Arrakis.py map script discussion

[Ahriman]

7 comes from spice on polar desert waste

I thought that spice wasn't supposed to be able to spawn on polar desert waste?

If not, it shouldn't be able to. Too much water. This should be part of the tradeoff for settling near polar terrain; less spice income.
I think 1w1h1c is too low for the polar desert waste, it should be noticeably better than normal rock tiles IMO.

There is not much use for polar terrain in dunipelago. You cannot found cities there because it is only one tile wide. (Or at least, the AI will not found there. I haven't tried manually.)

The polar terrain is plenty powerful in Dunipegalo; half the civs end up starting near the poles (because the location picker favors areas with high water income), and cities that are 2 tiles away (and so get 3 polar tiles in their BFC, and 5 polar desert waste tiles) still grow to be monsters. And they're actually better than a pure polar city would be, because they still have access to some normal tiles for hammer income.

The initial colonization race is normally the players at the poles colonizing east and west along the very edge of the map; they only start colonizing down towards the equator once the spots with access to the polar regions are taken.

Basically, the poles are very powerful on Duneipegalo.

See the screenshots from the last decent game I played 2 polar super-cities, population 18 with huge economies, when the rest of my cities were size 8-12.

 

[davidlallen]

I thought that spice wasn't supposed to be able to spawn on polar desert waste? If not, it shouldn't be able to. Too much water. This should be part of the tradeoff for settling near polar terrain; less spice income.
I think 1w1h1c is too low for the polar desert waste, it should be noticeably better than normal rock tiles IMO.

There is no special restriction for spice on polar coast today. But, perhaps it is a good idea. Rock is 1h and regular desert waste is 1h1c, so 1w1h1c seems sufficiently better than either.

Thank you for pointing out that there are polar supercities on dunipelago too. I suspect this is the source of the huge commerce incomes you are seeing, so reducing polar waste and removing spice may help.

 

[Ahriman]

But, perhaps it is a good idea. Rock is 1h and regular desert waste is 1h1c, so 1w1h1c seems sufficiently better than either.

That is not really a valid comparison; no-one works unimproved rock tiles. Whereas rock tiles with a basic improvement are 3 hammers, or 2 hammers and 1 commerce, or 1-2 hammers 1 commerce with growing commerce over time, and they increase significantly over time with improved tech.
You can't build improvements on desert waste tiles, and the only bonus the polar desert waste can get from tech is +1hammer from the weather scanner.

So no, 1w1h1c is arguably worse than a rock or graben tile.

And you're wrong, regular desert waste is 1h2c. Just look at the screenshots in post 401, they're giving 2h2c (+1h from the weather scanner, 1h2c base).

I suspect this is the source of the huge commerce incomes you are seeing

Not really, it was coming from cottagespam throughout my empire; I only had 2 polar supercities, and one of them is huge because it has my spice corp, which gave 125+ commerce, multplied up massively by all the gold and beaker boosting buildings.
And there was no spice on any of my polar desert waste tiles all game.

Btw; also note that reporting commerce yields and worked tiles alone from the economy screen can be badly misleading, because they don't include the yields from a specialist economy (specialists don't work tiles, and give beakers or gold directly, not commerce). By design, meritocracy + faufreluches can give very significant specialist economy yields.

 

[Deliverator]

Regular desert waste is 1h2c as Ahriman says. 1w1h1c makes sense for a buffer zone which is what polar desert waste is - as far as I can see there is nothing to stop us allowing some of the land-based improvements on it to increase the yield.

I think keep Dunipelago - variety is good and most mods provide at least two mapscripts, but I would like to encourage more people to try Arrakis.py. In 1.6 Arrakis will become the default option due to alphabetical order which will help.

 

[Ahriman]

1w1h1c makes sense for a buffer zone which is what polar desert waste is - as far as I can see there is nothing to stop us allowing some of the land-based improvements on it to increase the yield.

I'm fine with 1w1h1c if you can build improvements on it, but not if you can't, that seems too low for what is supposed to be a superior terrain type.
Think of it this way; desert waste is now the just 1c more than a basic mesa, but on mesa you can build mines or windtraps.

 

[Ahriman]

The Arrakis mapscript doesn't feel like it gives enough spice.

Can I suggest a secondary ring of small islands outside the primary main ring? So that cities can also be built on this outer layer, increasing significantly the number of desert tiles that are within cultural borders, and so increasing spice income?

 

[davidlallen]

The Arrakis mapscript doesn't feel like it gives enough spice.

Can I suggest a secondary ring of small islands outside the primary main ring? So that cities can also be built on this outer layer, increasing significantly the number of desert tiles that are within cultural borders, and so increasing spice income?

The amount of spice is proportional to the total number of plots on the map, ie width * height. Do you mean, that there is a lot of spice out in the deep desert towards the edge of the map which is unreachable?

If that is the case, it matches the Dune theme. The right way to fix this is much harder than adding islands. The right way is to allow harvesters outside cultural borders, *and* adjust corporation yields to include them, *and* change the AI so it will build and protect improvements outside cultural borders.

If the percentage of spice plots is too low, that can be easily fixed by a python change as described in the python how-to thread. There is a constant, 400, which can be increased or decreased to experiment with this.

 

[Ahriman]

The amount of spice is proportional to the total number of plots on the map, ie width * height. Do you mean, that there is a lot of spice out in the deep desert towards the edge of the map which is unreachable?

I mean that you have fewer desert tiles inside your cultural borders than you do in Duneipegalo, because your cities are in big solid blocks of land, so an equivalent sized empire will have many fewer spice resources than on Duneipegalo.
Its not that there aren't enough spice on desert tiles, its that you don't control enough desert tiles.

he right way is to allow harvesters outside cultural borders, *and* adjust corporation yields to include them, *and* change the AI so it will build and protect improvements outside cultural borders.

