Bribing AI to DoW another AI - is it really worth it?

[Fine_Distinction]

I have read about this strategy a lot, but never used it myself. I am currently playing Monarch, #1 in score, but with another civ close 2nd and Monte close 3rd and a clear leader in military. I am doing OK, but need some time to tech more and get military advantage over Monte.

Monte would currently DoW #2 civ (he is annoyed with them) for 3 techs, including construction. I decided giving Monte construction is not a good idea But two other techs he wants are non-military, so I wondered if it is worth haggling with him?

What are effects of bribing AI to DoW:

1) Does the victim civ knows and hates/DoWs me more?
2) Does AI often "fake it", i.e. take techs/money but makes peace after 10t without any fighting? (my specific example is Monty, he normally does not makes peace easily, I wondered if it is different if he is bribed?)
3) Do I get some positive diplomatic bonuses with the civ I bribed? (I do not plan to join the war)

Would be grateful if somebody has more experience with this and can clarify.

Or is it very situational?

One more question: when AI accepts my bribe, am I safe for 10t for him DoWing me? Can I DoW him (not that it makes much sense, just out of curiosity how this mechanics works)

 

[AcaMetis]

1. Yes, the victim will always know who did the bribing and you will get -1 (or -2, don't recall off the top of my head) for "you've brought in a war ally against us" as a result. That alone is not a guarantee that you're more likely to get DoWed, though it might bring your relationship down a notch, and going from Pleased to Cautious (or worse) does increase the chance of an AI plotting on you. In many cases it'll allow an AI to start potentially plotting on you in the first place, even.

2. They can, but in my (perhaps limited) experience it's fairly rare, especially when bribing AIs that have actual stacks of units ready to go. It is possible for things like the AP to force an early end to a war, though, so keep that in mind if it's in play.

3. Not if you don't share an active war, no. You might get Fair Trade bonus for trading with them, but that's not something unique to war bribes (in fact I'm not sure if war bribes can even give Fair Trade, but if there's a positive diplo bonus for bribing someone into a war you're not partaking in that'd be it).

As for whether it's worth it, it depends. The big thing to keep in mind is that AIs can potentially bribe in other AIs, to the point where one war bribe escalates into a full-blown world war. Keep an eye on what techs the people thrown into war have and whether they could potentially use that to drag their friends in.

 

[Capellan]

I'd say it is situational. It can definitely be useful for instance if you want to attack Civ A and are worried that they might bribe Civ B into the war against you. Bribe Cive B to attack Civ C beforehand and they will WHEOOHRN Civ A when you come after them.

(now if only I didn't forget things like this so often when I am actually playing!)

 

[Gumbolt]

On deity it's a great use. It means you can go about your business while the Ai destroy their stacks on eachother. You can also use it for shared war bonus which helps bring an AI to pleased or friendly. Of course you have to be at war with the Ai for this.

I used it on Harv Alex game. There was a 1 city AI left. Took the other Ai 20+ turns to reach the city. By which time I had 2-3+ duplo with GK who is a warmonger by heart. Eventually got GK to friendly helped by the shared war.
Pends on the AI peace wise. Some AI like Indians/Mansa have a very low resistance to being a vassal. Where others like Toku/SB may refuse to talk.

Generally your bribing on an AI you want dead anyway. So if your at war with the Ai it hardly matters if you take a 1-2 point hit. Pends on map and how many friends this Ai has.

Yes AI can bribe in other Ai but many you can beg 10G from to keep them peaceful. Need to have them at pleased for this.

 

[Fine_Distinction]

Thanks for detailed answer, G!

On deity it's a great use. It means you can go about your business while the Ai destroy their stacks on eachother.

Yeah, that is the general idea I have to think more whether it makes sense. The problem is I am generally preparing to DoW potential attacker (I mean the civ I consider bribing) but it will take me long time to get ready. On the other hand, I will need to give him some really good techs (like construction) to declare. Potential target civ is another aggressive civ that will become problem sooner or later However, my intended religion switch would probably normally got him to pleased, so he will not be a problem for a while, and I will get demerit for bribing on him. So that really depends on how much real damage war will do to these AIs ...

 

[Gumbolt]

It's cheaper to give techs once you are at war with the AI. I did on one game experiement using Monty as a war partner. The downside was come late game he had really done too well out of war and became quite a pain to deal with. You have to play the map and see what works best at times. Always check the military advisor screen as this will show who is willing to attack who.

