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

[aelf]

I know nothing about mods. Has there ever been a mod, not limited to Civ7, that makes the AI actually go for victory competently?

There are regular complaints about the AI not being able to win in more recent civ games, and in Civ 7 this seems starker because the AI is sometimes only failing to take the final step (e.g. building the World's Fair). Arguably, even in older civ games, the AI wasn't all that focused on winning either. And the response to this that frequently comes up is players wouldn't enjoy AI that plays to win.

So has there been any proof either way? Whether people find it fun to play against AI that knows exactly what to do to win at least one of the victories?

 

[stealth_nsk]

I know nothing about mods. Has there ever been a mod, not limited to Civ7, that makes the AI actually go for victory competently?

There are regular complaints about the AI not being able to win in more recent civ games, and in Civ 7 this seems starker because the AI is sometimes only failing to take the final step (e.g. building the World's Fair). Arguably, even in older civ games, the AI wasn't all that focused on winning either. And the response to this that frequently comes up is players wouldn't enjoy AI that plays to win.

So has there been any proof either way? Whether people find it fun to play against AI that knows exactly what to do to win at least one of the victories?

Civ5 vanilla AI was not "competent" in a way this word is usually used, but it had some play to win mechanics, like backstabbing (attacking player despite good relations). This caused huge amount of negativity and was later changed. But we clearly don't have any quantitative data on how many people disliked it.

P.S. Also, "competent" AI and playing to win in reality are independent things. Since AI is a game mechanics and doesn't play the game the way player do, the competency is something perceived by player. For example, many see Civ4 as having best AI, although it has very simple logic compared to later Civ games. The reason while it looked smarted is because SoD didn't allow AI to display tactical incompetence in front of the player and most of the bonuses were hidden. Similarly, Vox Populi mod is usually praised as the best 1UpT AI in all civ games, but it actually gives challenge by having more AI bonuses than unmodded Civ5. Appearing competent to player doesn't mean providing actual challenge.

 

[ThERat]

To answer the question, I am playing AW against Civ5 VP now, it's way more challenging, given the popularity of VP, I'd say, yes people like it. Civ4 had better AI and it was also enjoyed by certain people.
C3C now has a new exe mod which makes the AI more competent and I personally like it a lot, suddenly facing stacks of inf/cat threatening anything in sight or even AI using armies properly.
So yes, there is an audience, but I don't think thats what Firaxis has in mind. Civ6 and now Civ 7 seem to be concerned foremost with copious amounts of features, that the AI cant handle.

 

[aelf]

I don't recall Civ2 AI playing to win (it could win by space but seemed to take ages, often not before time ran out), and I may have abandoned games too much in Civ3 to know, but I can talk about Civ4.

Civ4 AI is a threat, and certainly SoDs is a big factor. But, at least when I was still playing, stacks were also the way to neutralise that threat. The standard early game strategy was to go conquering with a stack of axemen that would overwhelm and, eventually with promotions, outclass the archer-defended AI cities. So stacks are more of a disadvantage for the AI in the early game, the most important phase of the game. I feel like with 1UPT, since there isn't the same concentration of force, the AI has more an advantage in the early game when it can pump out significantly more units than players and promotions haven't really come into play yet.

So setting that aside, is the Civ4 AI better at winning? For one, it plays by the rules of diplomacy, which is more or less transparent to the player, and they can therefore be manipulated through the system. If the AI manages to run away, it could certainly win a space victory and it would be difficult to stop militarily (which is probably why it's perceived as better than AI post-1UPT). However, does the AI in Civ4 intentionally go for a victory and prioritise doing the things needed to win it? I don't think so.

There's a difference between the AI posing a challenge and the AI playing to win. The AI being a military threat poses a challenge to players, one to overcome in order for players to win. The AI playing to win means players can simply be left in the dust at higher difficulties. It would be like the runaway AI in Civ4 but worse. As it is, you might have a bit of time to muster a force to at least try and put the brakes on its victory, but I suspect that window would almost be non-existent with AI that plays to win. The only way to play would be aggressively from early game to keep all the AIs down, and that's assuming there's no other continent you can't touch till mid game.

 

[ThERat]

In Civ3, the agricultural civs were no joke on high difficulties -. They would routinely take out other civs and become very strong. The would pose at least a noticeable threat, so you would have to be on your toes at all times.

