Civ4 AI Survivor Season 3: Game One Writeup

This was the first game of Season Three, and I was hoping for an entertaining product after the long hiatus from Season Two. As it turned out, I shouldn't have worried about a lack of interest in the contest. We ended up with 115 submissions for the picking contest, and close to 100 live viewers while the game was taking place on stream. This particular game pitted a group of four leaders with a low peace weight ("evil" leaders) on the outsides of the map, against two leaders with a high peace weight ("good" leaders) in the center of the map. Ragnar, Stalin, Willem, and Kublai Khan would all be predisposed to like one another, and similarly predisposed to dislike Ramesses and Augustus. Most of the submitted predictions anticipated this map playing out in the same way that I did: the two leaders in the center of the map would get crushed, and then it would be a contest between the aggressive leaders around the edge to see who would come out on top. Ragnar and Willem were the two favorites, with Ramesses and Augustus overwhelmingly being the top picks in the First to Die category. Needless to say, things did not quite play out the way that most of us were expecting, and that's the fun of running these games.

One of the most unpredictable aspects of the AI Survivor games is where each leader will choose to send their second starting Deity settler. They will always use one of their two settlers to found their capital on the starting tile, and then it's essentially random where the other settler chooses to go. In this game, most of the AI leaders made logical decisions. Willem settled due south while Augustus went straight north, both of them immediately pushing to claim disputed parts of the map. Stalin and Ramesses also picked decent spots near their respective capitals. Kublai chose to send his initial settler down towards the tundra, however, and Ragnar made the worst choice of all, moving northwest onto the coast in the precise opposite direction from where he should have been going. This was a bad early sign for the Viking leader, the most-picked choice to emerge as the winner.

None of the leaders in this game began with Mysticism tech, turning the religious race into a complete wildcard. Kublai had the good fortune to pop Mysticism from a goody hut, and he founded Buddhism in his second city on Turn 16. This would allow the Mongol city of Beshbalik to dominate culturally in the southwest part of the map, and along with Kublai's Creative trait, the Mongols would control most of their western border with Rome. Willem would go on to found Christianity (the Polytheism religion) along with several other religions later on the tech tree. He had more of a religious focus than these other AI leaders, and as a result, there were very few competing religions in this game. Christianity would spread to most of the other AI civs over time, and that kept religious conflict to a minimum, aside from isolating Kublai somewhat from the other Christian leaders.

Fifty turns into the game, some early winners and losers were beginning to emerge. The biggest surprise was Ramesses and Augustus both doing considerably better than expected. Ramesses had five cities established, all of them pushing into contested parts of the map. He was encroaching on territory that likely should have gone to Willem, and had settled a preposterous city only four tiles away from Ragnar's capital that the Vikings surely should have claimed first. Augustus was in a similar position in the south, also out to five cities in a circular pattern around his capital. He even had a sixth settler about to grab the stone and iron at the bottom part of that screenshot. If the Mongols hadn't been so culturally dominant, Augustus might have been the game leader at this point. Kublai and Stalin were also doing reasonably well, with the Russian leader building the Great Wall and with the Pyramids under construction. Stalin also had that little backline peninsula to settle at some later date, once he could claim it from the barbarians, who would eventually get two cities in that back corner.

It was Willem and Ragnar who were struggling at this point. Seen purely from a score perspective, Willem looked like he was doing great. He had a self-founded religion and Stonehenge, plus was about to found a second religion from reaching Monotheism tech first. However, Willem's score was mostly an illusion, propped up by Creative culture and the Stonehenge build. The Dutch only had three cities and were getting crowded out of the important parts of the map. Ragnar was in even worse shape, stuck at four cities and already essentially out of room to expand. Both Stalin and Ramesses had planted cities near the Viking capital, while Ragnar was off settling the northern tundra. Ragnar had spent too much time building units and not enough time expanding. He was going to need to do something quickly or he would fall too far behind.

