Hints and Undocumented features

You've probably noticed that your Shogun manual isn't a lot of help when it comes to uniting Feudal Japan under your rule. 
Of course, the real warlords of the 16th century didn't have a manual either, so maybe this is Electronic Arts' attempt at realism. 
If you're interested in a fully realistic experience there are a few additional steps you can follow: take your turn only once every three months, ask your boss to pay your salary in portions of rice, and speak in gruff guttural syllables.
Preferably Japanese ones. If, however, you're simply interested in enjoying Creative Assembly's excellent game to the fullest extent, read on.

The main attraction in Shogun is the full campaign game. The key to success here is picking your battles so you only fight the ones in which you have the advantage. Fully developed farmlands and swarms of Ninja are all good and well, but it all comes down to winning the battles you fight.

There is one exception. The shortcut around pounding an enemy clan into submission is killing its Daimyo while you're allied with that clan. This is the single most effective use of Shogun's somewhat anemic diplomacy system. Drag a Ninja with a high honor rating or a Geisha onto the enemy Daimyo (each clan has one army marker with a samurai on horseback to represent the location of its Daimyo). You will get a dialogue box asking if you'd like to assassinate the Daimyo, along with the percentage chance of success. 
If you succeed without breaking the alliance, and if the Daimyo has no mature heirs, you will knock him out of the game and inherit some of his empire. 
Note that you can drag a unit onto a Daimyo from any territory and it will track the target and move, turn by turn, until it reaches it.

Otherwise, alliances are only good to forestall attacks from another clan. There is no economic or diplomatic benefit from an alliance. Note that clans are more than happy to attack even when they're allied with you, so an alliance is no guarantee that a border will be safe. To propose an alliance, drag an Emissary onto the other clan's Daimyo to bring up the necessary dialogue box. 
The Emissary does not have to be adjacent to the Daimyo. You can do this from any territory, no matter how far away your Emissary is, and he'll make his way to the Daimyo. 
A Daimyo might be reluctant to ally if you have large armies adjacent to any of his provinces, in which case he'll say so when you offer the alliance. Simply step the army back, propose the alliance again, and he's likely to accept.

Honor among thieves

To build up a Ninja's honor, let him run around the enemy's lightly defended areas where he'll find armies led by rank zero generals. Practice assassinating this easy prey to build up enough honor to take on the higher ranked generals. Geishas, on the other hand, are brutally efficient right out of the Geisha House. Special units are also useful for reconnaissance while you're attacking enemy clans. Send them ahead of your army to get a sense of the strength and composition of any defenses. 
Note that the computer gets the added advantage of being able to react to your movements after you've ended your turn. 
What this means is that when you attack the AI, he can move units in from adjacent territories before the battle begins to help defend the province you're attacking you might have thought you were overrunning a single unit, but when the battle begins, you might find yourself suddenly outnumbered. Keep this in mind while you're snooping around with reconnaissance units and note how many armies are located in adjacent provinces. 

Emissaries and Priests are good for reconnaissance, since they can't be caught and killed by traditional defenses like Border Forts and Shinobis. 
A quick way to estimate the size of an army is to look at the vertical banner hanging from the army's standard. The bottom of the banner is colored. 
The longer the color, the more units there are in the army. 

It's not necessary to keep border territories garrisoned with large armies if you use castles wisely. 
Castles are less useful for any direct defensive benefit during a battle than they are for stalling an attacker until you bring reinforcements in. 
This is how you can effectively park your main army one or two territories away from a potential threat. 
Keep a castle in a border province and garrison it with a token defense. 
When the province is attacked, hole up in the castle until your main army arrives to break the siege and hopefully destroy the aggressors. 

Installing a garrison in a castle, however, is one of the many interface quirks not described in the manual. You'll want to keep units in the castle not just to avoid cluttering up the map board, but also to keep you from having to retreat from a battle if you're attacked and need to hole up and wait for your main army. 
To move into the castle, click on the army with the units you want to use in your garrison. Then drag individual units, one at a time, from the display at the bottom of the screen onto the castle on the map board. 
If you try to pull the army from the map board onto the castle, even if there's only one unit in the army, it won't work. 
Note that cavalry units count double when it comes to space requirements. 
For instance, a Castle can hold four units, but since a cavalry unit will take up two slots, this means it can either hold two cavalry units, or one cavalry and two infantry units.

Sailing, sailing...

Building a Port in a province makes it effectively border every other province with a Port, no matter who owns that Port. However, you must verify the presence of the target Port before you can move an army there. You can do this with a recon unit, like an Emissary or Shinobi. Simply try to move one of these units from one of your provinces with a Port to another coastal province. 
If there's a Port in the target province, you'll be able to drop your unit there. 
The next turn, he will have confirmed the Port's location and you can move armies in. 
Using Ports as avenues of invasion is especially important in the endgame, when you have to do a lot of long range maneuvering and mopping up.

You can reinforce a depleted unit by dragging and dropping it onto another depleted unit.
The maximum unit size is determined in the Options-Performance screen.
There's a slider here labeled Default Unit Size. This is set to 60 unless you've changed it. This means that newly trained units will begin with 60 men and you won't be able to combine two units if they add up to more than 60 men. 
But when you increase the setting, Shogun will allow units containing up to 120 men. This takes effect immediately and can be changed at will with no effect on stacks you've already made. 
If, for instance, you change the setting to 120, stack two 60-men units into a 120-man unit, and then change the setting back to 60, you'll still have a stack of 120 men. 

