think* that it gives a more solid line instead of small 1-2 tile islands breaking off from the main Pangaean Island which I despise.

That definitely sounds like you chose Solid, not Pressed. (As you can see in my attached image, Pressed will sometimes create tiny islands off the mainland.)

The routine that produces Solid Shoreline is different, and has less error handling than the routine that produces the Natural and Pressed varieties. I wrote the latter, while Soren wrote the Solid routine. Since you are playing nothing but Solid, then maybe the error rate there is higher than I thought. (I've seen dozens and dozens of Solid pangaeas with no anomalies of the size you showed, but... maybe the error rate is higher than I realize.)

Again, if anybody sees this type of problem, report it here. If it's common enough, then it may warrant attention.


- Sirian

[Enigma_Nova]

Oh my god, a Developer! Not just a Mod, not just a Staff, a Developer!
:
In other news, I'll go generate loads of maps and see if I can find an anomoly similar to that.

[Addled Platypus]

could I bother you with a question Sirian

What is "tilted axis" map option ?

thank you in advance

[Enigma_Nova]

Take that. 6th map I generated.
Pangaea / Standard / Noble / Solid / Temperate / Medium sea

[Enigma_Nova]

And to prove I'm not lying, here's the save.

[Syllogism]

My fourth attempt on same settings produced similar results

[Aro]

Originally posted by Sirian


So that's still in the "insignificant" error rate category.


- Sirian

Not anymore, I see...

[SCarton]

Sixth map here had a significant island. The first map I generated had two medium sized islands off the coasts of the pangaea, but I don't know what "counts" and what doesn't.

After that point, I decided to look at it with a more analytical eye (i.e. size of island(s) and so on). For the sake of argument, an Ice tile in the ocean does not count, to me, as an island. Island sizes are (width x height) and is bound by a rectangle that encompasses all the land the island contains. In other words, the size will contain some water tiles, but I really can't be bothered to count each and every individual tile.

Map settings were Pangaea, Standard, Noble, Solid, Temperate, Medium sea. If for some reason it matters that I was playing as Mansa Musa, then I'll include that as well.

I generated 50 maps. Of the 50 maps, 20 had islands, for a rate of 40%. Even if you want to ignore all islands with either a height or width of 1, 16 of the maps still had islands.

Three maps had two islands, but in each case, one of the islands had a height or width of 1. No map had more than two islands.

Here's the raw data I collected. 12 (5x7) means that it was the twelfth map I generated and the island was 5 tiles wide and 7 units high:

No islands: 1, 2, 4, 5, 6, 8, 10, 11, 13, 14, 15, 16, 18, 19, 21, 22, 23, 27, 30, 31, 32, 38, 40, 41, 42, 44, 46, 47, 48 ,50

One island: 3 (1x1), 7 (13x9), 9 (1x1), 12 (5x7), 17 (7x9), 20 (1x1), 25 (18x13), 26 (4x7), 28 (7x9), 33 (14x11), 34 (8x18), 35 (1x2), 36 (8, 37 (1x1), 39 (9x9), 43 (7, 49 (9x24)

Two islands: 24 (3x1, 8x5), 29 (1x1, 6x13), 45 (1x1, 9x21)

More: none

[Solver]

The Pangaea map script is supposed to sometimes give islands that are small/not inhabited. However, a bigger island/continent with a civ starting on it constitutes a malfunction.

[Diadem]

I don't think there are civs on those islands. Even the big one in the starting post of this thread. Those are all barb cities!

However, it's still a malfunction of the script imho. A Pangea map having some small islands is okay, but there's a difference between an island and a continent.

Looks like the 'solid shore' script is bugged Sirian.

[Diadem]

By the way Sirian, could you take a look at this thread? It's a about a different kind of bug in the map script. Not a very significant one I guess, but still

[Solver]

Yes, the original post map is definitely a malfunction. A civ starting not on the main continent is a malfunction, and a second continent actually being somewhere is also a malfunction. I've never seen Pangaea malfunctions myself, I'd only see small islands... but it looks like we do have a somewhat reproducible problem now.

[Saurus]

I always play pangea and always solid and can say from my experience that it is uncommon, but not rare, that under these settings there is a large continent - where everybody sits on + one smaler continent. (not smaller islands but a real small continent)

When playing this usually becomes obvious pretty quickly when, as xxFluexx pointed out, everyone has made contact with everyone in a few turns already as if there were not enough space on the main continent witch, as a matter of fact, IS the case.

However what xxFlukexx had was extreme - I've newer seen that before.