Take the A-Train: Won!

I did it - I delivered the president, safe and sound and on time, from Washington D.C. to Hartford. Or maybe it was from Mar-a-Lago to D.C. The game is not specific about its geography or time. And not only did this trip have a body count, but I'm pretty sure it was supposed to!

 

In the south of the map, I have a stable cargo unloading loop-and-switchback system, and in the middle part, a stable revenue-generating passenger train loop. I'll need to make a few trips all the way up and down the map to retrieve more materials to finish the one-way presidential line, but I don't mind too much; the passenger loop already goes 40% of the distance, and three more full 490t loads ought to do it.

I could, in theory, shorten my travel time by incorporating the cargo train's route into the passenger loop and have it deliver rails closer to the front, but there's no way to do that without slowing down the passenger trains and therefore slowing down the revenue they bring in. The cargo train has to return to the yard, and the passenger trains don't, so I'd need to build more train stations and use time signals to keep the cargo train from crashing into the passenger trains as it enters the loop. Less trouble to just manually locomote myself back, using sidings to avoid crashing into my own trains, and pick up the goods myself, I think.

Plus, the south end of the loop is really cramped already. How would I even build more here?

Back at the front, all I have to do is keep building the line northward. For the most part this is not complicated.

A few screens up, and there's an apparent dead-end - but you can actually tunnel through the thin part of the woods. 

 

More track!

 

Eventually, there's another natural barrier, and this one's nasty.

This part actually ruined an earlier attempt, because the tunneling mechanics are so damn obtuse. It's really not clear where you can enter the woods, nor where you can exit, and once you go in, you're operating blind because the trees obscure the rails and your train. And since you can't see what you're doing, you'd better build your tunnel correctly the first time - which is a pretty unfair ask when you don't even know where it should exit! Create a junction here by accident, and you'll never know until the president's train goes off it and derails.

 

This will work, but you'd better not enter this tunnel during the day - you might construct a junction by accident. Best to only go back through at night when construction is impossible. 

90% of the way there. Gotta make one last cargo run.

Finally, the president's destination is in sight... but there's a problem. The way forward is blocked by some farmland!

 

Unless there's some other way of removing buildings that I don't know about, the only way through is violence. Violence and a carefully constructed series of junctions with the buffer stops removed.

 

This is why we needed so much money. As it turns out, I overprepared by about $100,000, but still, there's plenty of time left to complete the rail. It takes three train crashes full of passengers to clear a path through, but I do indeed make a path through.

Finally, I lock my cargo train into the switchback, wait for the remaining passenger trains to clear the southern loop, and connect the rails - this signals the president's train to start moving in 24 hours.

3x speed, transitions choppily edited out

 

GAB rating: Below average.

The first Take the A-Train is more interesting as a sign of things to come than a complete game experience in itself. It anticipates aspects of automation games, and largely achieves what it set out to accomplish, but the game just doesn't offer all that much to do, or much reason to explore the full extent of what its railroad engine allows, and between the interface friction points, the limited paths to victory, and the cramped map layout, actively discourages the sort of freeform play that these games would be known for.

Game 472: Take the A-Train

Artdink's initial entry in the A-Train series is weird. I was expecting a primitive version of Transport Tycoon, but what I got feels almost more like a primitive version of Factorio.

It hadn't helped that I couldn't find any full copies of the manual. The PC-8801 version of the game, and a digital manual, are available on the Nintendo Switch Store, but I don't have a Switch to view it on, so I'm stuck with the first page of it.

At least the in-game language barrier is thin. About half of the game is in English despite never being released outside of Japan, and the other half is terse enough that I didn't need a translation guide.

I chose the FM-7 version of the game, which launched first in December 1985 according to various sources (though the ingame copyright states 1986), using eFM77 to emulate.

Before writing this, I went through a lot of trial and error to figure out just what exactly I'm supposed to do and how to do it. Early on, there were a lot of crashes, derailments, and general frustrations with the interface. Then there were bankruptcies. And then there were crashes and derailments again. But now, I think I have a pretty good idea of what I'm doing.

