TW:S2 Diplomacy 201

[Maltz]

Note: This article is written for people who already have certain exposures to the game. It assumes that the reader already knows which diplomatic options are available and generally what they do.

***

Many people distinguish strategy and tactics this way: strategy is what happens outside the battlefield; tactics is what happens on the battlefield. With good strategy, we can simply auto-resolve all battles in the game and win. (But it is not so fun.) I do not think the opposite can be achieved due to the infamous Realm Divide.

Diplomacy lies in the core of strategy - right beside money making. In TW:S2, Diplomacy is a decision making process that takes place every turn, or every few turns if there is not too much going on. We go into the diplomacy view, and look at what is happening in Japan, and make appropriate deals.

The highest Bushido tech is called "Shi". "Shi" is the way how the things evolve and will become. This, and the pro-action in response to our perception of "Shi", is basically what diplomacy is. We make the best guess of what will happen, and make appropriate beneficial deals. We will plot a smooth route of expansion with good diplomacy - never lose a battle, never have to wait, never run into a wall, just keep expanding until the point before Realm Divide. And after some preparations, the unstoppable expansion continues post-RD.

So how to we plot a smooth route of expansion?

Chapter 1. Reduce enemies

All clans attempt to expand by attacking their neighbors. Every clan that borders us is a potential enemy. Some of them we might want to attack in the near future, some we want to leave until later. While you would welcome the clan that you wish to attack to actively declare war on you so you do not have to worry about their allies, you SHOULD expect those clans that you do not want to war with to attack you, sometimes in the most inconvenient times. According to Murphy's Law, they will! - if you do not prevent it diplomatically.

Here are the three questions that we can ask ourselves every time we look at a neighbor clan.

- Does the neighbor have another enemy (actively at war)?

If the answer is yes, take a look at the personality of this neighbor. Are they trust-worthy? Some clans tend to declare war on all of their neighbors no matter what. A good example is Ikko-Ikki. Luckliy they are rare. (I can only think of Ikko-Ikki and Otomo.) Most clans do not declare war to all their neighbors.

If the answer is no, there is a good chance that this clan is going to attack you. On higher difficulties, if an AI has multiple neutral neighbors, you should expect the AI picks you to fight. You are human! The AI has an obligation to prioritize on you on higher difficulties.

On the most extreme case, if an AI runs out of all other neighbors - all they border is you - very likely they will attack you even if you two are the best buddies in Japan. This is a scripted reluctant betrayal. In very rare cases, though, your buddy will launch an oversea invasion to somewhere else as it does not want to backstab you. I have seen Imagawa and Ashina doing that. The recipient of this naval invasion will probably have a "What the..." bubble on top of their head, though.

- Is the neighbor going to get killed by another clan?

Going back to the first question, does this new neighbor have another enemy?

- Does the neighbor have a spare army?

Send an agent there and take a look. If there is a stack standing right beside their castle - it is getting ready for an attack! If that castle is undefended - you are safe for now.

Now we know who might be an enemy soon. How do we delay the their attack and thus reduce the number of enemies? There are a few ways:

- Gifts

Each 1000 koku converts to +15 relationship, for a maximum of +100 achieved at 6633 koku paid at once. More payment does not give better result. Also, the more frequent the gifts, the less effective it is. Personally I do not recommend monetary gifts. The best gift we can give to an AI is a promise that we will not crush them in the near future. Which brings us to the following...

- Marriage

+100 and decays to +20. Marriage is a soft way to improve diplomatic relationship. There is no strings attached. Attacking your in-laws and force your daughter in-law to turn against their parents is so common in Sengoku Jidai. (And the rebellion of the heart-broken son is therefore also very common.)

Some clans have daughters to marry in early years of the campaign (such as Ashina, Shoni, Chosokabe, Date, Takeda, Ito), making them very easy to befriend. You can marry their daughter even if your clan does not have a bechalor in the immediate family - probably some distant second or third cousin? If you have a daughter, it is better to save her as an emergency friend maker. There seems to be no limit of marriage age.

Note that all enemies of your in-laws also dislikes you. So if there are two clans that are at war, and one is probably going to win, do not marry the losing clan.

