Realism Invictus

[Walter Hawkwood]

To be quite honest, you guys sound ridiculous - you acknowledge AI is better now and demand that it gets worse! One can never win with you!

In all fairness, though, I understand the demand for more flavour, so I will actually implement a large skew in tech preferences based on leader personality if "AI plays to win" is off. I'll upload it shortly. This close to a release, it's up to you two now to playtest the hell out of it and report back, as it was your protest that caused this change.

 

[AspiringScholar]

Fair enough! I wasn't complaining about optimization, but how that might curtail immersion through clamping out what is supposed to differentiate the shared AI calculation logic via the individual leaders and their at least nominal personalities.

Would this be compatible with current saves? I just rolled an amazing start and would be reluctant to scrap it.

 

[[Y]]

To be quite honest, you guys sound ridiculous - you acknowledge AI is better now and demand that it gets worse! One can never win with you!

Hah! We live to be difficult. But for what it's worth, the change in question feels less like "AI plays well" or even "AI is playing to win", and more "AI will maximize its beaker output regardless of practicality". As in, the decision, as far as I can tell, isn't being weight by whether the tech in question contributes to the civ's long term goal. If it was it would make more sense, but in the current iteration, it feels more like a Civ dropping whatever it's doing to concentrate on a tangent. But as you explained in the previous post, we're probably skewing the relevance of this one change since we aren't looking at it as part of the broader set of changes.

In all fairness, though, I understand the demand for more flavour, so I will actually implement a large skew in tech preferences based on leader personality if "AI plays to win" is off. I'll upload it shortly. This close to a release, it's up to you two now to playtest the hell out of it and report back, as it was your protest that caused this change.

@AspiringScholar our unit type has been updated to Charge Cavalry since we've been straddled with with this expedient responsibility!

 

[Walter Hawkwood]

Would this be compatible with current saves? I just rolled an amazing start and would be reluctant to scrap it.

Most likely yes, I don't see why it shouldn't. By the "current saves" though I mean the previous SVN update that was pushed out today - and it was definitely not save-compatible.

Hah! We live to be difficult. But for what it's worth, the change in question feels less like "AI plays well" or even "AI is playing to win", and more "AI will maximize its beaker output regardless of practicality". As in, the decision, as far as I can tell, isn't being weight by whether the tech in question contributes to the civ's long term goal. If it was it would make more sense, but in the current iteration, it feels more like a Civ dropping whatever it's doing to concentrate on a tangent. But as you explained in the previous post, we're probably skewing the relevance of this one change since we aren't looking at it as part of the broader set of changes.

Since it acts as a multiple on an otherwise perceived tech value, I wouldn't say it does that "regardless of practicality" - the base tech value that is being multiplied (and often quite modestly as I illustrated) is still based on the utility of the tech.

 

[[Y]]

Since it acts as a multiple on an otherwise perceived tech value, I wouldn't say it does that "regardless of practicality" - the base tech value that is being multiplied (and often quite modestly as I illustrated) is still based on the utility of the tech.

ahh, okay. I didn't realize that's how it worked. I just saw a "+%" and as far as I knew, that just meant it directly increased the odds of an AI researching that tech by that value.

 

[globosud]

I would like this change to be included

https://forums.civfanatics.com/threads/release-events-with-images-now-with-source-code.314250/

 

[shkitur1]

spy have 1 movement and can use enemy roads.
great spy (that should be better than spy) have 2 movement but cant use enemy roads. so when war begins great spyes become useless because they can cross only 1 (forests, hills) or maximum 2 (plains) titles. (I use spyes to follow enemy army to see the direction where they attack). May be great spyes also shoud have promotion 'use enemy roads'? Or that's how it was intended

 

[le_harv]

To be quite honest, you guys sound ridiculous - you acknowledge AI is better now and demand that it gets worse! One can never win with you!

In all fairness, though, I understand the demand for more flavour, so I will actually implement a large skew in tech preferences based on leader personality if "AI plays to win" is off. I'll upload it shortly. This close to a release, it's up to you two now to playtest the hell out of it and report back, as it was your protest that caused this change.

I don’t have a lot to base my perspective on but this change seems to have spread out the AI chosen tech paths nicely. It feels like the AI is not just following the herd and in turn… I can’t just follow the herd… I need to actually think about my own trajectory. Success? lol.

 

[JDCP]

GUYS WTH... 80????