Your diplomacy needs to be good for any of this to work. Get into habit of gifting spare resources early on to reach pleased with Ai quicker. Know which Ai will back stab at pleased.

Yeah I wouldn't give strong military techs to anyone you plan on attacking any time soon. Of course if your using HA construction can be quite good. (They will whip catapults for defenders.) If they have ivory not so good.

 

[dankok8]

I often do it to keep the AI off my back. In my forum game with Isabella where I won domination on Immortal, I bribed Qin to declare war on Gilgamesh who I was at war with because I was concerned that Qin could decide to attack me plus he could keep some of Gilga's army occupied. I eventually conquered all of Sumeria with 12 cities in 26 turns and then turned on Qin after.

 

[Jivilov]

Sorry for resurrecting an old thread, but can you ever bribe an AI to attack another AI when the bribee's "Declare War On" that particular leader is red? If so, how? Thanks.

 

[Qactus]

Sorry for resurrecting an old thread, but can you ever bribe an AI to attack another AI when the bribee's "Declare War On" that particular leader is red? If so, how? Thanks.

You can check out the required diplomatic relations here:
https://forums.civfanatics.com/threads/civ-illustrated-1-know-your-enemy.478563/

For example, in order to persuade the Persian leaders to declare war, they need to be friendly with you and they need to be annoyed with the target. People like Shaka are much easier to convince.

 

[AcaMetis]

If you hover over the red name the game will tell you the stated reason why that leader refuses to go to war with the highlighted leader, and if you address that they'll accept a bribe. Relationships can be one problem (either "we just don't like you enough" or "we couldn't betray our close friends"), but it can also be an issue of relative military strength ("we're afraid of their military might"), the AI already planning a war ("we've got enough on our hands right now"), or other factors.

 

[Whisker]

@AcaMetis
And what does my favorite reason for refusal "We would have nothing to gain" mean?

 

[krikav]

That one I really hate!
What do you mean "You have nothing to gain"?! You have everything to gain from weakening my target Ragnar!

 

[Jivilov]

If you hover over the red name the game will tell you the stated reason why that leader refuses to go to war with the highlighted leader, and if you address that they'll accept a bribe. Relationships can be one problem (either "we just don't like you enough" or "we couldn't betray our close friends"), but it can also be an issue of relative military strength ("we're afraid of their military might"), the AI already planning a war ("we've got enough on our hands right now"), or other factors.

OK. So how do you address stuff like "we just don't like you enough" or "we couldn't betray our close friends"? On our snaky continent, after kicking off Huayna to the west and taking his 3-city island to the southwest (culture from his capital had been encroaching on conquered Macchu Picchu) I tried to interest Saladin to the east in a war with Catherine (above Saladin to the northeast along an isthmus) but he didn't want to "betray" her. So all I figured could be done was to build the Spaceship, which was accomplished belatedly in 2028 CE (Emperor level). Was there any way my Julius could have convinced Saladin to go against Catty? Bribe him somehow? Rome had plenty of resources and some advanced techs but didn't want to give them away for nothing. Thanks.

 

[Nick723]

I’m sure @AcaMetis can confirm, but I take this as a softer version of ‘We fear their military might’. They are a bit powerful, but not so much so that we want to admit as such

 

[AcaMetis]

@AcaMetis
And what does my favorite reason for refusal "We would have nothing to gain" mean?

I’m sure @AcaMetis can confirm, but I take this as a softer version of ‘We fear their military might’. They are a bit powerful, but not so much so that we want to admit as such

I'm actually not sure . My interpretation has always been that the AI sees any potential conquests as being too far away from their home territory to want to keep it, Land Target status matters a great deal in terms of AI plotting probabilities, except I've been DoWed by AIs from the other side of an Inland Sea map plenty of times before so it's not as if the AI is shy about taking (and keeping) unprofitable cities. Reminds me of the time Wang Kong took and kept a ridiculously expensive city on the other side of an Inland Sea but still felt the need to go Gunpowder before Code of Laws ...

