AI behavior in 1.28

[GShock112]

Alright, the AI is very solid now.
At the Taisho level the match is even and so far 5-4 for me, the situation is VERY balanced (I only captured Osaka) and I can see a 1100 vs 1200pt clash coming in Gifu. At the Daimyo level, the AI will be looking for a higher advantage in points and that is a challenge orders of magnitude above what I can take right now (previous test 4-1 for the AI). If you remember the previous statistics you can tell everything has changed.

I have a few technical questions whose answer I saw in the detailed engagement reports but that are disabled now and I figured out I'd better ask for clarifications:
1) the unit size modifies the % of success AND the morale checks (bigger unit more likely to pass a check because the men perceive they can survive the engagement)?
2) are horses affected by missile (horse easier to hit)?
3) is the side the volley is coming from affecting the casualties taken by a unit (enfilade fire or fire in the back kills more and impacts more on morale)?
4) can a general be killed by missile fire?

In regards with the AI behavior I've seen teppo skirmishers and even archers sometime LURE me by appearing detached from the main body... I rushed them twice and I lost twice 2 units to rout those skirmishers. That never happened before so I guess it's intended and I love it. I have already pointed out in the other thread the "retreat" annoyance... I hope something useful can be done with a retreating unit so it doesn't come back unless it's ready to fight.

Finally: please... add the weather. We NEED those important changes to strategy according to the seasonal weather effects (muskets not firing, cavalry charging less effectively, etc).

Great, just great patch indeed!

[rbodleyscott]

I have a few technical questions whose answer I saw in the detailed engagement reports but that are disabled now and I figured out I'd better ask for clarifications:
1) the unit size modifies the % of success AND the morale checks (bigger unit more likely to pass a check because the men perceive they can survive the engagement)?

The trigger for taking a cohesion test from shooting is 5% losses in the turn, and 8% losses gives a -1 modifier. Obviously larger units can take more casualties before these thresholds are reached. (The calculations are in fact slightly more complicated than that, there is a size limit beyond which extra men do not count when determining % losses, but the principle applies).

2) are horses affected by missile (horse easier to hit)?

For bows this is so. For firearms not, because the rate of fire being so slow makes it harder to hit cavalry who (in reality) don't stand still long to be hit. Crossbows also get a bonus against cavalry, because the Chinese rated them highly against nomad cavalry and sometimes used magazine loaded repeating crossbows.

3) is the side the volley is coming from affecting the casualties taken by a unit (enfilade fire or fire in the back kills more and impacts more on morale)?

For artillery fire yes, enfilading fire (from the flank) causes more casaulties, because there are more men in a rank than in a file. (In the Italian Wars allegedly on one occasion 30 gendarmes were killed by one enfilading cannon ball).

Fire from the rear does not have this effect because once again it is firing down the files.

Therefore fire from behind the flank but not more than 45 degrees from the flank counts as enfilading.

4) can a general be killed by missile fire?

No, it would be too easy to concentrate artillery fire on generals.

[GShock112]

Yep I understand the problem of generals being easier targets for concentrated missile fire but that's exactly how it should be.
The AI is no longer idle, it attacks in coordinated fashion and you can't target a general easily with a concentration of missiles on the same target because most enemy units will be committed to fighting hand to hand and these units can't be targeted. Furthenmore... without the AI leaving isolated units idle it is virtually impossible to achieve a kill on a particular man (the general in our case) because if you do concentrate on that general's unit, the other enemy units will crush you in melee since your units ARE targeted and get weaker.

I think you should add the possibility of a general being killed or wounded by missile. Considering his bodyguard will take most hits, a separate % check with minimal chances of being killed is A LOT BETTER than having generals not killed by missile at all.

[rbodleyscott]

We will have to agree to disagree on this. While it is true that historically generals could sometimes be killed by distant shooting, not having them killed by shooting is a deliberate design decision, because it is easier for the player (with helicopter view of the battlefield and exact knowledge of the enemy general's position) to concentrate shooting on the general's unit than it would be in a real battle. Moreover, I don't know whether this was done in Japan, but in the west generals sometimes used decoy lookalikes to make it harder to target them.

[GShock112]

The opposite is true, most generals wore family heirlooms in battle and very visible armor to be recognized by friends and enemies, the big generals also had mons on their position so yes, the tactic was used and yes the head of a general was prized and rewarded by the winning daimyo so they TRULY were favorite targets for missile fire.

Out of the many examples, the most notable is Takeda Harunobu (Shingen) as well as the leader of the bakufu army besieging Shimabara whose name right now I don't remember (they were both picked off by musket fire... we also have movie examples with Darius the great shot by Temistocles and ah... the good of the 3 sons of the Daimyo of Kurosawa's RAN, the one dressed in blue...as well as Katsumoto from the Last Samurai movie. I can't relate to the latter because they are fictional and I don't know if Darius historically died to an arrow shot... we could press this further to the american civil war, the second of my historical expertises [in fact I'm very surprised I wasn't taken in for the Betas of slitherine's upcoming ACW game since I betatested every single ACW game ever released in the last 20 yrs]) but that's exquisitely for gunfire, of course all the generals who died... died from gunshots.

Your design choice is coming after being scared of generals being killed but generals DID get killed very frequently and were primary targets for missile fire. Just add the possibility and make it very slight because of the bodyguards. Also mind that generals obviously wore better, much better armor than lower rank samurai so it all concurs into doing that possibility and, taking into account bodyguard and improved armor, making it very slim.

Finally... that's another reason why there should be weather. On windy-rainy-snowy days missile fire is a lot less accurate and muskets could barely be used at all on rainy days.

[rbodleyscott]

The opposite is true, most generals wore family heirlooms in battle and very visible armor to be recognized by friends and enemies, the big generals also had mons on their position so yes, the tactic was used and yes the head of a general was prized and rewarded by the winning daimyo so they TRULY were favorite targets for missile fire.

I am aware of all this. Nevertheless, the format of the game makes it more easy to pick on them than it was historically.

Therefore, if there was a chance it would have to be very small.

[GShock112]

It was very small indeed.
The chances of hitting one particular man in a hunded is 1% IF the unit is hit... but for that man we're talking of a 66% armor (not 33%) factoring the role of the Hatamoto who took the arrows first.
Besides, I think you're underestimating the new AI. It's very very very hard to concentrate missile on one unit now because it advances in mass.