The ultimate goal here is to transport the president's train, seen here outside the white house and marked with the star symbol, several screens north within 365 days. You have $100,000 and 750 tons of materials, which isn't nearly enough, but it's a start.

The A-Train, the circle marked with the letter A, is your construction vehicle, and you directly control it. Or at least you will once the weird controls start to make sense. Rails follow the vehicle in insert mode, and are picked up from behind you in delete mode, but both modes are subject to a lot of edge cases which can set the rails in unexpected ways or just cause the game to yell at you that you can't do that. The A-Train can't derail, but a collision with any other vehicle is an immediate game over.

The first challenge is the controls. Even now, I often find myself fighting with them, not being able to move in directions I think I should, laying down tracks where I don't mean to, not laying down tracks where I think I should be able to, or not being able to remove tracks that I do mean to, wasting a lot of time and money. Compounding the problem is a day/night cycle; a full day last just under 30 seconds, and construction is impossible from 7pm to 5am, during which time the controls also immediately shift into an automatic rails-follow mode. Too often, I'd be struggling for several seconds to connection a junction or delete an erroneous piece of track, only for night to fall and force me to wait until morning to try again. And that's before the other trains start moving.

Also, inexplicably, signals at junctions and stations can only be changed at night. If dawn breaks before you're finished, tough luck. Hope you didn't need a train there today.

What the. I don't even want a junction there!

 

The next challenge is procuring more building materials. Money isn't a problem yet, though it will be. Trains IV and V are your cargo vehicles, and there's plenty of rail supplies at those stations, but they are inaccessible to you. You've got to use these trains to deposit those crates somewhere you can retrieve them from.

Planning this is crucial. You can put the receiving depot just about anywhere, but you need a loop that gives enough room for you to maneuver in and grab materials from the receiving depot, while also leaving you room to build a passenger line with all those materials you're about to get. Several of my games failed because I built myself into a corner here. Two rules to keep in mind - only three rails can ever meet at a junction, and never at 90 degree angles.

This design should work pretty well.

Once the rails are connected to a rail yard, the train parked there will be scheduled to start moving in 24 hours - just enough time to set the signals. Train IV will deliver cargo to the station on the west, loop around and return to the switchback, back up into the yard for more cargo, and repeat.

 

Now we have to worry about money.

This is the only place where you'll see how much cash is left.

The red window is the important one right now - we have $90,070, we've spent $390 today (so far), $1,190 yesterday, and $520 the day before. The numbers below are revenue, and so far we've made none. Everything costs money - moving, building, demolishing, operating trains, even switching signals. The more track you have laid down, the more you'll need to spend every day on maintenance and trackage fees, and the farther up north you build, the higher the rates get.

Your only revenue stream is passenger fares, collected by the trains marked O, I, II, and III. Passengers are picked up at stations in proximity to villages, and dropped at other stations - doesn't matter where, all passengers get off when the train stops, and the farther the arrival station is from the departure (measured by latitude, not track length), the more money you get.

I'm sure you can get fancy with an elaborate system of loops, junctions, and switchbacks, but I don't really see any advantage to anything more elaborate than a simple loop that goes between two villages fairly spaced apart. Fares seem to be directly proportional to the distance between stations, so having four stations in the loop instead of two means you get paid half as much twice as often, only it takes longer because the trains stop twice as often. And a complicated routing system that gets different trains to different destinations without any collisions is tricky to design in the map's often cramped terrain, and for very little benefit I can see.

Leaving the starting screen

Ah, a village!

Station positioning is important. Maximize the houses in proximity and minimize impassable terrain.

Periodically, I must return to the start for more track. Not a big deal.

Just passing through the next town.

 

Halfway to Hartford. I think this little hamlet will be my terminus.

Making a return loop

It's done!

Cash is down to $35,000. Weirdly, the does not show this on the main screen. But we're nearly ready to start bringing in some revenue.

With a simple two-point passenger loop, we can deploy all four passenger trains right away. First I wait until nightfall and divert the cargo train into the switchback so that it doesn't get in the way. Then in the morning, I connect all of the passenger trains, making sure to stagger them at least two hours apart so that they do not collide.

