Considerations on the Tenka Fubu Campaign (Takeda Side)

Sengoku Jidai: Shadow of the Shogun is a turn-based tactical and strategic game set during this turbulent time; primarily focusing on the Japanese Warring States period and Japanese Invasion of Korea. Other armies from East Asia are also made available to simulate different conflicts across the region.
Post Reply
GShock112
Administrative Corporal - SdKfz 232 8Rad
Administrative Corporal - SdKfz 232 8Rad
Posts: 169
Joined: Fri May 27, 2016 4:44 pm

Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by GShock112 »

I was able to merge my forces for a full scale attack in Gifu with all my combined armies against the AI's combined armies (15.000 vs 12.000) from which I emerged victorious by routing 40%+ and only losing 9%. After a single battle the Oda clan took a huge amount of casualties and its ambitions are already resized (this is still the Daimyo difficulty level).

Image

All my previous considerations about the strategic choices adopted by the AI are in place, Oda is now supposed to be on the defensive and it's being faced by a vastly superior army concentrated in Gifu. The AI's reply to this stunning major defeat is to keep its forces split and go for a siege in Ueda which will never succeed since it's going to be repelled by one of my soon-to-be-split forces much sooner than the Ueda garrison falls.

Oda loses to attrition and can't conquer Ueda. In some circumstances a decoy attack may prove useful to delay an assault on the Oda provinces but the defense of these provinces should be paramount.
In this case I won't need to rescue Ueda and I can split to beat the Oda army quite easily in Anotsu AND besiege another province.
Oda is going to be on the "mouse" with its only combat-effective army and have 2 of its home provinces besieged so the AI is adopting a losing strategy.

On the tactical side, instead, what I did observe in this major battle is the behavior of the AI with the flanking manouver, which is something I already pointed out in the other thread.
The AI engaged in hand to hand too soon and barely used its firepower at all. I concentrated my fire on the weaker units (sohei have no armor) before they arrived to contact and easily defeated the first line soon thereafter.

The second line was caught on its side by my pursuers (in this case I was lucky as I caught them and they didn't catch me) and the battle was a cakewalk from turn 1 to its end.
What caught my attention was this:

Image

The AI sent 2 Teppo units on a very very far flanking manouver and they never entered battle at all.
GShock112
Administrative Corporal - SdKfz 232 8Rad
Administrative Corporal - SdKfz 232 8Rad
Posts: 169
Joined: Fri May 27, 2016 4:44 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by GShock112 »

After beating Oda so badly, I could afford to split my forces again and decided to stop the siege at Ueda because 11% success was more I wanted to risk.
Oda was fortified but despite being weaker (it should have retreated... well... actually it shouldn't have tried to siege Ueda after that catastrophe in the first battle) it didn't and I totally slaughtered them.

Strategic priority for the AI should be to preserve the armies. This means if you are defeated you go defensive, you don't waste further manpower to do a siege you can't win. It's the winner that goes to siege and during the protracted siege time the defender reinforces and entrenches.

I think this was the very first time I met fortified enemy so I am showing you my battle plan and how the AI reacted.

Image

The already outnumbered AI left its fortified units completely idle for the whole battle so it was doubly outnumbered as I didn't even DREAM of attacking them. The yari ashigaru on the bottom let my ashigaru slip beside it without intervening (my unit was disrupted by the archers) and I could easily reach the Honjin and kill him.
In the next turn, as the routed % went to 59% they moved. The ashigaru unit up came to the center and the 2 yari units instead turned round to finally intervene. The battle ended immediately, they only had enough time to turn.
The Honjin was not rescued by the Oda cavalry and the other cavalry was pinned by my own cavalry (which lost the general) and flanked by my ashigaru unit.

From the beginning to the end the 2 idle units behind fortifications on the hill were bombarded by my missiles. What the AI did right instead was rushing out with the teppo and evading afterwards. The teppo did rout one of my ashi but were finally caught... we were just too many and those 3 idle units took half of the enemy away from me. I even willingly put my ashigaru with the side facing the yari samurai but they didn't charge. Totally idle.

