Version 1.8 Feedback

davidlallen · May 24, 2010

I can think of a few possible solutions. (1) Remove ranged bombard (2) Find somebody to debug why RevDCM AI isn't using it and then (3) Find somebody to write new AI to use it. (2,3) would be nice but I would not hold my breath. Now that we can see the AI not using it, is it a critical feature to make work?

Jester Fool · May 24, 2010

Deliverator said:
What would be a good way to strengthen Zeal I?

Zeal promotions as currently implemented :

Zeal I - +8% attack str; -10% defense str; +8% str in enemy territory
Zeal II - +8% attack str; +8% str in cities with more than 50% foreign culture
Zeal III - can attack multiple times per turn; +8% attack strength; +30% chance to heal on combat victory

I would suggest (while trying to stay within the flavor of the current implementation) :

Zeal I - +10% attack str; -10% defense str; +7% str in enemy territory
Zeal II - +8% attack str; +7% str in cities with more than 50% foreign culture
Zeal III - can attack multiple times per turn; +7% attack str; +25% to heal on combat victory

Thoughts?

Ahriman · May 24, 2010

Now that we can see the AI not using it, is it a critical feature to make work?

I have no strong objection to removing bombardment.

It would be nice to have, but its not high priority.

Even so I think we should tweak the siege units to make sure that they aren't more effective at city assault than melee.

Zeal I - +10% attack str; -10% defense str; +7% str in enemy territory
Zeal II - +8% attack str; +7% str in cities with more than 50% foreign culture
Zeal III - can attack multiple times per turn; +7% attack str; +25% to heal on combat victory

Zeal 1 and 3 seem reasonable, except that most units are 1 move so can't ever attack multiple times per turn.
Zeal 2 seems very weak; why would I ever choose this over combat (+10% attack and defense everywhere).
"Cities with more than 50% foreign culture" is a very rare occurrence.
Maybe kill 2 birds with 1 stone; have Zeal 2 give +1 movement, +10% strength in cities with more than 50% enemy culture.

Deliverator · May 25, 2010

Ahriman said:
Even so I think we should tweak the siege units to make sure that they aren't more effective at city assault than melee.

I already removed the city raider promos from them in 1.8.0.1...

Ahriman said:
except that most units are 1 move so can't ever attack multiple times per turn.

Thought about this. I'm thinking of setting the flag for granting an extra movement point on victory (<bOffensiveVictoryMove>), and then reducing the chance of heal on Victory to perhaps 20%. This means that the Blitz effect can be used by units other than Fedaykin.

Also, I'm planning to set bMustMaintain to 1 for both Zeal II and Zeal III meaning that any units with these promotions will lose them if you pick a State Religion other than Mahdi (representing the fanatical fighters becoming disillusioned).

Ahriman said:
"Cities with more than 50% foreign culture" is a very rare occurrence.

To be clear, this bonus applies when attacking cities with > 50% foreign culture so it shouldn't be that rare if you are a-conquering.

Ahriman · May 25, 2010

I already removed the city raider promos from them in 1.8.0.1...

Ok. Sorry, still haven't had a chance to do testing. Maybe this weekend. Life....

I'm thinking of setting the flag for granting an extra movement point on victory (<bOffensiveVictoryMove>), and then reducing the chance of heal on Victory to perhaps 20%

Ah I like this.

Also, I'm planning to set bMustMaintain to 1 for both Zeal II and Zeal III meaning that any units with these promotions will lose them if you pick a State Religion other than Mahdi (representing the fanatical fighters becoming disillusioned).

Makes sense... but risky for the AI. If the AI takes the promotions, and then gets bribed away from Mahdi temporarily, then its army gets significantly weakened.
But maybe not a huge deal.

To be clear, this bonus applies when attacking cities with > 50% foreign culture so it shouldn't be that rare if you are a-conquering.

Ah, understood. I'd thought it was only for defending, which was too niche.

Morlark · May 28, 2010

Is there any mirror for 1.8.0.1? The mediafire link doesn't seem to work for me.

Edit: Never mind, I got it working now.

