Game 469: Battle of Kawanakajima

Wargaming Scribe did this one before me - see his review and AAR too.

Here's a personal fact - Sid Meier's Civilization is one of my all-time favorite computer games, but the version I played first and by far the most was Koei's SNES port. You see, this was in the days when DOS compatibility was sketchy, but DOSBox didn't exist yet, and console emulation was often the more reliable way of playing games of the era.

Apart from that, I was aware of Koei as a developer, as they churned out a staggering large number of SNES and PS1 strategy titles, but apart from Civ I haven't played a single one of them. They struck me at the time as slow and boring, and even now the Koei brand seems remarkably niche compared to how prolific they were, and thanks to Mobygames I can better appreciate that all of those console games I dismissed in the 90's were localized ports from their true habitat; Japanese personal computers.

It's no surprise that their domestic output is even larger than their export catalog; from 1990 to 1999 they released over 100 games, and nearly half of them never left Asia.

I'm obviously not going to play all of them, but 1985 was the year of their first true whale; Romance of the Three Kingdoms, originally released for the PC-88 and later localized and ported to several platforms including NES, Amiga, and PC.

Before that, though, I will be playing their earliest known game, which translates as "Battle of Kawanakajima" and was never officially localized despite a recent re-release on Steam.

1845 triptych by Hiroshige, sourced by Takahashi Sangyo Co.

 

The battles of Kawanakajima were clashes between the Takeda and Uesugi clans, and the fourth in the autumn of 1561 was one of the bloodiest of the Sengoku period. Historically, these were indecisive, with neither side making gains, but became legendary and brought prestige to Takeda leader Shingen, who became one of Japan's most feared warlords and the biggest thorn in Nobunaga's side until his untimely death in 1573, and with it the decline and fall of the Takeda clan.

PC-8001 version

 

Originally written for a Sharp MZ-80C, this version is likely lost, but two images are floating in the Neo Kobe archives; a PC-8001 tape dump and a PC-8801 disk dump. The Steam re-release, which includes a PC-8801 styled "original" mode is not entirely authentic; crucially, it eliminates one of the game's most obnoxious mechanics, which you will read about (or already know about if you read Scribe's AAR already). Scribe did not suffer through that. I will, but at least I'll be able to play in English thanks to a translation by RPG Codex member Helly, provided to me by Scribe.

Helly's PC-8801 version.

Uesugi is somewhere out there, preparing to attack our position. This is a pretty good defensive position already, but Shingen has grander plans - destroy Uesugi first! Legend has it that Uesugi Kenshin personally lead a charge into Takeda's command post and dueled with Shingen himself before withdrawing, but that won't happen here; both warlords are confined to their immobile castles and protected by similarly immobile rearguards. We'll have to find his and murder him before he knows he's lost!

Crucially, the main visual presents a zoomed-in view of the battlefield; a 400x400 window of a 2000x2000 playfield centered on Takeda Shingen's position. I'm going to call these unnamed units of distance "Chō" (about 360 feet, or 110m) because that feels right. Men can march 200 Chō in a turn, horses 500.

It's really more like this.

 

The other crucial thing, and this is where it gets really annoying, is that the windowed view does not update when your units move! This is a discrete action that a unit spends its turn performing, and causes the view to recenter on the unit, showing any nearby units including enemies. Meaning that a unit can move or report its position, but not both! You look around, you wasted your unit's turn.

19th infantry is so dead.

Even stupider, attacking reveals the location of each enemy unit in range anyway, just not in a graphical manner. You'd never want to use it when an enemy is in combat range.

So I'm going to be relying a lot on Excel to keep track of where everyone is.

I need to know where Kenshin is, and I have just enough units to get a recon on the entire northern half of the battlefield with two reserves to spare, though it will take a few turns to get them into position. Units are moved freeform with a θ/Δs notation, and Excel proves useful in calculating these precisely.

Shingen's view on turn 2. Two gunner units are kept close.

  

In the middle of turn 2, some of my troops reported being attacked.

 

No good visual, but the 10th spearman, who I had moved 200 Chō northwest of his starting position, is under attack, so I rally the other spearmen and gunners to his position while everyone else keeps scouting forward. The 17th infantry also reports running into enemies, so I send the two closest infantry units to lend support.

Turn 2 overview

On turn 3, the horses keep scouting, and the spearmen in the southern conflict take a severe beating before I get the chance to counterattack, and see who is actually there.

Eight companies! Ignore the graphical view; it's completely stale and not centered on the action anyway.