This battle shouldn't have been fought because that battered Oda army shouldn't have rushed forward to siege Ueda but the way the AI played is bad. A possible solution is related to unit numbers: when an army is so small all the fortifications should be close and not separated.

So the situation is pretty dramatic for Oda and we're just in Early Summer 1570.

Image

One more defeat and they're gone. They just don't have the numbers to face the army in Ueda now and I could even split my 617 and go for 2 provinces at the same time. The only problem is with the auxilliary forces. I've had the impression the Oda were never a match so far from the first turn unlike the Takeda which made "mistakes" but hanged on till the last.
GShock112
Administrative Corporal - SdKfz 232 8Rad
Administrative Corporal - SdKfz 232 8Rad
Posts: 169
Joined: Fri May 27, 2016 4:44 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by GShock112 »

After careful consideration I decided my 409 Army in Ueda had no numbers to face the Oda 125 Army in Okazaki because of the strong auxilliaries there and my main 617 Army couldn't face the 422 Oda Army in Antoku either for the same reason. It seemed logical to wait and see and only then, push on the offensive. Oda switched and I faced the 422 + auxilliaries in Okazaki for a difficult battle in an agricultural area, which means rivers.... the main river in the CENTER of the map, cutting the battlefield in 2 and wherever one side had to cross would face the other.

I was outnumbered but mounted my defense line just behind the river. My previous considerations on the AI not using missiles and rushing too fast wouldn't have applied here since the Oda were inferior with missiles... Oda rushed to the river, which is what it should have done being inferior in missile power but this brings me to another consideration about the campaign system:

Behind attack and defense there's a map, there's a strategic situation... there are other armies and there's winter with its pause approaching. Regardless of who's the attacker and who's the defender, this is Oda's province so he MUST be the defender. There are no victory points here and the side that routs the other wins the battle. I have never tied (Once I did survive a fortified defense map because nightfall saved me but it was in my previous campaign) but I suppose if neither side wins, both sides bag their losses and the defender stands in its province.

Since this is the situation, the logical conclusion is that Oda shouldn't have attacked, despite its superior melee numbers (more numbers but of scarce quality vs river defenders with strong missile) and should have waited for me to cross the river (as Sun Tzu teaches with half of my forces before attacking) since I was the one who was trying to capture the province. Essentially the way the AI played tactically was right (rush melee to the river) but it was not the right situation in which to do it because Oda was not on the offensive but on the defensive here.

It was inconsiderate but it almost worked: I lost my entire left flank: 3 ashigaru units couldn't stop a seriously bombed yari samurai unit. It routed all three units and I slowly lost the center but I won on the right and swooped back hitting the 60% rout requirement having lost 52% myself. These numbers are enough to show despite the AI bad planning (and bad execution: Teppo again), it was a very close battle that could have ended either way: 1 rallied unit or 1 turn left instead of right by one of my cavalry pursuers and the battle would have been lost.
Do not forget the side who crosses over fights another side whose honjin's commanding presence is felt. Seriously, I did my very best and I think I pulled a miracle but I had had to be the one to cross the river I surely would have lost and this is what I'd like the AI to have done considering the strategic situation.

The AI Teppo AGAIN left to play their own game and reappeared almost on my Honjin when it was too late. The contribution of those teppo would have come right on the side I was winning and it would have really made a difference. Instead... they just didn't play at all. This is the situation now.

Image

... and I keep getting in the back of my head the same question: if I didn't know that by attacking that particular unit I would break it instantly or almost instantly, would I attack it and take the risk?
If I didn't know I would lose the first round but I'd still have time to give my unit a support would I risk the attack? :roll:
rbodleyscott
Field of Glory 2
Field of Glory 2
Posts: 28284
Joined: Sun Dec 04, 2005 6:25 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by rbodleyscott »

If you choose the "minimise tooltips" option you won't see the win:draw:lose chances.
Richard Bodley Scott

Image
GShock112
Administrative Corporal - SdKfz 232 8Rad
Administrative Corporal - SdKfz 232 8Rad
Posts: 169
Joined: Fri May 27, 2016 4:44 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by GShock112 »

This is the new situation here.

