Sengoku Rance:Tips

Be warned, this page is mostly opinion and very little of it has been subjected to debate or verification.

General

 * All non-foot soldier units with more than 300 troops will contribute to enemy troop scaling. Courtesy of KrossRoads/DarkMaster22's FAQ :


 * Enemy troops are scaled differently for different categories and difficulty levels:

Method 1:

Your average number of troops, calculated excluding footsoldiers and units under 300 troops. Difficulty level: Normal difficulty: You have the advantage in the scaling (enemies increase at a slow rate). 1-Star: The scaling is even between you and the enemy. 2-Star: Enemy has the advantage (same rate as you had in Normal difficulty). 5-Star: Enemy has a large advantage.

Method 2:

Number of provinces under your control. Difficulty level: Normal difficulty: You have the advantage in the scaling (enemies increase at a slow rate). 1-Star: The scaling is even between you and the enemy. 2-Star: Enemy has the advantage (same rate as you had in Normal difficulty). 5-Star: Enemy has a large advantage.

Units ingame are scaled as such:

Demon Army: Method 1 Enemy's new recruits: Method 1 All non-DA enemies: Method 2 Foreign reinforcements: Method 1 Map Recruitments: Method 1 Event Recruitments: Method 1 Event prisoners: Method 1 Captured prisoners: Method 2, max 1000 Tenshi's troop escape event: Method 1
 * Note that some commanders always seem to start with very large troop sizes (e.g. Sakamoto Ryouma) and enemy troops will also increase based on the number of provinces you control. Most enemy commanders will be scaled based on the number of provinces that you control, so you should try to increase them as much as possible when you have the funds to do so.


 * If you have too many satisfaction bonuses available, you will have to use up some to make the others available. The most common culprit of this is the one that turns Meda Toshiee into a kabuki actor. This can also hide the action fan bonuses.
 * In the end phase, if you click the skip button, it will skip all previously read text for the WHOLE end phase, not just that particular scene. Use with caution.
 * After you conquer a province, every turn the levy amount will go up by 100 gold, up to a maximum of 5,000 gold after 30 turns. If you are not in a hurry for books or gold, its worth waiting before levying a province.
 * Reinforcements take 6 turns to arrive, so its best to use the first few sat bonuses on them.
 * Chances of capturing prisoners decreases if the prison is too full (at approximately 20/30).
 * Starting from turn 6, you will get a random textbook (normal) the first time you fight a battle. Every 10 turns after that, you will also get another textbook for the first battle that turn. You can get speed textbooks this way, which are not available from levies.
 * The game encourages you to be in a permanent state of war. There are events designed to force you into war with other houses if you remain at peace for far too long.
 * Botan hunts become available again on turns that are a multiple of 4 (it doesn't matter when you do them) and panda hunts in Nakakuni (the Takuga gate) are available 7 turns after you do them. The one in Owari increases the botan's level by 1 every time you complete it and the one in Sado by 2 every time you complete it.
 * The cherry blossom event (pre-requisite to Nobunaga breaking the first gourd) only occurs after the first gourd house is defeated (typically Ashikaga) or on turn 12. The gourd event will occur shortly afterwards, or on turn 15 if no gourd houses have been defeated yet.
 * The dungeon events that unlock H-scenes with Sill and Suzume can only occur after the cherry blossom event.
 * The default game animation screen is VERY slow and turning off animations does not help much (time taken for damage numbers to finish displaying, time for windows to pop up, etc). One way of fixing this is to use cheat engine and set the speed hack modifier to something like 3, which will greatly speed up animation speeds.

