The 'Suicidal Daimyos' problem?
Hi to all old-Shogun-ers out there,
The notes to the 1.11 patch claim to have fixed the 'suicidal Daimyos problem'. Not so, it seems.
This has happened at least twice in my current campaign. Me as Hojo. Difficulty level 'Normal'.
Lord Takeda invades Musashi very early on, in about the third year. I slaughter most of his cavalry on the opposite bank but then forced to back up my units just a little from my end of the bridge and some of them spill out. So he crosses the bridge to press home the advantage. One of my unengaged YS kill half of his hatamoto and he swivels around so his back is towards me, at which point I send my Daimyo's unit after him and he, err... doesn't get the opportunity to run away. Dies without heirs, lands turn rebel and only have token forces left. Easy takover.
Later on, Uesugi is down to just Echigo and Dewa but has about 4 full stacks. I capture Dewa but he repeatedly lifts the siege each time by using his entire army to force me out (I abandon province). I eventually make a speculative twin attack and catch him out of position in Echigo with just a token force. He lamely retreats to the castle without a battle and dies by the next turn in the siege which followed. The multi stacks in Dewa therefore get no chance to respond and turn into Imagawa forces instead. I expect a long series of attrition battles (time-limit defeats) but having routed the initial units, somehow their reinforcements queue fails to work and 3500-odd just vapourise, with nowhere to retreat to. Wierd.
Later still, I spring a surprise attack on Oda (hitherto allied but he was clobbering Mori up the other end of the map and had to be stopped) who has been waiting in Sado, lacking the strength to take Dewa himself. Not a massive numerical advantage, maybe 800/650 in my favour but his defence isn't going too well and he lunges into a unit of H0 YS, which were hiding in trees and wasting his arrows. I send in a second unit and he's soon surrounded and killed. The rest of them scarper, only to be besieged. Must have been Nobunaga too because again no heirs were left and most of the map turns rebel. Oda left a near-hollow core west of there, allowing territories to fall in quick succession, mostly unopposed. Another lot of multiple stacks just sat there while I took over. Loads of cavalry and well teched-up compared to mine, too. The manual says rebels don't act in a coordinated fashion but TTTT, Rebel AI is just plain dumb.
Once shepherded into two provinces, each with no retreat path, the big stacks similarly vapourised in the battle-routs but one of them did take two attempts. Four stacks force me to abandon Owari but leave Mino empty, so I grab that and I have them surrounded. First attack cost me 900+ casualties and I ran out of both time and reinforcements. My Daimyo is 83 years old and I was trying to get him killed off. His unit is down to one man, and some No-Dachi have him surrounded. He should have bought the farm but that's when the clock ran out. I'd got about 1600 of them (probably the most I'd ever seen in a battle result) but it really spoiled the next battle. The AI failed to condense/rearrange its broken and full-strength units so the second battle (couple of years spent replenishing my losses) was against a selection of cavalry and YS cadres but only about 30 archers which then decided to make a run for it once I'd got a bridgehead. I'd stopped attacking temporarily, I was busy getting them to cross the bridge(s) and setup a formation. None of their reinforcements show up (potentially hundreds of archers, No Dachi and so on) and again 3000-odd disappear at a stroke. Official battle results about 180 of them and 90 of mine. Like I said - wierd.
I think I had over-mobilised to tackle this situation and I now have about half a dozen full stacks of my own. I should be able to finish the campaign without raising too many more troops. Taxes are low and I've had only one or two poor harvests in the last 20 years. If not linked to population loyalty, then maybe it was those years where no territories were taken or battles of any kind were fought? Kind of like saying the game 'rewards' you for battle successes but punishes you if you're inactive for too long. Of course, if you're inactive because you're skint and can't afford enough troops to attack anywhere, it's a vicious circle, if the theory's true.
Re: The 'Suicidal Daimyos' problem?
Suicidal daimyo isn't really fixed even in MI...
Re: The 'Suicidal Daimyos' problem?
I thought the Suicidal General problem was that he would lead towards the start of the battle... practically out front or the first wave of attackers... at least within the first third... the General attacked just like any other unit.
Re: The 'Suicidal Daimyos' problem?
Yes, that's the way it was in the early stages. The AI would engage with the Daimyo/Taisho very early on and often charge into a group of spearmen. It was not unusual for them to rush across a bridge first... ~:rolleyes: It got better with MTW. The AI started doing some more intelligent things like withdrawing the general's unit from the map when he was the last man in the unit (or sometimes a handful.)
A 5+ star general could pose real problems like this when he became a single man unit. He would often start the battle by withdrawing. So you couldn't target him, force him to run, etc. If you routed his first wave, later waves would still have the bonus of valour from being in his army and could crush your weary men. Then you would have to face him again...
