Total War: Three Kingdoms Discussion Thread

Discussion of other Three Kingdoms games (e.g. Destiny of an Emperor and Dynasty Tactics) and other games by Koei (e.g. Samurai Warriors).

Re: Total War: Three Kingdoms Discussion Thread

Unread postby Qin Feng » Fri Jan 25, 2019 10:21 am

Ideally they will add unique models to generic famous officers like Xun Yu later down the road. And not in a DLC I hope :lol:

The naval battle excuse is lame. I thought it was going to be because the rivers were made to small on the campaign map so there was no room for that, which could have been lame too, but more understandable. Same as above, being optimistic about what they said and the whole "we prefer to focus on land combat" means they'll add them down the road, which is strange since previous total war games had them by default anyway and next they'll charge for a naval battles DLC
User avatar
Qin Feng
Assistant
 
Posts: 125
Joined: Sat Apr 14, 2018 9:18 pm

Re: Total War: Three Kingdoms Discussion Thread

Unread postby Gray Riders » Fri Jan 25, 2019 2:10 pm

I did find it odd that after Liu Bei got Tao Qian's territory and officers he didn't have Mi Zhu or Mi Fang. I guess they don't start out with Tao Qian.
On the other hand I think the randomly generated officers they seem to have is something Koei should adopt to help out the later scenarios when officer numbers start dipping on the low side.

zirroxas wrote:https://youtu.be/38s3oIEKFqU?t=831

In a hilarious unscripted callback to the Shogun 2 opening movie, Guan Yu kills Lu Bu in a duel outside a fort, only to be shot in the back with an arrow seconds later and die...

......but not really, as both only ended up wounded on the campaign map.

(Legendary heroes have a confirmed higher chance of not dying in combat than non uniques)

Have we actually seen Lu Bu win a duel yet? He also lost to Xiahou Dun in an older video, and his class is Vanguard, which says they're weak to heroes.

Depite CA making the joke on reddit it looks like you can pursue Lu Bu as much as you want.

Dong Zhou wrote:Xiahou's defect so easily seems wrong, I had liked what they showed of the Guanxi system till they did that.

I'm not too surprised, as I had a feeling they'd make characters with very strong loyalties less reliable in the game to facilitate the officer micromanagement the game seems to be putting a large focus on (so you can't play as Cao Cao and just ignore the Xiahous or Cao Ren's loyalty).

I still suspect the latter era characters will be fewer, focus on initial game and initial FLC will be towards earlier era

With the game starting in 190 I suspect officers like Jiang Wei won't see much use, with the game being almost done when they appear.

*
I'm wondering if you can add units to captains so you can recruit into garrisons again? We really need that feature back.
Gray Riders
Scholar of Shen Zhou
 
Posts: 2174
Joined: Sat Jul 17, 2010 1:02 am

Re: Total War: Three Kingdoms Discussion Thread

Unread postby Qin Feng » Fri Jan 25, 2019 2:54 pm

Vanguard Lu Bu is a rather strange choice. I mean, yeah, color red = vanguard, but he is famous for being the strongest man in all the land. Kind of like the type of person a champion should be :lol:
User avatar
Qin Feng
Assistant
 
Posts: 125
Joined: Sat Apr 14, 2018 9:18 pm

Re: Total War: Three Kingdoms Discussion Thread

Unread postby Gray Riders » Fri Jan 25, 2019 3:05 pm

Qin Feng wrote:Vanguard Lu Bu is a rather strange choice. I mean, yeah, color red = vanguard, but he is famous for being the strongest man in all the land. Kind of like the type of person a champion should be :lol:

He was also great at breaking lines, though, and champions are bad against troops.
I feel the class system was not a great idea because there are many prominent characters that don't really fit into it.
Gray Riders
Scholar of Shen Zhou
 
Posts: 2174
Joined: Sat Jul 17, 2010 1:02 am

Re: Total War: Three Kingdoms Discussion Thread

Unread postby zirroxas » Sat Jan 26, 2019 2:55 am

A couple more random pieces of information that I've stumbled across recently:

  • The sliding modifier in provinces this time around is allegiance (like Thrones), rather than religion or culture. Commanderies will have shifting allegiances to different factions that have effects ranging from public order, to income, to replenishment rate.
  • Certain buildings requiring strategic resources is back. Example is the tea parlor needing tea (big surprise) and the blacksmith chain requiring iron and industrialists.
  • Places outside the battle map are visible in the distance (e.g. if you fight next to a settlement, you can see it just beyond the red line of death), so it seems that maps are generated for each tile rather than made individually.
  • Units are supposed to take a fatigue hit when using grappling hooks to climb walls. I have not yet seen the unit tooltip noticeably change however, though it only popped up long enough for me to read a couple times though.
  • You can place map pins! They have different symbols! You can give them funny names! We all know this is the most important thing ever!


Also, a bit of some personal commentary.

There's been some consternation regarding the battle AI in the game, with the words "completely broken" being thrown around in some circles, and lo and behold, the epicenter of it is everyone's favorite alt-right idiot, ArchWarhammer. You might know some of his works as the Rome 2 Female General controversy and "Han China didn't have cavalry," nonsense.

