Spoiler Thread for AU 203

[nbarclay]

I'm up to 1160 AD now, and I've been through my first (defensive) war. In the process, I made a serious blunder that I hope doesn't hurt me too badly.

Persia decided, for some unknown reason, to go through the territory Rome had previously captured from Russia and declare war on me. Their forces for the initial assault? One immortal and one warrior! I'd recently spent enormous amounts of surplus gold (without Leonardo's!) upgrading units that started life as warriors or archers to guerillas, a couple that had started as spearmen and one as a rifleman to infantry, and ones that had started as horsemen to cavalry, so I wasn't in as bad a shape as I might have been. But even so, my entire military consisted of five cavalry, five infantry, and a whopping eleven guerillas.

My response would have been completely sensible under most circumstances: mobilize and buy alliances with the entire rest of the world. (Only China and Babylon required anything other than obsolete techs as a payoff, and Rome was especially stupid to sign on since they had three cities between Persia and me and were totally outclassed in military technology.) But what I forgot was that not only can you not start non-military city improvements while mobilized, you also can't start non-military wonders. And I had two prebuilds in progress, one for Longevity and one for Theory of Evolution!

As events moved on, I figured out my mistake, but my only way out of the box I'd stuck myself in was to make peace with Persia and break alliances with six other civs. Worse, Persia wasn't inclined to talk peace for a few turns, and I lost hundreds of shields when China beat me to Universal Suffrage. The best I could do was to switch Berlin (which had been working on Universal Suffrage) to the Military Academy, abandoning work on the Military Academy as a prebuild in another city in the process. That set me back to square one as far as building Longevity was concerned.

I did, however, manage to delay my palace prebuild for Theory of Evolution long enough to make peace just as it was about to finish. That put me in a position where I now have a lead of four techs. I just hope none of the civs I've left furious with me makes me regret the move.

After making peace with Persia, I actually signed a ROP with them in the hope that they and China would do some damage to each other. But no such luck. Persia and China just made peace, and England seems to be Persia's current victim of choice - not that they've captured or razed any cities yet. (Come to think of it, I probably ought to build some settlers in case Persia decides to open up some territory for me.) In the meantime, I have my terrain pretty thoroughly improved, so I'm using a "wall of workers" technique to seal off my northern cities against a possible ROP violation (since even with the few units I built during the war, I don't have enough to garrison everything and still have a halfway decent border defense).

Persia and Babylon are both in Communism, and Persia is probably in their golden age. Babylon may be in a GA as well, but their bowmen don't hang around my part of the world. China stayed in Democracy, but they definitely used up their GA long ago (due to wonders if nothing else).

In regard to the wonder race, I'm now six turns from Hoover Dam and seven from Longevity. I just hope I have the production capacity to win the Longevity race. I'm also six turns from officially finding out how hard oil is to come by on this map, but I have some ideas for dealing with the situation.

(1) Research Radio before moving on to the techs that actually use oil for something. That buys time for a previously undiscovered oil deposit to show up if I'm really, really lucky. And it should provide time for me to sell both Corporation and Refining (see below) and get a halfway decent price for them.

(2) Unless it's been fixed, there's a glitch in how the AI values resources: it bases what it charges on what you can use the resource for, so if you have Refining but not Combustion, you can (or at least could) buy oil dirt cheap. I'm researching fast enough that it won't take more than six turns per tech by that point, likely just five and conceivably four, which can provide enough time to build at least a few panzers by the time the deal expires.

(3) If my research goes as planned, I should be early in the modern era by the time my early oil deal expires. Once the panzers started while I have oil available are finished, that could be a great time to switch to Communism. With my GA yet to come, I'll be within striking distance of MAs even operating under a Communist regime. (In effect, the extra gold from the GA will replace the gold lost switching from Republic to Communism.)

(4) Of course with me in a communist government and panzers available, obtaining further oil to build more panzers shouldn't be a problem.