I was expecting Ragnar to attack Augustus, his "worst enemy", or more logically his next door neighbor of Ramesses. Instead, Ragnar threw a monkey wrench in this whole competition:

He decided to attack Stalin! Wow. This came as a huge surprise; with Ragnar and Stalin having very similar peace weights, they were predisposed to be allies and a lot of the viewers were expecting them to go conquering together. This didn't even make sense from a religious perspective, as Ramesses had Christianity to compete with Ragnar's Buddhism, which should have driven them into conflict. I think this goes to show that no matter how much analysis we subject these game to, there's always a random element... usually caused by some overly aggressive AI doing something insane. This was the single biggest event that took place in Game One, and it was the falling domino that triggered everything that followed.

Needless to say, Ragnar's push with horse archers and axes against the city of Novgorod and its +50% city wall defenses did not succeed. Ragnar lost his initial striking force was forced to play defense thereafter, with the Viking city of Birka not even controlling its own first ring tiles. That's a recipe for disaster, and Birka was taken by the Russians on Turn 82. Making matters worse, Ramesses piled on and entered the war a half dozen turns later, pushing a stack of war chariots towards Ragnar's capital. The Vikings were reduced to a mere five cities, facing off against two other leaders with a combined fourteen cities. For the pregame favorite to win, this was a disastrous opening indeed.

Meanwhile, on the other side of the world:

Kublai Khan declared war against Willem with the goal of capturing Rotterdam. The Dutch city was hopelessly surrounded by Mongol culture, and it was a surprise that Willem had even managed to get a settler down to this location in the first place, lacking Open Borders with Kublai. As soon as that city was planted, war had been inevitable between Willem and Kublai. It was another unexpected conflict as the low peace weight of those two leaders should have made them allies, at least until Augustus and Ramesses were eliminated. This was not going at all how most of the entries in the picking contest expected. Rotterdam itself was toast, of course, and Kublai looked like he was in a good position to make gains against the weakly held Dutch cities to the north. Willem had a lot of cultural stuff but was dead last in military power. However, Augustus was pulled into the war two turns later, in what looked to be an alliance with Willem, setting up the dreaded two front war for the Mongols. Kublai's enormous culture allowed him to hold his cities for the moment, but any chance of making gains against Willem was now over.

This was the situation on Turn 100. Ragnar had lost another city to Stalin, and the vultures were beginning to circle around his capital. Which leader would manage to take Nidaros, Stalin or Ramesses, would be hugely impactful in terms of who would emerge from this war as the dominant AI power. The eastern war was stalemated for the time being, with lots of units traded on both sides but no cities falling. Augustus had two sources of iron though, and praetorians were going to make things difficult for Kublai the longer that war lasted. I continued to be amazed at how well Ramesses was doing, with eight total cities and a lot of land in the far north of the map that Willem probably should have claimed. The Egyptians were actually leading the game in score, at a time when Ramesses was on the verge of being eliminated in the first two season of AI Survivor. This was a much, much better showing for Ramesses than we had seen in the past.

Ragnar had gone for Construction tech at an early date, and the presence of two ivory resources allowed him to build war elephants long before anyone else. That kept him above water for a while, but eventually the weight of greater numbers began to take its toll on the Vikings. Nidaros was put under siege and eventually fell to Stalin, a huge pickup for the Russians. Stalin would win the race to Uppsala as well, at which point in time the Russians split off to the south while the Egyptians curled around to the north. Ramesses captured the one city up there in the tundra, while Stalin claimed the last remaining Viking city for the final elimination:

Ragnar was the First to Die on Turn 132. Only 10 out of the 115 predictions managed to get that scoring category correct - this was quite an unexpected result. It had all compounded off that initial settling decision that put Ragnar behind in the expansion race, followed by Ragar's reckless war against Stalin that backfired in the worst way. Once Ramesses joined the conflict, the writing was on the wall and Ragnar was sent home with an early elimination. The consequences of this war would also be far reaching. Stalin emerged as the dominant power in the west, not only claiming five of Ragnar's six cities, but also still having those barbarian cities in his back lines to conquer. Stalin also had a bevy of useful wonders under his control, including the Great Lighthouse, the Mausoleum, and the Statue of Zeus. He was now the favorite to win the game. Ramesses had also done quite well for himself, taking one city from Ragnar and establishing a very positive relationship with Stalin. Those two had Open Borders, the mutual military struggle bonus, and a whole bunch of resources traded back and forth. By forging this strong relationship with Stalin, Ramesses had put himself in an excellent position to sit back and tech to some kind of peaceful victory condition.