Ahriman · May 28, 2010

One of the problems with the tech tree as it stands are the highly specialized nature of the religious mandate and jihad techs.

I think by adding something extra to these techs, we will make the opportunity cost of ignoring these techs slightly higher, and make it slightly more valuable to try to grab them.

Some thoughts on things that might go here (none of these are great ideas I think, just some brainstorming):
a) A "generic" temple or religious academy type building at Religious Mandate, basically representing the standard OC Bible religion practiced by most galactic inhabitants. Basically this would represent state control over "standard" religion (separate from the major Religions as represented in the mod).

Think "pagan temple" from FFH for example.
I would see this as giving a priest specialist slot, a small % culture boost (15%?), and some other minor bonus.

This could also have UB versions for Fremen and Tleilaxu, who don't practice standard OC Bible religion.

Possible names: Religious academy, State Church, Orange Catholic Monastery,
Possible extra bonuses: +10% beakers, +1 happy but +1 unhappy if any of the other religions are state religion (ie net +1 happy only if no state religion), +1 culture, +1 culture from priest/noble/merchant specialists in this city

b) A conscription center building at Jihad that allows units to be drafted in the city.

c) Another idea for zealots; maybe the Mahdi Zealots should be *only* able to get the "zeal" promotions, not normal combat-type promotions? Would increase their role as cheap cannon-fodder attackers, and reduce their ability to be cheap/free city guards that frustrate enemies trying to take your cities.
And I think we should again consider adding an upgraded version of them in the late-game. [Or: have a promotion upgrade for them that requires zeal 3 that gives +3 strength and some late-midgame tech.]

We could create a "religious unit" class that included missionaries, Zealots and an upgraded Zealot, and have Spiritual promotion give something free to this class.

We could make the zeal promotions the only promotions available to zealots, and make zeal promotions available only to zealots, and then bump up their power a little.

d) Make the Raider promotion a selectable promotion at Jihad tech for melee/guardsmen units.

Thoughts? Other ideas?

Ahriman · May 28, 2010

More comments on latest patch.

such as Quads and Vulture Thopters (a Poorly Maintained -20% strength promo is applied to stop this from being too strong)

An idea; can we make this promotion be removed if the unit starts its turn in a city with a Factory?
So they're weak to start with, but lose the penalty once you reach the tech to build your own?

In the future, I want to make Siege Units unable to capture cities and may even make them defensive combat only

I don't understand making them defensive. I don't see these units as being like vanilla Civ Machine gun units (stack guards). Also, collateral damage doesn't work defensively.

+ Civilization Animosities - Harkonnen and Atreides leaders hate each other. Corrino leaders hate the Atreides and Fremen leaders hate the Harkonnen. Atreides and Fremen like each other as do Corrino and Harkonnen. A bit of experiment this so we'll see what happens...

I think this might be a little too much hard-coding, and too much removal of "what-if" games. I will test more.

Also; can we remove peaks from the settler evaluation algorithm? They should be treated as useless, not as mesa.

Deliverator · May 28, 2010

Ahriman said:
An idea; can we make this promotion be removed if the unit starts its turn in a city with a Factory?
So they're weak to start with, but lose the penalty once you reach the tech to build your own?

I thought about this too. I will think about what is a workable implementation.

I don't understand making them defensive. I don't see these units as being like vanilla Civ Machine gun units (stack guards). Also, collateral damage doesn't work defensively.

I probably didn't explain this well. Bombardment and Ranged Bombardment gives Siege units a way of damaging enemy units that is completely separate from conventional combat. What I was suggesting was removing the ability for Siege units to attack units via conventional combat (they would still obviously be able to defend). This would leave them with Ranged Bombardment as the sole mechanism for attacking enemy units. Ranged Bombardment can be used to first reduce city defensive and then to weaken units, but generally not completely destroy them (unless you have a Assault Cannon with the new Bombard III promo which I think allows complete destruction due to the raised Combat Limit). So, the difference with Machine Gun units is that Machine Guns can't bombard units 1 or 2 tiles and inflict damage that way.