Image

I did take Hamamatsu but Oda fell back from the siege in Gifu to attack me at Okazaki and, thanks to the auxilliaries, they defeated me and defeated me big time.
The AI sent the usual 2 Teppo far away... then their cavalry destroyed my right flank... took the same route and kept going until they actually arrived to the edge of the map... by the center, behind my Honjin. The Teppo instead cut back sooner to converge on the center where the battle was already lost to me.

Had the 3 cavs attacked the center immediately or had those 2 Teppo participated, I wouldn't have stood a chance. Thanks to this behavior, which I hope to see corrected, I stood HALF a chance. :-)

In the end 6500 for Oda and 4800 for me. Too great numbers but I made one terrible mistake with an ashigaru unit that got jammed in the woods, disordered and couldn't kill that one AI unit I needed to kill to regroup on a hill on the left flank. I think I would have lost anyway because of the numbers but also because I'm not yet familiar with the Takeda army. Well... you can tell how much I love this game by the way I enjoyed losing a few minutes ago.

Big campaign is coming in the new year. :-)
GShock112
Administrative Corporal - SdKfz 232 8Rad
Administrative Corporal - SdKfz 232 8Rad
Posts: 169
Joined: Fri May 27, 2016 4:44 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by GShock112 »

We've spent a year trying to get our armies to engage but again, in the end, we did it... even though I was unable to bear my second army, I did merge the 1st with the 3rd and my combined army gave chase, suffering attrition casualties for lack of supply and, in the end I caught him.

The battle was HUGE and on a very interesting scenario mixed with hills and woods on the sides but with a disappointing evolution.

As I approached the center, the AI charged in a disorganized way and I could shoot down so many units that by the second turn of missile range I had already killed almost 1500 men. Since we were 15.000 vs 12.000 that's already past the 10% losses... (not of much confort because my advantage comes from peasants) as I said, the AI doesn't know when to rush and when to camp and shoot, otherwise it would be a terrific opponent.

Even though there was this major mistake, the AI still flanked with 3 units in the woods on my right side but I was lucky enough to engage its muskets in the woods from outside with my peasants... with the help of a cavalry unit I managed to beat them and it was over. I repeat: I was lucky this time and this strategy was sound but could have never worked the way the camera and sound give away the exact number of units hidden in the woods.

I barely lost 3000 men and Oda short of 9000 (out of 12.000 it's a catastrophe for the AI) but I have to tell you this is one of those battles that could have ended much differently if not for the routing of the enemy troops that affects everyone else nearby. If the AI is taught to rush when it has less missile and camp when it has more a good step will have been taken. The problem here is the player concentrates its aim on the enemy elite troops... the AI can't really protect them if it doesn't attack fast enough but it seems to react only to some sort of positioning. If you stand idle it won't move... if you approach the center it will attack when you are already deployed (not a moment sooner). This means basically the AI is always playing offensive against an opponent who is on higher ground or well organized with missiles. In these cases it should attack in a rush to stop the raining of arrows.

As of the problem with the camera, giving away the hidden units, if you know how many no matter where they are exactly you'll be ready to meet them... which fools the entire plan. Considering how slow troops are at wheeling and reacting a round or 2 with units "lost" in a non-secret attack can be decisive. These small "mistakes" from the AI all add up to its weakness but I want to highlight again that this game is so complex that building a competent AI is already a huge achievement... and this AI IS competent.

Perhaps the best way to handle troops that are not in LOS is to have their sound disabled and their movement not followed by the camera when they start their turn in hidden location.

Anyway this is the situation now... Oda is on the run but it's not over yet.


