Any mod that proves whether players like/dislike AI playing for victory?

[itix]

In Civ 1 howitzers ignored city walls, so typical 3 x riflemen AI defence was easy to defeat. To survive, the defender had to take the initiative and destroy howitzers.

In Civ 6 I can park archers behind the city tile and shoot attackers on the other side of town Defending is made too easy.

 

[Marla_Singer]

In Civ 1 howitzers ignored city walls, so typical 3 x riflemen AI defence was easy to defeat. To survive, the defender had to take the initiative and destroy howitzers.

In Civ 6 I can park archers behind the city tile and shoot attackers on the other side of town Defending is made too easy.

Defending is too easy but it's certainly considered as a feature rather than a bug. Otherwise it would have been addressed at some point over the last 15 years.

I think Firaxis could learn from how Nintendo handles the difficulty curve. Nintendo games are very accessible at first, even reaching the end credits can be achievable without excessive difficulty. But gradually, without even noticing, you're invited to test yourself a bit more in optional contents. You then realize that's harder than expected, but it still feels doable, so you just keep trying. In the end, those challenges that you weren't actively looking for become the most memorable moments you had in the game. Those where you had the most fun.

Sure, many players don't launch a game with the intent to struggle, but that doesn’t mean they wouldn’t enjoy it if things actually turned challenging.

 

[Mark the Bold]

I thought Civ V handled it well. It usually comes after you've conquered your third capital, so you've made it clear to the entire world that you're a menace. In the diplo screens, you've been getting the "they have concerns about your warmongering" message. It's a plausible response on the part of the world to your actions.

If it comes again just to stop you from winning the game, that has built into it the historical implausibility that "how do these civs know we are reaching the 'end' of history?" But that implausibility is built into there being victory conditions at all, so it's one I've already swallowed, just to play the game, as a game, at all.

This. I heard people complain its too "gamey" when the AI's do this. But it mimics human behaviour perfectly. I hardly think that if you were playing RIsk with 5 buddies, and one guy takes over Asia / Australia, the other 4 are going to fight amongst themselves and let you steamroll. No way. At least my friends wouldn't.

Heck the entire history of Germany over 500 years has been that dynamic. When Germany becomes too powerful, all of its neighbors gang up on it and cut it down to size.

 

[Thormodr]

This. I heard people complain its too "gamey" when the AI's do this. But it mimics human behaviour perfectly. I hardly think that if you were playing RIsk with 5 buddies, and one guy takes over Asia / Australia, the other 4 are going to fight amongst themselves and let you steamroll. No way. At least my friends wouldn't.

Heck the entire history of Germany over 500 years has been that dynamic. When Germany becomes too powerful, all of its neighbors gang up on it and cut it down to size.

Yeah. You really want the AI to mimic human behaviour.

Do you remember the abomination in 5 when they had the AI rage quit?

 

[sixty4half]

Still, the biggest threat in civ 7 for your cities isn't the other players, but the coastal independent powers. If they arrive at your freshly founded town with 3 galleys... argh!

I would say that easily 20% of my games end by IP. Either 1 coastal or 2 nearby sending 6 warriors very early.

 

[JNR13]

Yeah. You really want the AI to mimic human behaviour.

Do you remember the abomination in 5 when they had the AI rage quit?

I am still trying to play a single match of Civ V. The AI keeps rerolling on turn 2 every time.

 

[Thormodr]

I am still trying to play a single match of Civ V. The AI keeps rerolling on turn 2 every time.

Lol. Nice!

 

[vorlon_mi]

I've played the most Civ3, though I got into the series with Civ2. My experience with both Civ3 and Civ4 -- especially on continents maps -- was that an AI player on the other land mass had the potential to runaway. They often, but not always, gobbled up some cities from their neighbors. They would build towards a culture or diplomatic victory. If they had a tech advantage, it was progressively harder to invade them and stop them from winning. The Civ3 AI (especially certain leaders) love building wonders; all the AI love constructing culture buildings. Both of those are key to culture wins.

I played the least Civ5, but I have seen some interesting behavior in the Civ6 AI players. They will pursue the culture/tourism, religious, and diplomatic victories. Each of those involve *building*, rather than conquering. Building wonders for diplo victory points happens, albeit at a slower pace. Certain leaders build lots of Theatre Squares and put Great Works in them. If I'm pursuing space (as I often do) and did not engage in military actions in the Medeival/Renaissance, the AI are contenders for the diplo victory by turns 350-400. Yes, yes, I know, don't let the game go on that long. But my point is: the AI do pursue a victory condition.