Commanders
"Depends on if you use bonus or not. If you do, you gonna have better units at the begin of the game, so upgrading Rance is not the best thing you should do (but you definitely need it if you want to get "fuck 300 women" achievement). I'm doing a 5-star no bonus at the moment and upgrading Rance's stats is actually the best thing I can do if I want to win the game by conquering (and completing all achievement as well). Here is what I do to up Rance to a godly unit at early-mid game:
 * Not all commanders are good. Some like Akihime are only worth using to see their events once.
 * Unlike most games, the insta-kill moves in Sengoku Rance have a fairly good probability of working, so they shouldn't be ignored.
 * Some skills are only available on certain turns, such as dungeon battle expert and counter attack 1. Since there is no way to tell before hand, save and reload if the skill you want is not available on that turn. The counter attack 1 skill seems to be available on turn 3 at the earliest.
 * Mikos except for Natori and Chinu are not worth it as the amount healed is too low, does not give you influence back in battle and prevents you from bringing another commander to kill the enemy quicker.
 * Keeping Fuuka around after you unlock her h-scene is not recommended unless you are trying to unlock the Orochi cave or get her character clear. Her usefulness is very low.
 * Raping Kanami does not grant you any SAT. You do not need to talk to her at all, just wait till Maria gives you the Kanami sex ticket. The H-Scene will occur in a beginning of turn event after that. She is required to find the Tora no Ana dungeon in Akashi so keep her around at least till then.
 * Monks (including Seigan), Archers, most tacticians (except for Sanada Tourin and Uruza) and Mikos do very low damage in commander battles, regardless of how high their level is.
 * Regardless of what items are equipped or how high your level is, the maximum amount of damage that may be done to Xavier or his apostles per hit is 20% of their maximum health. This includes Rance's Rance attack at level 99 vs a level 55 apostle (only 230 damage will be dealt, same as if you had used the normal attack).
 * If not doing the Kenshin route, Okita Nozomi can be dismissed after finding the Ero-Candle item in dungeons and unlocking her H-Scene, as well as finding the SM Dungeon in Mazo and/or getting the Charm of the Immortal Fox Deity. She is very slow in dungeons and only has 3 AP. The quickest way to do this is to talk to her once, talk to her at 7 affection to get a generic h-scene, find the Ero-Candle, then get her H-Scene.
 * More realistically, she's probably not worth recruiting at all. She's only good in dungeons, and costs too many fans to get.
 * In commander battles, only basic skills will be available. You will not be able to use charge, commander charge, etc and most passives like Hibachi or Lover will not work. Attack luck will work for both troop and commander battles however.
 * A points runner poster a guide a while back. On upgrading Rance:
 * At turn 6 and 16, try to get 2 spd textbooks (spd 4->6).
 * Conquer Mamushi oil field before turn 12 and impose levy to get 1 atk book and use it to Rance (atk 5->6).
 * At the end of turn 12, Rance's mood must be 7. The start of turn 13,Rance will get an event which gives him 1 affection. If his mood is at 7, it will give him an auto power up. Up his atk (6->7)
 * Conquer Miko Ins after get 2nd spd book at turn 16. H-event with Natori gives him an auto power up (spd 6->7)
 * After that, impose levy to get 1 def textbook and another def textbook at turn 26 (def 4->6). Power up him and up his def to 7
 * Give him elite troop. His stats should be 8-8-6-8 (give him int book if you want, but don't use power up to boost his int to 7. Not worth it)."