OK. So how do you address stuff like "we just don't like you enough" or "we couldn't betray our close friends"? On our snaky continent, after kicking off Huayna to the west and taking his 3-city island to the southwest (culture from his capital had been encroaching on conquered Macchu Picchu) I tried to interest Saladin to the east in a war with Catherine (above Saladin to the northeast along an isthmus) but he didn't want to "betray" her. So all I figured could be done was to build the Spaceship, which was accomplished belatedly in 2028 CE (Emperor level). Was there any way my Julius could have convinced Saladin to go against Catty? Bribe him somehow? Rome had plenty of resources and some advanced techs but didn't want to give them away for nothing. Thanks.

Check what causes AIs to like each other and/or dislike you, and to the extent that you're able, change them. If you're running a heathen religion, swap out of it. If AIs are running the same religion, convert one of them to another religion or swap them to Free Religion. If AIs are running each other's favorite civics, either use spies or bribes to swap them out of it, or use spies/bribes to swap them into it if you want the shared civic bonus.

It's entirely possible to end up in a situation where you'll never be able to bribe an AI no matter what, and timing can be crucial if you can only create an opening for one turn, but that's what makes AI diplomacy a complex mechanic. And, of course, this is where being Spiritual really shines. SPI leaders can pull off diplomatic stunts that other leaders literally cannot pull off without the Christo wonder, which is very late in the tech tree.

 

[Qactus]

I'm actually not sure . My interpretation has always been that the AI sees any potential conquests as being too far away from their home territory to want to keep it, Land Target status matters a great deal...

Exactly this.

"We're afraid of their military might" -> failed military check against a land target (requires 8 shared border tiles)

"We would have nothing to gain" -> failed military check against a non land target

edit: Found a post with both the actual code and its explanation for each of the four checks that are done in order to determine the AI response:

Just from a quick look at the code it seems "nothing to gain" is said in a couple of situations. One is when there's only 2 civs remaining and you offer a defensive pact. The more usual one would be in war bribes where it does a military power comparison. Call the proposed target "target" and the civ we are proposing war to "friend"...

T. Calculate target's military power, including any defensive pact partners they have
...If target is a land target of friend (i.e. has 8 or more adjacent land tiles), halve the target's power
F. Calculate friend's military power
....If we ourselves are already at war with target, add our power to friend's power

It's a simple comparison of whether F > T. If not, friend will reply either "We are afraid of their military might" (land targets) or "We would have nothing to gain" (otherwise). Which seems kind of wooly diplomatic talk because really they just don't fancy it

If the power test is passed, it moves on to attitude tests. The whole sequence is first WHEOOHRN, 2nd the power test, 3rd is friend's attitude to us (i.e. they must like us enough), 4th is friend's attitude to target (i.e. they must dislike them enough).

Spoiler Code... :

DenialTypes CvTeamAI::AI_declareWarTrade(TeamTypes eWarTeam, TeamTypes eTeam)
{
    PROFILE_FUNC();

    AttitudeTypes eAttitude;
    AttitudeTypes eAttitudeThem;
    bool bLandTarget;
    int iI;

    FAssertMsg(eTeam != getID(), "shouldn't call this function on ourselves");
    FAssertMsg(eWarTeam != getID(), "shouldn't call this function on ourselves");
    FAssertMsg(GET_TEAM(eWarTeam).isAlive(), "GET_TEAM(eWarTeam).isAlive is expected to be true");
    FAssertMsg(!isAtWar(eWarTeam), "should be at peace with eWarTeam");

    if (isHuman())
    {
        return NO_DENIAL;
    }

    if (getAnyWarPlanCount(true) > 0)
    {
        return DENIAL_TOO_MANY_WARS;
    }

    bLandTarget = AI_isLandTarget(eWarTeam);

    if ((GET_TEAM(eWarTeam).getDefensivePower() / ((bLandTarget) ? 2 : 1)) >
          (getPower() + ((atWar(eWarTeam, eTeam)) ? GET_TEAM(eTeam).getPower() : 0)))
    {
        if (bLandTarget)
        {
            return DENIAL_POWER_THEM;
        }
        else
        {
            return DENIAL_NO_GAIN;
        }
    }

    eAttitude = AI_getAttitude(eTeam);