Let me see If I get this... I need at least 40 cities to get a +1 happiness modifier?at that point one must have most of the world already owned, that's basically a domination victory. I am playing the Europe scenario and it's pretty entertaining so far, I like the fact you have two German civs fighting to unify their land, that's a nice detail although it seems you can also befriend them anyway, Teutons are epic!

 

[Walter Hawkwood]

I would like this change to be included
https://forums.civfanatics.com/threads/release-events-with-images-now-with-source-code.314250/

I did contemplate adding it at some time. If/when I decide to do anything major to events, this might make it in.

spy have 1 movement and can use enemy roads.
great spy (that should be better than spy) have 2 movement but cant use enemy roads. so when war begins great spyes become useless because they can cross only 1 (forests, hills) or maximum 2 (plains) titles. (I use spyes to follow enemy army to see the direction where they attack). May be great spyes also shoud have promotion 'use enemy roads'? Or that's how it was intended

A reasonable suggestion, consistency-wise.

GUYS WTH... 80????

[/URL]
Let me see If I get this... I need at least 40 cities to get a +1 happiness modifier?at that point one must have most of the world already owned, that's basically a domination victory.

Yeah, high default player numbers in certain scenarios mess up quests royally. I think I'll cap those.

 

[globosud]

I did contemplate adding it at some time. If/when I decide to do anything major to events, this might make it in.

A reasonable suggestion, consistency-wise.

Yeah, high default player numbers in certain scenarios mess up quests royally. I think I'll cap those.

if you insert the tag, I could create the .dds images, for current events, and for any future events, from what I read only a few strings need to be added

 

[sazhdapec]

Some feedback regarding recent SVNs.
1) At first I thought I wouldn't like the changes to ancient era unit strengths. However after playing for some time I don't mind these changes. Along with free wins barb archers stopped being obnoxious. I like that archers are less of a PITA in cities however they are pretty bad in the field both as an aid bonus and direct fighters. Right now I build them as cheap units to station in cities. Warbands being available later is quite a nerf to food heavy starts. On the one hand it's just one tech later, on the other hand it's the latest tech of an ancient era. This change also increases the value of copper substantially if one has to fight wars in ancient era. Skirmisher nerf is unnoticeable in an ancient era however this units becomes useless pretty fast. After horseback riding and iron working it is pretty much pointless to build them. I know a medieval recon unit was discussed a billion times but maybe explorers should become available just a bit earlier. I also liked the idea for changing the aid bonus of light charge mounted cavalry to recon. Almost two eras without proper recon units feels wrong.
2) Naval workshops that use cotton as an input require weaver's shops. Which you can build only three. So if you have more than three cotton you can't use them all to convert to naval supplies. Btw AI values naval supplies extremely poorly. There was a moment when a computer player didn't want to buy it for just one gold. And I'm pretty sure I was selling them to a civ with access to the ocean.
3) The buffed fortification building is making quite a difference. If you bring few artillery defending emplaced siege units can gain quite a few promotions during siege.
4) The cost of spy missions especially devastating ones is way too low. In my current game one civ (albeit with a politician trait) was constantly changing the state religion of three other civs during most of a medieval era. It shouldn't be so easy to disrupt whole civilizations. I'm not a historian but I'm pretty sure there wasn't such a historical precedent.
5) Some time ago the air unit evasion system was overhauled. The percentages were upped but some units of the era retained old percentages. For example, a unit Guided Missile has an evasion power of 100 while contemporary means to intercept are 200+. Paratrooper 75 vs Fighter 120 and even more after. Special Forces 150 vs Advanced Fighter 500. Tactical nuke 50 vs Air Superiority Fighter 475.
6) I don't know how historically accurate is this but I feel that Free Religion civic becomes available a bit too early especially how eagerly AI adopts it. Seeing most of the world secular in renaissance feels strange.
7) I don't know if it's technically possible but maybe add a separatism malus to a revolutionary trait when separatism is enabled since we've established that this trait is not that negative.
8) I don't know whether this is a bug or I'm misunderstanding how combat works. I had at least 4 first strikes but in actual combat only hit once before the opponent. Can anyone explain?