The current AI don't min/max to achieve that victory condition. But if I don't keep up with them, they will beat me to the Statue of Liberty. They will spam Apostles to convert their neighbors and me, if I don't stop them.

 

[civ4-advciv-oracle-bug]

Hello,

It is a bit spammy of me to mention it here as well, especially as this is not a civ4 thread per se, but since it seems mentionned and pehaps relevant, i have released a mod based on AdvCiv called AdvCiv-SAS (Simple Advanced Strategy) (Modpacks downloads section link) that greatly increases AI efficiency and competitiveness, with otherwise overall minimal additions but heavy adjustments, even though it should feel familiar for a base advciv or even bts player.

It rekt me at monarch although i was bit rusty but still (see a few screenshots of how it went in the AdvCiv-SAS discussion thread), which i would win comofrtably/easily on (sometimes even on emperor as long as pangea at least last time i tried xd) in base advciv 1.12.

As this thread seems to be about min maxing, i'd be looking forward to hear from players that attempt to play my mod and share how it went and their thoughts on it. I would just advise to lower difficulty to get adjusted to the AI and changes at first.

I want to insist that penalties are overall a bit lower than base AdvCiv especially at higher difficulties (on lower difficulties the advantage is smaller than in base advciv as well), so whatever AI competitiveness or efficiency only comes from it playing smarter xd, although it could be improved more in many ways, thanks,

(edit: fixed discussion thread link that was redirecting to downloads page rather)

 

[kaspergm]

I think this is an interesting discussion. I have a few thoughts, which I think I've shared before.

1) I don't think having an AI that solely focuses to win would be fun. Given the current system of AI production bonuses, the AI could always spam warriors from turn 1 and then steam-roll the human player as soon as they meet. That would not be fun. I've lost a number of games to Sumerian War Cart rushes in Civ6 and it was ... meh.

2) I don't think having an AI that completely fails to grasp a victory condition that is within reach is fun. I remember many cases of Civ6 where an AI is run-away in science and then builds all space projects but never finishes the last one. That's like ... WTH?

3) I hated the "the dislike you because you are about to win" mechanic in (early?) Civ5. That was very immersion breaking. Having all Civs DoW you is not fun and kills any purpose of having a diplomatic aspect of the game.

4) I do think it's possible to have a sane situation where AI will dislike you if you are competing for victory. Basically I think game needs to divide (AI) players into two groups: A "game-leading" group of 2-3 (maybe scaled by map size) who are leading the game and thus contenders to win, and a "spectator" group, i.e. minor AIs who are falling behind and basically are playing defensively to just stay alive. I think it's perfectly fine to have game-leading civs envy you for your success or power (particularly if it's phrased like that) - think USA, USSR, China, etc. from real world modern times - and have them engage actively and negatively with you to prevent you winning. On the other hand, minor civs who are completely inferior should obviously not go on suicide missions against you.

 

[civ4-advciv-oracle-bug]

The way i see it it doesn't seem necessary to split AIs into different roles, as this happens naturally due to different starters, barbarian, tech trade etc, so it is rare that all AIs are exactly at same pace anyway, but it does happen quite often that say half are still competing in the quite late game (say turn +/- 300 in my mod which is near to future era in autoplay), so as far as variety goes i think it happens naturally.

About your point 2), it is maybe tricky though, as going say for culture and build wonders, when opponent is going to conquer them/you is also quite idiotic, so sometimes you'd have to prioritize some victory conditions over others (in particular war) if you'd want AI to be competitive/challenging. In my mod i make sure they don't build wonders unless it doesn't take too long (+/- 20 turns as of now adjusted to game speed), but otherwise don't interfere too much with their wonder interests unless it is harmful to them (enemy is stronger, war, etc -> redirect hammer to military, or such other examples/cases if any anyways etc), but in the end in the later game they build a few and more which is fine as their production is better, survivability is gained for the most part so far at least but always stay alert anyways etc.

About the 50+ warriors you mentionned, it actually happened experimentally, i was fixing a no production bug (at least tremendously reduced it), and during one of the experimental code autoplays i tried to fix it, AIs would produce stupidly non stop warriors (called ancient macemen in my mod) since turn 1 xd, in all cities, so at turn 100 they'd have like 50 warriors stack xd. While it seems like a great idea, actually i'm sure it's probably harmful, because you can't kill all players so fast, in my mod longbows are fairly soon at tech_construction, they can slave a bit if needed, so by the time you get such a stack, you are massively behind in tech, and also only have one city, needless to say even if you somehow bruteforce kill one or 2 rivals, this warrior stack will melt with later units or simply defensive strategy of rivals that would make them obsolete i guess (although i didn't test it, in my mod ancient macemen are bad at attacking cities, can't rely on that). So i think favouring military early is fine for efficiency, but AI is bit competitive as a result, and rekt me due to not expecting that and being bit rusty too (perhaps AI somehow was bit stronger in the map i played as my starter land was bit worse so aybe theirs was bit better relatively who knows, but mostly on me still), however it's not like they don't focus in(on?) other areas as well like building cities, economy, techs, etc (at least they seem to do so well enough that i didn't need to rein in if i may say their military thirst xd)