Troop battles

 * After every battle, there is a chance that one of the commanders in the battle will gain an affection point. There is also a chance that you will gain an affection boost item.
 * Since you need to attack and defend multiple times a turn, arrange the commanders in the order that you want to use them to make it more convenient. You can still use "arrange by relevant stat" when exploring, etc as it will not permanently change your sort order.
 * You don't always need to put ranged units in the back. Sometimes its better to put them in the front row to get better protection from foot soldiers without all guard.
 * Not all troops move at the same speed. Monks are very slow, while diviners act quickly on their first turn but are very slow on their second turn. Increasing the speed stat for slow units such as diviners will not do much. Turn order is determined before the start of the order, so speed buffs in battle from tacticians appear do nothing unless they obtain it before the battle via speed luck or items such as Field Battle Weekly.
 * You can only kill or capture a max of 2 enemy commanders a turn, although light attack allows you to bypass this limit (for example, if you hit two enemy units with light attack, you can potentially capture or kill up to 6 commanders after the battle, although this requires lots of light attacks).
 * Instant kill attacks do not guarantee that the commander will be killed after the battle. Troop count of the defeated commander will also stay the same after the battle.
 * Unique commanders, such as house leaders required for the plot, undismissable commanders like Suzume and commanders like "Ogawara Guard" that only show up in certain battles cannot be captured, although some of them may be killed after the battle.
 * Exception : Rance always dies when his troop strength is reduced to 0.
 * Commanders that are captured then released from prison are removed from the game and will not join other houses. Combined with light attack, this can be a great way to remove troublesome commanders like Goemon rather than waiting for the random capture chances of defeated warrior hunt.
 * Tacticians are very effective. Their buffs make a huge difference in battles. Correspondingly, enemy tacticians should be priority targets as well. However in FFA mode, when you mostly have access to generic tacticians, they are much less useful as they only have battle strategy 1.
 * Since tactician buffs are randomly assigned before the battle, it is sometimes better to use less troops in conjunction with a tactician. For example, a foot soldier, Omachi, and a tactician will ensure that Omachi has a very high chance of receiving an intelligence boost.
 * The number of pre-battle buffs a tactician can give is 1 per INT, and the effectiveness is 5% per INT.
 * Guard shikigami can block any attack except instant kill attacks, no matter what the strength of the attack is. Use weak units to remove the barrier, or wait for them to attack and have it removed via counter attack damage. The first portion of the damage from commander charge will also remove the barrier before the rest of the unit attacks.
 * Footmen guard halves all physical damage done to the unit and killing large stacks of footmen can take a very long time unless you bypass it with magical attacks, guard break, guard cancel or instant kill attacks.
 * On top of influencing the Battlefield Effect, fields, towns, and castles affect initial Battle Rating of the fight.
 * Field: 10% advantage to defender, no Battlefield Effect change.
 * Castle: 90% advantage to defender, Battlefield Effect +20%.
 * Town: Around 30% advantage to defender, Battlefield Effect +10%.