 

The turn isn't great. I lose all three infantry in the northern clash, all three spearmen in the southern, and one of the gun companies is nearly destroyed.

Turn 4 isn't great either. Both gun companies are destroyed, and Shingen's guards are fired on from the northwest. But my horsemen have arrived at the north end of the map!

 

As turn 5 begins, and Shingen starts shooting blindly at the horsemen camped outside, I get a curious alert - horse company #8 is being attacked by enemy unit #1! This means I found him - at the north end of the map, which shouldn't have been a surprise at all.

Shingens' guards concentrate their fire on the attackers, destroying one horse company and weakening a gun company, while Kenshin's guards return fire on horse company #8, destroying it.

 

Turn 6 starts. That's... a lot of targets.

Dang.

 

New tactic. Horses stay put, and everyone else moves to Shingen's northwest guard position. We'll wait a few turns, and then my mobile units are to converge 500 Chō south of where I think Kenshin is, and move in for the kill all at once.

Turn 2 start from Shingen's POV

Uesugi horsemen arrive ahead of the main. We waste them easily.

 

They're here!

On turn 6, my tightly clustered troops inflict heavy damage on the Uesugi units who stepped into range on turn 5, and the survivors who didn't step quite as close (which is why they survived) have difficulty hitting me with their return fire. Two of their infantry companies are destroyed by my guards, a third by the horsemen, and as more of them step closer throughout the turn, we easily take out a spearman and a fourth infantry. All they manage in response are some grazing shots on one of my infantry... and Shingen. I'm sure he'll be fine.

Turn 7

Turn 7 is less exciting. A few new Uesugi units step into range but most continue attacking from their combat ineffective ranges, and few of them are hit by us. We only manage to destroy two more infantry unit, and I expect they have seven units left, plus their guards.

On turn 8, I break the standoff and rush them, and this costs me a horseman and a bowman company.

Turn 9

I don't have a lot of time left. The horses and spearmen stay and fight, taking out three of the four units on Shingen's northwest flank; a gunner, a bowman, and another infantry. Everyone else moves due north, and suffer a few potshots en route.

Turn 10

The guards ineffectively shoot at the barely-in-range infantry units, who ineffectively shoot back, but my horsemen chop up one last infantry unit on the northwest flank. Now there's just the four units on the north.

On turn 11, I move my horsemen and spearmen right into the thick of them - this gets one nearly killed and another completely killed, but I still have three healthy units in the middle and ready to cause mayhem. My bowmen, two gunmen, and seven infantry units keep pressing north.

Turn 12

The survivors of the last jump inflict three heavy blows on the north flank, but they finish off my wounded horse company, leaving one horseman and two spearmen.

On turn 13, I have the horsemen disengage to catch up with the company moving north, who are now halfway to Uesugi Kenshin, leaving the spearmen behind to fight four companies, three of them heavily wounded.

Turn 16

Uesugi is reduced to one heavily wounded gun company at my gate, and his guards. My horsemen have caught up, and in one turn we will rally at his gate.

Turn 16 from the horsemens' POV

 

On turn 17, forces at my gate are eliminated, and my forces are stacked up 230 Chō south of Uesugi Kenshin. The southmost guard lands a lucky shot on one of my gunners, nearly killing it. We all move forward.

The guards let loose on turn 18 and kill two infantry, nearly kill a third, and finish off the gunner. But this will not save Kenshin. He dies today.

Again, ignore the stale graphical view. We're all within bashing distance of Kenshin.

 

GAB rating: Below average. I don't hate this game, and I admire it a little bit more than Nihom Falcom's inaugural Galactic Wars for being somewhat original, but this is primitive and shallow. Kawanakajima might as well be one-dimensional for how irrelevant the gameplay map turned out to be - during my second attempt where I had some inkling of what I was doing, I barely noticed the lack of a useful graphical overview at all.

But this was just an appetizer before the grand strategy of Nobunaga's Ambition, which I'll have to attempt in Japanese as it was never translated by anyone. Wish me luck!

Paradroid: Won!

Profile view of the Paradroid, current deck (staterooms) highlighted

The stateroom deck map

 

My macro strategy for clearing Paradroid is simple - know where the strongest droid on the ship is, have a plan to kill it, and try not to waste too many stepping-stone bots along the way. Repeat until there's nothing left but you.