Spoiler :

9) In keeping the tradition of nerfing doctrines can we please reduce the duration of a new one - Art of War - to be obsolete earlier? It lasts noticeably longer than other doctrines.
10) The maximum number of cities needed to gain all era arts is seven. Except Romantic Art. It requires nine cities (three operas). All other era arts require buildings that can be built in any city but an opera is a conditional and a restricted building. I know it fits thematically but this exception requires the player to get additional cities and/or place operas in a hurry instead of waiting for a better city to build them.
11) I like the recent changes to AI behaviour regarding jealousy and open border agreements. AI is more sane and sensible. It's very stressful to be a chart leader.
12) I like the somewhat recent change that allows steel ships to attack multiple times per turn. No need to match the number of ships AI throws at you if you have a tech lead.

 

[[Y]]

I've also been liking the new early game unit strengths, but I think that militia should have their archer bonus nerfed (or removed entirely) since they now cut through archers too easily.

 

[sazhdapec]

I've also been liking the new early game unit strengths, but I think that militia should have their archer bonus nerfed (or removed entirely) since they now cut through archers too easily.

Nooooo, not going back... Let me have payback for some time!

On a serious note though as long as we don't return to suffering from barb archers it'd be probably fine.

 

[Walter Hawkwood]

if you insert the tag, I could create the .dds images, for current events, and for any future events, from what I read only a few strings need to be added

We'll see; I'm perfectly capable of producing those myself as well, as you can probably guess. But in any case not before the release.

Skirmisher nerf is unnoticeable in an ancient era however this units becomes useless pretty fast. After horseback riding and iron working it is pretty much pointless to build them. I know a medieval recon unit was discussed a billion times but maybe explorers should become available just a bit earlier.

Yeah, the more I think about it, the more apparent the need for an intermediate medieval recon unit, and I'll likely create that class.

I also liked the idea for changing the aid bonus of light charge mounted cavalry to recon.

Makes a lot of sense.

Naval workshops that use cotton as an input require weaver's shops.

I'll remove the requirement; it's not really justified.

Btw AI values naval supplies extremely poorly. There was a moment when a computer player didn't want to buy it for just one gold. And I'm pretty sure I was selling them to a civ with access to the ocean.

Well, I'm not sure it's wrong. It is currently a resource with a rather limited utility. I have a couple ideas to mitigate that...

4) The cost of spy missions especially devastating ones is way too low. In my current game one civ (albeit with a politician trait) was constantly changing the state religion of three other civs during most of a medieval era. It shouldn't be so easy to disrupt whole civilizations. I'm not a historian but I'm pretty sure there wasn't such a historical precedent.

Agreed. I'll double the costs of all active missions except for counterespionage.

5) Some time ago the air unit evasion system was overhauled. The percentages were upped but some units of the era retained old percentages. For example, a unit Guided Missile has an evasion power of 100 while contemporary means to intercept are 200+. Paratrooper 75 vs Fighter 120 and even more after. Special Forces 150 vs Advanced Fighter 500. Tactical nuke 50 vs Air Superiority Fighter 475

Paratroopers are supposed to be low - they should be fairly helpless if you didn't establish air superiority before airdropping them. Missiles are an oversight. Special Forces... I don't know - perhaps they should be increased as this is a small covert operation that is likely to go unnoticed.

6) I don't know how historically accurate is this but I feel that Free Religion civic becomes available a bit too early especially how eagerly AI adopts it. Seeing most of the world secular in renaissance feels strange.

Free religion is not exactly secularism (or it would have been named as such), it's a broader notion of not persecuting people for having a different religion - and while Renaissance Europe wasn't a particularly tolerant place, contemporary Ottomans, for instance, embody this civic to a full extent despite being very far from secular.

7) I don't know if it's technically possible but maybe add a separatism malus to a revolutionary trait when separatism is enabled since we've established that this trait is not that negative.

I wouldn't want a trait to have a radically different balance depending on whether a game option is turned on or off.

8) I don't know whether this is a bug or I'm misunderstanding how combat works. I had at least 4 first strikes but in actual combat only hit once before the opponent. Can anyone explain?

Don't take my word for it, but it feels as if all the damage from first strikes is simply rolled into the first round.

9) In keeping the tradition of nerfing doctrines can we please reduce the duration of a new one - Art of War - to be obsolete earlier? It lasts noticeably longer than other doctrines.

It is a pretty weaksauce doctrine, and I didn't want to obsolete it somewhere without a logistics bonus - and there aren't many relevant techs before Food Preservation. I guess Cavalry Tactics can work.