 * The damage cap of a unit is it's actual troop size modified by skill modifiers and battlefield effect. For example, a unit with 1000 troops can only kill 1000 troops assuming it is using a normal attack and the Battlefield Effect is 0%, regardless of how high it's stats are.
 * It's not necessary to increase all stats to 9. Some unit types like musketeers can easily reach the damage cap against most enemies with only an attack of 6. See here for more details.
 * As a general rule, it is always worth it to max INT for diviners/mages/tacticians, attack for Ninjas/Warriors/Footsoldiers that you are using for damage and an attack of 6 or 7 is sufficient to reach the damage cap against most enemies for cavalry/musketeers/canons, assuming they receive tactician buffs.
 * Archers are rarely worth it, as their poor stats (they do about the same damage as ninjas but are much slower) and high replenish cost make them ineffective unless you use them in very large troop sizes, which is expensive and affects demon army troop scaling.
 * Unless doing a point run, it is not necessary to use as many operation permits as possible. It is very expensive to replenish the non foot soldiers ones (unless you use SAT bonuses to lower the replenish cost) and it also affects enemy troop scaling.
 * Every time you deploy a non-generic cleared character with Rance, he or she will get +1 affection point. This is an easy way to build up lots of affection for commanders that can benefit from many level ups, such as Uruzu. Only one character can gain affection this way and some take priority over others.
 * Since most defensive battles will be fought on fields, giving a commander in a defensive group Field Battles Weekly is a good way to make full use of it.

Vassaling Advantages
Vassaling is generally a bad idea, but I believe the following to be a complete list of the advantages that you can get from it:
 * 1) It's slightly less hassle for getting Seigan from the Tenshi Sect.
 * 2) It lets you avoid a war that isn't worth the time, effort, or resources to win. For example, the Hara House is sometimes vassaled in challenge runs due to its castle being fairly difficult despite rewarding the conqueror with little more than 5 SAT.
 * 3) There's a free item, the Blue Seal Case, if you vassal Tokugawa and go through some events.
 * 4) You can attack through your vassals but can't be attacked through them, this can give you some tactical advantages. For example, you can beat the Ran route without ever fighting the Shimazu House.
 * 5) It's an easy way to relieve some pressure when things go badly or you're in a hurry.
 * 6) Vassaled houses give some of the same benefits as conquered ones, allowing for some obvious advantages whenever you're getting desperate. For example, if you need NP and need it right now, then you can vassal someone. Note that you cannot increase the NP of a vassal.
 * 7) The reinforcements units are sometimes worth something.
 * 8) It's good if you need to stop a gourd auto-breaking. For example, if your war with the Ashikaga House is not going to end before turn 30, then you're often better off just vassaling them. This is because turn 30 will have the Takeda House's gourd auto-break, and you don't want to have to deal with that on top of Ashikaga's gourd breaking when you finally conquer them.
 * 9) There is a glitch that lets you avoid breaking up to 3 gourds despite having vassaled some gourd-holding houses. It's detailed near the top of the Walkthrough page, but I think that the MG translation patched it?
 * 10) +5 SAT from H-ing Naoe Ai if you know what you're doing. It's also one of the ways to H Hibachi.
 * 11) Some vassaling events are on the Clear Game Report.
 * 12) In some extreme points runs, it is optimal to let a vassalable faction make the Coke Bottle drop for you because vassaling will save fans that would otherwise be spent on reclaiming the lost province(s) and defeating the faction in question.