    for (iI = 0; iI < MAX_PLAYERS; iI++)
    {
        if (GET_PLAYER((PlayerTypes)iI).isAlive())
        {
            if (GET_PLAYER((PlayerTypes)iI).getTeam() == getID())
            {
               if (eAttitude <= GC.getLeaderHeadInfo(GET_PLAYER((PlayerTypes)iI).getPersonalityType()).getDeclareWarRefuseAttitudeThreshold())
                {
                    return DENIAL_ATTITUDE;
                }
            }
        }
    }

    eAttitudeThem = AI_getAttitude(eWarTeam);

    for (iI = 0; iI < MAX_PLAYERS; iI++)
    {
        if (GET_PLAYER((PlayerTypes)iI).isAlive())
        {
            if (GET_PLAYER((PlayerTypes)iI).getTeam() == getID())
            {
               if (eAttitudeThem > GC.getLeaderHeadInfo(GET_PLAYER((PlayerTypes)iI).getPersonalityType()).getDeclareWarThemRefuseAttitudeThreshold())
                {
                    return DENIAL_ATTITUDE_THEM;
                }
            }
        }
    }

    return NO_DENIAL;
}

 

[Kaitzilla]

Exactly this.

"We're afraid of their military might" -> failed military check against a land target (requires 8 shared border tiles)

"We would have nothing to gain" -> failed military check against a non land target

edit: Found a post with both the actual code and its explanation for each of the four checks that are done in order to determine the AI response:

This is really useful, thanks

 

[Qactus]

Dear diary,
today has been a special day. I posted something that @Kaitzilla found useful. The previous 5325 times it has been the other way around.

 

[Henrik75]

I have read about this strategy a lot, but never used it myself. I am currently playing Monarch, #1 in score, but with another civ close 2nd and Monte close 3rd and a clear leader in military. I am doing OK, but need some time to tech more and get military advantage over Monte.

Monte would currently DoW #2 civ (he is annoyed with them) for 3 techs, including construction. I decided giving Monte construction is not a good idea But two other techs he wants are non-military, so I wondered if it is worth haggling with him?

What are effects of bribing AI to DoW:

1) Does the victim civ knows and hates/DoWs me more?
2) Does AI often "fake it", i.e. take techs/money but makes peace after 10t without any fighting? (my specific example is Monty, he normally does not makes peace easily, I wondered if it is different if he is bribed?)
3) Do I get some positive diplomatic bonuses with the civ I bribed? (I do not plan to join the war)

Would be grateful if somebody has more experience with this and can clarify.

Or is it very situational?

One more question: when AI accepts my bribe, am I safe for 10t for him DoWing me? Can I DoW him (not that it makes much sense, just out of curiosity how this mechanics works)

Stirring up trouble between the AI is totally worth it if you can. You can declare on someone behind another AI thats next to you (think of it like this: 1 | 2 | 3 - where you are 3 and declare on 1, if you bribe 2 in on 1 also, they can't get to you and you're perfectly safe while they are constantly fighting and slowing themselves down in tech). This is extremely useful in high difficulties to slow down techers. I had a recent game with 5 good techer AI on deity; willem, zara, rameses, de gaulle and wang kong, but kept them all tied up in wars pretty much and the tech pace was extremely slow allowing me to peacefully get myself to lib without getting involved in wars.

In other situations it's also extremely crucial if you're getting plotted on or think you might be a target, to bribe someone in on that person, then they will stop plotting on you and have to defend/attack that person that just declared on them. Zara demanded a heap of gold off me and i refused but he instantly declared on me despite he was already at war with 2 people. Knowing that he was already busy, I didn't have to defend much at all except whip a longbow on my two frontline cities then continued teching and building as he's not going to march a massive stack over to you until he makes peace with the other 2 people.

As to answer the questions altho im pretty late, there's no 10 turn treaty after bribing people in but they wont continue to plot on you if they suddenly get declared on or declare on someone else, and curiously enough, when you bribe AI to make peace they don't have a 10-turn peace treaty as if they made peace themselves naturally where they would have 10 turn treaty.

 

[krikav]

curiously enough, when you bribe AI to make peace they don't have a 10-turn peace treaty as if they made peace themselves naturally where they would have 10 turn treaty.

I'm starting to question that they even have a 10turn treaty even when they make peace themselves.
Have been seeing AIs start plotting on the target they have made peace with two times the last month or so, when I went into preparation mode because I was sure that they had a peace treaty, but then they suprise me abit later by simply declaring on their target again.

Maybe it was just ceasefires? Do the AIs do that?