Now I just have to survive long enough to implement my plans.

Nathan

[Dominae]

One question Nathan: did you find Mobilization useful? In my game, the extra production in each city was typically not sufficient to reduce the number of turns to completion of Infantry and Tanks (the two main units at the time). For instance, a city producing 60 Shields produces ~75 Shields during Mobilization, but an Infantry unit costing 90 Shields still requires 2 turns to produce. Thus in a vast majority of cities Mobilization would only increase their Shield waste. Did you work around this somehow?


Dominae

[nbarclay]

Originally posted by Dominae
One question Nathan: did you find Mobilization useful? In my game, the extra production in each city was typically not sufficient to reduce the number of turns to completion of Infantry and Tanks (the two main units at the time). For instance, a city producing 60 Shields produces ~75 Shields during Mobilization, but an Infantry unit costing 90 Shields still requires 2 turns to produce. Thus in a vast majority of cities Mobilization would only increase their Shield waste. Did you work around this somehow?

Almost all of my cities are laid out in such a way that they can't hit size 12 without using a mixture of land and aquatic tiles, and I'm not sure any of them were actually at size 12 when the war started. Further, the only two cities I had power plants in were working on wonder prebuilds, since Hoover Dam is intended as my main source of power. Those factors combined to put my cities that were actually building military units in a much lower production zone than what you describe. So to the extent that I actually built military units rather than leaving cities on what they'd been building, yes, mobilization tended to provide an advantage. (I built a whole nine infantry in the course of the war and counted on Persia to live down to the usual level of AI stupidity in managing a military campaign.)

Nathan

[Dominae]

I believe I considered using Mobilization at a much later point than you describe in your game. I built almost no military units from Steam Power to Refining, focusing on the big improvements of the era instead. When war broke out, I used Draft to get Infantry. So when finally came the time to mobilize, my economy could do much better without. Just a matter of timing, I suppose.

But even if a city is producing somewhere around 30 Shields, it needs to be size 15 to jump up to the 45 Shields required to build an Infantry in 2 turns (and all tiles worked must be producing at least one Shield). For this to happen empire-wide is rare, I would think. I'm not criticizing your decision, I'm just wondering about an ability (Mobilization) that I often neglect.


Dominae

[nbarclay]

It's a matter of percentages. Unless most of your cities have practically identical production capacity, some cities will be able to build things a turn faster with mobilization and others won't. Except that you can tilt things a little in your favor by building more expensive units in the same time (e.g. panzers instead of infantry in three turns in your hypothetical 30-shield city) or cheaper units in less time (a city with 35 shields before mobilization building infantry in two turns instead of panzers in three, assuming you don't already have more infantry than you want).

I rarely use mobilization myself because I always have cities where I want to focus on infrastructure rather than cranking out units. But if I get caught in a tight spot, or if I want to crank out a lot of panzers in a hurry, it can come in handy. (Note: I love three-move units, so while chances are high of my mobilizing as Germany to fight a war in the "tank" era, I'd almost never do the same thing as any other civ. And by the time MAs come along, I generally have enough tanks to upgrade that mobilizing seems superfluous.)

Nathan

[nbarclay]

It's now 1315 AD, and the War of Chinese Aggression is over. Hoardes of cavalry appeared just inside my borders, and I ordered them to leave or declare war. Guess what? They declared war, costing me the benefit of a just-signed tech sale worth almost 200 gpt (plus some lesser deals that were nearing expiration).

I'd recently detoured through Radio on my tech path, and I was really glad I did. I quickly built a radar tower and started attacking the stacks of cavalry. Many survived to retreat since most of my attackers were guerillas and infantry, but many more died. China did pick off that radar tower in their turn since I hadn't bothered to defend it (I was spread thin since I still had a ROP with Persia and didn't want to leave cities Xerxes could get to completely undefended), but all three cavalry who went to the radar tower's location died; not the worst trade I've ever made.