I strongly disagree. I think it will be incredibly hard to get the AI to intelligently contest areas that are outside their cultural borders (the standard AI doesn't even intelligently contest its cultural borders; at best it defends its cities).
And I don't think it will be fun for the human player to being having to control and contest areas outside their cultural borders. The limited line of sight will be frustrating (AI doesn't suffer from this), and the long supply lines and distance won't be fun either. The combat would be boring, since there is little variety in the deep desert units (just a suspensor gunship line and the thopter line), and since the AI is bad at using aircraft carriers. Plus, Fremen would have a huge advantage since they can bring their normal soldiers out onto the desert, while everyone else can only use suspensors and thopters.

I do not think that adding conflict to uncontrolled areas is desirable or feasible.
I think we need to design the game around spice income coming from spice controlled within cultural borders. We need to use the civ engine, not fight it.

 

[Deliverator]

@Cephalo: Please can you either make an option in the map menu for more broken up, archipelago-like landmasses or point out some parameters in the Python that can be changed to achieve this?

 

[cephalo]

@Cephalo: Please can you either make an option in the map menu for more broken up, archipelago-like landmasses or point out some parameters in the Python that can be changed to achieve this?

I can but it might re-create the other problems we had before I made the AI change. There will be very random size islands and starting plots that are too close together etc. If some people want an Archipelago, it makes sense to keep the one we have and make both maps work.

In my opinion, the continental configuration fits the fiction alot better, but I don't think I understand what the complaint is. Whatever the commerce rate, the map should be the same for everyone.

 

[davidlallen]

In my opinion, the continental configuration fits the fiction alot better, but I don't think I understand what the complaint is. Whatever the commerce rate, the map should be the same for everyone.

I have not played vanilla archipelago much. But, balancing vanilla archipelago vs vanilla pangaea, with variable water level for each, must have been a challenge. For us, the challenge is slightly worse, because "water" is highly useful for commerce income and relatively useless for anything else.

His comment is that when you switch between the dunipelago map and the arrakis map, the play feels very different; your commerce value is much higher on dunipelago than arrakis.

When we get comments on tuning the economy, we will also have to be sure to ask, which mapscript they were using. We will likely start to find that a tuning change which makes one map better, makes the other worse.

If there is a water level option in arrakis, we can experiment with that.

 

[Deliverator]

For a purely aesthetic point of view I do think the current script matches the fictious map quite well. The issue is that a certain amount of desert tiles need to accessible from land to get a good income from spice harvesting. I agree with Ahriman's point that allowing spice harvesting and contention outside your cultural borders is probably an unrealistic stretch for the Civ 4 engine for the reasons he describes.

 

[davidlallen]

We need to use the civ engine, not fight it.

I agree it is hard to draw the line between an "easy" sdk change and a "hard" one.

 

[Ahriman]

In my opinion, the continental configuration fits the fiction alot better, but I don't think I understand what the complaint is. Whatever the commerce rate, the map should be the same for everyone.

The number of spice resources you control is a function of the number of desert tiles that you control that are within your cultural borders.
On big blocky continental maps, the only desert tiles you control are a few tiles expanding out from the edge of each continent. On thinner, islandier maps, your cities aren't just concentrated on a big continent, they're spread out over archipelagos, and so you have many more desert tiles, as almost every city you have is coastal.

This has serious balance implications, particularly for the Arrakis Paradise vs Spice Economy decision. Arrakis paradise civic weakens spice income but allows terraforming of land tiles, so it it is much more powerful on big blocky maps with little chance of getting spice income anyway, and lots of land tiles to terraform. Arrakis spice civic slows the degradation of spice resource, and so favors maps with many more desert tiles with spice on them, but doesn't do anything for land tiles.

The current configuration of Arrakis (corrected) means that spice income will never be a major part of your economy.

 

[Deliverator]

I think you meant 'The current configuration of Arrakis means that spice income will never be a major part of your economy.'

 

[cephalo]

It might not be that difficult to count the city workable desert tiles and adjust the spice level to that rather than the map size.

 

[Deliverator]

The amount of spice isn't the problem. It's that only a small % of the spice that there is can be covered by cultural borders.

 

[Ahriman]

The amount of spice isn't the problem. It's that only a small % of the spice that there is can be covered by cultural borders.

Yes, well put. The proportion of desert that has spice is good. And we don't want to just ramp up the power of each spice resource: we want a large number of spice resources so that you have to devote significant time and workers to building harvesters, and a larger number of desert tiles so that a "law of large numbers" effect comes into play, reducing the volatility of the spice income you have.

(Example: suppose we multiplied the value of each spice resource by 5, to compensate for on average having only 1/5 as many spice resources. This wouldn't work well, because with only a few desert tiles, the variation in the number of spice resources you control would be relatively much larger, and your commerce income could swing wildly depending on the random appearance of spice blows. Whereas with the much larger number of desert tiles on which spice can appear, the variation in commerce from the mean is smaller.)

 

[davidlallen]

It might not be that difficult to count the city workable desert tiles and adjust the spice level to that rather than the map size.

I considered that, but from the ""reality"" standpoint, the amount of spice blows produced by an ecological process should not really depend on how many plots of cultural control there are in a particular game. I think the main point is that with subcontinents rather than islands, the average number of coast/ocean available to each city is less.

 

[cephalo]

Are you sure this isn't just a matter of perception? I've only played on Arrakis, and I haven't noticed any problems, maybe because I just don't know any better. All the civs seem to progress just fine.