arrakis terraforming victory?

[Psychic_Llamas]

having just completed a terraforming victory i just want to say that it might be a bit too easy to win. i *never* finish a game, let alone win, though i admit i was only on noble difficulty.

id suggest:

• making the 7 final buildings more difficult to get (i was able to rush them all with great techmen)
• increase the rate the land terraforms and have the victory trigger only when x% of the world is terraformed. as it is i had only terraformed about a third of my territory before i won.
• make another step *after* the 7 buildings to complete a project which slowly physically turns the world into grass, including dunes and such. it should be after this step that victory can be accomplished. give those folling the Spice Future an opposing project and their own counter to terraforming.
• or a combination of all of these. i think a final project is my favourite option.

 

[davidlallen]

Waking up this thread again. With a suggestion from The_J, I have now figured out how to add a terraforming victory condition by percent of land terraformed. See this thread for details. I can now set a number of plots as a threshold, and award a victory to the first player who has that many plots terraformed to grassland. The AI will not intentionally pursue this. Since Reservoirs of Liet have a high AIWeight, I think it will happen unintentionally well enough.

I did a few runs of the current mapscripts to find out the total area and what percentage of the map is land. The percentage of "dune" terrain is quite small, actually, and although I excluded it from land it really doesn't seem to matter. It may be that the arrakis mapscript broke, but both archipelago and arrakis scripts generate a single digit number of plots of dune, like 3-8 plots, regardless of what settings I tried.

The attached spreadsheet shows my runs. There are a lot of variables and I don't have any way to automatically execute the runs. So I only picked a few variables and gave up when I got bored of repeating "click on mapscript options, launch game, exit to main menu". I am sure that the exact numbers will vary a little and we should have more runs to get any kind of statistical significance, but I don't think it matters that much.

I need a little bit of help on deciding the right threshold. Since the books refer to a "three percent solution" I am tempted to just set the threshold to 3% of the map area. The "three" column of the spreadsheet shows this.

I would probably alert the player to an AI pursuing this, by giving a popup when the first player reaches 1%. I have had problems in the past where an AI pursues this victory, and the first time player suddenly and unexpectedly loses, without even knowing they were in danger.

What do you think?

 

[Ahriman]

I think running some simulations, either in-game or with a toy model, will help us here.

Its very hard to think about an appropriate number of tiles without doing that.

Also, we need to finalize whether or not wells will give fresh water access (and allow terraforming) before we even think about tuning this. It will make a huge difference.

[Example thought experiment:
As a thought experiment, suppose:
Standard map size, so around, what, 3600 tiles? Call it 30% land, for 1080 land tiles.
3% of the total map tiles = 108 land tiles.
8 groundwater tiles with 4 tiles on average, with wells on average by turn 120.
6 catchbasins, with 3 non-mesa non-well tiles on average, completed on average by turn 220.
5 reservoirs, with 5 non-overlapping tiles on average not irrigated by wells or catchbasins, completed on average by turn 300.
Transformation rate = 1% + 0.25% per catchbasins + 0.5% per reservoir.
Tiles must be transformed twice to get to grassland.

This allows for 75 potential grassland tiles, and these tiles will be reached somewhere around turn 340.

 

[Deliverator]

On the graphical side of terraforming, I'm wondering whether we could replace the Sword Grass bonus with Saguaro (it will be quick to make art from the Scrub graphic) and then use the current Sword Grass graphic as a feature representing Anchor Grass. The final Grassland state would change the underlying terrain. What do you think?

 

[Ahriman]

So, the "anchor grass" *terrain* type would be rock with the anchor grass graphic?

That could look weird for sink/graben tiles that terraform to anchor grass.

Or are you abandoning anchor grass as a terrain type, and trying to make it a feature, that can show up on multiple terrain types?

How would that work with tile yields?

 

[Deliverator]

Or are you abandoning anchor grass as a terrain type, and trying to make it a feature, that can show up on multiple terrain types?

How would that work with tile yields?

That is what I'm suggesting. If it's not doable, then I'll have a go at making the terrain art more distinctive.

 

[davidlallen]

The python code for terraforming is nice and symmetric with two levels of terraforming improvement which are both terrains. It will be much messier if I make the first level a feature and the second level a terrain.

I think the main goal is to have the two levels more visually distinctive; for example, make anchor grass yellower, or give it brown patches so it doesn't look as extensive.

If you also want to use the saguaro from the FF scrub forest terrain we found as a replacement for one of the food resources, that's great. Swapping those is easy. I haven't thought of a useful way to add the scrub forest as a feature. It doesn't seem like a hammer bonus or a chop bonus makes sense for saguaro.