But the key is maybe your 1) point, indeed getting rekt was not fun, and it involves a lot of min maxing to be competitive, but somehow it lingers in my mind, and i'm passively thinking if i did this that or retry this or that then maybe xd what would happen, in my mod at least it doesn't seem unreachable, but difficulties like settler or such are no placeholder (and like i said penalties are bit lower overall than base advciv so totally within reach, only need to refine/tighten strategy and skill which takes time/skill too xd but anyways etc), if you can win at settler you already did decently well as it should be hard enough (something like +/- 20% advantage vs AI which considering how well it plays (in my mod it adjusts to handicap settings too, so if difficulty is low (i.e. penalties it has high enough), it would rely for example more on spying and less on research xd, more on war as a human would if playing deity xd i guess although i didn't play deity myself to tell too much xd)). So maybe this is more about what *most* players would want in a game, which i assume most of xd but anyways etc (and perhaps me too if too brutal *maybe (not sure)* but anyways etc) don't look forward to the stomping, but as for me, even if i don't play i'd love to see a player trying to minmax and defeat the odds, it's very enjoyable and like watching sports/soccer/esports whatever competitive game or such, you don't necessarily have to play it yourself to enjoy it xd, but if opponent is too weak it's boring though for sure whether playing or spectating i would guess.

As for the rest it's maybe a matter of preference and maybe fine as such but just my opinion i mean thanks for feedback as well i added my thoughts though (and fixed upper link to screenshots where i got wrecked xd), in my mod optimal strategy is not strictly military, it would backfire to be too military so they are competitive enough at techs too although i didn't tweak it too much, but just efificency gains on better settling cities and worker improvements and city specialists and such tweaks/fixes of what was sometimes massive issues (no production many turns, citize specialist city stagnant at pop 2 not working the wheat in city radius, etc.)

 

[GeneralZift]

Yeah. You really want the AI to mimic human behaviour.

Do you remember the abomination in 5 when they had the AI rage quit?

Uhhh what are you talking about? Or is this an ancient scrapped feature

 

[GeneralZift]

AI needs to play somewhat for Victory and somewhat in character. I think the worst thing they can do is have it not be competitive.

Competitive AI like Civ5, and so I've heard, pre-Civ5, are important to have the player actually involved in the game.
I'm not sure whether newcomers prefer the steamrollable easy AI that the later games have, but maybe they do.

Almost every Civ5 game I play with Alexander, he conquers half the map, enslaves all the city states, and shafts my victory. Caesar is usually also potent. I just don't feel any threat from AI in later games, unless it's from their artificial bonuses

 

[civ4-advciv-oracle-bug]

Yes in base advciv it is precisely what it it very frustrating for me, like the game was tough in monarch/emperor, then they'd do something MASSIVELY stupid, like abandon >80-90% of the units guarding their strong capital, to be totally exposed on a roadless tile, with my 25+ stack of cataphract that was incoming and that can now freely nuke/ambush them or simply ignore them and capture their 3 unit guarded capital, and many other such blunders.

I don't think i fixed all such issues and i didn't especially test it, but in theory these specific issues should happen a bit less in my mod.

But point is yes, after that game was massively less challenging, it's in fact one of the main if not the main reasons that pushed me to develop AdvCiv-SAS xd

When i was playing regularly advciv, around turn 150-200 is where i would quit games as i was ALWAYS in one way or another ahead and it was just already decided for most grinding to actually finish it.

But in my mod? You'd better stay tense till you win xd, i mean it's not perfect but a long way from how it was xd, stay on your toes if i may say which i think is exciting but very frustrating if you just want to coast easy as i thought i could after i had a decent production and a 10 reserve romaing stack xd, nope it wasn't won nor safe actually, real challenge xd and lesson learned, but for now i'd rather look at players that aim to try the challenge xd, will see if i actually play it

About caesar and such, they are surprisingly weak in base advciv as well for some reason, i made sure to boost their buildunitprob as well as of other leaders for accuracy, see (leaders (ai) note in my very quick get started guide) for details hehe if interested.