The biggest stupid thing the AI would do in MTW was put all the heirs in one army under the king then attack. It was a powerful, small, fast force but it would kill off the heirs quickly even if it won, and a loss was catastrophic.
In STW/MTW the AI used the Daimyo/king/prince as a battering ram since it was so powerful. You were pretty much stuck killing him once he charged into the middle of things, else he would break your center. I often tried not to target the enemy general/daimyo/king since the AI would not be as good about protecting him as I would be--and killing him made the battle/game to easy. That all went out the window when the dude rushed the center though.
Re: The 'Suicidal Daimyos' problem?
Hi Red,
Quote:
Originally Posted by Red Harvest
Yes, that's the way it was in the early stages. The AI would engage with the Daimyo/Taisho very early on and often charge into a group of spearmen. It was not unusual for them to rush across a bridge first... ~:rolleyes: It got better with MTW. The AI started doing some more intelligent things like withdrawing the general's unit from the map when he was the last man in the unit (or sometimes a handful.)
I can't say I've ever seen the AI voluntarily withdraw any units myself, in the sense of enabling elements of the reinforcements stack to come into the battle. More often than not, when they depart the map it's because they're routing (cursor-hover describes them as 'running away'). I'm interested in how you even got the opportunity to see this happen, what with all the hectic goings on in the typical battle.
Quote:
Originally Posted by Red Harvest
A 5+ star general could pose real problems like this when he became a single man unit. He would often start the battle by withdrawing. So you couldn't target him, force him to run, etc. If you routed his first wave, later waves would still have the bonus of valour from being in his army and could crush your weary men. Then you would have to face him again...
I can see how that would be irritating. It seems that units with less than 10 men end up being high-honour units because some sort of averaging is going on. That's to say if you merge 10 H3 with a unit of 45 H1, you end up with 55 H1, not H2 or H3. I think the battle log file records the number of kills for each individual and hence man-by-man honour but it's averaged for the whole unit. Of course, once placed in a bigger unit, they're part of a bigger target and stand a greater chance of dying by missile fire (and small units probably don't attract arrow fire in the first place). I tend to send these units 'back to base' until I've got equal-honour (or better tech) cadres to merge them with as they're of limited use on the battlefield.
Now a single man unit probably has ridiculously high honour level plus a bonus of 2 or 3 from the general status, so he's probably near invincible in single combat. If you look close enough, if you send two units to attack a 1-man unit from in front and behind, the fight is still 1-on-1. The unit coming from behind just sits and watches. No doubt, it's only when the defender is totally exhausted that he finally gets beaten. I see this often in castle assaults (current assault-CTD problems aside) where the last defender lasts a good minute or so, sometimes taking out 2 or 3 of my men in the process.
Quote:
Originally Posted by Red Harvest
The biggest stupid thing the AI would do in MTW was put all the heirs in one army under the king then attack. It was a powerful, small, fast force but it would kill off the heirs quickly even if it won, and a loss was catastrophic.
Yes, like you said, it tends to make the rest of the battle too easy but, for me, the most disappointing thing was the effect it had on the rest of the campaign. Catastrophic is the word.
Takeda's lands were won all too easily and they were also undeveloped, so I had to bear some of those costs. If you wait for the first ten years or so, you stand to gain more when you win territory. For example, Imagawa (allied) took Kai and built the gold mine, which I promptly stole ~:smoking: since the garrison was ludicrously small (more AI stupidity). Uesugi's areas were a more tricky prospect because he hadn't been in the customary position of duelling with Takeda. It took some time but again they were well developed and worth the wait.
Re: The 'Suicidal Daimyos' problem?
Hi to all,
STW1.11
Actually, I think the whole honour-level business is the core of the problem. The AI will go on the attack within the first few years of the campaign (in very predictable ways*) but the regular troops are all H0, or a max of H2 with the Daimyo's bonus. Since the battle goal seems to be more to do with triggering a mass rout than inflicting casualties, it will tend to send it's highest honour unit into the attack at an early stage, even though the 11 heavy cav is such a small unit. It's as if a 'rule' has been programmed in to pick the most effective unit for a specific attack.
The behaviour does change, even if the rule does not, however. In the late phase of the game, the AI opponent will have plenty of improved honour, improved morale, improved weapons/armour unit types and it will attack, or counter, using those, with the Daimyo's unit hanging back, like it should.
I've read some threads about campaign completion (1.02? or WE or MI) in under ten years (starting 1550, instead of 1530) but, from the screenshots, I see the starting position is just one territory, surrounded by rebel zones and it's possible to make very rapid progress and crush the AI clans before they've made much of a start. In the Takeda example, the player has Heavy an Naginata Cavalry right from turn 1, so I gather that you are given a handful of teched-up or improved honour units as a starter. Assuming the AI gets a similar supply of good units I reckon that this would alter the behaviour and tend to make it more conservative about where and how it deploys the Daimyo unit.
