Accepting capitulations or not when playing for domination victory

[Todelotti]

When a civ I am at war with offers capitulation is it a good decision to accept when aiming for domination victory?

I'm currently playing a game where I have captured 3 cities from De Gaulle. He offers capitulation now but still has 5 cities (including Paris). The game started with 6 opponents. One is dead, De Gaulle is the second war target and I own only 20% of the land mass right now but I need 64%. (2nd war target has already 8 cities? Yeah, I'm going to lose this game! ) It feels too early to me to accept a vassal at this stage and I'm inclined to continue the war but I'm not sure.

I see some Pro and Contra for accepting a capitulation:

Pro:
- Saves time and units and next target can be attacked earlier
- Vassal contributes to next war - with his remaining units (that I otherwise would destroy) and with new units
- With capitulation it's often possible to get some gold and a tech or two on top

Contra:
- Less tiles count towards domination limit compared to capturing all cities
- Unhappiness in already captured cities of that civ
- Less cities for my own unit production or economy
- It's easier to win cities (and land tiles for domination) against an already weakened opponent

Are there some rules of thumb you apply to decide whether you accept a capitulation or not in a domination game?

 

[Lexicus]

A few addendums to the Pro side. Have to be careful, because the AI will only contribute units if they have a certain surplus. I forget the exact numbers but the AI will always divert a certain number of units to city garrisoning, anything left over above and beyond that will go into their offensive stack (this is not an exact but rough description of the AI behavior AFAIK - better players will correct me as necessary).

You shouldn't expect much help from AIs whom you've capitulated, particularly if they've capitulated very recently.

Now, there is not really a "correct" answer to your question here. "Rules of thumb" is more like it - you want to ask yourself if you think you have enough land of your own, and if the answer to this question is "no" you should keep taking cities, if the answer is "yes" then capitulate away!

As you point out in the Contra section you get unhappiness from "We want to join the motherland" (though you get happiness in cities of your own culture from "we influence other civs!"). You also get IIRC -1 relations with the AIs for each vassal you have (probably not something you particularly care about if going for conq/dom).

Optimal play for domination is definitely to capture everything and not capitulate the AI as not only do you add more % to your dom totals you add more cities which can be whipped for more units, which speeds up your victory date.

But ultimately it comes down to your preference. I don't play for the fastest victory dates, I enjoy the roleplaying aspect of having puppet states, so I like to capitulate as soon as I'm comfortable with the amount of land I have.

I would also look at how much of a hassle you expect from taking DG's remaining cities. This depends on a variety of things like what tech he's researching, terrain, how far away they are and so forth (if he was 1 turn away from longbows with a bunch of hill cities I'd definitely capitulate him, for example).

 

[InovA]

I would say it depends on when you are planning to win the game. If you don't foresee yourself winning or being in a certain winning position in the next ~100 turns, I wouldn't accept! Otherwise, do.

The reason being that accepting a vassal has more effective short-term benefits but overtime the long-term disadvantages can really take a toll on you.

Short-term benefits:
- Less army wasted. (On both sides)
- Immediate production of cities. (vs. you capturing and waiting ~8 turns of anarchy, and having destroyed a few buildings.)
- Some techs and .
- Steal one resource from vassal. (long-term benefit, too.)
- Increased in your cities. (long-term benefit, too.)

Long-term disadvantages:
- (-1) diplomacy with all other civs.
- Increased # of cities maintenance cost without generating or from the cities.
- City revolts of captured cities happen pretty often.

So let's say in 150 turns you had ~6 French cities to pay extra maintenance on,

Pc is the population of an individual city.
If France goes up to 9 cities, the 0.855 multiplier becomes 1.283, scaling the cost by 50%.
Every city you own will cost 1-1.8 more per turn, so 12 cities at an average of 12 pop can cost 21 more than usual per turn, becoming somewhere in the ballpark of 3000 in the 150 turns.
And as you conquer more, gets higher, bla, bla.

*These numbers are with Immortal difficulty modifer (0.95) and standard size map modifier (0.3).

So if you plan to stop teching in the near future, go ahead and claim a vassal! Otherwise kill the French!

 

[Gumbolt]

It's all a question of percentages. A vassel's land is only worth 50%. So if they have 5% overall land, it will only adds 2.5% to your land % for domination. More vassels mean you effectively have to vassel most of the AI.

Of course if you plan to quickly capitulate 5-6 Ai then it is another matter. Also how close you are to this. If I have 30-40 cuirs and it's clear the Ai is going to roll over quickly then I may go for the quick capitulation route. Especially if you are fighting archers and LB.

Of course if an Ai has a vassel then maybe it's quicker to roll over several Ai cities to reach the domination limit. The other downside to vassels is the game uses average powers (Including vassels.) to work out when an AI will capitulate. Often with the final 1-2 Ai requiring power ratios of 3-4 before they will cap.