 

[a pen-dragon]

However, does the AI in Civ4 intentionally go for a victory and prioritise doing the things needed to win it?

Actually, it does. This is mostly visible for cultural victories, where it will concentrate culture in the three required cities and actually favour culture over science. An AI going for culture is usually very fast. For a science victory it is less focused, but also definitely a threat. For domination/conquest victories, I think it behaves more aggressively, but more in a general manner that is not very targeted.

IMHO the biggest AI failures in that regard are not in pursuing a victory condition, but choosing one. For instance an AI might choose to go culture while sitting just below the domination threshold and lose the game due to this.

 

[thecrazyscot]

I'll wager my eggs and toast this morning that most civ players don't actually want an AI that "plays to win" and is good at it.

Look at all the hysterical screaming about forward settling. People don't like the AI encroaching too much on "their" space, despite it being a...pretty good strategy. Hem in, deny resources, establish a base of operations. If the AI were actually good enough to drop ship large armies in their FOBs and then surprise attack? The level of rage would be through the roof.

What most people want when they say "give me good AI" is actually "give me enough of a challenge so my victory doesn't feel hollow".

 

[Fluffball]

A few things have been mentioned which really should be brought together rather taken in isolation to get to the root of like or dislike of AI 'challenge'.

To summarize i feel it is fair to say players like it to feel challenging but not artificial.

If we take the two examples, firstly the mechanic where you have the AI sees the player is winning and everyone just DoWs the player to slow them down. This is completely predictable and false difficulty that is not challenging but just adding additional grind. The big joke is that it usually involves every AI DoWing the player, including tiny neighbour's with one city and no army which really just makes the player stronger as it can absorb all those easy targets and then the player knows it is always coming so is always prepared while also leading the player to be disappointed by the 'diplomacy' aspect as they know the AI will just always hate them eventually so why bother trying to be friends and the game is essentially always won by being military strong and aggressive.

Secondly the mechanic of just massively buffing the AI to make up for their shortfalls. In worst cases giving the AI extra settlers and military right at the start. Civ 6 was a classic case of this where the AI could start with as many as 3 settlers, sometimes making it impossible to even place another city once the influence/loyalty system was introduced as any placed city would automatically be overpowered by the influence from the AIs starting 3 cities.

In general this mechanic usually just ends up being a case of surviving X number of initial turns and you know you have won the game because you managed to get over the hump or took advantage of all the bonuses the AI starts with ymto get an even better head start, stealing their workers, stealing their cities etc.

This also leads to the player having no options in the start but to simply survive the AI rush, you can't build any wonders as the AI always gets access to them first, has much higher production than you, starts with more cities and you have to spam military to fight off the know AI rush that is coming.

This is them made worse by the fact you know the AI is not focusing or going for a strategy but gets all the early wonders as it runs out of other thing to build, which is made obvious when it comes to winning when the AI has all the cards to win but just doesn't (probably to as it would just always win due to all its bonuses if it was allowed to) which just makes the AI look even more stupid.


The strategy games/mods I have found most enjoyable in recent years allow the player and AI to start on an equal footing and slowly ramp up the AI bonuses as a rubber band mechanic. If you don't push the AI will still run away, if you get ahead then stop pushing the AI will catch up and take over. If you consistently push and focus you can keep up, get things like wonders you want if you actually focus and have to keep going until the end of the game.


Civ 7 has tried a version of rubber banding with the era reset but it just seems even more false and obvious, it also loads the AI with bonuses while making things more expensive for the player yet it is even more easy and the AI looks even more stupid as it is passive, you are never really in a race with the AI so giving it loads of bonuses doesn't matter and it can't even seem to start a victory conditions, never mind win one.

Because the AI is no challenge and you don't even have to interact with them really to do 'diplomacy', while the AI doesn't just inevitably hate you there is still no diplomacy

Civ 7 in the zenith or boring, bland, unchallenged gameplay and as I think about it, it would probably be a better game if there was no AI 'competition' and was just some sort of builder game with the use and entertainment the AI brings to the game.

 

[Gori the Grey]

One indirect way of getting at @aelf's question (mindful of @thecrazyscot's axiom, which I think highly probable) might be a poll that asked "what percentage of games do you win at your favored difficulty level?"