I was going to make a point about the group formation buttons being a part of the problem, in that some of them place the general's unit in a frontal position but there's no way of knowing if the AI is programmed to employ those formations. I can't say I've seen it use them, other than that it does a better job of keeping all its units together than I generally do ~:wacko: I've an awful habit of sending a unit off to do a task at the start of the battle and find they're still sat in that position, doing nothing, when it's all over, while the rest of my army is at the far end of the map, polishing off the rout - but only just - and those extra men might have made a difference.
-------------
* Footnote
Predictable AI first moves..
Imagawa attacks Shinano within the first 3 seasons.
Mori attacks Inaba, even though it's farm output is a paltry 95 per year. Sometimes takes several turns to put down the consequent peasant revolts but never seems to get around to making use of the Shinobi bonus by building a Tea House there.
Mori then attacks Oda in Tamba, even though Kawachi would seem a better choice in terms of income. Seems to be purely to get an opportunity to attack Yamashiro (for Kyoto).
Hojo goes for Sagami, even if Kai is worth more with the gold mine, plus the chance of demolishing Takedas dojos.
Uesugi tries for Hitachi but, instead, will waste time trying to regain Shinano if Imagawa has won it.
Takeda will attack Musashi but, if well garrisoned, will go for Shinano instead.
Re: The 'Suicidal Daimyos' problem?
Quote:
Originally Posted by EatYerGreens
I can't say I've ever seen the AI voluntarily withdraw any units myself, in the sense of enabling elements of the reinforcements stack to come into the battle.
He is talking about MTW in which withdrawing (ctrl+W, and not routing: ctrl+R) is a regular feature, unlike STW. In MTW the computer will withdraw depleted units and horse archers with no ammo left. In STW the computer doesn't do that, and many STW players were not aware that there was such a feature.
Quote:
I think the battle log file records the number of kills for each individual and hence man-by-man honour but it's averaged for the whole unit. Of course, once placed in a bigger unit, they're part of a bigger target and stand a greater chance of dying by missile fire (and small units probably don't attract arrow fire in the first place). I tend to send these units 'back to base' until I've got equal-honour (or better tech) cadres to merge them with as they're of limited use on the battlefield.
Correct: honour is kept track of on individual basis. That's why you can have a 'legendary swordsman' event while none of your units is of high honour. This honour is not lost when you retrain or merge the unit. The men are just as experienced, but there experience is compensated by the inexperience of the other men. Don't worry, they are still there. You can notice this because sometimes the honour of a unit drops when it is taking casualties under missile fire: the experienced men have died.
Re: The 'Suicidal Daimyos' problem?
Quote:
Originally Posted by Ludens
...and many STW players were not aware that there was such a feature.
Strange, it's there in the pop-up menu for a unit. You would have thought they would have at least tried it once, to see what it did.
I certainly use it myself. If I'm sending in more than one stack, it's the only way to guarantee that the reinforcements will make it onto the field.
If you've taken gun units into battle and it's raining on the day, they have to be withdrawn right away. Routing might get them off the map a little faster but I like to see things done in an orderly fashion. ~D Also, there's no knowing what would happen if they routed through an Ashigaru unit, say, in the rear and caused them to rout too. Not sure if this happens in STW but I gather it's a significant issue in MTW.
I'm very wary of the rout command itself, preferring to let units rout spontaneously. There's always the risk of forgetting that I've got an entire group of units selected at the time.
You also have to watch it with the formation buttons for groups. An accidental double click on the 'square formation' button could be disastrous because the Rout button is positioned directly underneath it! And there's no 'are you sure you want to do this?' dialogue box afterwards...
Re: The 'Suicidal Daimyos' problem?
Quote:
Originally Posted by EatYerGreens
Hi Red,
I can't say I've ever seen the AI voluntarily withdraw any units myself, in the sense of enabling elements of the reinforcements stack to come into the battle. More often than not, when they depart the map it's because they're routing (cursor-hover describes them as 'running away'). I'm interested in how you even got the opportunity to see this happen, what with all the hectic goings on in the typical battle.
I keep an eye on where the AI general is to avoid nasty surprises. They tend to hold back when there are few men left in the unit. They do withdraw at times to bring on big reinforcement units. You see this some with other units as well. You hammer a unit of skirmishers or light cav badly and they may or may not rout. However, it is obvious they are spent as a unit, and it is not unusual for the AI to to withdraw them to bring a full unit on the field. It will happen with infantry too. Missile units sometimes just withdraw when out of projectiles so that the next unit can enter. (I use the same strategy myself in big battles at times.) We are getting a bit far afield into MTW though.
