Version 1.8 Feedback

[Ahriman]

As it stands, the only leaders that *ever* adopt Paradise are those specifically hard-coded to use it as favorite civic, and the human player.

We should realyl investigate the civic selection AI and figure out how to tweak them or hardcode so that most Leader AIs will sometimes adopt either.

Never adopt Paradise: Bene Gesserits, Ecaz
80/20 spice: Corrino, Harkonnen, Ix, Ordos
80/20 paradise: Leto I, Alia, Tleilaxu?, others?
Never adopt spice: Liet-Kynes, Stilgar, Leto II, Muad'Dib

 

[Teodosio]

Another bit of feedback:

Expansion too powerful and easy for human player?

I am playing a Bene Gesserit game: Monarch difficulty, archipelago map, normal speed. I am playing a custom game were I added two more factions than default for more challenge (small map: from 5 to 7 civs).
Let me specify that I am a decent Civ player but not exceptionally good: I usually play at monarch or lower difficulty.
The fact is: I played the early turns trying a very strong territorial expansion, now I am at turn 179 owning 17 cities. The second higher-ranking civ has 8 cities and half my score.

It seems to me that strong and quick expansion is too easy for the human player and it provides too much advantage.
What do you think? Have you experienced the same?

 

[Jester Fool]

It seems to me that strong and quick expansion is too easy for the human player and it provides too much advantage.
What do you think? Have you experienced the same?

You should probably play on Emperor. There is a big jump in difficulty from Monarch to Emperor. If you play on Deity, the situation is reversed and you end up having half as many cities as the AI and your science rate is pathetic. It's pretty hard to expand if you don't have good military tech (otherwise you're just building cities for the AI).

 

[dammitbiscuit]

What else could we put there that would make these techs more valuable? It stinks that they're useless to non-mahdi factions.

If you add enough generally-useful things to the techs so that they become valuable to non-mahdi, they will probably become overly powerful for mahdi.

It's sort of a hack, and would be difficult to make sense of lorewise, but I think the most efficient game-balance solution is for the techs to include benefits that only apply to NON-mahdi.

I can't really think of any good ones myself, but maybe something like
"Skeptic's Wisdom: Your cities are 20% more resistant to the spreading of Shai-Hulud, Mahdi, Qizarate, and Imperial" or "Sponsored Debates: Due to heavily-publicized, sham debates with puppet 'proponents' of other religions, non-state religions are 10% less likely to successfully take root in your cities"

Or, I dunno, maybe some alternative special units that aren't quite as good as the Mahdi units. A reduction in the production cost of Inquisitors. Something.

 

[Ahriman]

If you add enough generally-useful things to the techs so that they become valuable to non-mahdi, they will probably become overly powerful for mahdi.

I don't think this is a major concern, anymore than Academies is over-powered for Qizarate, or Feudalism is overpowered for Imperial.

 

[Deliverator]

Someone called Yokto has posted quite a positive review of the mod over at the Dune Wars ModDB page.

I'm wondering whether anyone else has experience the following issue described in the review:
"Annexing cities also crash the game for some reason. Wile i have managed to do so once all subsequent tries crashed the game for me. Wile it did not ruin the game for me i still wish i had the option to do so. Especial considering how much i had invested in my spy network at that time."

 

[Jester Fool]

I know that annexation worked fine in 1.7.x. I was able to annex multiple cities (3 or 4 if I remember correctly). Haven't tried it on 1.8 though.

 

[Ahriman]

I know that annexation worked fine in 1.7.x. I was able to annex multiple cities (3 or 4 if I remember correctly). Haven't tried it on 1.8 though.

Same. Sorry, I haven't had a chance to do any testing or design work for a while; too busy with RL.

 

[Teodosio]

Same. Sorry, I haven't had a chance to do any testing or design work for a while; too busy with RL.