The problem with this at the moment is that I don't think the AI is using Ranged Bombardment effectively even in apparent no-brainer situations. I have posted on the RevDCM thread about this and I'll keep playing around with the AI code to see if I can fix this.

I think this might be a little too much hard-coding, and too much removal of "what-if" games. I will test more.

It only affects four factions and I can't live with the Harkonnen and Atreides being best buddies - even allowing for what-ifs. This way they can be on reasonable terms, but relations will be more strained. It probably wouldn't be too hard to make this a selectable game option.

Also; can we remove peaks from the settler evaluation algorithm? They should be treated as useless, not as mesa.

In 1.8.0.1, I fixed the Arrakis.py code so that this is taken into account for the starting plot, but I have not yet coded this into the found value calculation for subsequent cities. It's on my list (the list that never gets any smaller that is!).

fjordan · May 28, 2010

This is a truly amazing mod! It has a very unique atmosphere and some great concepts. Also I think you caught the essence of the books very well. Great work! I will certainly try out all the civs and religions.

Some minor issues (path 1.8.0.1):
- Dune Messiah is getting out of hand: I now have +1055% military bonus. Luckily the overflow was broken in the last BTS patch

.
- Some effects have a delay. I built the Guild Research wonder but did not get the beakers. Save/reload did not help. A few turns later it suddenly started working. I had the same issue with spice silo's (showed the bonus with the building but did not get the gold, also not showing in the F2 screen) and a mentat XP bonus. I have no idea what triggered that it suddenly started functioning.

Ahriman · May 28, 2010

This would leave them with Ranged Bombardment as the sole mechanism for attacking enemy units.

Ok, understood, this sounds good as a design goal. Obviously heavily AI dependent.

It probably wouldn't be too hard to make this a selectable game option.

I'm leery of too many options in general, but we could consider that for such a mechanism.

Dune Messiah is getting out of hand: I now have +1055% military bonus. Luckily the overflow was broken in the last BTS patch

Hmm. Did you really have ~130 cities with Mahdi, or is the mechanic broken? We might have to fix this.
Or not though. I have no problem with the Mahdi shrine city pumping out a military unit every turn.
Remember to compare this to how big the bonus would be from, say, a standard +1 gold shrine. If you have 30 cities and +100% gold, thats 60 gold per turn. Which can let you run 100% science and not bother building gold boosters anywhere else.

How exactly *does* overflow work these days anyway; what is the bug?
You can't overflow more than the cost of the next unit?
Does a massive overflow on military units allow you to insta-build a unit or Wonder the next turn?

I built the Guild Research wonder but did not get the beakers. Save/reload did not help. A few turns later it suddenly started working

Odd. We should investigate.
I thought it always just took a single turn to process it into the city screen? But you've had it take multiple turns after construction?

fjordan · May 28, 2010

Dune Messiah: no surely there are not that many Mahdi cities. Even after I just founded it it was certainly more than 7% per city, but it increases exponentially. Regarding the overflow, patch 3.19 "fixed" that boni where not applied to overflow. There was an exploit with whipping protective walls for gold, but many on the forum think that the current implementation where you get no gold at all is broken. It is a good thing here though, 1 unit/turn is not that big an exploit, getting hunderds of gold too would be.

The delay was more than just the next turn, it was three or so. The strange thing is that the next spice silo I built started functioning right away so I am sorry to say this is a "Sometimes it does not work" issue bug-fixers like so much :crazyeye:

.

Ahriman · May 28, 2010

I probably didn't explain this well. Bombardment and Ranged Bombardment gives Siege units a way of damaging enemy units that is completely separate from conventional combat. What I was suggesting was removing the ability for Siege units to attack units via conventional combat (they would still obviously be able to defend). This would leave them with Ranged Bombardment as the sole mechanism for attacking enemy units. Ranged Bombardment can be used to first reduce city defensive and then to weaken units, but generally not completely destroy them (unless you have a Assault Cannon with the new Bombard III promo which I think allows complete destruction due to the raised Combat Limit

Another thought;
we could replicate a similar effect by giving high withdraw chance, collateral damage, and a maximum damage level (ie can't damage enemy units more than ~70% - so they can't drop below 30% strength).
Then, no AI issues to worry about.

