Friday, April 24, 2026

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!

17 comments:

  1. "I will be playing their earliest known game"

    This raised my eyebrows as there is old conventional wisdom that their earliest releases were eroge, but I gather more recent scholarship may have pushed past them back to still earlier strategy works.

    "a unit can move or report its position, but not both!"

    Shades of the Heisenberg uncertainty principle!

    ReplyDelete
    Replies
    1. This one isn't even listed on Mobygames yet, but if the attributed release year of 1981 is correct, then it predates the eroge that are.

      Delete
    2. That's correct. Koei was founded in 1978 as a wholesaler of dye and industrial chemicals, and, quoted from Japanese wikipedia;
      "In October 1981, they released their first title, “Simulation War Game: The Battle of Kawanakajima” (distributed on cassette tape), and it generated three times the annual sales of their dye business alone. Along with the simultaneously released “Investment Game,” other titles such as “Underground Exploration,” “Combat,” and “Operation Normandy Landing” were also launched. These five works became popular and were collectively known as Koei’s “Red Box Series” due to their package design. Around this time, the company was also involved in business software and adult games (later known as the “Strawberry Porno Series”).

      Delete
  2. One thing to note about Koei's early strategy output is that it was very nearly a one-man show for a long time. Kou Shibusawa/Yoichi Erikawa (he goes by both names for whatever reason) was plugging away at the design and programming of these titles himself until the studio really took off with Dynasty Warriors. They were boutique games for a limited audience, but really impressive accomplishments all the same.

    ReplyDelete
    Replies
    1. "(he goes by both names for whatever reason)"
      "Kou Shibusawa" was originally a fictional "person" credited as the producer of all of Koei's games, representing the entire dev team, and different Koei devs would get interviewed as "Shibusawa" by magazines depending on which game "he" was being interviewed about at the time. Erikawa eventually adopted it as his personal pen name in 2000, but "Shibusawa" would still get credit as executive producer on games Erikawa had no actual involvement in, so "he" still lives somewhat on as this personification of Koei as a whole.

      "Eiji Fukuzawa" is another fictional person you'll see credited in many of Koei's non-war games.

      Delete
  3. Let me know if you need any help with the Japanese.

    ReplyDelete
    Replies
    1. Thanks. I expect the game is mostly a lot of charts and figures, which Google seems to be pretty reliable at translating, so it's mainly the tedium of constant screenshotting/uploading that I'm dreading. But I'll certainly ask if something with nuance or contextual sensitivity comes up.

      Delete
    2. There are options to translate while playing that can be helpful, text output in separate window or even full screen. I used the agth text hooker for PC-98 games many years ago, nowadays I occasonally use retroarch's service with Ztranslate which works with different platforms, there seem to be quite a number of tools these days.

      Delete
    3. I haven't had great luck with these tools in the past when it comes to PC-88 games. I suspect it's a combination of the low-res font and the nonstandard katakana. AGTH seems to be dead, and a successor project ITH just doesn't work at all.

      I did find and give Ztranslate a try, but the output is chaos.
      https://imgur.com/a/c6fECAd

      Delete
    4. It looks a bit better on my end (just tried the Sharp X1 version), but it's still way too crowded and not very helpful. Apparently, it's supposed to be possible to do a more restrictive area OCR on the Ztranslate site once an image is uploaded there, but I don't understand how that works if it does.
      Last time I messed around with new developed real-time game translators, most of the focus seemed to be on visual novels, which probably provide a much friendlier text layout (though this is a genre I wouldn't bother with machine translation, but alas it's apparently good enough for people)

      Delete
  4. Congratulations on winning this one "blind" - I knew it was the authentic way to experience the game but really couldn't be hassle (and I actually have a rule of "best version of the game that feels authentic-ish" instead of "historical experience").

    I am curious to see how you experience Nobunaga's ambition. I kind of vibe-played it without understanding all the numbers behind it, so maybe I'll understand a few things thanks to you.

    I am also going to accelerate Romance of Three Kingdoms so we release our article within a short span of time. I reached 1985 anyway.

    ReplyDelete
  5. regarding "first game" status - according to the page here - https://en.namu.wiki/w/%EC%BD%94%EC%97%90%EC%9D%B4

    they actually had two games both release in October 1981, the other being a Stock Investment game 投資ゲーム. Since they launched with two titles they both were their "first release".

    We need a Simulation Addict to try the stock market one

    ReplyDelete
    Replies
    1. I have two findings motivated by your mention of this.

      First, a great article on the history of Koei, and not written very long ago either. It also mentions "Investment Game" as an October 1981 release.
      https://www.gamingalexandria.com/wp/2026/04/general-excellence-origins-of-koei/

      Second, the transliterated title is "Toushi Game" and can be found in the Neo Kobe PC-8001 collection. It's copyrighted 1982, not 1981, but that could be the date of the conversion rather than the original which I assume was also on the Sharp MZ-80C (and also assume is lost). And it looks half decent! Like something SSI might have published in 1981 and sold just under 1000 copies of.

      Delete
  6. First and foremost:
    Thank you for this blog!
    It is highly fascinating to me and especially exciting, when obscure and old topics like this game are covered.

    I have several comments to make.