After that, China decided that cavalry might not be the best way to attack me, so they sent in a huge stack of foot soldiers. I decided to borrow a page from Theseus's playbook and create a line of forts along the three-tile-wide border with Persia's English conquests that China was moving through. And of course I rebuilt my radar tower, adding still more to my defense. After some dilly-dallying around, China threw an assault at my line, but when their first few attackers died and then one of my infantry was promoted to Elite without injury, Mao decided he'd had enough. The rest of the stack withdrew. I think this is the first time I've ever seen an AI break off an attack in the middle when it wasn't going well. A small amphibious assault force also attempted to strike, and was destroyed at the cost of one veteran guerilla unit.

China then started maneuvering around to try to get to the weak end of my line (which a city blocked direct access to), but by the time they got there, Mao had reconsidered his "No peace negotiations" stand. He agreed to a cease-fire on even terms.

That war was a massive victory for me in military terms, but in economic terms, it was a disaster. The broken trade deals cost probably over 4000 gold and ended my ability to research at 100%. Worse, Babylon took advantage of the diversion to demand Steel from me and I didn't dare refuse with one war already under way. I did, however, quickly trade Steel to Persia for oil so I can start building panzers in a few turns. (It looks like Firaxis probably fixed the bug of dirt-cheap oil before Combustion; neither Persia nor Babylon was willing to sell me oil at any price until I had a use for it. That's really a good thing, but it's annoying that I built some trade plans around an AI behavior that seems to no longer exist.)

Anyhow, I'm at peace again, and three turns away from being able to build panzers. After that, I'll have about twelve turns of oil left before my deal with Persia expires, so I will have Panzers available for whatever I decide to do beyond that.

Nathan

[Nor Me]

Originally posted by nbarclay
(It looks like Firaxis probably fixed the bug of dirt-cheap oil before Combustion; neither Persia nor Babylon was willing to sell me oil at any price until I had a use for it. That's really a good thing, but it's annoying that I built some trade plans around an AI behavior that seems to no longer exist.)

That's interesting because I had no problem doing this (Yes, PTW v1.14 MODded.) You did offer them 1 gold for it, didn't you?

[Catt]

Originally posted by nbarclay
I think this is the first time I've ever seen an AI break off an attack in the middle when it wasn't going well.

I have seen it happen recently, but never so clearly as your situation (a massed force hitting once or twice, and looking for another target).

(It looks like Firaxis probably fixed the bug of dirt-cheap oil before Combustion; neither Persia nor Babylon was willing to sell me oil at any price until I had a use for it. That's really a good thing, but it's annoying that I built some trade plans around an AI behavior that seems to no longer exist.)

I was still able to trade for oil for a pittance pre-combustion. I wonder if other factors could be at play (your broken alliances? AI knowledge of your tech pace and tech choices available to you?)

Catt

[nbarclay]

Originally posted by Nor Me

That's interesting because I had no problem doing this (Yes, PTW v1.14 MODded.) You did offer them 1 gold for it, didn't you?

I offered a lot more than that, and even asked them what they would need for a deal. Maybe the situation was a result of my having broken the alliances earlier, although the fact that I was able to make a deal for oil later really had me hoping the problem was fixed.

[nbarclay]

Originally posted by Catt


I have seen it happen recently, but never so clearly as your situation (a massed force hitting once or twice, and looking for another target).

Actually, it was more complicated than that (or than I described initially). First, a huge stack (about 35 units, including a couple old medieval infantry) came up to the northwest corner of the line. I responded by moving in every infantry I felt I could afford to, about nine total, and hoping it would be enough with the fortress and radar tower bonuses. The AI stack backed up momentarily without attacking, but then came in again, albeit with a few less units for some unknown reason - maybe withdrawing some to load on ships for naval invasions? Rather than attack the northwest corner, they moved down into position to attack the middle of the line. By then, I'd built a few more infantry, and I scrounged a little deeper as well, so I had (I think) ten infantry in one defensive stack and nine in the other. It was probably something like the fifth straight unit lost by the AI that resulted in my infantry's being promoted to elite without being injured. That left both of my stacks that the AI could get to immediately with an elite unit rather than a veteran on top.

