Shadow Warrior units represent 25,000 samurai not subject to the normal movement rules - they can strike far and fast, but are no good in defense, unless you wish to dismantle them into a normal samurai army, thus losing their mobility. Your Daimyo are your leaders, and like many wargames your army is largely incapable of movement without the presence of a Daimyo. Army units move from point to point; Daimyo and Shadow Warriors move from province to province, and affect all units in their province. Each province has multiple, border "checkpoints" - these are the places where the movement paths cross the borders, and control of them is crucial for a couple of reasons. First, you can only move a Daimyo or Shadow Warrior into a province if you control the checkpoints between where the unit currently is and where you want to send it. Secondly, control of a province gets you 20,000 koku of annual income, and you must control all checkpoints of a province to collect that income. While samurai can sit at a checkpoint between two provinces, they cannot move beyond that checkpoint - into a province - unless one of their Daimyo is in that province.
     Each player starts with 100,000 Samurai. These can be organized in any denominations of army units you wish, within counter mix limitations. (There are only 15 of each of the aforementioned denominations, which means the most units you can deploy for that 100,000 man army would be 47.) As the smallest increment of loss in battle is 3,000 samurai, defending with a single, 1,000-samurai counter can often be as effective - and more efficient - than defending with a 2,000- or 3,000-samurai unit. Normally, a player will pick a small number of "meaty" units (10,000 and 25,000) to form the cadre of his forces - something with which he can try to play shell games. Since border checkpoints must be covered, a large portion of your army will be spread out at any time; this requires the use of numerous, smaller units. The overall effect of these factors is that the army composition for the various players will probably differ less than you might otherwise think.
     Movement is handled by the roll of one, ten-sided die. Each Player, on their turn, rolls the die, and has a number of movement points equal to the die result. Each movement point can be used to move a stack of armies (or a single army) from one point on the map to an adjacent point. The movement rules allow for "dropping off" units. Obviously, there are going to be some large disparities in movement allowance, from one turn to another, or (worse) from one player to another. If I roll a "1" on my turn, and you roll a "10" on your turn, guess who is getting the upper hand for the turn. When the designers were confronted on their forum about this, one of them responded that this forces a player to be more "efficient" with his movement, and to be flexible in their planning. This type of response is all fine and well, but rather glib and not exactly to the point. To use an analogy, if I get 3 moves in a game of chess for every 1 move Kasparov gets, even I can kick his butt. At that point, superior strategy is unequivocally overwhelmed by the disparity in levels of activity. (Imagine if the Army of the Potomac had trucks, during the Civil War. From a results point of view, McClellan would have come off as a military genius, and Lee would have seemed the hesitant nincompoop.) Our group has adopted a compromise of sorts - we roll one ten-sided die at the beginning of the turn, which applies to everyone, and each player adds 1 six-sided die roll of his own to that number, during his or her turn. This allows for some variation, but establishes a common minimum of movement. If you're completely dead-set against any type of disparity, you could just apply the same, one roll to all players for the turn. (An even more interesting possibility might be to give each player the 1 through 10 of one suit, in a deck of cards, and simultaneously place one facedown at the beginning of the month - revealing your choice on your turn. If no card can be re-used until all cards have been played, then this creates variety that never unfairly affects one player over another.)
Your method for expansion is somewhat scripted, based on my experience. Your initial moves will be to secure the borders of your starting province (or provinces), and you'll then expand into the province(s) between your castles and the central province, where the resource sites are found. As you deploy garrisons to the various border checkpoints, you'll whittle down your "screening" units, and thus restrict your ability to deceive your opponents. (Fewer shells increase the odds for the mark, in a shell game, assuming someone isn't cheating.) Your decision at this point is just how much to throw into the resource province. The way the game is set up, you'll need to capture at least one resource center, early in the first year (12 turns), in order to have enough Koku to support all of your samurai, at the end of the year. There are 5 resource centers, so in a 3 player game, this isn't difficult. In a 4 player game, however, it could be a struggle, particularly if your opponents are inclined to be greedy.
     After 12 turns (i.e., months), there is an "annual" maintenance and income phase, in which each player must first pay for his troops (by paying maintenance for his castles), then collect income and cards from his castles. The resource centers pay on a monthly basis, so you should have some of that stored up, going into the maintenance phase. The options for collecting from your castle cards (as described above) are critical - it's tempting to try and emphasize just money or just ring cards, for example, but you must balance your needs, to a certain extent. You also draw a Fate card, at this time, and these cards represent the vagaries of life, as it were.