I play Civ V at deity level and I suspect it would be stretching the truth to say that I win 1 in 10 games.

Everyone who answered < 50% would presumably be ok with an AI that "plays to win."

Edit for x-post: Also, I feel differently from @Fluffball on one point. I think the AIs banding together to keep you in check when you are the runaway 1) is plausible and 2) does represent a legitimate game challenge. (That's setting aside the question of how well they execute it). If you are going for conquest victory in Civ V, you can be sure there will be a time when 3-4 civs will all declare war on you at once. I find this the most exciting moment in the game. My troops are often spread thin or concentrated in the spot where I am attacking, and it can be a real scramble to get enough into position to defend my previous conquests/home turf. I actually wish the games' competitiveness and rubber-banding was based on this principle: the civs keep an eye out for runaways (in any dimension of the game: economic, scientific, military, cultural) and they have mechanisms for banding together make things more difficult for the runaway.

 

[stealth_nsk]

One indirect way of getting at @aelf's question (mindful of @thecrazyscot's axiom, which I think highly probable) might be a poll that asked "what percentage of games do you win at your favored difficulty level?"

I play Civ V at deity level and I suspect it would be stretching the truth to say that I win 1 in 10 games.

Everyone who answered < 50% would presumably be ok with an AI that "plays to win."

That's a pretty good metric. Threshold could be discussed though. Also, I don't know where to put restarts - they could range from just not wanting to play the game to avoiding immediate loss.

Finally, there are players who claim Deity is too easy (not sure if it actually is or it's because of mods), so they'll have high percentage regardless of whether they want challenge or not. Maybe a separate option for them is needed.

 

[Gori the Grey]

Theoretically, the "Civ-Ideas and Suggestions" section should be a place for discussing such matters--that don't pertain to any particular iteration. But because it's in the Civ 7 main section, I don't think it really operates that way. It's too bad. I would often like to engage more general discussions of the design of 4x games.

Also, I don't know where to put restarts

I keep all of my games, saves every fifty turns. (I have no idea why. I never go back to them). I put them in folders. I don't finish every game. In fact I don't finish most games. But I force myself to categorize every quit game as a win or a loss. I have three main folders: "loss" (for when I've played 150 turns in and it's clear to me I'm not going to win), "call it a win" (for games I feel certain I would win, but don't want to bother grinding out), and "quits" (for games I quit early because it's clear to me that I'll never get any traction). But quits count as losses in my own accounting.

 

[AriochIV]

I'm not sure how a mod would "prove" anything since most players don't use mods.

I think that most players want to be challenged, but they also prefer that the AI players behave in a manner consistent with the internal narrative of the game (rather than as if they were human players playing a boardgame). I don't think these two goals are mutually exclusive... especially with all of the non-military victory conditions in the game today, there's no reason that AI factions can't "play to win" while still friendly or even allied with the player.

I do recall that Civ2 had rubber-banding mechanics; the AI civilizations would cooperate against the player in the late game and rapidly catch up in technology if you were ahead.

 

[Gori the Grey]

(rather than as if they were human players playing a boardgame)

That might be another thing worth polling, because if you are correct, then I am in the minority.

In my ideal civ game, each AI would at some point pick a victory condition. This would come about 100 turns into the game and the decision would be based on the victory for which the civ was best suited in general and their accomplishments up to that point in the game. They would play out the rest of the game min-maxing toward that victory.

There would be ways of my knowing which one they picked and how far along they were. Not with absolute accuracy, but with a certain degree of reliability.

That would make the late game interesting to me, because I would feel the race (that Civ games fundamentally are) as a race. I'd have a sense of what I needed to do to reach my chosen victory condition more quickly than any of the AI reached theirs.

In short, it would feel a lot like playing a boardgame against human players.

 

[Taefin]

I'm not sure how a mod would "prove" anything since most players don't use mods.

I interpret the question as working on two premises:

1) Vanilla AI doesn’t win games most often
2) In previous iterations of civ, there are stories of mods and iterations of mods where the AI would win games (semi-)regularly.

Therefore, memory of those mods is a source of some evidence whether a likely-to-win AI is fun in the context of civ.

I don’t have any such experience within civ, having joined in 6, but Old World has a combination of difficulty settings, game mechanics, and AI where I can set up a game I know I will lose, or a game where I have a good chance of losing.

