CHRONICLES

RBCiv Epic Forty-Seven

In 1545AD, Korea sneak attacked India. They originally sent an attack force against the Indian core, but not nearly enough to get anywhere. Indian counterattacks wiped these out, and before Korea could send more, Indian troops were dispatched to attack weakly-defended Korean colonies in former Egypt. This pulled Korea's puppet strings, diverting their attention to responding to Indian threat to their cities.

Quite nearly the entire war between India and Korea was thus fought over a bunch of junk cities.

I feared that India would buy in Persia against Korea and then Korea would fall to a new dogpile, but interestingly that did not occur. Korea bought in Persia. Five turns later, India bought in Incans and it was 2v2, the strongest and weakest AI vs the two in the middle. Incans lacked rubber and were the lowest AI on the tech totem pole, of the four survivors.

The most interesting thing of note here is the way that Persia and Korea were chained to attacking weak Incan targets, giving strong India a complete pass. India had bought themselves a meat shield and were now completely insulated from harm, except where their attacks on Korean or Persian cities pulled puppet strings and drew a reply.

Here you see Persian tanks ignoring Indian offensives into their core, instead shipping the Persian tanks across the continent to attack irrelevant stacks of obsolete Incan units merely because the Incans looked like the easiest pickings on a one-unit-to-one-unit mathematical basis.

Persia has lost three core cities but doesn't care to try to take them back, because the Indian defenders are "strong" and Persia is chained to targeting the weakest options.

So now Indian cities are all but invulnerable, not being targetted much, while both Korea and Persia concentrate on attacking Incan forces and cities. INDIA IS GUARANTEED TO WIN by the hardwiring of the AI, and it has been so ever since they bought Persia in on their side vs Carthage.

Instead of the AIs forming up realistic fronts along their main border, both sides have sent most of their armies to the wasteland of former Egypt to fight over strategically useless bits of land.

I click next turn and all of Persia's brand new tanks arrive, still chasing Incan pikemen while India thrusts into the Persian core.

There are three primary strategic imperatives to the Civ3 AI that plunge it into a tactical quagmire.

1. The offensive force picks one city at a time to target. The selection is made on the basis of the current strength of the best defender, without regard to distance between the attacking force and the target. Therefore, the human can make the AI dance to any tune he likes by shuffling his defenders. Move a strong defender into a threatened town and move all strong defenders out of a town somewhere else and the AI's ENTIRE STACK will abandon its attack on the town with the stronger defender and shift its target to the one that is now guarded by weaker units. In theory, a player could keep an entire AI army locked in a frozen dance forever by setting up oscillating targets, just by moving defenders in and out of towns.

2. When one of the AI's cities is threatened by enemy forces moving in to attack position, the offensive force "turns around" from whatever it may have been doing and the whole thing tries to go rescue the threatened city. The human can pull these puppet strings by parking a few otherwise nonthreatening units near an AI's city, and moving them in and out of attack range.

3. The AI prefers in general to attack targets with the weakest defense strength. The human can exploit this by baiting the AI with soft targets, to various ends.

Thus these are the AI's objectives. Lacking objectives of strategic value, the AI haphazardly throws its forces around the board chasing tidbits of tactical advantage. This can appear to work well in AI-on-AI action, but even that turns out to be an illusion.

Inca has no Rubber and so can only make cavs and rifles. They've got all kinds of obsolete units that never got upgraded because they were fortified away from cities at the end of earlier wars.

Poor X-man is slave-chained to wiping out all these "soft targets" in the west. Turn after turn after turn, his tanks are sent out to kill yet more pink Incan musketmen, pikes, and rifles. These targets present the best tactical odds for winning, from the view of one unit to one unit. X-man is ripping poor Incans a new one in terms of number of units killed.

Unfortunately for him, while he off chasing pink units on the other side of the world to rack up a high body count, HIS CORE IS FALLING APART.

India doesn't have any obsolete opponents. India is fighting Persia and Korea, who both have rubber and are making infantry and tanks, or at worst (in Korea's case) cavalry. There are no musketmen running around that India would be COMPELLED to go wipe out, simply because they are weak on defense value.

So India is driving at Persian core cities and attacking. WHEN THEY FAIL, Persia has its puppet strings pulled and will wipe out any Indian units near its cities. WHEN INDIA SUCCEEDS in taking a city, if they can manage to move a strong infantry unit in there and fortify it, Persia WILL NOT COUNTERATTACK because they will instead target cities whose best defender is not as strong... namely either Pink's cities or some just-fallen city with a wounded enemy unit on top.

Persia is therefore doomed, and there are no two ways about it. They will continue to send their forces across the planet to chase pink targets, because pink's units are weaker, while India gradually rips out their throat.

This war started when Korea declared on India. Since these two civs are neighbors, human logic would dictate the bulk of their armies would clash at their border.