 

And they're off!

At this point, the best thing to to is sit back and wait. I will need a lot of money, and these trains are money losers at first. I am spending $3000 per day just waiting, and the trains are pulling in about $1500 each on a round trip that takes two and a half days. That's a $6000 loss on each trip. But keep waiting; with each delivery, the towns grow, and the revenue grows with them.

This hamlet is getting bigger!

 

A bustling metropolis.

 

Revenue starts to cap at about $16,000 per trip, certainly profitable, but not amazingly so. By day 100, I've got just over $100,000 in the bank. And I keep waiting because I will need more. By day 150, I have $200,000, and this, I feel, is enough.

No need to expand the loop - from here on, we are building the presidential line.

Romance of the Three Kingdoms: Won!

My greatest foe is dead. My next-greatest hasn't got half the forces I do. Time to finish this.

#37 goes down by the end of the month.Yuan Shu had been killed in combat, and his subordinate Wang Ping was already on the ropes.

 

I'm not going to spend any more time micromanaging provinces or babysitting generals. First thing, Sun Ce's red provinces in the middle are going down. I can mobilize crushing forces easily; I just need to get generals there and transfer soldiers to the loyal ones (i.e. away from the ones I won in battle) before going in. 

This can also be tedious.

Sun Quan's (Ce died) last stand

 

Next target -  Yuan Tan. Regrettably, he breaks through my wall, but I know he won't be able to expand quickly enough to escape a multi-point invasion.

Sending reinforcements to #50

Which isn't to say he (or his eventual successors) make it easy for me. Where he is weak, he withdraws into unoccupied territory when attacked. Where he is strong, he attacks where I am weak. Where he must defend, he utterly abuses smoke and fire to delay the inevitable. I have so many useless generals that I wind up sending them back into southern China for some personal development and recruitment. And one summer, Cao Ang dies, depriving me of a loyal general, and worse, 20,000 soldiers.

The state of China in 203AD.

 

But they're fractured in a few months.

 

Looking grim for them by September.

 

I finish by conquest of south China by winter, which delays my final conquest of the north a few months.

Spring 204 brings more disease, and Yuan Shao dies, leaving his son Yuan Xi in charge. So does one of my generals and another 20,000 soldiers with him. By now I don't really need them.

I begin mobilizing troops northward.

Summer brings poor omens - locusts for me, which are bad news, and rebellion for Yuan Xi, which I didn't even know could happen.

In May, I begin my final attack on the northeast. 747,000 soldiers against a combined roughly 200,000 between both remaining masters, and this time I don't need to reserve all that much to protect my own lands from counter-invasion.

The old monkey rice trap formation. It never fails!

 

Yuan Xi's last stand. He dies in September after foolishly charging into the rice trap.

 

Winter. So close to victory! The inaction is painful.

 

Finally, at the end of February, I catch and behead my last rival in his castle at Youzhou.

 

Sure! But not this one.

GAB rating: Average.

I am so glad to be done with this one. 

On paper, this is a much better game than Nobunaga's Ambition (the 1983 version). It's a more complete and cohesive game, better designed and balanced around its expanded systems. And you no longer get wiped out or set back irrecoverably in the first turn. War, in particular, is a massive improvement, with meaningful tactical options that give the underdog a chance, and the economic and diplomatic layers are meaningfully broader. There's still some illogic carried over from its predecessor (arms/training are independent variables of army size, provisions travel anywhere instantly but armies travel one state per rotation, etc.) but overall this aspect is still a major improvement.

I also have to give some credit to the game's localization. I can't speak to its accuracy or efficacy in capturing nuance of the original, but this is, by my estimation, one of the first Japanese video games with a significant language element to be translated into English, and shows none of the awkwardness that plagued so many of its contemporaries and games released well beyond.

