Civ4 AI Survivor Season 3: Game Six Alternate Histories

(Note : "A" column tracks the number of war declarations initiated by the AI, "D" the number of times the AI is declared upon, "F" the points for finish ranking, and "K" the number of kills.)

Amicalola: Coming into this game with different starting techs and without the Apostolic Palace, I wasn't entirely sure what to expect. Would Caesar be the dominant leader consistently, or would someone else turn out to be favoured under the new setup? The answer to that question was a resounding 'yes,' as Elizabeth turned out to win these games more than half the time, with no one else able to achieve more than a smattering of random victories. Not only did Elizabeth win frequently, those wins were also quite dominant, almost all of them coming before Turn 300 (quite early for AI Survivor), before anyone else was even close. Often, the warlike AIs in Caesar, Brennus and Cyrus would devour Pericles and Qin (or each other!) and have snowballs cooking, but Elizabeth was able to stifle whatever their plans were by just sort of... winning. With Elizabeth winning around half the time, that left everyone else fighting for scraps, and no one could pull above the rest of the pack. It also meant there were few kills, as many games ended with 4-5 leaders remaining, instead of the usual ~3.

So why was Elizabeth so strong? In brief, it was because she was left alone early on. Zara Yaqob and Caesar near-inevitably fought a lengthy duel in the west, and regardless of who won, this gave Elizabeth a safe western flank in most games. Cyrus and Brennus would often clash similarly in the east, and when they didn't it was Pericles or (infrequently) Qin who fell to their combined efforts. Only Qin attacked Elizabeth semi-consistently before Turn 150, but as the results show he was extremely unlikely to perform well in that scenario, with Real Game Six apparently an outlier in this regard. Finally, Pericles was a total no-show on this map: with nearly zero strong performances, he was generally massacred in a Celto-Sino-Roman-Persian feeding frenzy.

When Elizabeth faltered, pretty much anyone else was capable of putting up a strong performance, depending on what was happening across the rest of the map. Sometimes, it was Brennus that had eaten Pericles, or Cyrus that had eaten Brennus, or either Caesar/Zara who'd won their little western duel. The only leaders that put up few strong performances were Pericles, who was just dogpiled into oblivion, and Qin, who had a tough central position and was unlucky to be Elizabeth's primary opponent early on. This left none of the other leaders able to claim a convincing second place over these 20 games, and other than Pericles/Qin they were probably all plausible results for that 'honour' - it just happened that Cyrus got it this time.

Finally, Elizabeth's strength in these games was fascinating from a viewer's perspective. Time and time again, there would be interesting developments across the map, and several leaders jockeying for the pole position in what looked like an unpredictable game... only for Elizabeth to claim victory before any of that promising setup could reach its payoff. This was most apparent during the wacky Game 17, where Pericles donned his best Alexander cosplay and bum-rushed Caesar on Turn 49 with phalanxes against warriors and chariots, only for Elizabeth to win before he could make anything of that remarkable feat. It happened repeatedly though, with nearly every leader getting screwed at some point or another. Thus, over the course of these games, particularly in the final stretch after Game 11, Elizabeth transformed from a plucky underdog into the game villain. It was a reminder to be thankful for the pacifists' general weakness - one Elizabeth win is a fun deviation from the norm, but 11 of them became not only boring, but actively frustrating.

Now for a look at the individual leaders:

Leader Summaries

Elizabeth of England
Wars Declared: 22
Wars Declared Upon: 61
Survival Percentage: 65%
Finishes: 11 Firsts, 1 Second (57 points)
Kills: 4
Overall Score: 61 points

Elizabeth was the strongest leader on this map, and by the end it wasn't very close. She won over half the time, and those games were rarely nailbiters. Elizabeth wasn't necessarily left alone either, with the most defensive wars of anyone in the game at 61, and a not-insignificant 22 wars started herself. Almost all the time though, her only early opponent was Qin Shi Huang, one of the weakest leaders on the map, and he was generally incapable of derailing Lizzie's cultural battering ram without outside assistance, as evidenced by Elizabeth only dying first in one game (where she was dogpiled unusually early). Julius Caesar and Zara Yaqob were too busy with each other to provide any assistance in stopping the English cultural juggernaut, and the same was true of the Cyrus/Brennus/Pericles trio, who invariably fought each other in some combination. Interestingly, Elizabeth usually did not delay Rifling as many other economic leaders do, perhaps because of her Redcoat UU. Thus, these other conflicts had to end quickly, or Qin would find himself on the wrong end of some shockingly early redcoats.