That said on most games if you can get an Ai to friendly and have the land and power right the first Ai will normally peace vassel. On my last game I captured my entire continent first.

Play a game as you feel it. Of course this is all based on immortal level. For all I know you could be playing Noble with stacks of 9-10 units come 1500ad.

 

[AbsoluteZero]

More things to consider:

You're diplo relations with AI's is ( I believe ) an average of their relations with you and your vassals. If you have a vassal that they are friendly toward and they are cautious with you then I think it averages to them being pleased with you.

On the other hand if you have a vassal everyone hates then that will drag down your relations with each AI that hates your vassal. This can be a real problem.

/////////////////////

Regardless, in my games I almost always take vassals ASAP for a few reasons:

The quicker I conquer one AI the quicker I can start attacking the next / or start recovering from war. This is probably the #1 reason to vassal an AI.

If you do it right and you have enough time you can gift warriors / chariots to your vassal and they can upgrade those to higher level troops like rifles / cavalry at a fraction of the cost and become a huge factor or even the only factor in your offensive force. I've literally won games by gifting my vassals troops and letting them conquer my enemies for me. It's very situational ( and can even backfire because of AI stupidity ), but it has saved me in games where I would have lost otherwise.

 

[Todelotti]

...though you get happiness in cities of your own culture from "we influence other civs!"...

Oh, good to know! I never noticed that.

...Formula... ...So if you plan to stop teching in the near future, go ahead and claim a vassal! Otherwise kill the French!

Hmmm, that formula, interesting! Thanks for this detailed analysis! "nc" is the number of cities I own, is that right? But that would mean that a vassal city actually adds only half of the "Num city maintenance" compared to a city that I own (because of the nt=2 denominator). Keep in mind that I don't want to raze the captured cities but keep them (domination, not conquest is the goal). So 6 captured cities means: (nc + nvc/nt) increases by 6. And 6 vassal cities means: (nc + nvc/nt) increases by only 3. Also if I own those 6 cities they add to my total empire maintenance while the 6 vassal cities don't (only "indirectly" by increasing the "Num city maintenance" in my other cities, but that's also the case - even twice as much - when I own the cities).

Maybe I misunderstand the formula but it seems to me that the total empire maintenance for a new city that I own increases significantly more than for a new vassal city. Which would make sense somehow because a city that I own also contributes to income while a vassal city does not.

So, the question is basically if the new captured city can generate enough income (trade routes, working commerce tiles, producing wealth, etc.) to compensate the higher maintenance compared to a vassal city. Perhaps some cities can (a well-developed Paris probably), others can not (like a small stagnating Tundra city or so). Anyway, for domination I want to control land tiles and have to keep them all and live with a mix of wealthy and poor cities, I guess.

But your final conclusion means that (at least on average) despite the higher maintenance a captured city will have a positive effect on the economy compared to a vassal city, right? Hm, important point, as I probably cannot afford to stop teching completely (it's Immortal and some civs are running techwise...)!

 

[Todelotti]

The quicker I conquer one AI the quicker I can start attacking the next / or start recovering from war. This is probably the #1 reason to vassal an AI.

Yes, that's also my main reason why I consider to capitulate De Gaulle. Plus, I seem to have a window of opportunity right now: Roosevelt, my next target, just went to war as an ally of Ghandi who got war decced by Peter, and he moved a big amount of his forces away from my border towards Peter's borders. It might be the right moment to attack him.

If you do it right and you have enough time you can gift warriors / chariots to your vassal and they can upgrade those to higher level troops like rifles / cavalry at a fraction of the cost and become a huge factor or even the only factor in your offensive force. I've literally won games by gifting my vassals troops and letting them conquer my enemies for me. It's very situational ( and can even backfire because of AI stupidity ), but it has saved me in games where I would have lost otherwise.

Wow, what a cool strategy! Doesn't exactly work with De Gaulle at the moment (he would downgrade my units probably being so backwards in tech as he is ) but with another civ or later in the game that sounds really interesting to try out.

 

[civfanchambers]

Another point to consider is the after war usefulness of said capitulated Ai. How many cities will you be leaving your vassal with? 2-4 cities? Better to just conquer them completely and make them your cities. Now if you leave the capped vassal with 5-7 or more cities they can be a production powerhouse in terms of both building units, and effectively doing research. A weak vassal does you no good. A strong vassal can help you. On a side note---- good to see you around AZ, love your Civ4 LP's. I've learned alot from watching your vids.

 

[InovA]

Thanks for this detailed analysis! "nc" is the number of cities I own, is that right? But that would mean that a vassal city actually adds only half of the "Num city maintenance" compared to a city that I own (because of the nt=2 denominator). Keep in mind that I don't want to raze the captured cities but keep them (domination, not conquest is the goal). So 6 captured cities means: (nc + nvc/nt) increases by 6. And 6 vassal cities means: (nc + nvc/nt) increases by only 3.