- Hostage

It is good to have lots of sons, so you do not regret if you lose one. You can attach a Hostage in trade and alliance negotiations and obtain a 8-turn +100 relationship with the target clan. However, this is probably a very bad way to maintain peace with a neighbor, as the risk of losing a son is probably too big. Now I am a father myself, I just can't do this to my virtual son. :)

- Military Alliance

+100 subject to decay to +40. Military Alliance is the tightest bond the game offers between two clans, but it also comes with the worst side effects. In order to keep your daimyo's honor, you will have to join your allies' side if they are being attacked. So you will probably end up making more enemies. However, if your powerful ally can handle one border for you very nicely, and your enemies are not going to launch naval attacks on you (you are human!), why not? It is how Oda and Tokugawa emerged as victors of the Sengoku Jidai.

- Trade

+15 and slowly increases to +60. Trade acts too slowly if there is an emergency, but if an attack is not imminent it will eventually become significant. However, after the FOTS patch, whether we can trade with a clan is not depended on whether we want to, but whether they want to. So it is not really a strategy now, but more of a luck thing.

- Selling Military Access

Each turn of military access can be sold for about 200 koku. More if the AI likes you, less if not. The AI always love to buy your military access if it has a land-route access to your territory.

Selling military access does not make a clan like you more. In fact, you are giving them the right to march an army all the way to your castle and if they wish to betray your trust, they can. (Fortunately, I have not seen AI clans doing this. If they want to declare war on you while they have your military access, they declare war while their army is on the border.)

Selling military access helps delay an attack by making the AI poorer. Even Legendary AI does not have infinite cash. If they do not have money, they can't hire soldiers. In very early game, the AI operates on very limited budget. So you can delay the AI's readiness of launching an invasion by quite a few turns by stripping 4000 koku away from them to buy 20 turns of your military access. (Try sell them 5 turns, then next turn 10 turns, then next turn 20 turns.) This could make a huge difference where you successfully destroy your first neighbor without getting back-stabbed by another.

But selling military access definitely has its drawback - if you attack this clan within certain turns (it is now displayed in game), you suffer a great hit of diplomatic relationship against all other clans. So it is not a good idea to sell military access to some targets that you want to attack immediately next. Fortunately, this financially weakened neighbor might become a victim of some of their own neighbors, so you might be able to simply take on the new neighbor instead.

That's the end of chapter 1! We will continue with chapter 2 (Shi - the prediction of what is going to happen) at a later time...

[Harleyquin12]

I'm not sure if you read the guide stickied at the top, but there is no risk of your castles being taken by a former client of military access when war is declared. All armies enjoying military access when revoked are bumped back to the nearest viable border. In addition, the traitor clan gets hit with a diplomatic penalty across the board. Everything else in the military access column is more or less correct.

[Maltz]

Actually I think that guide is incorrect (for the current version of the game) - I have been declared war by AI clans who have my military access, and their army is not bumped to the border.

[Harleyquin12]

That would qualify as a bug or intentional feature change (the programmers probably realised military access selling was exploited too much at higher difficulties).

I take it you don't have mods or the like installed when writing this guide?

[Maltz]

No mods.

[Maltz]

Chapter 2 - Predict what will happen

Now we know how to reduce the number of enemies, but that only puts our strategic depth at the current turn. A good player of chess or go (East Asian traditional chess, "go" is its pronunciation in Japanese) thinks many steps ahead. A good diplomat also plan for the future.

The best way to make an educated guess of what will happen is gaming experience. By observing the AI's behavior, you will have a good idea of what they are thinking, and know what they will be doing. Here are the AI patterns I have observed (on Legendary):

(1) AI is 100% offense and 0% defense

The AI simply has nothing defending their home. The no.1 cause of death for AI in STW2 is NOT losing a great battle and have all their soldiers killed, but to have their castle stolen by another lucky clan when their army is away from home, and their soldiers vaporized out of their border.

A prime example of this is the Shimazu clan, the easiest clan for human to play because of their close proximity to the richest trade nodes and secured location. But Shimazu clan is very easily destroyed by the Ito clan when the AI controls it. Why? If you ever watch Shimazu's actions, you can see that they always prefer to attack Sagara's Higo. When the powerful full-stack Shimazu army is away attacking Higo, Ito, their starting enemy, is bound to steal their home Satzuma with a crappy force. So Shimazu loses to Ito not because they lost a battle, but because Shimazu failed to take care of Ito's very close base in Osumi before marching onto Sagara's Higo.