It isn't a great game, not even at its best - even though it's deeper than its predecessor, the systems and their interactions are still more broad than deep, and outside of combat, many gameplay options just aren't worth using. Castles are the only decent revenue source, so you just build them whenever possible, and then your economic concerns become a lot simpler. I never saw any point in levying taxes, searching for gold mines, or borrowing from the merchant or rivals. I also never felt it necessary to explore diplomatic options beyond "pay/marry off rivals who you don't want to fight yet" (though admittedly these options - including some subterfuge strategies, might be underexplored on my part). And for all the stats on your generals, none mattered nearly as much as the two loyalty stats, which are trivial to maximize.

But Romance of the Three Kingdoms' biggest fault is an incredibly tedious endgame, where you know you're going to win, but you still have to manage 30-50 states and make decisions for them every single turn. For the most part, these were extremely trivial decisions too, especially in winter when the game's already slow pace comes to a halt. I'd estimate it took me at least nine hours to get through the final twelve months.

The game is aware of this and gives you the option to authorize a state to make decisions for you, but the implementation is so flawed that I didn't want to do that.

• For one, only the master can automate (and de-automate) states, so you still have to wait for his turn before the option is available. And you might want the master to do something else, since there are options only available to him.
• Second, there's no way to know what an authorized governor is actually doing, so the option might as well be an auto-skip function. That's bad on a newly conquered territory, or one held by a general with even the slightest hint of disloyalty, because mutiny and rebellion are ugly things. Toward the end, that described most of my territories and generals, because I needed the loyal ones for war.
• Third, if an army moves through an automated state, that army ain't going anywhere else until you de-automate it.

The game was somewhat enjoyable at the start, but never all that satisfying. Building up a state isn't much of an accomplishment; only conquest actually feels like a major progress event, and even that part takes forever.

Romance of the Three Kingdoms: First scenario won!

199AD

 

Dong Min, successor of the late master Dong Zhou, remains my most well-armed foe with 24 generals commanding 370,000 soldiers, and occupies both of China's capital cities Loyang and Ch'agan. I want them.

I have... quite a bit more than that, with 49 generals and 633,000 soldiers, but loyalties and arms levels are all over the place, and I do need to reserve much of my army to defend the regions I already have. By my estimation, I will need about 540,000 soldiers to take any of Dong Min's territory, hold it, and guard my own from retaliation, and the more equipped they are, the better.

Phase 1 of my attack plan. Spread out. Retake as much free territory as possible, and scour for weapons-grade metal and soldiers. A lot of it is territory I previously abandoned; it lies deteriorated, disloyal, and fallow, but meticulous redistribution of gold and rice and several months of maintenance gets it back into shape.

New generals are recruited too; Cao Cao has a field day charming Guan Yu's disgruntled servants in region 2 over to his side, and more free agents are found in the territories I've taken. With the playing field narrower, not to mention the inevitable victor of this conflict obvious, they're not in nearly as much of a hurry to leave my employ any more.

Phase 2 is ensuring that each general gets what they need from wherever it is available. The process is... not exciting.

It's three times longer than this and I have another one for provincial needs.

 

By spring of 200AD, I have 54 generals, mostly loyal, half of them fully armed, and more than enough metal to arm the rest albeit spread all over China. And almost 750,000 soldiers.

Phase 3 begins. I begin moving troops toward Dong Min's domain, picking up weapons along the way.

In January 201AD, I launch my first invasion, on state #18.

 

Even with all that fortification he can't hold out against my overwhelming forces, and I don't hesitate to use fire to keep him moving. It's a rout, and Dong Min's commanding officer withdraws before the end of the month.

Next, I invade #20.

This one's a bit disadvantageous for me, but liberal use of trick attacks followed by fire forces a lot of retreats and evens the odds.

It spills into the months to come, and Dong Min sends reinforcements, but this cannibalizes #21's defenses which I also invade with my spare generals.

I'm outnumbered, but now they truly have nowhere to run.

 

#21 runs out of rice and falls, and I immediately have the surviving generals, now in my employ, invade Yuan Shu.

I'm quickly reminded not to do that, as multiple generals defect during the fight, adding their ranks to Yuan Shu's and making the battle drag out far longer than it need be.

Before I'm done with Yuan Shu, #20 is taken by attrition.This triggers the victory conditions for scenario 1, and simultaneously, of 2, 3, and 4.