    If you are still looking for the game's entry to Mobygames, here is the link:

    https://www.mobygames.com/game/105288/kawanakajima-no-kassen/

    Since I am a Japanologist/Anthropologist, who has been researching the history and development of the japanese role-playing video game genre for some years now and is more or less preparing a doctoral thesis on the matter, I thought I'd chime in to state a few points.

    "Kawanakajima no kassen", from the current standpoint, is indeed the first Koei game published. The eroge games came later in 1983 and 1984. It is a vital topic to my research, since I see very early elements of the tactical rpg genre in it (i.e. a unit's morale growing, making it stronger and more resilient, which I view as a parallel to level growth in later games). I have played through it multiple times, as it highly fascinates me.
    From what I have experienced, Uesugi Kenshin's army is not only positioned as a mirror image to your army, but his way of going about his (soon to fail) attempt to assinate you also mirrors the way the player should do it to succeed. When I use the first 3 or 4 turns to stack my units in three equally spread out positions as a vanguard to the leader, the cavalry of the enemy will just jump to your leader. This is the moment the leader and his bodyguards should destroy them.
    The rest of the hostile army will be soon to arrive. I prefer to take them out with my cavalry units until at least one of them has reached a morale level of a minimum of 20. I will then make this unit jump to the enemy's base during the next 4 turns. Being at level 20 will make it invulnerable to Uesugi's and his bodyguards' attacks, but will enable it to kill the leader in 1 to 2 strikes.
    This is the point at which the game likes to glitch out or cheat, though. Uesugi and his bodyguards might decide not to attack the cavalry unit in their base, but any unit, even the ones out of reach. Doing that, they might be able to kill any unit on the field, but will continue to attack that unit even after its defeat, which would give them a tremendous, perpetuant stat boost, making them invulnerable. If you are unlucky, they might even kill your leader from a distance, so murder Uesugi fast!

    Regarding the "Investment Game":

    You are given 1.000.000 $ in the beginning and have to multiply that amount through trading.
    You can buy stocks, invest in currency or trade in resources.
    You should also research the probability of stock values rising or falling, but each inquiry will cost you 100.000 $.
    Current world events are displayed in a window on the left side of the screen and will influence the stock market.
    It sounds dry, but like "Kawanakajima no kassen" and also the other games from the Red Box Series it will get pretty carthatic with growing success.
    I never would have thought so, but I actually started liking it quite a bit.

    "Combat" and "Normandy jouriku sakusen" are similar to "Kawanakajima no kassen", as they are tactical, turn-based games reminiscent of the tactical rpg genre.
    In the former you need to capture the enemy's base and kill the commander occupying it. The commander is just displayed as unit "A". Should he be killed early and, for example, unit "C" takes the base, it will become the commander and has to be eradicated as well.
    In the latter the roles are reversed, as you have to defend your base in Paris now.

    Sorry for the ramblings. I just thought I'd try to answer a few questions arisen from this blog.

    Thank you again for the pleasant and informative read!

    Best wishes from Germany
    Phil

    ReplyDelete
    Replies
    1. Great comment indeed! I did not experience the AI cheat you mention, possibly because I used the Steam version of the game (not the modernized one however) and they may have corrected the bug.

      You mention "Combat" and "Normandy jouriku sakusen" - by any chance do you have tehe manual for those games? I wanted to cover them (them and a few others: Hokkaidou Sensou, Rungaoki Yasen, Dasu Buto, Ginga Senryak, Hoi Hoi, Dogfight, Sansetto in Radikku and Chitei Tanken - they were prolific I guess).

      Delete
    2. I did notice some weirdness toward the end of my successful campaign with Uesugi bodyguards shooting at my own, to no effect.

      Delete
    3. @Narwhal
      Now, that is a coincidence!
      You would have been the next person on whose blog covering this game I would have commented. Guess I can spare you the long read now. But thank you for also covering this game and for your work in general. I enjoy it very much.

      And thank you for your compliment! Sorry for responding so late!

      Unfortunately, I do not have the original copies or any documentation to the games you mentioned. I had to emulate them, but should you be willing to cover "Combat", "Normandy jouriku sakusen" and "Chitei tanken", I'll be glad to lend any help I can give, gameplay-related or language-related. I finished all of them several times and think I know the gist of them.

      If I might be so bold as to share a link here, this is a really excellent video on "Chitei tanken":

      https://www.youtube.com/watch?v=luk6vJ0ZqUw

      I'll have to admit, you mentioning of the games "Hokkaidou sensou" and "Dasu buto" has left me both fascinated and stressed out. As does the creator of the aforementioned video, I always assumed the Red Box Series by Koeir just consisted of the five titles discussed in my post. Now I had to discover, that there are actually a 6th and a 7th title. I found "Dasu buto" in my rom archives. The title is actually a wrongly transcribed version of our German "Das Boot" (The Boat), which in this case relates to a submarine.
      "Hokkaidou sensou" on the other hand still eludes me. You would not happen to know where to find a disc image? I would be very grateful.
      Then again, I already am grateful to you for leading me to "Dasu buto".

      Please let me know, if I could assist you in playing those games.

      @Ahab:
      I am glad I am not the only one discovering this. Since it did not happen to me the first time I beat the game (or maybe I just didn't notice), I thought I had somehow broken the game permanently.

      Delete

Commenting with signin or name/URL is encouraged but not required. If the spam filter deletes your legitimate comment, apologies - it does that sometimes.

Most popular posts