Weaponsmiths or armourers?

[Jacobin]

Once I capture provinces like Sagami or Echizen i try and recruit all of my new units from them but haven't quite yet worked out which upgrade path produces the best results.

As I play SP at Legendary and tend to favour ashigaru armies until very late in the game (in late game I find almost all of the battles I actually fight on the battlemap are sieges where a full stack of ashigaru and garrison units defending a castle or citadel can beat two or three besieging samurai stacks - the sieges where I am attacking or field battles tend to be so one sided I usually autoresolve) so I first went down the road of maximising armour to gold for the melee units and attacks for missile units to gold.

This meant that my yari ashigaru were much better at surviving the storm of arrows AI bow samurai tend to produce in mid-late game battles but that my main troop killers were my bow and matchlock ashigaru units who regularly notched up 500, 600 or 700 kills each in siege defences but with their pathetic armour ratings got themselves very easily slaughtered in field battles.

But in my current campaign I am building armouries in all cities (having silver armour garrison units and the other benefits of encampment+armoury seems worth it) and going for the master weaponsmith and master bowmakers in my troop-producing cities which gives me silver upgrades for all units armour and for the melee and missile attacks.

Main difference is that my yari ashigaru die significantly faster under missile fire but are more likely to win melees against the katana and yari samurai the AI throws at them (at least if these have been decimated by my matchlock and bow ashigaru units first - as in sieges they usually are due to the AIs habit of having them run or stand around uselessly before they find a wall to climb).

This seems to be an acceptable compromise but I am open to persuasion - what upgrades do other people choose?

[faker01]

Normaly weaponsmiths are better for mellee and armourers are better against ranged units. Yari ashigaru have more def(armor+def) and cause mellee is typically much more important, i choose weaponsmith.
Against ranged units my ashigaru nearly never had problems. Typically I attack them and often there are other units in the way, so mellee bonuses are still usefully.

[Maltz]

If you want to train tank infantry (Yari/Naginata), armor is better as they will be put in front, and take the enemy's arrows. If you look at your units' armor value, you can see that +5 armor is usually a huge step up.

If you want to train shock troops (Sword/melee Cavalry), sword is better as these units will rarely be shot at (if you use them with caution). If you look at your units' melee attack value, you can see that the bonus does not make a big difference for many, except for Ashigaru. But they are terrible shock units.

In many situations, shock units can be replaced by more Bow units. Bow units are afraid of bad weather, which you can easily avoid by being the aggressor.

So +armor is my top priority, followed remotely by +sword. I... actually have not won a Legendary campaign with any +sword unit ever in the army.

[Fagar]

I tend to find armouring up is the better option..
The troops at the front are the ones that NEED to hold that line..
I use them to bottle neck the bulk of the enemy while I position my swords and horse in flanking positions and co-ordinate them to smash into the enemies unprotected flanks this shock value is easily as good as having +'s to weapons and will normally break most units just as fast ..
My armoured up spear therefore are vital to buy the time for this to happen..

[Jarmam]

In values alone it seems on paper like armor is pretty much always the better choice unless you go heavy into cavalry. But what exactly does +6 attack or +5 armor (or +10 charge for that matter) do? Are the unit values broken down somewhere so they're easier to value against eachother?

For that matter is there any way to customize them for self-testing? Am I just re... cretinous and incapable of finding that option in custom games?

[Jacobin]

Ludens did a spreadsheet https://forums.totalwar.org/vb/attac...9&d=1319571126 which tells you for instance that applying +5 armour to a standard yari ashigaru gives them better armour+defence than say katana samurai (YA armour+5=7, defence =4, total =11; KS armour =5, defence =5, total =10).

Working out the cost-benefit effect of improved attacks/accuracy/morale is rather more complicated....

[Fagar]

The basic additional values are not too hard to work out just adding the bonus to the basic unit stat..
To me this is a good enough indicator for the relative worth of the koku outlay for improvement.

It does get a lot more complicated for accuracy or morale and really I find it impossible to try and be too accurate with this at any rate considering the various applications of generals bonus/traits, clan indicators, mission bonuses and objectives met..
Let alone trying to measure the individual changes in units experience as they progress through the campaign...

[Jarmam]

I dont mean what the stats will be before or after, thats pretty straightforward.

I mean what is the difference between 2 armor and 3 armor? And is it better, worse, equal to or something otherworldly when compared to 3 armor against 4? Against what? Is it a percentage or a flat bonus? If so then to what extend? How about attack skill vs defense skill? What does charge bonus do - increase the efficiency of a charge, sure, but by how much? For how long? What is the difference between 30 charge bonus and 35? What time is it? How come I have two faucets?