I had a problem with annexation. I placed a Reverend Mother in a city of a faction I was at war with. The annexation order succedeed and the game didn't crash BUT... the enemy garrison remained in the city! The city was mine but enemy troops were inside! Of course next turn the enemy automatically took it back

 

[davidlallen]

This is an area which hasn't been playtested much. Initially, annexing a city didn't change your relationship with the former owner. If you were at peace before, you would still be at peace. So, you could annex multiple cities and the former owner would never declare war; the cities would stay yours. The initial feedback was that this is not realistic and the AI should do *something* when you start taking cities. So, there is some chance that the AI will declare war when you annex a city.

That is what happened to you, probably. You were at peace, you annexed the city, the AI declared war, and you had no units to defend the city. It may be helpful to have your own stack in the city (assuming you are at peace with open borders). Then when the AI declares war, *his* stack will be ejected and will have to attack your stack.

If you can suggest a different way for this to work, which does not make annexation too powerful a tool, please let us know.

 

[Jester Fool]

That is what happened to you, probably. You were at peace, you annexed the city, the AI declared war, and you had no units to defend the city.

That happened to me (luckily I had saved prior to annexation). You definitely feel the effects of the 'you manipulated us' negative bonus, especially after two or three times. That is why you have to be careful when trying to annex other civ's cities. I personally like how this has been implemented. It wouldn't make sense to be able to annex cities without any negative consequences (and it would be too powerful as well). Another point, it makes annexation more of a strategic mission (e.g. annex a city to get a resource you don't have or takeover a strategically located city) and that fits the BG theme to a 'T'. Just my two cents.

 

[Deliverator]

The 1.8.0.1 patch for Dune Wars 1.8 is available here. This adds a number of new features to the mod.

 

[Ahriman]

If you can suggest a different way for this to work, which does not make annexation too powerful a tool, please let us know.

Annexation should kick any occupying troops out before triggering the diplomacy change and possible war.

Its one thing to annex a city, have them declare war and immediately re-conquer it on the next turn if you fail to bring in defenders. This is fine.

Its quite another thing for that reconquest to happen instantly before you even get a chance to move in troops.

You shouldn't need open borders and pre-placed troops in order to use annexation.

 

[Jester Fool]

Annexation should kick any occupying troops out before triggering the diplomacy change and possible war.

Its one thing to annex a city, have them declare war and immediately re-conquer it on the next turn if you fail to bring in defenders. This is fine.

Its quite another thing for that reconquest to happen instantly before you even get a chance to move in troops.

You shouldn't need open borders and pre-placed troops in order to use annexation.

I totally agree. I seem to remember that the troops did get kicked out before the AI declared war (I had total espionage superiority BTW). However, I was not at war when I tried to annex the city so maybe being at war prior to annexation allows the AI to stay in the city (which would be a bug IMO). Seems like some playtesting is in order.

 

[Jester Fool]

General impression... Favorite change so far - "There is a bond of friendship between us" (Fremen-Atreides). Nice touch (and makes sense given the books). Same goes for Corrino-Harkonnen, etc. The cave feature is a welcome addition (my friend raves about this). I am enjoying the new promotions as well. Zeal I seems rather weak (but in combination with Zeal II and III is probably made up for) in comparison to the other low level promotions available. Favorite added promotion so far - Cave Fighter. The new implementation of withdrawal is better than the old. One odd thing I noticed though, is when 5 razzia raiders attacked a stack of quads, every quad retreated when attacked. It's possible that they got lucky, but I would not have expected all of them to escape. Anyway, enough singing to the choir, time to get back to conquering Dune.

 

[Deliverator]

Glad your enjoying the new features JF. I will test defensive withdraw again to see if I can recreate your issue with the Razzia Raiders vs Quads.

What would be a good way to strengthen Zeal I?

One thing that is bothering me since I noticed it is that the AI is not using Ranged Bombardment. For example put a stack of enemy Missile Launchers down and put some of your own units 2 tiles away and the Missile Launchers should attack IMO, but generally they move to a better defensive location or do nothing which is not a good use of the Ranged Bombardment feature. In the current RevDCM codebase in cvUnitAI there is additional AI code for Ranged Bombardment which is more sophisticated than the original DCM code. I tried using both versions of the code, but I've only seen the AI make the decision to using Ranged Bombardment on very rare occasions. This is something it would be nice to resolve to make Siege units more of a threat in the hands of the AI.