The main mechanics helping AI win in old world is that victory is a score based on number of wonders and cities at each level of cultural development, which AI bonuses directly contribute toward, creating a real urgency to stop a leading AI (or effectively creating a domination victory where stealing X cities from the leading AI will decide the game).

A likely-to-win AI in civ7 would just need to reliably build its victory projects as soon as they became available, and the AI would be winning on science often, just using its science bonuses, which each player could scale to their preference.

I am curious whether this is a conscious or accidental design feature that AI seems to stall this key stage of victory.

 

[aelf]

I'm not sure how a mod would "prove" anything since most players don't use mods.

I'm not thinking about quantitative data.

 

[stealth_nsk]

Well, the

I'm not thinking about quantitative data.

The sample will still be quite specific. I think players who heavily mod their games have different behavior patterns from the rest.

 

[Krikkit1]

I think there are 2 separate questions
1. should the AI be Trying to win in the same way as a human would
2. How often should the AI succeed at that

They are 2 separate questions
If I make an AI that is just some top civ Deity players locked in my computer…I can still win 100% of the time at my preferred difficulty level if that difficulty level gives them big enough penalties.
or I can win 10% of the time at my preferred difficulty level. (if that’s what I prefer)

So an AI that plays to win will utilize the mechanics the best possible given the bonuses/penalties it has on a certain difficulty level.

To make a game stay challenging, the idea of adjusting the bonus to the AIs based on the players strength is good. (and hopefully

I think the answer to #2 for most players is 90+%….The problem is too many players think they should be able to do that on the highest difficulty level.

Which is why it’s a mistake to have a highest difficulty level, they should be effectively infinite, since all they do is change bonuses.

 

[Fluffball]

Edit for x-post: Also, I feel differently from @Fluffball on one point. I think the AIs banding together to keep you in check when you are the runaway 1) is plausible and 2) does represent a legitimate game challenge. (That's setting aside the question of how well they execute it). If you are going for conquest victory in Civ V, you can be sure there will be a time when 3-4 civs will all declare war on you at once. I find this the most exciting moment in the game. My troops are often spread thin or concentrated in the spot where I am attacking, and it can be a real scramble to get enough into position to defend my previous conquests/home turf. I actually wish the games' competitiveness and rubber-banding was based on this principle: the civs keep an eye out for runaways (in any dimension of the game: economic, scientific, military, cultural) and they have mechanisms for banding together make things more difficult for the runaway.

To clarify, there is nothing inherantly wrong with the AI DoWing you to slow you down or because it sees you as weak but it is the obvious, false and poorly implemented way it happens.

I have had strategy games (not just civ) where the "we have to kill the player" check box is activated and everyone just DoWs me even if they are my friend or they are the civ i left alive with one city so I didn't get the 'wiped out a civilisation' massive debuff (even though they probably attacked me [repeatedly] ) like it has a final suicide wish.

A couple of big AIs ganging up on me to beat up on me seems fair enough. The whole world DoWing me because we are X turns from me winning, including a load of small AI who I can roll over in 1 turn is obvious, easy to deal with, boring, unchallenging and often just makes the victory more assured as I can now wipe out all my main competitors as everyone hates me anyway.

It is the lazy implementation that is the problem, not the concept.

 

[civac]

Civ4 AI does go for the culture victory in a planned manner. (Which is not to say they are good at it.) The AI culture victory strategy has several phases. They can enter the first preparatory phase VERY early in the game. In these preparatory phases they found and spread religions, construct culture buildings (in particular the religious culture multiplier buildings like Cathedrals). Then in the last phase they use the culture slider, directing all of their empires commerce into culture for the last sprint to victory. Civ4 AI will not go after domination or spaceship victories in such a cohesive manner. They can win either victory when they snowball hard enough. In particular, they build spaceship parts with a very high priority when they become available. So when they do reach the end of the tech tree they will launch a spaceship in short order (2-4 turns or so on high difficulties). They do not prioritise spaceship techs over other techs.

 

[aelf]

Well, the

The sample will still be quite specific. I think players who heavily mod their games have different behavior patterns from the rest.

Well, pretty much no one discussing preferences for AI gameplay on this forum is relying on quantitative evidence. I guess for the purposes of this thread, you can say the proof I'm talking about is personal testimony - i.e. people playing such a mod and reporting how they feel about the topic.