If anything, accuracy and reload skill are the two stats that are fairly easy to determine the effect of as they are both visibly clear when archers shoot. It seems the archers start predicting enemy movement better at around 40 or so accuracy and start becoming unrealistically insane snipers (in terms of the shot itself connecting with the target, not whether it actually causes a casualty) at 75ish. Firebomb throwers work quite a bit differently, but the +accuracy Art is very easily visible in combat. Reload skill is also really obvious - take a decent archer like, say, any Ashigaru Yumi and pit it against the worst unit in the game, which would be the Ikko Ashigaru Yumi. The difference is mindboggling.

Attackskill, defenseskill, armor and charge bonus I have virtually no clue on what their effective function is and how its calculated.
To make a crude example lets say that armor is a linear chance to avoid getting killed by an arrow (arbitrary value of, say... 6% per point). If that's the case then 2 armor is a ( 0.06 * 2 ) / ( 1 + 0.06 * 2 ) = 0.12 / 1.12 = 0.107 = 10,7% chance to avoid death upon arrow to the face. So 4 armor would equal twice the amount of connecting arrows required to obtain same lethality. But what if armor works with a flat percentage point chance per point of armor? Like... 5% (again a random value for effect). Then the increase from 1 to 2 armor would be significantly less effective than the upgrade from 8 to 9 armor, as 1 to 2 would increase arrows required to kill by 12,5%, where as the effect of going from 8 to 9 would be an increase in arrow requirement by 100% or twice as many. It could also stack with a base % and have each +armor add x % points, effectively the same except making 1 to 2 even less appealing. It could also be diminishing?

What Im pondering is what do the stats mean? Its impossible to gauge 6 attack and 5 armor if I have no idea what 6 attack or 5 armor (or 10 charge) means. Thats what I meant by "has anyone tested what these things mean and how they work or is it common knowledge and I just dont know where to find it"?

[Maltz]

Here is a quote from Jack Lusted from Creative Assembly:

http://www.twcenter.net/forums/showt...09#post9237609

"Armour applies against both melee and missile attacks. So units with high armour are much more resilient against bows. Heroes have multiple hit points to compensate for their small size and it also makes them take less damage from bows. Armour is powerful hence the lower bonuses for it."

[Jarmam]

I guess that confirms that getting +armor is virtually always the correct choice. I just wish there was a more clear explanation of what the stats do ;)

I used to get +charge bonus for my Light, Yari, Fire and Great Guard Cavalry, but I stopped as I realized I had absolutely no idea what I was doing and whether it had any effect at all. I couldnt tell the difference.

[Jacobin]

This is why this game is so frustrating as compared to say the Elder Scrolls series where all of the algorithms for determining results are well documented and you can calculate exactly what the impact of changing a weapon or armour or spell is.

But that is just the OCD numbers geek in me - as the great pen and paper game designer Greg Stafford says 'x (x being whatever fictional or historical people the game represents) don't know the numbers'.

Having an element of mystery is in its way more realistic - Oda Nobunaga didn't know what the precise impact of increasing his ashigaru's armour or lengthening their spears was - he just knew that more is usually better.

[Nelson]

Having an element of mystery is in its way more realistic - Oda Nobunaga didn't know what the precise impact of increasing his ashigaru's armour or lengthening their spears was - he just knew that more is usually better.

Bingo! The various upgrades improve the men somehow, either offensively or defensively. (Not that there’s anything wrong with sorting out the math.) Ultimately it comes down to piling morale deficits onto the enemy force until they break. I always want my guys to have the greatest staying power. In game terms that means high morale and armor. I don’t much care how long it takes me to kill them as long as the other people quit fighting first!

For more offense I rely on my generals to get the infantry commander attack bonuses.

[Jarmam]

Any simple roll of dice in combat would provide plenty of randomness.

Let me give an example of why I get frustrated by being ignorant: In the good old 1995 days there was an excellent turn based strategy game called Warlords 3, wherein your units would have a "str"-rating of x (between 1 and 9). They could move x spots per turn and take x (1-4) hits before dying in a fight. Now the combat system was that a dice was rolled, the unit's str was added to the result of the dice and the loser took a hit. The dice being rolled was the randomness - a dragon str 9 hp 3 could get flattened by a simple group of light infantry str 2 hp 2 if they were lucky enough with the dice rolled.

However, as it always was with such things, the dice being rolled could be anything you wanted in the particular game - from... I think it was a 8-sided dice to a 50-sided if you like a bit of randomness to spice things up. Naturally this meant that in campaigns the dice would vary in size depending on the map. Thus, you might wish to improve a certain unit's str to increase their chance of winning massively in a 8-sided dice-map, but you would much rather add more hp or more movement to a different unit to increase its chance of winning in a 50-sided dice game (as its almost a 50/50 who takes the hit, a hit point is about 25 times better than a str point in such a case on a mobile unit working alone, as a surviving unit gets all its hit points restored).