Uesugi Kenshin route
All of the methods below are so fast that they can win the Emperor Race before Xavier becomes able to kill anyone. In roughly descending order of speed, suggested paths are: For all of the above, it helps to do some preparation before route entry (e.g. get the third action fan and introduce the Shimazu House early by conquering Izumo early). In some cases, it also helps to have: Mouri Motonari uncursed, obtained the Ryuuguu Sword Urashimaru, and already recruited any of the commanders that your plans require. As a general point, it's also worth having the Miko Institute conquered, or at least ready for you to declare war on, before route entry.
 * 1) A three duel route with major abuse of Motivational Paper. This is possibly the fastest possible method, but requires extreme preparation. This method completes the race in five turns.
 * 2) A four duel route, with level loss:
 * 3) *Before route entry, without taking the Mouri Family Bonus, save up four SAT Bonuses, give Kenshin the Ryuuguu Sword Urashimaru, and recruit Gon, Okita Nozomi, Yuzuhara Yuzumi, and Mouri Motonari. It will help to be at war and have one Happy Pokkuru.
 * 4) *Throughout the route, starting immediately after route entry, have Kenshin lose at least 7 battles while above level 20, giving you an extra 21 duel-free points.
 * 5) *After entering the route, have Kenshin lose her levels as soon as possible (talk to her twice), then restore them quickly using the saved SAT Bonuses and at least one level-increasing item or event (even with the sword, without either Attack Luck or a level, it appears that Kenshin cannot beat Mouri at level 41). Duel the four racers at your leisure (probably one a turn, so Kenshin can heal).
 * 6) *The advantages of this method are that you do not need to do any start of turn event optimization, difficult duels, or +2 level events. Furthermore, defeating the Mouri House early will quicken the Demon Army's introduction and give you a border with them that can be used to conquer them once Kenshin becomes the Emperor.
 * 7) *This method completes the race in eight turns. For a seven turn race, win two duels in one turn and, probably in a different turn, use any means of letting Kenshin lose two battles in one particular turn.
 * 8) A four duel route, without level loss:
 * 9) *Do not talk to Kenshin and never allow her to lose her levels. Instead, get 80 points from Suzutani Musou, Kumano Gorou, Shimazu Kazuhisa, and (uncursed and in his own army) Mouri Motonari. Musou and Gorou's duels are so easy that they can be done on the same turn as a different duel.
 * 10) *As in the previous method, the other 20 points will come from having Kenshin lose battles. However, to accelerate the triggering of the three needed racer introduction events, you must stay below 20 points until Kumano Gorou and Shimazu Kazuhisa's racer discussion events have been seen and avoid the equivalent events for as many other racers as possible (see the bullet points on start of turn events in the walkthrough; the main tip is to not be at war with other houses that have racers). This means that you will have many turns where your only objectives are to have Kenshin lose a battle and to not lose the provinces needed to battle the four racers listed above.
 * 11) *Make sure that you are familiar with Kazuhisa's complex event conditions. Again, check the walkthrough.
 * 12) *The disadvantages of this method are that you need to keep the Mouri House standing, not lose control of several provinces, and that you need to wait for the start of turn events for Musou, Gorou, and Kazuhisa. You could swap Kazuhisa for another racer, but that would add an extra unneeded start of turn event. The advantages of this method are Kenshin not losing her levels, having a good border with the Demon Army, and not needing to recruit any racers.
 * 13) *This method completes the race in eight turns.
 * 14) A five duel route, with level loss:
 * 15) * Before route entry: Recruit Gon, Yuzuhara Yuzumi, and Okita Nozomi, save four SAT Bonuses, and try to avoid being at war with any houses that have racers. It will help to have taken the Protection Paper as a Second Game Bonus.
 * 16) * After route entry: Give Kenshin the Ryuuguu Sword Urashimaru and avoid letting her lose her levels until both Suzutani Musou and Kumano Gorou have been found. Beat them both in the same turn (doing otherwise will have an event where Rance congratulates Kenshin take priority over Gorou's introduction), have Kenshin lose her levels, and restore her to level 41 with the SAT Bonuses. Finally, beat the three other racers that are in your army.
 * 17) * Depending on how your fan count works out, you may find it best to defeat Gon on the same turn that Kenshin reached level 41 and defeat both Yuzumi and Nozomi on the next turn. If Gon is fought in this way, then you will have to beat him while Kenshin is very low on HP. The Protection Paper is recommended for this, but the other four duels are best done with the Ryuuguu Sword Urashimaru.
 * 18) * This method has the benefit of simplicity, but its overwhelming disadvantage is that you need to wait for both of Musou and Gorou's events and then spend several turns restoring Kenshin's level. Aside from keeping control of Naniwa and Saitama, you will have many turns where you have no goals at all.
 * 19) *This method completes the race in nine turns.
 * 20) Any number of duels, fully exploiting the Akashita Dog:
 * 21) * Give Kenshin the Akashita Dog, restore her levels only if they fall below 21, and, whenever possible, have her lose two battles a turn (e.g. one attacking and one defending) until you have won the race. In this case, you mostly do not need to bother doing any army battles at all, other than losing with Kenshin. To speed things up, you can try to beat Suzutani Musou, Kumano Gorou, and even Shimazu Kazuhisa before Xavier does. In extreme cases, you could also consider any of (uncursed) Mouri Motonari, Baba Shounen, or Dokuganryuu Masamune. Given that you probably do not need to do any conquering, it should be easy to get Kenshin up to a practical level.
 * 22) * This is by far the easiest method, but it is also very slow. A major advantage is that it works almost independently of how well you have played the game. Even if you only have Owari under your control, you can probably make this work.
 * 23) *If care is taken with your start of turn events, allowing you to duel Musou, Gorou, and Kazuhisa without ever having Kenshin lose her levels, this method completes the race in ten turns.