This isn't necessarily the most optimal strategy - it takes awhile, and loses out on the bonus points (and heightened security!) scored by killing enough droids in rapid succession to raise the alert level to red, but I feel it is the safest. Even so, my success rate is probably around 25% to 33% - getting ambushed or just fumbling the hacking minigame can end your run in an instant.

To aid myself, I put together a ship map to show the deck layouts and what robots might be encountered on each - numbers in brackets represent the best droid on the deck, which isn't always the highest class! Notably, the 500 series have weaker weaponry than (and decay twice as fast as) the 476 droid, and the 999 droid decays so fast that you really want to get rid of it for a 711 or even a 600 series as quickly as you can.

Some further gameplay notes I discovered:

• Your health is an invisible stat, but the game seems to track health and maxHealth separately.
• MaxHealth is always ticking down, and the higher the droid class, the more quickly it decays. Consequently, the longer you've had a droid, the less damage you can survive. The inverse does not hold; taking hits does not reduce the amount of time you have. 
• 999 lasts about 30 seconds, 500's-800's last about two minutes, and anything else (including your 001 at the game start) lasts about four minutes.
• You can get a rough sense of how much time is left by how fast the player sprite is oscillating. It gets noticeably slower at the 50% mark, and you start flashing at the 25% mark.
• A successful hack completely restores both your health and maxHealth.
• A failed hack restores your maxHealth but not your health, and reverts you to the 001.
• A fatal hit leaves you in a very bad state where you are reverted to the 001 and put into flashing critical mode. You have about 30 seconds to find some prey before your batteries run out.
• Droids with antigrav are the fastest, and often the best in their class in other ways. Droids with tracks or wheels are the slowest and can be a liability.
• I still have no idea what the various sensors do. 

I've cleared the ship a few times now, and made a video recording of my most recent success. It's not my best or quickest run - in fact, I get off to a pretty lousy start and nearly ruin it, but it is the one I recorded and will AAR.

The game allows for pausing at any time except the hacking minigame, and I use this regularly to check my maps. These pauses are edited out. 

 

Each game randomly picks a starting deck and it's either research, stores, or staterooms - this is the staterooms, and the most dangerous of the three, because there are armed 476 maintenance droids here.

My first goal is to pacify the 999 command droid, on the deck, and there's a lift, but going there would be suicidal. Most of the droids there have powerful weapons, and would be severely disadvantageous to try hacking. As a general rule, if you try to hack a droid more than two classes above you, you are taking a risk. More generally, avoid being even on a deck that has droids more than two classes above you. Such as the staterooms right now.

I immediately go to the nearest lift and ride it up to the stores.

Oh, hello there!

 

I can take a 296.

When I first started playing Paradroid, my hacking strategy was to pick the side with fewer dead pins. This is still a pretty good strategy, but it isn't always the best one, especially not at low levels where you don't have enough pulses to hit all of the live pins on your side. Splitters can make your limited pulses go farther, and in this situation, the yellow side has more splitters; the middle wire will split your pulse twice and hit three bits. But one of those bits is guarded by an inverter, which means that at best you are flipping two bits with that pulse, and possibly flipping the third to the enemy's color!

So I went with purple - in retrospect this might have not been the best choice, but Paradroid does not give you time to think about it; you get ten seconds to pick a side and another ten to act on it.

I also learned that it's a good idea to wait a second or two before acting. Bits are flipped to whichever side hits them last, so if you allow the enemy to fire a few pulses first, you can flip their bits back instead of giving them the chance to flip yours. But you have to learn to move quickly; when you're out of time, you're done. And at the higher levels I usually finish with unused pulses because I just didn't have time to use them all.

But anyway, I win this one no problem, and now have command of a beverage dispenser on treads. It's not great, but I'm in a better position to go back to the staterooms and hack a maintenance bot - if I can catch one!

You're not a 476, but you'll do.

 

 

But I failed the hack, because I wasn't paying attention - the 420 snatched two of the lower bits while I was focused on the upper side. Oh well - it happens! Back to the store deck I go to hack another servant bot, and then have another go in the staterooms - this time I get a weak but speedy 302 messenger bot.

And with it, I hunt down and catch another 420.

 

My next step up is the observation deck where I can typically find a 614 sentinel or at least a 500 series crewbot - either one an upgrade.

Getting up there! But the bots are getting stronger; from the 600 series on, they shoot back, and hacking starts getting pretty chaotic.

The quarters deck is populated exclusively by 700 and 800 series droids, both quite dangerous. The 700 battledroids most commonly have screen-clearing disruptors, and the 800 series security droids move fast and shoot accurately with maximally powerful laser guns.