The thing here is that while its random, there is logic behind the madness. I cannot control whether my units do well or not, but I can understand the randomness and how it works - all the mechanics are clear. Its like playing poker to a certain extent. Sure I don't know what cards you have, but I know you have a combination of the 52 in the deck.

Not knowing what charge bonus does (does it affect 1 attack? For 3 seconds? What if you just run into a unit and don't specifically order a charge) or how attackskill works vs defense+armor (why do Katanas beat Naginata Samurai when the Naginatas have more attack+defense+armor - is it attack being more effective, defense or armor being less effective or what gives?) just means Im drunkenly stumbling around in the dark as to what Im doing. If I think attack+defense+armor equals chance to win a direct melee fight, but your Katanas humiliate my Naginata, it feels like you just pulled out a joker and beat my full house with a "lolapallozza"-combination that works in this particular casino. I wouldnt know what just happened or why it happens consistently (aka not as a function of randomness). Such things bother me :p

But until someone finds out how these things work I will settle for improving armor on anything that cant shoot. Except maybe Yari Cavalry... that depends whether or not charge is any good, but since I dont know I guess giving them more +attack would make them significantly better against enemy samurai archers and katanas that wont easily break and rout due to morale shocks.

[faker01]

I always want my guys to have the greatest staying power. In game terms that means high morale and armor. I don’t much care how long it takes me to kill them as long as the other people quit fighting first!

When the other have to quit first, wouldn't it make sense to try to balance the stats?

Not knowing what charge bonus does (does it affect 1 attack? For 3 seconds? What if you just run into a unit and don't specifically order a charge) or how attackskill works vs defense+armor (why do Katanas beat Naginata Samurai when the Naginatas have more attack+defense+armor - is it attack being more effective, defense or armor being less effective or what gives?) just means Im drunkenly stumbling around in the dark as to what Im doing. If I think attack+defense+armor equals chance to win a direct melee fight, but your Katanas humiliate my Naginata, it feels like you just pulled out a joker and beat my full house with a "lolapallozza"-combination that works in this particular casino. I wouldnt know what just happened or why it happens consistently (aka not as a function of randomness). Such things bother me :p

It seems a Katana has a bonus against infantry and a yari has a big bonus against cavalry.
I also don't like not to know, how everything is calculated, but my tests seem to confirm, that one factor of the chance to kill in mellee is :attack/(def+armor) and for ranged 1(unkown factor(bow?)+armor)

[Jarmam]

It does seem like Katanas have a bonus against infantry, but what's weird is that the units with bonuses against cavalry have that listed as a specific individual stat - if Katanas have the same vs infantry, why isn't this stat listed? Is it hidden or integrated in how attack skill works vs defense+armor? It could be (edit: No it couldn't apparently) - I can give nothing but a half-qualified guess.

Argharhghahraha, ignorance is bliss? Who said that!?

Edit: I went and tested Katanas vs Naginatas and the results baffled me - especially vs Monks

The Katanas and the Naginatas broke completely even during the charge (within 2 kills of eachother) in all 3 tests I ran and all three times the Katanas would slowly crawl away in terms of kills until they were about 10-12 ahead of the Naginatas at which point the Naginatas gave up and ran with their remaining 55ish guys against the 65-70 Katanas. Expected results despite the Naginata having superior combat stats at first glance.

The monks, however, I did *not* expect to lose vs the Katanas and they didn't... exactly. They won by morale alone. They won the charge by about 8 guys (7, 8, 8) but then started losing troops despite having more 2 attack skill than Katanas as well as 1 more defense+armor. Only when Warcry was ready did they turn the tables around after the Katanas had more than equalized (being ahead by around 5 guys at the time of Warcry in all three scenarios). At Warcry the monks pulled away again and ended up with 40, 40, 39 against 35, 33, 38 Katanas. And lo and behold, when Warcry ended on the Katanas they started catching up... and catching up... and you guessed it, they actually overtook again. The Katanas do not have 15 morale like the monks, though, so with a headcount lead of 3, 2, and 6 their remaining 24, 21 and 23 guys gave up and ran despite the monks having only 21, 19 and 17 guys left.

IGNORANCE! I DONT GET IT!

Edit 2:

I tested again with me having the monks as to avoid Warcries and it's a travesty to say the least. Three times I ran it and the results make no sense to me:
Breakeven at charge (20 died from each) = 46 Katanas rout the 32 Monks
Monks win the charge by 10 guys = 32 Katanas rout the 19 Monks
Monks win the charge by 6 guys = 36 Katanas rout the 24 Monks

[faker01]

Maybe there where 2 hidden stats(percentage, one against infantry, one against cavalry) and it didn't work as wanted, so a 3rd or maybe also a 4th stat was added. Maybe there were no bonusstat for mellee against infantry and it was added later(and not changed in the ui?)
Maybe someone, who reads the files can say more.