But the third fortress in the line only had two infantry guarding it. It was on the other side of a lake with a city behind the lake, (see the screenshot below) so China couldn't get to it immediately (at least unless they got a ROP with Persia all of a sudden). Rather than keep throwing forces into what was clearly a losing fight - China might possibly have won the battle, but only at a cost of the vast majority of their stack at a horribly unfavorable loss ratio - China decided to go around to my weak side. The stupid thing about it was that by the time they could have gotten there without a ROP agreement with Persia, I would easily have rebalanced my forces so it wouldn't be weak anymore, but I'm impressed that the AI recognized that the initial attack was going badly enough that continuing it wasn't particularly viable.

The Chinese forces seen in the picture below are just two units, but I think the rest of the stack was right behind them.

Nathan

[Catt]

Originally posted by nbarclay

Actually, it was more complicated than that . . .

I've really only seen (1) AI forces react to strengthening particular defenses (i.e., AI chooses a new target when its initial target's defensive profile has changed considerably) and (2) seemingly call off an assualt when the first several units die without much success. Your example is still more clear than my experiences.

I obviously have no idea if its just coincidence or something else at play (those little AI tweaks that make it into patches?) but I like to believe that the "education" of barbs in PTW -- i.e., attack only with reasonable chances otherwise look for a more inviting target -- may be influencing AI civ wartime target selection and battle initiative, so far at least for the better (though with long strides yet to go -- the shifting of offensive force in your example over the course of several turns even though your potential counter-maneuvers can be done in 1 turn).

Catt

[Jawa Jocky]

It sounds like the AI re-evaluetes after every move.

At the start the attack was good because the overall odds were in its favor. After a couple of losses the odd are no longer in its favour and it stops throughing units at the line. At some point the AI crossed a line in its evaluation and the switch to attack was toggled off.

[Sinapus]

Originally posted by Jawa Jocky
It sounds like the AI re-evaluetes after every move.

At the start the attack was good because the overall odds were in its favor. After a couple of losses the odd are no longer in its favour and it stops throughing units at the line. At some point the AI crossed a line in its evaluation and the switch to attack was toggled off.

Which actually sounds rather good. I know I'd break off an attack if I kept losing tons of units after an assault with no real damage to the enemy.

[nbarclay]

In the early 1300's, the Two Centuries War started when Persia violated an oil trade deal and invaded. Germany quickly made arrangements for an alternative oil supply from Babylon, and the first panzers went into service in 1340. Wave after wave of Persian infantry crossed the border over the ensuing decades, only to be softened up by artillery and mowed down by panzers. Southern Germany became a training ground for what would eventually be over two dozen elite units.

With the war came Germany's golden age, just as Germany was about to enter the modern era. Germany set aside its plans to change to a communist government and conquer the world with panzers, and chose instead to research toward an even more advanced line of armored vehicles. First came Computers and mechanized infantry. Then a detour through Miniaturization for increased production and the Internet (since getting a leader would be no problem at all). Next, Fission for the wonders that come with that (except it looks like you can't build the U.N. with diplomatic victory turned off; is that right?) And then on to Ecology and Synthetic Fibers.

While all this was going on, though, the world changed. China decided to attack, and wave after wave of bombers punched massive holes in the German infrastructue, joined shortly by large numbers of warships. China never mounted serious assault on the line of forts (although an occasional solitary unit asked my mechanized infantry for help committing suicide), but a couple attempted naval invasions had to be wiped out.

With the new Chinese problem, peace with Persia started looking like a good idea. But not too long after that peace was established, Babylon got into the act. That gave me two major naval and air forces to contend with instead of just one. Babylon pounded my infrastructure from the eastern side while China continued to do its dirty work from the west, and even with my jet fighters shooting down some of the bombers, I lost infrastructure faster than I could repair it.