Image
rbodleyscott
Field of Glory 2
Field of Glory 2
Posts: 28284
Joined: Sun Dec 04, 2005 6:25 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by rbodleyscott »

When I get back to the office I will see if these clues can be coded out.
Richard Bodley Scott

Image
GShock112
Administrative Corporal - SdKfz 232 8Rad
Administrative Corporal - SdKfz 232 8Rad
Posts: 169
Joined: Fri May 27, 2016 4:44 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by GShock112 »

rbodleyscott wrote:When I get back to the office I will see if these clues can be coded out.
Simply put, in the campaign the object is still to rout 60% of the enemy or 40% with 25% advantage.
This doesn't take into account that in the campaign one side is defending a province and the other one is attacking. The AI will try to rout you anyway as soon as you approach the center of the field (unless it's vastly superior in numbers, in this case it comes to you immediately).

I think it would be a good idea to change the rout % so that the attacker routs sooner than the defender (alright boys we can't win anymore here, let's regroup and fight another day) and we know the loss also means loss to attrition during the retreat from the invaded province. If the routing % to achieve are the same, whichever side camps has the advantage... because of the missiles hitting the guys who advance before they get to contact... but we know the attacker can't camp otherwise he'll be defeated (fall back from invasion) and lose men to attrition (draw = strategic defeat).

Also... these requirements seem too narrow... many times I hit the requirement and won but I was in such a bad shape myself I could have lost the following round. This last battle didn't give me that feeling but many others did. Considering it's hard to rally units... considering I've never ended a battle in a draw... there's something wrong here. Too tight imo, in fact battles last too short and you can almost never make up for a bad start, which is wrong also.

Finally... the AI needs to know if it comes to the player it will be skewered by archers. The player concentrates fire on a unit and when it routs, it disrupts the others... so when the AI attacks it must attack in force without hesitation and reach the melee as soon as possible so its melee troops can't be targeted and the AI can shoot the player's missile units. This bad calculation of firepower is what makes the AI use cavalry in the middle or never actually make it with a flank manouver that actually hits the missile troops. If these handicaps are softened I think what is already a very credible AI could become a fearsome opponent. So far, every time I camped I won.
GShock112
Administrative Corporal - SdKfz 232 8Rad
Administrative Corporal - SdKfz 232 8Rad
Posts: 169
Joined: Fri May 27, 2016 4:44 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by GShock112 »

I've lost a big battle by facing an army that was numerically too vast and had to fall back... but in one year of chases nothing changed much. We both reinforced and met again in the Oda capital provinces.

This time 15.000 vs 15.000 promised to be a big battle. Unfortunately the AI started at disadvantage with part of its forces not yet deployed. Despite of that, as I was closing in, it decided to attack me (AI was defending the province), reached out in disorder, got pounded by missiles and when I routed a unit on its flank my pursuer got 2 of its second line from the side routing them instantly. It's only partly luck because the Oda are down on skill after the big defeats they took. Last time their numbers were too high and they pulled it off but this time they didn't.

Image

43% victory and a great disappointment for what promised to be a big battle, possibly conclusive. This is another of those battles where if the AI had waited for me to go I would have taken the missiles and I would have suffered. Really if the relation with the strategic situation is glued to the AI and it will be made to attack only when it needs to conquer a province then defend when it's the player who needs to conquer it, I think at least 50% of the problem will have been solved.
GShock112
Administrative Corporal - SdKfz 232 8Rad
Administrative Corporal - SdKfz 232 8Rad
Posts: 169
Joined: Fri May 27, 2016 4:44 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by GShock112 »

Oda is on the run and his armies are shattered but the auxilliaries could be a problem.
I chased from Gifu to Okazaki and gave battle.... 15.000 to 10.000 because I knew if I was defeated my second army would pick up.