Even after I just founded it it was certainly more than 7% per city, but it increases exponentially

Ahh. That sounds like a bug. It should increase linearly. 10 cities means +70%.

The strange thing is that the next spice silo I built started functioning right away so I am sorry to say this is a "Sometimes it does not work" issue bug-fixers like so much

Hmm. Well, as long as it works *eventually*..... ;-)
thanks for the feedback btw, forgot to say that before. Always very appreciated.

Ahriman · May 28, 2010

Caves feature is far too common, this should be rare, and the graphic... doesn't look great. Is there any way to have a cave mouth appear on the side of mesa, rather than on the top?
And 1-2 entrances would be enough, 4 just looks pockmarked.

Spoiler :

Also, I think the botanic station graphic is cool, but looks a little large. Can we shrink it down, so it doesn't take up the whole tile?
Maybe 2/3 the size, or less?

Deliverator · May 30, 2010

This post for patch 1.8.0.2.

Thanks for all the feedback folks. Keep it coming...

Ahriman · May 30, 2010

1.8.0.2 feedback.

Awesome!!

+ New Unit: Mahdi Mujahid - As Mahdi Zealot but Strength 12 and Cost 60 (instead of 6 and 30)

60 seems too cheap for strength 12; we can playtest, but I'm guessing ~90 might be more appropriate (what is the base cost of a shield trooper; this should be ~2/3 or 3/4 of that).

Slavery Civic is now Harkonnen only

Love it. I think this will be a big AI boost.

Rebranded Spiritual trait as Religious: now only boosts production of Qizarate, Shai Hulud and Mahdi missionaries and buildings

Great. We might need to find another minor boost though to avoid being underpowered, or not. Hard to tell.

Also, are either Tleilaxu leader Spiritual? I forget. If so, we should extend Spiritual to cover Zensufism buildings.

Deliverator · May 30, 2010

I think the problem of the AI not adopting Mahdi as State Religion is not only due to the Fanaticism/Divine Mandate/Jihad branch being a bit weak.

I see this scenario quite often in autoplays:

By the time the Mahdi religion appears the most of a player's cities have a religion already. Due to Mahdi having a very low natural spread rate plus the fact that religions don't naturally spread to cities that already have a religion, most commonly the Mahdi founding Civ only has one city with Mahdi and that is the holy city. This means it will never adopt Mahdi as its State Religion because the majority of cities don't have it. Mahdi has no missionaries and can only be spread by conquest so it will now never spread beyond the holy city.

I have seen the Fremen with no State Religion despite the fact that they have have founded Mahdi hundreds of turns previous. When you look into the reason for this, what jumps out is that only the Holy City has the religion. Very often the Holy City is on some remote landmass or island which makes it even less likely to spread naturally.

Possible solutions:
1. Give Mahdi a missionary unit with National Limit 1 like Imperial. My feeling perhaps that a religion without a missionary unit goes too much against the grain of vanilla civ.

2. We could write some custom code that puts Mahdi into other cities some other way. Perhaps on discovery of Mahdi 75% of your cities flip to Mahdi only or something like that.

Basically, it seems that for Mahdi to be chosen at a state religion by the AI it needs to be present in more of that civs cities that any other religion. This rarely happens at present.

This also means that is highly unlikely that you will get two Civs united in jihad.

Ahriman · May 30, 2010

I agree that this is a problem and I have similar observations.

I really don't like the idea of adding a Mahdi missionary; it would reduce the level of "differentness" for Mahdi. No missionary (no peaceful spread) is a key characteristic of the Jihad religion; the whole idea of Mahdi is that unlike other religions you can't spread it your neighbors *except* by the sword, or a small chance of adoption from the spread factor.

However, something like instant conversion of cities for the religious founder is much more promising, and seems in flavor (the word of the mahdi spread like wildfire through the land).