[frogbeastegg]

Weapon types do get different bonuses. Katana gets a strong anti-infantry bonus. Yari get a strong anti-cavalry bonus. Naginata gets a mild anti-infantry and a mild anti-cavalry bonus. Thus, as specialised killers yari and katana units are superior, whereas naginata are capable of performing well all-round. If the game didn't set bonuses in this fashion, naginata units would render the rest obsolete.

Naginata warrior monks are set up as shock troops. Katana are designed to kill steadily, and they get a bonus versus the monks. Wave versus rock; if the wave cannot drown the rock swiftly, the rock will outlast it.

[Maltz]

Interesting. Can someone confirm the existence of the "anti-infantry" stats maybe by looking at unit information (as a modder)? Not sure why the developers would include anti-cavalry but leave out anti-infantry.

[Jarmam]

Weapon types do get different bonuses. Katana gets a strong anti-infantry bonus. Yari get a strong anti-cavalry bonus. Naginata gets a mild anti-infantry and a mild anti-cavalry bonus. Thus, as specialised killers yari and katana units are superior, whereas naginata are capable of performing well all-round. If the game didn't set bonuses in this fashion, naginata units would render the rest obsolete.

Naginata warrior monks are set up as shock troops. Katana are designed to kill steadily, and they get a bonus versus the monks. Wave versus rock; if the wave cannot drown the rock swiftly, the rock will outlast it.

We can agree on the necessity of a katana-to-infantryface-stat, but its quite strange and quite unnecessary that they've hidden it.
Also it makes sense to label Monks as shock troops simply through their terrifying Warcy ability, but Katanas bring a bigger charge bonus and lower defense stat (even including their innate 5 armor) - its simply impossible to know what the design philosophy and effective role is by looking at their stats as Katanas scream shock trooper more than Monks do, which really beckons the display of their assumed innate bonus vs infantry. At a glance I would think the Naginata Samurai as the slow killer and the No Dachi as the insane man's version of Katana Samurai for hitting the right spot hard enough.

Also the Naginata Monk vs Katana charge results baffle me again now that I think about it... why do the Monks consistently win or at worst break even if they have 3 less charge bonus? If charge bonus is a simple +attack skill for x seconds, x attacks or first attack one would think the Katanas would either win through their +infantrystat or at the very least break even with the Monks - but they don't. What gives? I am really getting curious as to what exactly charge bonus does.

[feelotraveller]

My betters will correct this inferior post. Please forgive an old beggars musings...

From memory charge continues until the man (or is it a certain proportion of the unit?) stops winning combat/pushing the opponent back. Morale effects attack in combat and drops with fatigue and loss in battle (amongst other factors).

So the Warrior Monks come in with elevated morale and continue charging longer than the Katana?

I've probably got it all wrong.

On the wider topic my suspicion is that we are meant to learn from playing the game rather than trying to game the play... if that makes any sense. Doesn't stop me trying to, though.

As to personal preference I tend to go with Armour except for bows where I'll go Accuracy although I'm torn on the latter... if rich/slow I consider having different set-ups in different provinces... armoured archers up front backed up by accurate archers, for example

[faker01]

When I have tested high-end katana cavalry, I have expected charge would be much less effective. I have testet it with very big units sizes. When the cavalry charged into a weak unit (80 strong vs 200 weak units) it got about 100 kills at impact. after that the killspeed still was fast.(but much lower)
It seems that, that are to many kills for a simple bonus on attack, especially when I didn't use a appropriate formation. I had no experince with formations, so I simply used wedge-formation.
I think it gives a bonus to attack, but that can't be all.

[faker01]

I have testet it again. One time with yari cavalry(because of the high charge). They where not effective(even with wedge formation). They got about 3 times more losses. I tested it one time with katana cavalry with standard formation. This time they also were not very effective. And I tested it one time with katana cavalry with wedge formation. They breaked the other formation in the middle and overrun nearly everything there, while with normal formation the charges where ineffective. And because they where in the middle of the other formation, they couldn't be flanked(nice logic, btw) except from back, while other units still could charge the splitted unit.

[GoldenToad]

I always go for armour as I remember reading or hearing that armour increases exponetially therefore in my eyes it is much more useful as it keeps troops alive longer. Furthermore, less is more in regards to armour.

N.b. I can't be one hundred per cent sure that this is the case but I seem to think that it is. Hence the viability of a melee general in multiplayer.