 

[Ahriman]

Ranged bombardment is definitely an issue. We need to tinker with it and see if we can get it working, and them do a redesign of siege units.

If we can get bombardment working, then I think we should consider making siege units primarily ranged bombardment; drop their city attack bonuses, keep them fairly low strength, but with a good bombardment attack.

Otherwise we should consider removing the ranged bombardment.

I've never really noticed the AI using the attack.
I don't understand the AI for the bombardment or the actual mechanic itself (what are the actual parameter values that determine how many units get hit by the bombardment attack, and how much damage they take? How is unit strength of the target and of the siege unit incorporated?).

But one thing to consider is that the AI is not necessarily doing anything stupid, given current parameters; the ranged bombardment attack does very little damage, often ~4-5%, whereas attacking/suiciding with the collateral damage siege unit itself can do significant damage.

 

[davidlallen]

Is there any proof that the AI ever used ranged bombardment correctly, in any revdcm mod?

 

[davidlallen]

I seem to remember that the troops did get kicked out before the AI declared war (I had total espionage superiority BTW). However, I was not at war when I tried to annex the city so maybe being at war prior to annexation allows the AI to stay in the city (which would be a bug IMO). Seems like some playtesting is in order.

"Annex city" seems to work the way I designed it. I did the following using a fresh install of 1.8 with no patches. Start any game as BG and autoplay 200 turns so that some other target AI player has > 5 cities. Using WB, for each city of the target player except the capitol, put 2 RM's. Put 10 spymasters into one of that player's cities. Put a heavy carryall somewhere on land, outside the target player's borders but within one move of a target city and put 6 kindjal soldiers in the same plot. Exit WB, pop all 10 spymasters so you have a huge number of espionage points, load the kindjal soldiers onto the carryall and save the game. Now do five experiments.

1. While at peace with the target, and with open borders, annex a city. Unless you are unlucky, the target will not declare war. You can see that you get control of the city; it is in revolt; and the target player's units stay put. This seems fine.

2. Reload the game and drop open borders. Then annex a city. You can see that the target player's units "bounce" to the nearest plot owned by the target player, just as other situations in which it is suddenly no longer legal for their units to be in your territory. This "bouncing" can be unexpected but this seems fine also.

3. Reload the game, declare war on the target, and then annex the city near where you have your heavy carryall. You get control of the city and the enemy units are moved into the next plot. Although you have your carryall nearby, do not move it. Hit next turn. The enemy units will take back the city. This also seems fine to me. If you are at war, and you take a city but cannot defend it, the enemy should take it back.

4. Reload, declare war, annex the city, and move in your troops. This is the recommended approach when you are at war. This probably guarantees you will keep the city.

5. Reload, and annex each of the cities one by one. Look into the diplomacy screen and you will see the "You manipulated us" penalty grow 1 point per annex. As in #1 or #2, the enemy troops will stay put or "bounce". At some point, randomly but probably after 3-4 annexations, the target will declare war. When this happens, look back at the cities you already annexed, and you will see that all of the defending troops have been moved out of the cities.

As you see, there is no case where the target automatically takes back the city. If you don't have units to defend your new city against an angry enemy, you should lose it.

Are there any suggestions for something to change, or disagreement about any of the cases?

 

[Deliverator]

Is there any proof that the AI ever used ranged bombardment correctly, in any revdcm mod?

Good point. I tried it out in RevDCM 2.721 using the Artillery unit which also has 2 tile ranged bombardment and the behaviour looks pretty similar. Perhaps no one has actually succeeded in coding good AI for ranged bombardment over a 2-tile distance. I notice that DCM_RANGE_BOMBARD is disabled by default in RevDCM - not sure whether that is significant.