 

[Deliverator]

I think the main goal is to have the two levels more visually distinctive; for example, make anchor grass yellower, or give it brown patches so it doesn't look as extensive.

OK, I'll do that.

 

[Ahriman]

It will be much messier if I make the first level a feature and the second level a terrain.

I think the main goal is to have the two levels more visually distinctive; for example, make anchor grass yellower, or give it brown patches so it doesn't look as extensive.

I agree.

I haven't thought of a useful way to add the scrub forest as a feature. It doesn't seem like a hammer bonus or a chop bonus makes sense for saguaro.

I don't think the FFH scrub forest as a feature on grassland would work visually.
But there must be some other forest art out there that might make sense with a hammer bonus. Maybe the New Forest from FFH? We can still call it scrub forest, but have art with some green plants.

 

[Deliverator]

But there must be some other forest art out there that might make sense with a hammer bonus. Maybe the New Forest from FFH? We can still call it scrub forest, but have art with some green plants.

The Light Forest from Colonization could be worth a go. I'll give it a try.

If not there are loads of trees/bushes in other games. If you can specify what a Dune +1 hammer forest might look like I can make something.

 

[davidlallen]

Let's take one small step back. What are we trying to solve? What I want is more graphical differentiation between anchor grass and grassland. A long time back, like 1.2, we were discussing if initial, unterraformed terrain should have "something" on it like vanilla forest, to give a hammer or chop bonus. But we could not find anything. If you find a bush or light tree graphic, how would we use it in the game?

 

[Ahriman]

My feeling is a scrub forest would be a features with a low chance of spawning on grassland tiles, probably in grassland sink, in a similar manner to the oasis/lake.

I feel no need for a chop bonus, the goal of these forests isn't to chop them down.

The design goal of such a feature would be:
a) Increase the terraforming feel, changing from a dry harsh landscape to one with thriving living things
b) restore the balance of sinks being superior to regular flatland terrain. At the moment, graben is superior to rock, but Grassland Sink is the same as regular Grassland.
c) Make sinks (with their shelter from winds) feel more special

 

[Deliverator]

How's this for the anchor grass texture?

Figuring out the tessalation for the blend texture files is complicated so there are one or two slightly square looking bits, but I don't want to spend all day figuring it out.

 

[davidlallen]

That looks great.

 

[Psychic_Llamas]

out of curiosity, is there any reason why the actual dunes dont terraform as well? it looks a little odd having grassland butting directly onto desert. im guessing there is a game play reason for it?

graphically itd be nice if turned into something like 'reclaimed desert' or something, even if its just a graphical change and not affect gameplay.

 

[davidlallen]

out of curiosity, is there any reason why the actual dunes dont terraform as well? it looks a little odd having grassland butting directly onto desert. im guessing there is a game play reason for it?

Do you mean the terrain called "dunes", which is very rare? Or do you mean the terrain called "desert waste" and "deep desert"? Desert waste is really coast, and deep desert is really ocean. It would cause major changes in pathfinding, etc, if water plots changed to land plots during the game. That is not fatal, but I think it would be kind of weird. Also, having grass appear on deep sand does not seem ""ecologically possible"".

Perhaps we could add a "chrome" terrain which is desert waste plus a little bit of grass, so that there is a more subtle transition. But it would not have any effect on gameplay; it would be equivalent to different types of tree in vanilla forest terrain.

 

[Hived]

The discussion turns very "graphically" now.
Deliverators "new" anchor-grassland looks much better, although imo it could be a bit more "yellowish", too
Also making desert waste adjacent to terraformed tiles look different is a good idea...

...but more important (to me) is redefining the terra-forming victory. I've also just finished a Fremen terraforming victory on prince-level. As you can "time" the building of the resrvoirs of liet, the diplomatic malus, they are giving, doesn't matter that much. I also rushed the last reservoir, but with privat property. In my game there were 4 or 5 turns between the finishing of the first and last reservoirs, so I had just a few turns diplomatic mali. I could have timed even better, letting them all finish in one turn, but I didn't matter about that before.
IMO there should also be "something" after having built the reservoirs. Maybe the "green-planet-project" or something.
I'd also love to see the terraforming victory as a possible team-victory, especially, when victory depends on a terraformed-land-percentage.

Greetz, Hived!

 

[Slvynn]

I agree that terraforming victory is too easy to achieve without any sagnificant malus.

 

[Ahriman]

How's this for the anchor grass texture?

This works fine. Easy to distinguish.

I agree that terraforming victory is too easy to achieve without any sagnificant malus.

We're all agreed on this. The issue is designing the alternative.

 

[davidlallen]

@ deliverator, please upload the art and terraininfo file for the new anchor grass so I can include it into 1.6.4 today.

@ others, I have implemented the new victory condition as described above. The next step is to playtest it.