Some ways to implement this:
1. When a player founds Mahdi, each of their (non-holy) cities has a 2/3 probability of instantly adopting Mahdi.
2. When a player founds Mahdi, ~3 of their cities instantly gain Mahdi religion.
3. Create a Mahdi missionary unit which is unbuildable, but when Mahdi is founded an event triggers that creates ~3 Mahdi missionaries.

I wonder if we should do something similar for Technocracy and/or Qizarate, where a similar problem occurs (AI hardly ever adopts them).

Qizarate:
When a player founds Quizarate, an event triggers that gives that player ~4 Qizarate missionaries.
The AI will then go and use these missionaries in its own cities, which will have the added benefit of wiping away other religions, and so will make adoption of Qizarate even more likely.

This also means that is highly unlikely that you will get two Civs united in jihad.

I'm not really sure that this is a problem; this isn't supposed to be a diplomatic religion. The only way this should really happen is if you conquer/vassalize. There Can Be Only One... Mahdi.

davidlallen · May 30, 2010

deliverator said:
+ Slavery Civic is now Harkonnen only. Added PrereqCivilization tag to CivicInfos in order to acheive this. Population Hurry aka Whipping now yields 45 production rather than 30 - this can be easily tweaked in the file XML\GameInfo\CIV4HurryInfo.xml by changing the tag value for iProductionPerPopulation.

This may be good news, but I am a little worried. Before this change, it seemed that the Harkonnens were always lowest ranked in autoplays. This seemed to be because their population per city was less, which seemed to be because they always adopted slavery and whipped a lot. If a human player uses whipping wisely, it can be a benefit, but the AI might not choose as carefully.

Now that no other civs can adopt slavery, the Harks may get further behind. It's possible that increasing the hammer benefit from a whip may counter this. I am sure that by increasing the hammer benefit "enough", we can make it an advantage. But after you selected 45, did you do some autoplays to see if Hark still lags behind? It is a fair amount of work to do this, but I have done economic tuning by running the 9 starting positions in the thread Pick Your Leader Challenge, with related analysis around this post.

Since the maps in the PYL attachment are WB save files, they "should" still be valid to load as scenarios in the current version of DW. The mapscripts have changed a little. It may be worthwhile to re-generate a map to be used for this. I didn't automate this, but it should be simple enough to generate one map and make nine copies with players in different positions.

Ahriman · May 30, 2010

Before this change, it seemed that the Harkonnens were always lowest ranked in autoplays.

I didn't observe this, but they were often behind, yes.
I think it was also because Rabban's insanely offensive AI is not very effective. Which doesn't necessarily mean we should remove it, because it can be fun to have a crazy-offensive AI personality.

Now that no other civs can adopt slavery, the Harks may get further behind

While this is possible, I think we need to be careful about the logic here.
In order for this to happen, it must be the case that removing ability to adopt slavery does in fact improve AI performance (as we suspect it will).
In which case... it would be worth Harkonnen falling a little behind in order to boost the peformance of all the other AIs. So, no problem.
Remember that in an absolute sense, Harkonnen are strictly better off with this change than they were without the change (50% higher yield from whipping); the only problem is if other factions are better off by even more.
Even if if Harks fall slightly behind in an absolute sense (and I'm not sure they will) I think its worth it.

To boost Harkonnen, I strongly suggest we implement the long-intended changes of making slaves 2 movement and able to enter deserts like regular workers.
If slaves are able to build harvesters, this will be a BIG boost for Harkonnen.
[Also; somewhat amusing for the human player to be able to make worms go away by feeding them slaves....]

Also; can we make the slaves zero maintenance cost? We don't want the AI to ever get messed up by maintenance costs from slaves.

Version 1.8 Feedback

Deity

Emperor

Tyrant

Graphical Hackificator

Tyrant

Warlord

Tyrant

Tyrant

Graphical Hackificator

Warlord

Tyrant

Warlord

Tyrant

Tyrant

Graphical Hackificator

Tyrant

Graphical Hackificator

Tyrant

Deity

Tyrant

Similar threads