Taking the 751 feels like a close call!

A good strategy at the higher droid levels is to target the repeater diodes first. Even if the corresponding bits are already your color, this effectively locks them down so the enemy can't flip them, and then you can use your next moves to react to whatever the enemy does. At higher levels, you have pulses to spare, but not time.

I upgrade to a 834 before leaving for the bridge, where I have a big fight on my hands.

I lose the fight. I should have hacked the 711 near the lift and used its disruptors to clear the bridge, but instead I kill it with my lasers, and then get destroyed by the other 711's on the bridge.

Back to the lower decks! There are still enough droids left alive that I can work my way back up to something combat effective.

I come back some time later with a 516 crewbot from the robostores deck, and snatch an isolated 711 battledroid before it can disruptor-blast me.

With the 711, I waste a few sentinels and easily hack the commander. Who I then promptly take back down to the quarters deck and downgrade to something that will last more than a few seconds - an 834 security bot. And while I'm here, I take out the rest of the bots too - now the most powerful droid on the ship is the 821 security bot, I know they can be found here, and the only way to ensure the deck hasn't got any is by clearing it. I defeat the droids through hacking rather than by shooting, since they shoot back.

The lights dim after you've killed all the droids on a deck.

 

Commander down, quarters cleared. But there are still some 800 series droids to be found in the cargo holds, which are accessed by a lift behind the staterooms - I go there next.

There are two cargo holds, and they are both large, open decks with a mixture of dangerous security droids and cannon fodder. Most of my failed runs ended in the cargo holds; with nowhere to hide, you can get sniped offscreen, but hacking a level 8 droid is almost as dangerous as fighting it; usually I win easily, but if the sides are roughly even, then the level 8 droids stand a chance of beating you through sheer brute force; they get enough pulses to just flood the board with their color, they can spam pulses faster than you, and it almost doesn't matter that they don't have any strategy. You either beat them by picking the better side of the board and exploiting it, or you beat them by reacting to their moves and ensuring the board flips to you at the last moment.

 

I nearly get killed stepping off the lift to the lower cargo deck!

 

The 883 (ex-term-in-ate!) is a tank but moves like one. Ditch it ASAP.

 

The cargo holds take some time to clear out, but I manage without incident, and I can safely say that the ship is free of anything stronger than a battledroid.

Next I clear out the vehicle hold and shuttle bay, where I again must rely on hacking to take out the majority of the battle droids, as they are immune to their own disruptor weapons, and I did not manage to find and hold onto any 821 security droids, which are otherwise the best 700 killers.

After clearing these decks, I am in command of the sole 711 droid on the ship, and attempt to murder some weaker droids in quick succession to see if I can get the ship's alert level up for some bonus points. I am not successful; the maze-like layouts make it too difficult to track down and kill with the required speed, and eventually I'm forced to downgrade the 711 for a fresh 302 on the reactor deck.

At this point, all of the truly dangerous droids are destroyed, and I'm not at any real risk of losing. A cautious approach will clear the rest of the ship. I take over a sturdy, well-armed 476 droid on the engineering deck below, and this is enough to take on the rest of the decks, which I do in descending order of difficulty, killing the rest of the droids with my laser, and occasionally swapping for a fresh 476.

The last droid is seen cowering on the research deck.

A freighter-clearing bonus is awarded and we prepare to board the next ship, Metahawk. But it's the exact same layout as the Paradroid, just with slightly stronger robots. I wish the game let you pick the starting ship, because I could see myself revisiting this game and picking a harder ship to start with for more challenge, but having to go through the repetition of clearing the "easier" ships each time first would absolutely kill my incentive to bother doing the harder ones.

Especially given how unforgiving Paradroid can be - I had not mentioned this yet, but control mishaps are another means of losing - the one button interface where shooting, hacking, and taking lifts are all performed with the same button can cause you to perform one action when you meant another, and the hacking minigame can be fiddly at just the wrong moments and cost you the game because the cursor moved past your target or didn't move at all.

Despite that,

GAB rating: Good

Paradroid is great. Technically solid, well crafted and designed, and just a great blend of action, exploration, and strategy. Easily the best UK title I've seen so far, and not a half bad showcase for the fairly new Commodore 64 platform that developers were just starting to learn how to tune - the smooth four-way scrolling, the bold, clash-free colors, and ambient whirring and beeping sounds just wouldn't be the same on a ZX Spectrum. I recommend and promote this one to the ivory deck, with a well-deserved harpoon.