Shortly after Babylon entered the war, China was willing to make peace for about 200 gold. That let me focus on Babylon for a few turns until they would be willing to make peace (also for some gold), but by the time that happened, Persia decided to attack again. So the end of the Two Centuries War was much like the beginning, with Germany fighting against Persia.

(Thiis is where my earlier broken alliances really cost me. Had I been able to, I would have gotten alliances with China and Babylon when Persia first attacked and tried to get those two powers to waste a lot of their forces slugging it out. But China and Babylon have consistently refused to even consider my attempts to propose an alliance. They'd both rather fight against me.)

Of course the new war wasn't quite like the old one. The German people were tired of always being on the defensive, and if the leaders of their republic were not interested in changing things, well, there were others in Germany who would. Almost four decades of anarchy erupted as the Communist Party made its bid to wrest power away from the foolish peacenicks who had been in charge, but at long last, they succeeded.

The first item on the new government's agenda was securing an oil supply. Fortunately, Persia had previously conquered all of England and Russia except for a distant English island city, but Persian culture had yet to take hold in most of the area. That made it a simple two turns' work for elite panzers to sweep down and capture the cities on either side of the Russian oil field, and a band of German workers established a colony to operate the oil wells until the area could be brought more securely into the grip of German culture. And faster than Xerxes could blink, over forty panzer batallions (elites as well as veterans) were reequipped and retrained to use modern armor.

From that point, terrain and cultural borders were the only real limiting factors on Germany's conquest of Persia. New units went into operation faster than old ones could be destroyed, and city after city fell before the German onslaught. Just forty years after Germany's new government took power, Persia was no more. (With Persia's culture somewhere in the neighborhood of one and a half times mine, I didn't dare leave them alive.)

That was the easy part; now comes the hard part. While Germany was busy conquering Persia, China and Babylon all but caught up in technology. As of the time of the Persian conquest, Miniaturization and Synthetic Fibers were the only technologies I had that China and Babylon didn't, while they had Space Flight (which I was four turns away from) and Amphibious Warfare. That means I can expect to be up against an opponent with Mechanized Infantry and an economy that rivals mine for the first time ever. Worse, China has Leonardo's for cheap upgrades and Universal Suffrage to help it combat war weariness (although it could be worse; at least the wonders are in cities relatively early on my target list).

Against those odds, I'm trying a gamble that I hope pays off. I traded Miniaturization to Babylon for the techs I was missing, just over a thousand gold, wines, and 249 gold per turn. The gold will help me on my way to Satellites so I can compete in the ICBM race if it starts. (I'm starting to regret having built the Manhatten Project earlier, but I had a leader with nothing better to do.) But the sneaky part is that I'm actually hoping Babylon will turn around and sell the tech to China. I can't bleed off China's gold per turn after the war starts, but Babylon can. So if my plan works, when I invade after the current peace treaty expires in another four turns (or possibly sooner if I provoke a war with a failed spy mission), maybe, just maybe, China will find itself a bit short on gold to upgrade infantry to MIs. It seems worth a try, anyhow.

By the way, mobilization during my GA seemed superfluous, but shortly after my GA ended, I went into mobilization to crank out units faster. With the Persian war ended, I'll spend a couple turns out of mobilization to get other projects started (especially libraries in the former Persian territories) and then mobilize again for the upcoming war with China. Against an opponent that has MIs, may get MAs soon, and outnumbers me, I figure I need every edge I can get.

Nathan

[nbarclay]

By the way, during the Persian war, I noticed a serious oddity. Often, Persia's radar towers were defended by tanks rather than by infantry. I'd be curious to know what kind of programming logic results in such behavior. One possibility that occurs to me is that Xerxes might have thrown away just about all the infantry he could spare from city defense attacking me so tanks were what he could scrounge up to guard the towers when I blitzed through his territory so quickly.

Nathan