So, the question is basically if the new captured city can generate enough income (trade routes, working commerce tiles, producing wealth, etc.) to compensate the higher maintenance compared to a vassal city. Perhaps some cities can (a well-developed Paris probably), others can not (like a small stagnating Tundra city or so).

You understood the formula right! And yes, nc is the number of total cities, and for increased number of vassals the "nt" denominator increases, which can throw a wrench in the point I was trying to make, so one could argue that multiple vassals > no vassals > 1 vassal if you don't care about the diplomacy issues. But that's another story.

One thing though, a captured city does not need to generate enough to compensate for ΔCost, the , , and drafting are what you're really after for future wars.

I don't know how many cities you currently own, but I'll guess around 12? More?
For 12 the Num City maintenance would increase by the aforementioned 1-1.8 per turn based on its own population and the number of vassal cities. So in the worst case scenario, France quickly gets to 9 cities and your 12 cities at 12-pop cost 21 more. We are saying it will be >150 turns to win, so you're at ~3000 deficit for vassal cities.

When we say you instead gain 5 cities and kill the French, 12 cities at 12-pop and 5 cities at an average 8-pop (we're assuming their tiles aren't as good and they are delayed for the early revolts, etc.) so if we plug in the numbers for the 12 cities now and after gaining 5 new cities, their maintenance increases:

The 5 cities at 8-pop increase costs:

so their deficit goes up to 51.53 per turn overall.

at 150 turns, 7729.5 lost, more than twice as expensive as the vassal cities!
But it gets worse! That's just the Num City Maintenance, the Distance also plays a factor. No clue how far the cities are, but I'll make a guess that it will be 3.33 per turn for each city.
Add 2497.5 to the expense column.

However, let's say each of these cities having an average of 8 population would likely be working 9 tiles (including city) to produce, say: 16-18 ~14 ~8. We'll limit most of the cities to just a Granary and Library, so no bonuses other than 25%. This still yields 22 per turn when building "Wealth" and 0%, at 150 turns that's 3300 per city. 5 Captured French cities look a lot better for a long-term goal when you see an income of 16,500 with only 10227 expense. Should these cities get to a higher population and greater yield, they'll generate a much higher gpt and surely be worth it, even without drafts/whips you'll use them for when the real wars begin!

Also noteworthy:
The more cities you have, the larger your empire population. This influences your unit cost in your favor:

So an increase in 40 population will save you 15 per turn in unit costs if they would have otherwise exceeded the limit! Another 2250 in your pocket

TL;DR:
Worst case, a vassal can cost you -21 per turn, capturing can yield you +56.82 per turn after revolts, etc., and give you the option to whip/draft.

 

[DrCron]

One more factor that is hasn't been mentioned here yet is that you can tell vassals what to research and where to attack. This can also be useful:

1) Telling them what to research means you can b-line and then trade with them for the techs you left behind but were at least a bit useful to you. This is especially useful if the vassal has at least 5 cities, on a high difficulty level (hence they can tech fast enough to get you some good techs).

2) Telling them where to attack can be a useful distraction when you go for your next victim, because they can't keep all of their units waiting only for your advance. This is especially useful if your vassal has a border with your next victim that is not in the same side in which you will attack.

 

[yena]

Having a vassal between you and your next target can be better than taking cities bordering to the target yourself. The target will focus on your vassal initially since his cities are closer, and you can sweep down and clean up once their forces are engaged. When this is the case I prefer to vassal.

 

[Lexicus]

Yep, having buffer states is cool. The "land target" rule also makes for some weird outcomes sometimes. I'm in the middle of a game right now where I capitulated 3 AIs, then declared war on Victoria. Her 3 main cities including the capital were on the other side of Mansa Musa's land (we've been buddies since the BCs and he peacevassaled to me a while ago), so I captured those three cities. Her empire was reduced to 2 island cities and 2 barbarian cities on the opposite side of the pangaea, and she wouldn't capitulate!

But then, she sent a small stack of infantry and captured one of my vassal's cities, thus giving her a border with me. I went into the diplomacy window and capitulation was suddenly open.

So basically, she capitulated to me immediately after capturing a city. Kind of funny if you think about it.

 

[gaash2]

For me (before I started playing with vassal states off) I think it depends how much more powerful you are. If I can just mow them down in a few more turns I just mow them down. At the levels I play (prince/monarch) I am usually way more tech advanced that tech trading and what to research has no value to me and it's annoying if you do decide you want to kill them at some point to have vassalized them.

 

[Seraiel]

I usually don't create rules of the thumb, but I'd simply say conquer the 1st two civs completely and then vassalize everybody.