1: Additional Grivalance against the Civ you pillaged.
2: Additional Grivalance against all Civs, as they witness your cruel behavior. (after any player researched Mass Media)
I can see this turning badly for the following reason: the modifier that almost every AI seem to get which is "grievances you have caused others". Pillage the hell out of one AI player (and accumulate grievances) without taking cities and I can guarantee you all AIs will hate your gut even before reaching Mass Media.
You can also question the whole rational for other AIs/players hating you for pillaging others. As an example, contemporary people care very little about ongoing genocides currently taking place in Africa, let alone pillage wars. It seems people care very little for "tragedies" which do not concern/impact them directly.
3: Additional loyalty towards the pillaged city when owned by your enemy, Loyalty penalty if you capture that city. As the citizens remember the harm of the pillage.
There is already a loyalty penalty linked to grievances against the founder civilization when you capture a city (until wiped out). If you add a loyalty penalty for pillaged tiles/districts and add grievances for pillage, you effectively double count the impact of pillage. You can do one or the other, but both seem like overkill.
I however think at the very least pillaging should impact the target with war weariness/unhappiness.
4: Reduced yield when conducting further pillages in that city (or Civ?), as people there start to learn to protect their wealth from pillage.
You can argue that it's the job of the player/AI to make it more difficult for others to pillage their land by building higher tier walls, forts, encampments, fortifying units, etc. and learn from past attacks. By doing so, the rewards from pillage stay the same (high), but the associated costs related to pillage increase and the probability of success decrease making it less worthwhile to do so.
However, since the AI is currently incapable of doing any of this, this could be an avenue worth exploring...
***
The Main Problem of Pillaging
My main gripe against pillage is that pillage yields are a function of your own tech/civic progression and not of the target's. This is the main reason why fast science victories currently involve pillaging: your own tech/culture progression fuels your pillage yields. There thus exists a positive feedback loop where you can use your current science edge to increase it even further. This goes against the "spirit" of pillaging which naturally lends itself as an "equalizer" (or rubberband mechanic if you will): you steal what others have already researched. Pillage yield should ideally not help you research anything new/undiscovered. (Also, the fact that AIs accept city "gifts" which they have absolutely no chance of holding through loyalty is a related, but separate issue that contributes to the same problem.)
As an example, what could current preeminent countries ("tech leaders") possibly learn from pillaging a Library/University/Research Lab located in a developing country? Certainly not much... However, under the current civ6 rules, because they are the tech leaders with a lot techs (or civics) discovered, they would find an incredible amount of science by pillaging a campus in this developing country...
A simple solution would be to use the target's tech/civic progression to calculate the pillage yields, but an even better solution would use "the number of techs unlocked by the target and not discovered by the raider" as the base factor entering in the pillage science yield (similarly to Peter's UA). You can do the same for civics and culture pillage yields. Thus, pillaging the amphitheaters of the culture leader would provide incredible culture yields, but pillaging the amphitheaters of a civilization lagging culturally would not provide much. This factor could also enter in the calculation of AIs' decision of who to attack (ie potential rewards). (An alternate, but more complicate solution to implement would be to get a fixed percentage, like 10%, of science for all techs you have not researched yet, but the target has when you pillage.)
For improvements and districts that provide faith or gold, the total faith/gold output per turn of the target civilization could be used instead as the main factor. One should salivate at the idea of pillaging Mansa Musa's mines and Sugubas, but not so much for the mines and CHs of an AI with a weak gold output.
***
Pillage Wars vs Conquest Wars
This topic is infinitely more complex as it mainly relates to how loyalty is calculated which is a completely different can of worms in itself. I will however attempt to discuss this subject without starting a debate on loyalty.
- Since an empire's yields are currently mainly a function of the number of cities the said empire owns, there is a very strong incentive to own as many cities as possible. Currently, the only reason you should refrain of conquering a city when you have the means (army) to do so is if you believe you cannot hold it due to foreseen loyalty problems. The current rules governing loyalty however make this unlikely to happen.
- Since pillage yields are high and the downside to pillage are low (movement points, repair district/building time), when you are in a position where you are able to pillage there is very little reason not to do so.
- To conduct either a pillage or conquest war, you need in most cases a similar investment in troops. Add 1 battering ram/siege tower to your melee army and you can take down a walled city. Add 1/2 siege unit(s) to your light cavalry army and you can do the same. The difference in the amount of troops needed to conduct a pillage only versus conquest war is thus very little. The size your army needs to be at is mainly a function of the opposition it has to face.
It thus follows that most wars end up involving pillage followed by conquest.
If a strong loyalty penalty linked to pillaged tiles/districts was to be added to conquered cities, the incentives would only shift to conquest (with little or no pillage).
Currently, there are only a few cases where doing a pillage only war make sense
- Opportunistic war: When you have an inferior army to your opponent, but this army is occupied dealing with barbarians or another enemy at the other side of their empire, you can use this small window of opportunity to pillage their unprotected lands as your army is not strong enough to deal directly with your opponent's or the defensive capabilities of his cities.
- Foreseen loyalty problems: You are in an age in a tier below your opponent (ie dark vs normal or dark vs normal/golden or normal vs golden) and your army is not powerful enough to conquer cities quickly enough to create stable population pressure. Since loyalty is a function of population pressure and the type of age you are in and that it is currently easy to chain golden ages, these situations rarely happen. If you however find yourself in such a situation, you can mitigate this problem by first focusing on wiping your opponent's armies and secondly focusing on the more populous cities of your opponent and the cities closest to your own.
Analysis of proposed changes
- Changing the mechanics related to loyalty could create situations where holding cities with low loyalty would provide low return on investment to their conqueror, making pillage only wars more valuable.
- Adding a loyalty penalty to conquered cities in which you pillaged would nerf the currently most powerful approach which is to pillage and then conquer a city, thus adding a risk to the most rewarding approach. It would create a separation between a pillage only versus a conquer only approach.
- Changing the pillage values to my previously outlined changes would create interesting decisions about who you decide to target, when you decide to target them and what improvements/districts you should choose to pillage.