I'm not one to claim that CA's AI track record is anywhere near excellent, and I have incredibly bitter memories of being burned by Rome 2 on launch. However, Arch is usually wrong, so I decided to take a look. I dug back through as much of the battle footage as I could find to form to form my own opinion on the matter, paying special attention to the AI's movements in each engagement.

Here are a couple of videos with good battle footage, though I used basically every video I could find:

https://www.youtube.com/watch?v=uHKAx95vufc
https://www.youtube.com/watch?v=fn_h-N23iRs
https://youtu.be/n9EzBuFN5VU?t=796
https://youtu.be/3N55nSVlWec?t=953

For the most part, I think the AI is functional, and even has some improvements, but there are some very definite hangups that I think the AI is stumbling on, though none of them seem to be problems with the underlying system and more just implementation issues.

With the fact that we're dealing with low level armies with deliberately overpowered heroes in mind, the open field AI does mostly resemble Warhammer's, which is mostly alright. It'll try to engage most of the units you have, and then use its reserves (if it has any) to overwhelm something. It's ranged units attack anything that's out in the open, and the cavalry does a mixture of things depending on how much room is available. It's nothing spectacular, but barring the player executing some really strange maneuvers, it won't do the Thrones thing where it meaninglessly reshuffles its line. It's also not totally cognizant of the unit strengths and weaknesses, but it doesn't have to be to put up a fight. While it lost mostly every battle, but it was also almost always outmatched, and only did a couple things I'd say were just outright stupid (mainly regarding hero use).

In the case of siege battles, CA's said back around the first gameplay video we ever saw (siege of Xiapi) that the AI is supposed to withdraw further into the settlement when it thinks its lost a particular part, rather than just fight to the death, and that seems to be the case. You can see that usually, the Ai will draw its units back towards the settlement center as time goes on.

The main issues tend to arise here. Firstly, the AI abandons its outer defenses way too easily. It may have still had a chance to hold, or at least exploit the walls and gate, but began fleeing inside almost immediately after the player drew close. It also does the thing that the field AI has avoided and is constantly moving units around in an inefficient way, which diminishes its combat ability. It's not quite as bad as in Thrones, but it's still noticeable that it barely is able to put up a fight for the walls.

The good news is that as it draws back, it's able to recognize and use chokepoints appropriately. The bad news is that it still changes its mind too quickly and will constantly try to reposition after only a small time fighting even if the battle isn't going that badly. This is usually caused by the situation that the AI is reading changing so fast that it keeps jumping between tactics, which can be remedied remedied by either slowing down its reevaluation or just increasingly fuzzing the states that its in to prevent it from changing plans so quickly.

It does create some decent last stands, and its certainly a better use of the sprawling cities than the nothing but wall fights that previous titles turned into. In Turin's video, the detachment that was defending against Zhang Fei drew back to defend another gate rather than fall back to the center immediately since there were troops on both sides of it. However, when Zhang Fei circumnavigated them, they began slowly falling back towards the center to help out.

There are other issues. Some of the unit assignments in sieges are off (using only melee units to defend the walls while keeping the archers on the capture point), it still hasn't grasped that a unit in a formation isn't quite the same as a unit out of formation, and some units get stuck guarding an approach the player will very clearly not use. However, these aren't as critical to the AI's capabilities as it's need to actually commit to a plan for more than 10 seconds.

At the end of the day, I can usually see what the AI is trying to do, which is definitely above Empire (where the AI is just completely illogical) and Rome 2's launch (where the AI was actually incapable of doing battle and would sometimes even just spin in place), and even a lot of other Total Wars (at the very least, when it charges its general into a line, I understand why). There's still some major issues to iron out, but they're not something that's prohibitively hard to fix. After that, the only thing that would need help is balancing. The system is pretty clearly made for battles that last longer than this.

TL;DR The AI isn't utterly broken. It's the Warhammer AI (which is mostly functional) that's just having a bit of trouble pulling off the new tricks it was just taught.
I don't underrate the value of military knowledge, but if men make war in slavish obedience to rules, they will fail. - Ulysses S. Grant
User avatar
zirroxas
Initiate
 
Posts: 74
Joined: Sun Dec 02, 2018 5:05 am
Location: The Lands of Always Drizzle

Re: Total War: Three Kingdoms Discussion Thread

Unread postby Gray Riders » Sat Jan 26, 2019 3:41 am

zirroxas wrote:[*]Units are supposed to take a fatigue hit when using grappling hooks to climb walls. I have not yet seen the unit tooltip noticeably change however, though it only popped up long enough for me to read a couple times though.

I suspect this like in Warhammer where since fatigue drains over time the first troops that get over are fine but once the whole unit is up they're usually very tired.

There's been some consternation regarding the battle AI in the game, with the words "completely broken" being thrown around in some circles, and lo and behold, the epicenter of it is everyone's favorite alt-right idiot, ArchWarhammer. You might know some of his works as the Rome 2 Female General controversy and "Han China didn't have cavalry," nonsense.

Han China didn't have cavalry? Did he make an entire video claiming that?