Nanjou Ran route

 * This is the most relaxed route other than the Kill the monkey route, so it is sometimes recommended for your second playthrough.
 * After you clear the oil dungeon before 5 gourds are broken, you have 6 turns (assuming all purple plot events are chosen as soon as possible) before Rance will automatically head to Honnouji. It is not necessary to fight Xavier since it is a scripted fight. You can simply right click at the commander selection screen to "lose" the battle automatically.
 * Keikoku will first appear 10 turns after you clear the oil dungeon. The Shimazu family will declare war on you after that.
 * 21 turns after you clear the oil dungeon, Keikoku will disappear with Souun. 2 more turns later, the first holes to hell will open up in Mazo, Shinano, Sado, both Dokuganryuu provinces, Saitama, Kazusa 2000, Edo, Kai and Toukaidou. The ogre units defending the holes are very weak so can be dealt with easily.
 * To avoid Shimazu from automatically taking over Senkan Nagato, you can leave the Mouri house alive with Senkan Nagato as their last remaining province. It will not be automatically taken over then.
 * Using a throw away generic commander at hate can be a good way to sacrifice a commander in order to recruit the Shimazu brothers.

AI exploitation
The precise details can be found here. Be warned that exploiting these tips gives such a large degree of control over the AI that it will forever ruin your enjoyment of the game.
 * The Tenshi Sect will not attack a province that they do not own on even numbered turns.
 * Losing more than one battle during the end of turn phase will prevent every faction's AI from attacking you again on that turn.
 * Don't overlook how powerful this tip is. It essentially lets you control the enemy AI. If you're fighting someone, then they won't even try to reclaim their own land if the end of turn phase starts with you surrendering two territories to someone else. Imagine being able to conquer Mouri without them ever attacking you back, just because you let Hara steal some territories.
 * This is extremely exploitable with Pursuit Battle or Battle Permits.
 * You don't even need to enter the battles - just right click.
 * The Demon Army is always the last faction to take their turn, so as long as you are at war with some other faction, their AI can be almost completely disabled with this tip.
 * The game's code also suggests that losing just one battle during the end of turn phase should be enough to give the AI a 50/50 chance of not being able to attack you again.
 * The above tip does not apply in the following situations:
 * Normal difficulty (0 Star), uses a slightly different mechanic that should overwrite the above one. I don't think that I've ever tested it properly. The mechanic for that is something like "no house can fight more than 1 battle per end of turn".
 * Battles that the game treats as events, such as the Shimazu Blitz, Takeda Blitz, Apostle Shikibu's attacks on Cairo, or the secret end of turn event attacks of the Tenshi Sect and the Ashikaga/Houjou/Tokugawa houses ignore this limit.
 * The AI won't attack a province that it doesn't own if it doesn't have more than 19 commanders at over 59% troop count that haven't already acted this turn. For reclaiming territory in provinces that it does own, this number is 9.

Conjectured tips
Whilst diving in to the game's code, I've spotted a few tricks that I've not looked in to enough for them to merit them being put on a more formal page. Some of these I've put to the test in the game, but most of these should be taken with salt. if in battle vs nobunaga without kentarou or rance: return kentarou or rance
 * If you're on 2 star difficulty, then after turn 40 most houses will get a boost that makes them able to do more in their turns. The boost is bigger on 5 star.
 * It appears that the Dokuganryuu House cannot have generics of certain classes. For example, I don't think that youkai Mikos or Monks exist.
 * It appears that generic Ninjas cannot spawn with upgraded first or fourth skills, e.g. Leaves Hiding or Assassinate.
 * Female units have 90% of the HP as an equivalent male unit.
 * Allegedly, the below is pseduo-code for how units are swapped in during commander/dungeon battles:

if row == back && sill is available: return sill

randomly pick number from 1-30 up to 39 times if player has eligible character in that slot with max action flags: if character troop type isn't suboptimal for the row: return character

randomly pick number from 1-30 up to 39 times if player has eligible character in that slot with any action flags: if character troop type isn't completely unsuitable for the row: return character

return failure

Character is eligible if not in battle and has at least 2 HP.

suboptimal front: tactician, diviner, miko, mage back : warrior, musketeer, ranceLike, knight, nonCombatant, darkLord, youkai completely unsuitable front: nonCombatant, tactician, diviner, miko, archer, cannon, mage back : nonCombatant, warrior, musketeer, ranceLike, knight, darkLord, youkai

Free For All
What follows, with some minor corrections added, was taken from a forum.

If we're talking about default FFA, my most notable achievement is beating it with Akashi. Imagawa runs brought me nothing but pain and misery. There is probably someone here who did that though. Lately I'm having fun with the DHR mod, in there I beat it as Ice without custom units and as Shangri-La with customs. Haven't tried Pencilcow runs yet.

So for tips. You're probably playing as a weak faction without good uniques (why would you need tips otherwise?). This almost always means you have only 1 starting territory, low income and strong neighbors. There are people who enjoy challenges and try to make no-neutrality pact runs, but with weak countries your only chance is to play defensively and upgrade your barrier to max level. Speaking of barriers, they are good. I don't know the formula, but even a level 1 barrier can protect you from many attacks. Good to build them at chokepoints if you're struggling, have borders with many different enemies, or don't have enough good units to fight many battles in 1 turn.

Back to what I was saying. 1 territory, weak army, strong neighbor. Okay, if I recall correctly, your starting generics have 5 or 6 in stats, so they're not that bad. Enemy uniques will crush you. The key is your starting Tactician with Defeated Warrior Hunt. You really need a good unit to start rolling in FFA, so if you don't have it at the start, you have to capture it.

For about first 10 turns enemy armies will deploy 3-4 units in battles. Formations will be wacky, sometimes they won't even have a front line. This is a great chance to shine for Diviners, if you have them. Enemy numbers will be similar to yours, so you can take chances and grab some land even from your stronger enemies. But you'll be open to counterattacks and your other enemies will try to take you starting territory. Last locations will have all 6 slots deployed, so if you are not able to take it, you've just wasted a lot of time and money.