You cannot divide units once you've combined them. Also, the Default Unit Size setting will have an effect on the amount of time it takes to build a unit, since you're building larger units. It will not, however, change the basic cost; by changing default unit size from 60 to 120, you're simply paying twice as much to build units that are twice as big and take twice as long to train.

Your income is collected once a year in the interval between the autumn and winter turns, so when a province promises an income of 200 koku, you'll only earn this once every four turns. The rice output from your farms randomly varies by as much as +/-50 precent. Although it's cheating, you can quick save before ending the autumn turn and then reload until you get a more favorable harvest. 

This is also the time of year when you collect income from Mines, Ports, and Cathedrals. These sources of income are important supplements to farming because they're not subject to the vagaries of the yearly harvest. Mines are almost always a good idea, since they'll generate 200, 400, or 600 koku depending on whether they're producing copper, silver, or gold. Since these sites are worth defending the gold in the provinces of Dewa and Kai make them attractive targets, especially later in the game once they've been developed it's usually a good idea to upgrade to a Fortress so you can build the Mining Complex that doubles output. Only improve farms where the land is already rich it's not worthwhile spending just over a decade and 3000 koku to double a province's output if it's only scraping together 100 koku every four turns.

Brother, can you spare a grain?

Between the autumn and winter turns, you also pay the amount listed under each province's Expenses column in the Economy Info Parchment. 
This is the support cost for the units located there. 
Unit support costs one koku per person in your armies. There are three exceptions: 
1) your Daimyo and his retinue are free, 
2) Yari Ashigaru are half cost, and 
3) some clans have reduced support costs for certain types of units (e.g. the Takeda clan's 25 percent discount on cavalry).

Training and building costs are paid in full when construction begins, so pay attention to how much koku you have available, and give orders and fill queues accordingly. 
When koku is tight, be careful with queues as they might interfere with specific plans for unit training. 
Unless you enjoy an embarrassment of riches, it helps to use the winter turn to plan your yearly koku expenditures. 
Give build orders in your provinces as needed and then fill your training queues based on how much koku you'll have left. 
It helps to keep a written list of when each province will be finished building structures, arranged by year and season. Refer to this list every winter when you're planning what to build that year. 
The undocumented up and down arrows in the bottom third of the Unit Training and Building Production windows can be used to cycle through each eligible province, making it easier to manage your empire.

The Annual Harvest report can be a little confusing at first. 
"Last Year's Treasury" is what you had in your storehouse at this moment last winter. 
"Farm Income" is what you've harvested in rice, with "Total Income" being the amount it's boosted when you include Mines, Ports, and Cathedrals. 
"Unit Support Cost" is the yearly salary you just paid for your troops and is likely to change over the coming year as you train new units and suffer casualties in battle. 
Finally, the entry under "Building/Training" is a little misleading, since it doesn't figure into any immediate costs you paid this amount over the previous year when you ordered construction and training, so it's been deducted already from the amount listed at Last Year's Treasury. 

If you click on the little box that tells you have much koku you have in your treasury, you can access a Set Tax Rate screen. This lets you set the empire-wide level of taxation that will apply during the next harvest. 
It also shows you what the loyalty will be next season at the given tax rate. 
Note that although the loyalty change takes effect immediately, the actual tax rate only matters in the interval between the autumn and winter turns when you collect your income. You can keep your tax rate at Very Low throughout the year to enjoy a loyalty bonus and then boost it to Punitive once every autumn and you'll still make the same amount of money.

Loyalty is most important when you're annexing provinces from other clans. 
You have to nurture a province's loyalty by keeping a garrison or a Shinobi or two present until the loyalty rating has been boosted. 
It's worth the additional koku from a Punitive tax rate if you can keep your loyalty in the 100-150 percent range. If, however, you're being attacked, drop your tax rate to Very Low. The increase in loyalty will make it harder for the attacker to hold captured provinces; when an attacker takes a province that had a loyalty of 200 percent or higher, you might get an instant rebel army to strike back at them the next turn. 
Similarly, if you move a cluster of Shinobis through your enemy's newly conquered provinces, you can quickly whip up an army of rabble-rousers to keep his hands full.

Decisions, decisions

To get an overview of how the other clans are faring, look at the color-coded map in the throne room. Use special units to find a clan's important production provinces. One of your first priorities in conquering a clan should be finding out where he's making his most powerful units and taking these provinces from him.

When you're planning the building construction that determines which units you can build and where you can build them, pay attention to the presence of iron sand deposits; these are the only places capable of ultimately building Heavy Cavalry and Naginata (the latter are especially important for their defensive value).

When the Portuguese finally show up, you'll have to choose whether to embrace Christianity by aligning with them for early guns or whether to wait for the Dutch. The early guns aren't a powerful enough incentive to convert to Christianity, especially considering the loss of Warrior Monks and the potential widespread chaos of the loyalty hit you suffer when you alienate your Buddhist subjects. 
A far more compelling incentive to convert is the Cathedral income. 
If you're fighting a clan that holds the richer farmlands and more valuable mines, Christianity can be a good way to even the playing field. To handle the conversion shock, drop your tax rate to Very Low and get ready to seed your empire with Churches and Jesuit Priests. Otherwise, just wait for the Dutch to arrive.