I won because, as usual, the AI left many troops behind and fought too slow. As a general guideline, if the second line of the AI attacked with the first one or, even better, flanked either side... the AI would have won (it would have won at least 50% of the battles it lost). The issue is that the second line is camped behind the first line taking arrows and bullets... and it also takes the routing morale impact from the defeated first line... that's why it can't enter battle properly. It's not flanking but it's too close behind the first line, taking the hits. Not in reserve, not flanked, not rushing in... that's the problem and when your pursuer hits the second line in the flank you know the second line is also on a bad facing. It happens (you have no control over the pursuers) but when it happens it's a catastrophe for the AI.

Aside from that, I noticed 3 cavalry units kept behind, they never entered the fight I don't know why. I had a flanking force from my left far side but it was just 1 teppo and 1 cavalry. One of the enemy cavalry was pinned there and that was enough because it was my strong side.

I won 6% to 40% and Oda retreated to Ueda... as I attacked the campaign ended with victory (6 wins 2 losses).


Here's the final situation.

Image

I can finally move to another campaign. :-)
rbodleyscott
Field of Glory 2
Field of Glory 2
Posts: 28284
Joined: Sun Dec 04, 2005 6:25 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by rbodleyscott »

GShock112 wrote:I can finally move to another campaign.
Try increasing the difficulty level one step. It won't alter the AI but the battles will be tougher because the enemy will have more points worth of troops.
Richard Bodley Scott

Image
GShock112
Administrative Corporal - SdKfz 232 8Rad
Administrative Corporal - SdKfz 232 8Rad
Posts: 169
Joined: Fri May 27, 2016 4:44 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by GShock112 »

I know that but I think the best answer would be to teach the AI the things I wrote in this topic... needless to say, it's a huge task, I know, I really do, but maybe the simplicity of this system makes things easier both from the strategic campaign point of view and from the tactical battle point of view.

The AI, with all its mistakes of this campaign would have still won half of the battles at the Daimyo level. In my 6 victories, many came after my hurt units were still standing despite the AI numerical superiority not because of troop quality but because of the problems I listed (many AI units totally out of the battle for its entire duration).

Increasing the difficulty means fighting even more at numerical and qualitative disadvantage... it's an artificial escape to the problem I never liked.
rbodleyscott
Field of Glory 2
Field of Glory 2
Posts: 28284
Joined: Sun Dec 04, 2005 6:25 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by rbodleyscott »

GShock112 wrote:I know that but I think the best answer would be to teach the AI the things I wrote in this topic... needless to say, it's a huge task, I know, I really do, but maybe the simplicity of this system makes things easier both from the strategic campaign point of view and from the tactical battle point of view.

The AI, with all its mistakes of this campaign would have still won half of the battles at the Daimyo level. In my 6 victories, many came after my hurt units were still standing despite the AI numerical superiority not because of troop quality but because of the problems I listed (many AI units totally out of the battle for its entire duration).

Increasing the difficulty means fighting even more at numerical and qualitative disadvantage... it's an artificial escape to the problem I never liked.
Fair enough, but improving AI and testing the improvement takes time. So, in the meantime....
Richard Bodley Scott

Image
GShock112
Administrative Corporal - SdKfz 232 8Rad
Administrative Corporal - SdKfz 232 8Rad
Posts: 169
Joined: Fri May 27, 2016 4:44 pm

Re: Considerations on the Tenka Fubu Campaign (Takeda Side)

Post by GShock112 »

I still told you that I would have lost half of the battles had I not had the lucky pursuer catch the enemy second line in flanks or back.
It could have gone either way in these cases (lose the pursuer)... the problem seems more related to the AI not being aggressive enough in exploiting its numerical advantage (no real flanking taking place) and not camping when having missile superiority.

The campaign system may come handy in these cases ... when a side is defending it should be on the defensive... which is where missile really hurt. This is lacking in the battle system where the only objective is to rout the enemy, without any consequences being involved on either side being on the offensive in the campaign.

6-2 is not an easy victory... when I get to 7-1 perhaps I can choose the higher difficulty level but for the time being I need to see the mongol units first. :)
Post Reply

Return to “Sengoku Jidai: Shadow of the Shogun”