If you're into minmaxing and hard opponent and want to perhaps try my mod please tell me how it goes, ideally if you have video even better if not screenshots or textual feedback is fine too i mean, if not and you know someone that tried it please tell me as well how it goes/went, in all cases thanks i mean anyways etc

edit: about identity/character, yes too i mean, i didn't go out of my way to erase their base preferences or make them ultra efificent, most of my changes are more like pre-filters of: "do this / keep doing this as long as it doesnt harm you, else do that rather in such cases" (that happen to be quite often, but differences between leader profiles should still exist although a bit less than in base advciv due to the more focused mindset (gandhi may go miltiary for example xd if circumstances dictate for it or/and it is the best play (say he has a nice land and high unit count or is ahead in tech or whatever, it did happen in autoplay although justin superseded him in the end if i am not mistaken but anyways etc), all in all i kept some flavour.

But more generally for the how ai should be question i'd say that some identity is still nice, just they should ideally not play stupid or obnoxiously so and then fine, so i didn't interfere too much if a leader prefers canons vs say mix of grenadiers or muskets or whatever (in fact i didnt interfere at all, only the global strategy i adjsuted rather no expansion if in danger, no naval units if threatened, more like hard rules / pre-sets in my mod for as effective as they are which is not always so but often enough i'd say/guess but test to be sure interested i mean but anyways etc))

 

[aelf]

4) I do think it's possible to have a sane situation where AI will dislike you if you are competing for victory. Basically I think game needs to divide (AI) players into two groups: A "game-leading" group of 2-3 (maybe scaled by map size) who are leading the game and thus contenders to win, and a "spectator" group, i.e. minor AIs who are falling behind and basically are playing defensively to just stay alive. I think it's perfectly fine to have game-leading civs envy you for your success or power (particularly if it's phrased like that) - think USA, USSR, China, etc. from real world modern times - and have them engage actively and negatively with you to prevent you winning. On the other hand, minor civs who are completely inferior should obviously not go on suicide missions against you.

Doesn't Civ6 have a negative diplomatic modifier for this? I seem to recall getting one when building space projects.

I think Civ4 had something like "They fear you" as a big negative modifier too.

 

[vorlon_mi]

Civ3: my experience most often was that the AI liked to build things. Certain leaders/civs loved wonders, some built lots of units, but all built buildings that the really didn't need. Since it was possible to win some of the victory conditions (culture, space, UN / diplo) by simply building, the AI were working for victories if the game lasted long enough. At the highest difficulty settings, the AI players would field a large army and invade, both each other and the human player.

Civ4: Similar to Civ3, in that AI players would build stacks and invade each other. Religion made some civs/leaders enemies; weaker tribes would sometimes "peace vassal", basically capitulate, to stronger factions, making it more challenging for the human player. I always felt that if I was not actively fighting an AI, then I should be preparing for the next war. I saw the most runaway AI in Civ4, especially on continents maps.

I didn't play enough Civ5 to really get a sense of their aggression. I never installed / played Vox Populi, sticking to the main expansions.

Civ6: In contrast, the Civ6 AI players were much less aggressive than either Civ3 or Civ4. Since I love the space victory (in all the franchise games), I often play games that last 350-400 turns. Militarily, I can't remember seeing many AI that conquered an original capital. Yes, they compete for diplo victory points. Yes, they spam their apostles towards me and other weaker AI. Since they also like building things, I've seen some build lots of Theater Squares and museums with great works, accumulating tourism. The Civ6 AI may have build a spaceport district before me, but rarely launched their exoplanet ship.

I do want the AI to understand what the victory conditions are and to work towards them. If I make a major screwup, or simply do nothing for (say) 50 turns, I want at least one AI player to gain ground, move ahead in score or science or culture. I would expect the AI player to recognize an opportunity to invade / take advantage of a weaker neighbor (human or AI), to advance its own empire. I want to have the risk of losing -- literally, another player achieving a victory condition -- if I do nothing.

I would not find it fun to be targeted from turn 1 for invasion. The barbs in Civ6 had some of that flavor. I understand that some players like "coming from behind," digging out of a hole, and that's part of the starting bonuses. Me, not so much. I don't want the AI players to pursue the victory condition "eliminate the human player in SP" as their primary VC. That's a unique VC that I can't pursue. My fun is increased when they compete with me for the same VC's.

 

[socratic]

Doesn't Civ6 have a negative diplomatic modifier for this? I seem to recall getting one when building space projects.

I think Civ4 had something like "They fear you" as a big negative modifier too.

It does in Civ 6, but it doesn’t affect anything too much until you’re like 10 turns from victory. I’ve only really seen them act when I’m going for science victories. Civ 6’s AI serves only to be annoying imo. The biggest threat you’ll face are barb camps spawning a billion horse archers on turn 15.