(This is how it works in GalCiv, for instance. The AI space navies charge at one another and fight in the space BETWEEN their two civs, and the first one to push through to the other side's core is going to win, because as soon as production centers start to fall, a side is going to collapse.)

However, in Civ3, all the core cities are "well defended" and the AI will pass them up to target the enemy city with the weakest "top of stack" defender. Since Korea had several scattered colonies in the far west, where Egypt got wiped out, India DID NOT ATTACK THE KOREAN CORE AT ALL. They sent all their forces via Right of Passage with Pink around to the west to attack the softer targets of the Korean colonies. Korea -did- attack India's core, but with Indian units pulling their puppet strings in the west, at their colonies, on most turns, Korea never managed to make a sustained push into India's core.

Thus:

1. India was compelled to attack Korea's colonies because they had weaker defenders on top, and because they were smaller and lacked city defense bonuses.

2. Korea attacked India's core, because ALL Indian cities were large and well defended. However, Korea had its puppet strings pulled and was compelled to go defend its threatened colonies in the west.

3. Korea and India clash in the west, and each turn of fighting leaves some wounded units (survivors) out in the open. These weak units draw attacks (which, generally, is good) and so the armies are no longer even fighting over cities, but mopping up one another's wounded survivors from the last turn. Only when all the wounded are dead do the city attacks re-commence. India has stronger and more numerous units so they are slowly making gains.

Thus nearly all of the fighting is centered in the west. Both sides have their entire armies over there, clashing over junk cities. To the human eye, it's one of the silliest ways to play a strategy game that one could imagine.

Once the dominoes start to fall, the next link in the chain becomes easy to predict. Who lines up against whom is largely a matter of dice rolls. An AI makes a threat, or it rolls up a surprise attack. Once two AIs are at war, they will buy in allies. Usually this means one side dogpiles another, but every now and then a more even matchup will occur. YET EVEN THEN, the phenomenon seen above will doom somebody. One side will be driving at another's core while that civ's army is off on some wild goose chase. Then one city at a time, the loser will be wiped out, unable to respond.

I gave Korea a Right of Passage to help them out. They have better mobility than India as a result, and that has prolonged a lopsided fight.

Now that Pink is in the war, Purple's cities are completely safe. Both Green and Blue are compelled to attack Pink's cities because of the tactical odds. Thus Purple has pulled a strategic coup by accident. They have obtained a meat shield to feed to the bear that is Blue+Green, and can attack more or less with impunity when their puppet strings are not being pulled in the west, at the captured colonies.

However, even India is still chained to the coding logic. Their territory is bisected by mine, at the orange circle shown below.

Their homeland is in the east, and they single-handedly captured all the lands of a now-dead civ in the north. Keeping an eye on that orange circle shows MASSIVE numbers of purple troops moving back and forth across that narrow isthmus of my culture (just south of Troy) almost every turn. Their troops are trying to attack green's core in the east, but are forced to respond to any puppet strings pulled in the west. As a result, a fairly large number of Indian troops are stuck in the west, near their colonies, and so even now most of the war is occurring in the west.

The six blue circles shown above were Korean cities when this started. They have changed hands so many times... The two yellow circles were Indian cities when this began.

Human logic would lead to more rational strategy.

In real life, the white arcs in the east would represent the natural front, with armies forming up to face one another. India would gather a stack and try to make a thrust along one of the yellow arrows, hoping to catch Korea with its forces evenly spread and reinforcements unable to arrive in time to prevent a thrust from taking (or destroying) a city. Korea would try to make a thrust along one of the blue arrows. Or perhaps one side or the other would try a flanking maneuver "around the corner" of the main lines.

Instead, the armies are off in La-La Land chasing the AI's coding logic all over the map.

In the absence of strategically chosen objectives, coding logic and other haphazard forces will choose objectives anyway, and players will decipher the elements involved and figure out ways to exploit them.

This is as far as I got in my game. The only combat I saw to date was my one-city attack on OCC Carthage. India now has bombers, so I dare not make a move until I can get some air defense. On this map with armies of this size, a couple of dozen bombers could wipe out my defenders and their cavs waltz in to take over, carving off chunks of my territory, me unable to do anything about it. So... I'm just watching.

I may decide to continue this game. Although I had nearly caught up in tech when I entered the Industrial age, the AIs have left my self-research no-tech-trading butt in the dust. The DEGREE to which I've again fallen behind while the AIs were not warring is the biggest surprise (to me) of the entire test. X-man is in the modern age now, and I'm still researching Refining. I'm half an age behind again.

As Kylearan said, this scenario turned out to be harder than expected. I knew I was in for a rough road with minimal tech trading and avoiding wars, but I enjoyed dissecting the various AI behaviors.

If I do play some more, I may or may not report on it, depending on what I learn. Thanks for tuning in. I hope you enjoyed it.

- Sirian

| Back to Chronicles | Return to the Main Repository |