10) The maximum number of cities needed to gain all era arts is seven. Except Romantic Art. It requires nine cities (three operas). All other era arts require buildings that can be built in any city but an opera is a conditional and a restricted building. I know it fits thematically but this exception requires the player to get additional cities and/or place operas in a hurry instead of waiting for a better city to build them.

Agreed, will reduce.

I've also been liking the new early game unit strengths, but I think that militia should have their archer bonus nerfed (or removed entirely) since they now cut through archers too easily.

Agreed, will review.

 

[urbestfriend]

Note: I'm a few svn versions behind due to ongoing games so apologies if this has been changed

I feel like the alliances system kind of takes over the entire game. While this is likely partially influenced by only playing scenarios with like 30 or more civs, it feels like as soon as alliances come online, every civ in the game descends into endless war with everyone else for the rest of the game. You'll regularly see civs even at war with their own allies due to all the interlocking webs. And when there's like 15 different wars going on, the moment a civ does make peace and get out of war, they get bought into one of the existing ones the next turn - which means being pulled into like 7 wars because of the alliances. It also feels very random - sometimes it results in the weaklings being wiped out and consolidation of power, but sometimes the fact that everyone is getting backstabbed means no one makes any progress and the whole world stagnates.

 

[Walter Hawkwood]

Sounds like World Wars to me! I may add a game option to disable them if they're too chaotic, but currently it doesn't feel like they're doing anything wrong - I was afraid of the other extreme, eternal hugboxes that never declare war on each other.

 

[ConquestCola]

I've just spent several days catching up on the conversations in this thread. I am really, really looking forward to the next stable release. Since I play late era, sometimes with time victory only, I am eager to see the differences in the AI use of aircraft and bombing.

I have two things to say and they both refer to comments from many pages ago:

1) In the SVN's so far, how are gladiators and circuses implemented in the slavery civic? What effects do they have? If there was any elaboration, I'm afraid I missed it and I haven't even attempted to keep up with SVN's at all.

2) Also, to add to a very old conversation about espionage and its worth: Did you know that using a spy to cause a revolt in a city eliminates all defensive bonuses during the turn of the revolt? I discovered this while trying to besiege a well fortified border city on a hill, with walls and lots of archers, that kept reinforcing itself every time I tried (and failed) to take it. I sent a spy to cause a revolt on a whim, mostly out of spite, and then I realized my siege weapons didn't have anything more to do for that turn. I was able to take the city that turn, in what basically amounted to a war of attrition after several waves of attack. Had the revolt not caused defenses to fall, I never would have been able to take the city and they likely would have reinforced the city beyond my ability to conquer if I had waited for defenses to be reduced to my own, self-declared "comfortable" levels.

 

[Sinocpm]

I've read through the changelog for the upcoming 3.7 update and I will echo what others have said here before: I'm also very much looking forward to this update! Lot's of exciting and sensible changes and tweaks!

I've attached a screenshot which highlights two issues from the current official version, I could not find anything about that in the changelog, apologies if it has been discussed here before:

First from a UI perspective, if there are more than a certain amount of units on one tile (seems to be around 16) the list grows too long, so that the first unit on that list on the top gets hidden underneath the science/culture/espionage sliders and the upper left part of the main UI. Is there a way to remedy that?

And secondly as evident here the AI still loves to create enormous doomstacks. I took this screenshot after this particular stack had already been decimated to about half it's size, meaning it started with more than 80 or so units. I've seen bigger ones yet as well.
Overall I think the system of stack aid bonuses and logistic problems penalties are an elegant way of discouraging the player to just cram all available units into one huge stack. As I and others mentioned however it seems that if you have a big (or even huge) but diverse stack you will still end up with a net positive (although maybe small) bonus for units in that stack. Maybe that is why acording to cold AI logic it still makes sense to create those doomstacks. Or maybe the AI just can't handle/understand this system of bonuses and penalties because it wasn't part of the original Civ IV code?
In any case I would suggest two things: 1. ramping up the logistic problems penalties, especially for the higher tier ones. 2. On the other hand I feel like the "free" unit numbers are quite low, across all eras in my opinion. So I would suggest raising these a bit, maybe 1-2 units for the ancient era, 2-3 for medieval, 3-4 for industrial and beyond.

 

[RezerCuid]

I am also in favor of increasing the penalties for the number of units on 1 tile