The same is true for Otomo. You probably seen many times that Otomo getting destroyed by Shoni. This is not because Otomo lost a war. That's because they prefer to attack Sagara or Ito. Likewise, Shoni loves to march onto Chugoku before taking care of their neighbor in southern Kyushu. The result? They lost everything. Very often you see Ito emerge as the winner of Kyushu. This is not because Ito is a great clan, but because their enemies all march their army away and leave an empty home for Ito to steal. I have seen so many times that Ito steals Higo because Sagara attacks Shoni, and then steals Buzen because Shoni attacks Chugoku... Ito will move on to take Chugoku following the exact same pattern, because Chugoku daimyos are stupid enough to leave their end undefended, too.

So we can generalize the Ito phenomenom this way: AI that is located in a corner, or a virtual corner backed up by allies (such as Hojo), has an huge advantage over those who have many neighbors, simply because their single-minded aggression works extremely well when they do not have somebody preying on their defenceless back.

Suppose you have three clans who seem to have equal power, the clan that borders everybody is almost always bound to be ganged and killed, and the clan that is backed by sea is almost always bound to emerge victorious. Of course, there could be other factors covered below that may complicate things a bit.

(2) If all others equal, AI's who acts first wins

Besides the load game option, we human have a really unfair advantage against the AIs... We act first! Because we act first, we can steal AI's castle, sabotage their army, assassinate their daimyo, kill their freshly recruited agents... before they act once!

When two AIs are at war and have their army faces each other in an apparent stalemate (not one clan going after another clan's defenceless end), you can see whoever acts first wins. This is because the AI who acts first gets to decide which castle to steal, and the AI who acts later are forced to passively adapt to what happens, only to find that their reaction half a turn too late.

In one game, I saw Takeda Invading Hattori. Both have about 3-4 full stacks but they avoided fighting that highly anticipated huge battle. Takeda instead decided to steal Hattori's Ise castle. Hattori, who acted after Takeda, reacted by following Takeda. The next turn, Takeda successfully stole the defenceless castle, and Hattori clan decided NOT to take that castle back, but back tracked! So basically Takeda made two meaningful actions, but Hattori wasted two turns. Takeda ended up taking out Hattori without ever facing Hattori's 3-4 full stacks literally right beside them.

Oda acts before Ikko-Ikki - no wonder they always beat Ikko-Ikki! Takeda and Hojo both acts very early - they are just bound to succeed even if they do not meet another clan's defenseless end. The only exception for these early actors to lose is in situation (1), where they show their defenceless end to an enemy. They will happily do it if they are given a chance to.

(3) AI's daimyo and generals

We already covered that AI's 100% about offence. So if you know where the AI's army is, we can often predict what is going to happen. For example, Oda now has 5 provinces. They can either expand east or west. So if your ninja reveals that Oda has 2 full stacks in the east and no army in the west... where are they going to expand? What's going to happen if they get attacked from the west? It is very easy to predict.

AI's generals are the soul of their army even outside the battlefield. For example, if an AI general is leading a full stack towards you and you assassinate the general, you can often see that the entire stack suddely becomes immobile. And you might be able to see that the AI is sending another general from far away to lead this army.

AI's daimyo's loation is a particularly good indication of where this clan is going to attack. If the daimyo is coming your way - that's really bad news! Once I was able to change Oda's mind by wounding their daimyo - and the stack he used to lead actually walked all the way back to Owari and ended up attacking the opposite side.

(4) AI's trade partners

Just like human-AI relationships, a good trade partnership may upgrade into an alliance. Often you demand an AI to break alliance with another clan, only to see that they re-signed the alliance next turn - because they are still trading! But if you also manage to break their trade, you can see that they soon turn into enemies. Looking at an AI clan's trade partners can be a good indication of who they do not want to attack.

After the March patch, trade partnerships are particularly hard to establish. This makes trading an even better indication of whether two AIs might become allies or declare war in the future. But of course, the friendship from trading can be completely ignored if one of the clans have nowhere else to expand.