With the fact that we're dealing with low level armies with deliberately overpowered heroes in mind, the open field AI does mostly resemble Warhammer's, which is mostly alright. It'll try to engage most of the units you have, and then use its reserves (if it has any) to overwhelm something. It's ranged units attack anything that's out in the open, and the cavalry does a mixture of things depending on how much room is available. It's nothing spectacular, but barring the player executing some really strange maneuvers, it won't do the Thrones thing where it meaninglessly reshuffles its line. It's also not totally cognizant of the unit strengths and weaknesses, but it doesn't have to be to put up a fight. While it lost mostly every battle, but it was also almost always outmatched, and only did a couple things I'd say were just outright stupid (mainly regarding hero use).

Warhammer's field AI had several major flaws but it's functional, so this isn't terrible news.

In the case of siege battles, CA's said back around the first gameplay video we ever saw (siege of Xiapi) that the AI is supposed to withdraw further into the settlement when it thinks its lost a particular part, rather than just fight to the death, and that seems to be the case. You can see that usually, the Ai will draw its units back towards the settlement center as time goes on.

The main issues tend to arise here. Firstly, the AI abandons its outer defenses way too easily. It may have still had a chance to hold, or at least exploit the walls and gate, but began fleeing inside almost immediately after the player drew close. It also does the thing that the field AI has avoided and is constantly moving units around in an inefficient way, which diminishes its combat ability. It's not quite as bad as in Thrones, but it's still noticeable that it barely is able to put up a fight for the walls.

Sieges are where I was worried. Warhammer's siege AI was borderline non-functional to the stage that some factions have an easier time winning a siege assault than a field battle against the same army. This at least sounds like an improvement over that.

I like a lot of Throne's ideas (and it helps that it's similar to the setting of Viking Invasion, which I loved) but it's AI, yeeesh. The things I've pulled off with Mide's javelin cavalry is sad.

Thanks for the evaluation.

Edit: Empire building with Ma Teng
https://www.youtube.com/watch?v=F4Y4RvthA_M
He has Qiang troops as special units.
Rush-building (pay extra for instant construction) is now in.
Gray Riders
Scholar of Shen Zhou
 
Posts: 2174
Joined: Sat Jul 17, 2010 1:02 am

Re: Total War: Three Kingdoms Discussion Thread

Unread postby zirroxas » Fri Feb 01, 2019 6:46 pm

Gray Riders wrote:Edit: Empire building with Ma Teng
https://www.youtube.com/watch?v=F4Y4RvthA_M
He has Qiang troops as special units.
Rush-building (pay extra for instant construction) is now in.


Nothing particularly new I see. At least Qiang units are in the game, even if there are no Qiang factions. Having "early access" seems to imply that other people can get later access, but it could be meaningless too.

The unit cards are more readable now, but I wish the kept those counting rods.
I don't underrate the value of military knowledge, but if men make war in slavish obedience to rules, they will fail. - Ulysses S. Grant
User avatar
zirroxas
Initiate
 
Posts: 74
Joined: Sun Dec 02, 2018 5:05 am
Location: The Lands of Always Drizzle

Re: Total War: Three Kingdoms Discussion Thread

Unread postby LiuBeiwasGreat » Fri Feb 01, 2019 10:02 pm

This video was rather nice, like zirroxas says it doesn't really show us anything revolutionary but it is still nice to see a bit of campaign perspective from Ma Teng.

I for one look forward to playing Ma Teng and helping my good friend Dong Zhou restore peace and prosperity to the Han Empire. :lol:
"If you can't drink a lobbyist's whiskey, take his money, sleep with his women and still vote against him in the morning, you don't belong in politics."
LiuBeiwasGreat
Scholar of Shen Zhou
 
Posts: 2663
Joined: Mon Mar 22, 2004 4:13 pm
Location: Raleigh, NC

Re: Total War: Three Kingdoms Discussion Thread

Unread postby Qin Feng » Sat Feb 02, 2019 6:14 pm

I like how there seems to be some more thought put into campaigns. Surprised seeing the AI actually offering vassalage, I thought it didn't have much use in Shogun 2 since having vassals basically means another open front.

Also, Pang De is a unique officer with a generic appearance. Maybe they'll add unique models later down the road, and hopefully they'll be free :lol:

Also, in some days they'll start a series of streamings. I'm looking forward to that since there are some concerns about battles and CA hasn't really shown much of them other than letting other people release footage of some presumably early build.
User avatar
Qin Feng
Assistant
 
Posts: 125
Joined: Sat Apr 14, 2018 9:18 pm

Re: Total War: Three Kingdoms Discussion Thread

Unread postby Gray Riders » Wed Feb 06, 2019 1:59 pm

Gameplay livestream today:
https://www.twitch.tv/totalwarofficial
3PM GST, 7AM PST, 10AM EST.

Vote for which of the Yuans we want to see played.
Gray Riders
Scholar of Shen Zhou
 
Posts: 2174
Joined: Sat Jul 17, 2010 1:02 am

PreviousNext

Return to Other Three Kingdoms and Koei Games

Who is online

Users browsing this forum: No registered users and 2 guests

Copyright © 2002–2008 Kongming’s Archives. All Rights Reserved