Money is a problem. You should start with a hospital or a shopping district. I prefer hospitals, but you should run the numbers for yourself (e.g. a district may be better if most your losses are on cheap-to-replace Foots or you really need to buy troops for a good commander). If you have Mikos, they're great money-savers for the beginning. If you capture Monks, you can recruit them, then sell all their units on the next turn. You probably will even need to use flags on "heal 25 troops on every unit" actions.

You probably won't have enough units to scout until you recruit some prisoners, but you probably won't need it. AI always tries to send uninjured units (except when he defends his last province), so dealing enough damage to his important uniques may make sure you won't see them for some time. AI has trouble healing troops, so prolonged wars will weaken it greatly. As will capturing and killing commanders.

About captures. Strong unimportant uniques are easily noticeable and great targets for capturing. They won't die while on AI's side and can only be captured. Important uniques of course cannot be captured until they become ronins. But mostly you'll be capturing generics. You need to watch everyone's damage-dealing closely in battles to see if someone's stats are good. You can easily see the difference between attacking a 3 DEF unit and a 7 DEF unit. Same with ATK. Now, for the types:
 * Warriors - Get units with at least 6 ATK. High SPD is great, DEF is preferable, but you can make use of glass cannon Warriors too.
 * Foot Soldiers - Don't release them, Foots die more often than any other type because they are in the most danger of being annihilated. In mid-game you should have at least 4 Foots with at least 5 DEF because you'll need to protect your back line. Not always though. When fighting really strong opposition, where you fear you Foot WILL get annihilated, it may be better to sacrifice worse Foots. That's when your 1,000 troop 3 DEF stacks may come in handy.
 * Archers - Garbage. I struggle to use even 8 ATK Archers, Volley is bad, Foot Soldier Snipe is usable, but situational. No, really, sometimes Foot Soldier Snipe is the only thing to deal good damage to guarding Foots.
 * Ninjas - Good enough ranged units. They're fast, so can cripple strong enemy units before their turn, may have more turns than other units, giving more turns to your side and giving less to the enemy and can break preparations. This is the main reason to have them in reserve. Assassinate is good, but don't count on it. On the downside their damage is pretty bad. I probably won't take Ninjas with less than 7 ATK.
 * Monks - Skills are situational. If they have good stats (mostly ATK) then take them, otherwise Warriors are better.
 * Mikos - Good in early game and defensive battles, will alleviate your money problems. At least 6 INT. All of the above Miko discussion applies.
 * Tacticians - Really good. I'd even take 5 INT Tacticians. I haven't seen generics with Battle Rating Down, but we all know how buffs change the flow of battle. If you're playing DHR, where you can upgrade them, then all the better.
 * Tebasaki - Godsend. You can compare them to Warriors with 2 ATK higher then theirs. And this is without taking in to account better speed, ability to attack backrow and chance to have Battle Rating Up. Even 3 ATK Tebasakis can be used, you'll probably get better ones.
 * Musket - Wouldn't advise. Having a unit with only 1 action is bad, most of your damage often comes from the front line and Warriors should out-damage Musket users in 2 attacks. If you really need someone busted down then you could. I mostly just don't like them. Except Special Cannon.
 * Diviner - Amazing in early game, where they'll fire off all of their Shikigamis. Later, they'll be used only for Guard Shikigami.
 * Mage - Generic Mages exist only in DHR. Can be used against Foots because of their magical attack. Damage is fine. INT 5 will probably be enough.
 * Peasants - Once upon a time I captured a Peasant with 8 in all stats. I didn't recruit him.

Don't forget to look at the non-combat stats. You should have some units with good Search on scout duty. Negotiation and Construction will help too. If after selling its troops it will have 0 cost - even better.

The breaking point is getting 10 territories, after which you'll be able to upgrade buildings. You see, mansions are strictly better than other buildings money-wise, you just don't get the value from them while struggling in the beginning. But AI's troops are always increasing, so you'll need mansions to keep up. So it's finding the balance between present value of healing troops and recruiting troops to good units and future value of having many large usable troops.

FFA ends at any instance where you don't have any enemies. So if you are left only with houses that are neutral to you, you instantly win. That's mostly the case when there are 2 enemies left and you neutraled one of them, because double-front wars are stupid (challenging) in FFA.