(5) Hostile agent actions & revolt incited

AIs are very sensitive to the precense of other clan's agents in their territory. Often they will hurry to hire their own agents and go after the detected forieng agents. If the AI succeeds, no bad feelings (at least for them). If they don't, the bilateral diplomatic relationship takes a huge hit from "hostile agent actions". This is particiularly important if our own agents are drawing attentions of an AI clan. The extra hate may break a trade, which escalte into a war later. A similar but slightly smaller effect is induced by "Revolt Incited". And even if you succeed to incite a revolt, there could be a -40 modifier.

(6) Naval invasions

When an AI feels that it has no land target to attack, it will launch a naval invasion. Very often you can see them as your trade route also lights up the fog of war. If it is coming your way, chances are it IS going to attack you. You are human, anyway. So get ready for it!

Now we know who is going to be become an enemy for not only the immediate future, but also maybe in the far away future. The situation changes all the time, so your predictions should be updated very frequently. In the next chapter, we will discuss how to respond when a imminent war probably cannot be avoided.

[Maltz]

Chapter 3 - Last minute effort

So it will eventually come down to this - an AI clan is highly likely to attack you - they have an army coming at you. There are still things you can do besides readying an army for it.

(1) Grab allies

Allies have the moral responsibility to join a defensive war. So if you think you can use some help and have some extra money in the bank, go out and make alliance that can make a difference. When your military power is great, it is easy to pursuade clans to ally with you. Or you can try to include payment in the deal.

You do not have to ally with a neutral clan before war is declared on you. This is because your ally might actually decide to abandon you in a war - this can happen quite frequently. It might be better to bribe others to become your ally and request them to join your war. This way you guarantee to have clans on your side.

However, we should not expect much as most AI clans are quite useless allies. They often have their own agenda. Only after they kill their original target would they consider helping you.

(2) Convert the enemy's ally

Things could turn really ugly if you are declared war by two clans at the same time. While allies have no obligation to join an invasion, the target of attack is unable to ally with the invader's allies, thus opening up the risk of being attacked by two or more allies in the future.

In the best scenario, just before the war breaks out, if clan A and B are neighbors and allies, you can make clan B your ally while asking them to BREAK alliance with clan A. As soon as clan A declares war on you and you call for help, clan B has the honor responsibility to join your side, not clan A's side. Now suddenly clan A has two enemies! To make sure clan B does NOT re-sign alliance with clan A, you can even declare war on clan A yourself.

Sometimes, clan B will not agree to break alliance with clan A, and when clan A declares war on you, clan B will turn down your call for help and thus break your alliance. In this situation, it is obviously better NOT to call for help, so you can at least keep your alliance with clan B, and avoid being attacked by two clans at the same time.

(3) Ally with them!

In some cases, you can actually ask the imminent aggressor to ally with you. This works especially well for the AI clans with a nice guy personality. The AI will then immediately break the alliance, since it wants to invade you. But on the next turn you might be able to ally with them again. There is probably a loophole in the AI's logic that goes like:

"Destroy human -> must break alliance first"
"Human asks for alliance -> human is pretty powerful and friendly, why not."

The AI will never be able to get out of this loophole and get stuck in this forever. And you might never get attacked! This might be a bad thing too as the alliance also forces you not to attack them for a while.

(4) Selling infinite military access

The last chance you can make money from this clan without violence. This has the same side effect as (3), as you won't be able to attack them should they suddenly decide not to attack you.

Bonus: Not diplomatic but useful - create a rebel buffer zone

An effective way to block the AI from invading you is to create a rebel buffer zone. The AI army is very predictable; if it sees a rebel castle in their route of invasion of your land, it will try to take it first. So what you can do is to recite the AI's frontline province to be taken over by rebellion, and march up one general (just the general) to lay siege on the castle. When the end season comes and the rebels sally, just select retreat (or run around for 20 minutes for a heroic victory) and return to siege the next turn. The AI army will patiently wait beside your general, probably hoping that you will be defeated and it will be their turn to lay siege.

But the AI does not realize a fatal flaw in their logics - you act first, and the rebels act after them! So every time the AI acts, you will be the clan laying siege on the castle. They have absolutely no chance to take that castle even if they outnumber it 1000 to 1, and you are therefore aboslutely safe behind this buffer zone.

Maybe not, the AI sometimes launch a naval invasion to get you.