     Based on what you roll on the 20-sided die, combined with what side of the board you call your home (East, West, North, or South), these cards can give or take things such as money, Element cards, Ring cards, Samurai, castles, or Daimyo. In the case of the first three, this isn't usually a big plus or minus. In the case of the losing or gaining of Samurai, castles or Daimyo, however, the cards may have a disproportionate impact on the game. I say disproportionate, because one "Community Chest" card shouldn't have the power to cut your force's mobility by half (when losing one of two Daimyo, for instance), your income by nearly half (in the case of the castle), or your armies by a roughly equal amount. A negative result on some of these cards, at the end of the first year, can really cripple your chances of winning, all other things being equal. Somehow, this doesn't seem appropriate in a game that touts its element of strategy.
     The last of the card types to discuss are the Element and Ring cards. You start with some of each, at the beginning of the game, and are eligible to collect more, during the annual maintenance phase. (And eligible to lose some, too, through a bad result on a Fate card.) The rules describe the two types of cards as follows:
     "Element cards represent tactics of war, and Ring cards represent the strength of your tactics." Hmm. I'm not sure that I completely agree with that assessment, but judge for yourself. There are five types of Element cards:
     "Fire" allows you to send a ninja against an enemy Daimyo or family member, anywhere on the board.  If successful, the ninja does, well, what ninja do.  Woosh.  Thwack.  Grunt.  Gasp.  Thud.
     "Wind" allows you to give one group of your Samurai potentially incredible (dare I say "ludicrous") mobility, for one turn. Just how much they move depends on how many of your Ring cards you want to toss in the deal.  Consider that the most an army can move during a turn is 10 (on a die roll of 10), and consider that - provided you have the Ring cards - an army can move an additional 27(!) with a Wind card. If this reminds you of certain, gravity-defying scenes from a certain, recent, Oriental-legend movie, then you are not alone. However, the scale is far grander here, as literally tens of thousands of Samurai can hang from hidden wires with this card. For those who scoff at such fantasies, consider the Wind card as giving 16^th century Japanese armies thousands of troop trucks. Feel better with that image?
     "Water" allows you to bypass any number of consecutive, enemy-occupied points with one stack of Samurai. I suppose this can be rationalized as moving quietly down a stream or river, while the dumber, bad guys watch only the road.  Ahem.
   "Earth" is - simply put - a card with awe-inspiring effects. The Earth card destroys anything located at one point on the board, provided you have Samurai adjacent to that point. I guess this is the 16^th-century equivalent of DUNE'S "Family Atomics".  If you can stare menacingly enough, and make convincing enough guttural sounds through gritting teeth, then presumably the very ground beneath your enemies yawns open to swallow them into a - no doubt - eternally-painful oblivion. Sweet.
     "Void" can be played to negate the effect of other Element cards, except for the "Wind" card. However, you must play a higher total value in Ring cards with the Void, than your opponent played with his card. (Your teeth-gritting sounds are more convincing than those of your enemy.)
     Needless to say, you want to (recite after me) always keep a Void card on hand.  As indicated above, "Ring" cards (each having a value ranging from 1 to 5) dictate how convincing your side of the issue is. It is the gaming equivalent of a nay-saying shouting match; I shout "Earth!" at ninety decibels, and you shout "Nuh, uh!" at one hundred decibels. Clearly your tactics are superior to mine. Q.E.D. You win.
     Battle occurs between adjacent units, and can be both a perplexing and frustrating affair. It involves computing of modifiers, and rolling of dice, and favors the defender. As the rules explain it, each battle (or what they call a "skirmish") requires one of your movement points.
     Now, you get a modifier to your roll for having more Daimyos present than your opponent, and for having numerical superiority measured in multiples of the enemies size (that is, twice as large as the enemy - or 2 to 1 - adds 1 to your roll, 3 to 1 adds 2 to your roll, etc). The third modifier is one added to your roll for each location adjacent to the defender that you have armies, not counting your attacking stack. So, if you have 3 stacks adjacent to the enemy (regardless of their content), you're attacking with one of them, and the other two are adding 2 to your die roll.
     The perplexing part is the decision to use a 20-sided die for combat resolution.  Since the defender wins ties, this means that in a "straight-up" (that is, no modifiers, or equal number of modifiers) battle, the attacker has a 45% chance of winning (180 out of 400 results). That's all fine and well - in a straight-up battle, the defender should have an advantage. Now, let's look at 3 to 1 odds (that is, the attacker has 3 times the troops the defender has), which in the history of combat is usually the "magic" number at which an attacker can reasonably expect victory, all other circumstances being equal. At 3 to 1 odds, the attacker adds 2 to his die, however his roll cannot exceed 20, according to the rules. What this means is that the attacker now has 179 out of 360 possibilities of winning, or a 49.8% chance of winning. Not exactly reassuring, is it? So, how about 6 to 1 odds? At 6 to 1, historically, a defender will almost inevitably get crushed,