The east had remained a military stalemate for a long time. Kublai signed peace with Willem after a while and we were left with a lengthy struggle between Mongolia and Rome. Augustus had the edge with his praetorians but couldn't break through and take any cities because of the immense culture pushing out from Kublai's western cities. Eventually Augustus managed to unlock elephants thanks to an ivory tile at the far edge of his borders, and that seemed to swing things in his favor. Kublai was fielding large armies of keshiks, and they were no match for the elephants and their bonus against mounted units. Augustus eventually managed to take Beshbalik, the Buddhist Holy City, which greatly damaged Kublai without adding much of a benefit to the Romans. (The city was culturally crushed and had no tiles it could work in its 21 tile radius.) On Turn 146 the Dutch jumped back into the war again and recaptured Rotterdam, but that was followed by Kublai signing peace with Augustus, and a new stalemate emerged between Mongolia and the Netherlands. Lots of units dying on both sides, no territory changing hands.

The long term effect of this warring was to cripple all three of the eastern civs. Stalin and Ramesses stayed at peace for long centuries, teching along and building lots of infrastructure. They quickly ran out to a huge tech lead, eventually getting more than an entire era ahead in research. Since both of them also had larger empires to begin with thanks to absorbing former Viking territory, the gap between east and west continued to grow as the turns passed. Kublai in particular was really taking it on the chin, trapped in these wars that never ended against first the Dutch, then the Romans, then the Dutch once again. Mongolia was falling badly behind in science, even compared to the two eastern leaders. Kublai staged several long sieges of the Dutch city of Utrecht, which inflicted ghastly casualties on both sides but never managed to capture the target. While we kept watching this endless war in the east, there was very little of substance taking place aside from the slow bleeding out of everyone involved.

The game finally began moving again when Stalin declared war on Willem on Turn 216. He crushed the Dutch border cities and the march of the cossacks was on. Willem was just entering the Renaissance and had absolutely no answer for the swarms of cossacks and rifles and grenadiers that overwhelmed his territory. Stalin cleverly used his Open Borders through Egyptian territory to speed up his advance, and the Dutch resistance simply collapsed in a matter of turns. Once city after another fell before the tidal wave of red units. Amsterdam was taken on Turn 227, and Willem was completely gone by Turn 235:

That was fast. After such a long time spent watching the grinding inaction of the eastern civs, this lightning campaign suddenly had the game in flux once more. Several of Stalin's new conquests were under heavy pressure from Egyptian culture, and we wondered if that would draw Stalin into potential conflict with Ramesses. They had been great friends throughout the game thus far and still enjoyed "Pleased" relations, and yet Stalin will declare war at "Pleased" so it wasn't out of the question. If Stalin did choose to pick a fight with Ramesses, we didn't see it ending very well for the Egyptians, not with the power bar graph looking like this.

Further to the south, Augustus had renewed his war with Kublai one more time, and with Kublai greatly weakened from so much fighting, the Romans pushed through and captured the Mongolian capital on Turn 225. Augustus was using cuirassiers and was about to graduate to rifles and cavs, while Kublai remained stuck with medieval troops. The Mongolians were too beaten down from endless warring, and as Augustus continued to capture cities, they were only growing weaker still. It soon became clear that Kublai would be the next one eliminated from the game unless he could get one of the western civs to intervene in this conflict. Ramesses was unlikely to do anything, but Stalin was "Annoyed" with Augustus, and it looked like only a matter of time before he entered the fray. On Turn 248, the Russian leader sent his cossacks raging across the Roman border:

Needless to say, this was not good news for Augustus. Ravenna fell on the initial turn of the war, with Antium and Mediolanum both dropping on the subsequent turn. The cossacks were off and riding again, and even though Augustus technically had units at the same generation of military technology, Stalin's enormous edge in territory and production gave him an insurmountable edge. For that matter, the remnants of Mongolia were also still fighting against the Romans, snapping at the heels of Augustus. The Roman leader needed Ramesses to intervene on his side if he were to have any chance, and Ramesses didn't seem interested. The Egyptians were building Research in every city and pushing ahead in technology, not caring about the turmoil raging to their south. Lacking allies, it was only a matter of time as each Roman city was run over one at a time. The city of Rome was gone by Turn 260, and it was all over on Turn 275:

In the end, I was correct that Augustus' high peace weight would prove to be his undoing in this game. It just happened at a much later date than I expected, with this whole match thrown off kilter by Ragnar's early declaration of war against Stalin. That had led to the low peace weight civs on the coasts fighting against one another rather than teaming up against Ramesses and Augustus, allowing the latter to survive until the Industrial era and the former to emerge as the tech leader. That was the surprising result of this most recent conflict: Stalin now had a massive advantage in production and territory, but Ramesses had surged far out in front in technology. He had built the United Nations and all of the Modern wonders, and was about a dozen techs ahead of Stalin. We were all stunned when we looked at the respective Tech Trees and saw how far ahead Ramesses was as compared to Stalin. In other words, Stalin's victory was not a foregone conclusion after all. If this game stayed peaceful, Ramesses was going to win via Spaceship before Stalin could get there himself.

The drama now was whether Stalin would make a move against Ramesses before the Egyptians could finish the rest of the tech tree. Stalin remained much higher in military power, but that was a bit of an illusion. Stalin had built a massive navy, close to 200 destroyers by our best count, and all of those ships would be completely useless against Ramesses' landlocked empire. Stalin had built all those ships because there were two barbarian galleys trapped in pockets of ice, and something in the AI causes them to go nuts building ships so long as there are barb ships on the map. Amusing though this might have been, it represented a gigantic waste of resources by Stalin. The chat discussed this and we decided that future maps will be set up to avoid any such pockets of ice to prevent this from happening again. We left the current game alone though, and Stalin kept on building those destroyers at a frenzied pace.

As Turn 300 approached, Ramesses continued to inch closer to the spaceship. He had already completed the Apollo Program, and his research had carried him to Robotics and mechanized infantry. It was starting to become questionable as to whether Stalin could win a military victory at all, even with his big advantage in territory. Russian infantry wouldn't do very well against Egyptian tanks and mechs if it came to another war. Then on Turn 295, Stalin decided to call a Diplomatic victory vote in the United Nations. Ramesses had built the UN earlier and Stalin had won the Secretary General vote, but a diplomatic victory had been out of reach because Kublai wouldn't vote for Stalin. They were still practicing different religions and that dragged down their relations a fair amount. However, Stalin had gone into Free Religion since the time of the last vote, and that had boosted his relations with Kublai into the "Friendly" range. Would that be enough to put him over the top?

Yes it did! We therefore wound up with a very unexpected Diplomatic victory for Game One, with Stalin taking first and Ramesses taking second. Kublai Khan survived to get another chance in the Wildcard game after the end of the opening round. Only 6 of our 115 entries chose the Diplomatic victory condition, indicating how surprising this proved to be. For some reason my screenshot failed to take, and I had to pull this ending image directly from the Livestream footage. What an odd game this had been.

Strangely enough, there had been a lot of movement of Russian units on the interturn before the victory message popped up, and I wondered what was going on. When we watched the replay after the game ended, we discovered that Stalin had actually declared war on Ramesses on the same interturn that he achieved victory. Wow! We had never seen that happen before in one of these games. Because the war declaration technically occurred on the turn before the victory message appeared, we decided to include that war declaration in the scoring, ending up with 11 total wars instead of 10 wars. Out of curiosity, we also played out another dozen turns after the victory to see what would happen. The result was a stalemate, as it turned out, with Ramesses and Stalin each taking some of their respective border cities from one another. Ramesses had the quality while Stalin had the quantity, and they were trading power fairly equally as far as the bar graphs were concerned. Since there was no immediate winner here, we eventually stopped this exercise to set up the map for the next game. I genuinely don't know which of these two leaders would have won the game if the Diplomatic vote hadn't taken place.

Special congratulations to Commodore, who won the picking contest for this game with a score of 13 points. It was a bit of bloodbath for everyone as far as scoring was concerned, with the average result coming in at only 4.50 points. Thanks for reading, and we'll see what the next game holds.