As a result, there were multiple games where Elizabeth defended against macemen or grenadiers with Infantry, and even Elizabeth could win those wars pretty handily (even if she had no interest in finishing them, with a measly 4 kills). Elizabeth's bad games typically occurred when multiple of these leaders attacked her before infantry, and once she got to Assembly Line (around Turn 200-220) she'd typically be a safe bet. There were plenty of games this didn't happen (Games 2, 3, 12, and 14 among them), but usually she was just too fast. Lizzie nearly always won before Turn 300, with her earliest result coming ludicrously early at Turn 248. The only exception was the bizarre Game 10, where Elizabeth lost her capital around Turn 240, only to simply keep the slider running and win with a random fourth city... 120 turns later! For reference, the earliest non-Elizabeth wins were two random diplomatic victories and a runaway Brennus cultural win on Turn 300... over 50 turns after Lizzie's earliest result. The only option for the remaining contenders was to dogpile her. For what it's worth, when this happened early enough Elizabeth would still collapse like a wet noodle, dying in nearly every game that wasn't won - her opponents just often lacked the time.

Cyrus of Persia
Wars Declared: 56
Wars Declared Upon: 18
Survival Percentage: 80%
Finishes: 2 Firsts, 5 Seconds (20 points)
Kills: 15
Overall Score: 35 points

Cyrus graded out as the second-strongest leader on this map with a pair of wins and a series of strong second-place finishes, although 2nd to 5th were all so close that they could be considered roughly equal. Cyrus' game typically revolved around Brennus', as these two fought each other in most games. When the war went heavily in Cyrus' favour, he was a strong contender to place, and if things stalemated he and Brennus would typically both be insignificant for the game. Cyrus was helped here by Pericles, who declared war on Brennus far more often than expected, and sometimes played a crucial assist in Cyrus' route to victory; Cyrus generally thanked Pericles by devouring him next. Cyrus' other path to success came from attacking the weak Qin Shi Huang to his east, which happened less often. This was a double-edged sword though, as it also removed Elizabeth's most common roadblock, and it usually just led to Elizabeth winning.

On the rare occasion that he snowballed out of control, Cyrus' route to victory (as with all other leaders) then revolved around whether he could kill Elizabeth in time. He managed it twice, both times with Julius Caesar's help; those two were a bit of a dynamic duo, and if one won, the other always came second. Cyrus did recognize the English threat, and declared war on Elizabeth in the lategame nearly always, but often these wars would end with Lizzie losing 1-2 cities before winning, as Cyrus was just a little bit too slow.

Julius Caesar of Rome
Wars Declared: 55
Wars Declared Upon: 17
Survival Percentage: 75%
Finishes: 1 First, 6 Seconds (17 points)
Kills: 14
Overall Score: 31 points

Contrary to the real game in which he fought Pericles first, Caesar's performance almost invariably revolved around his early clashes with Zara. Sometimes, Zara would squeeze out the Roman leader with his Creative trait, and stifle any attempts at aggression; this led to Caesar being a third-rate power that did nothing. Other times though, Caesar would solo-conquer Zara, and become a terrifying force to behold. This happened in Games 14, 16, and the hilarious Game 12, in which Zara attacked Caesar early, only for the Romans to pull an uno-reverse card and make Zara first to die. Less frequently, Caesar would fight Pericles instead, but this usually went worse for the Roman leader, who would often stalemate against the Greek walls - perhaps he got some lucky combat results in Real Game Six that snowballed out of control. Occasionally, Pericles attacked Caesar and caught him with his pants down, which happened most noticeably in Games 17 and 20; this was, of course, a disaster.

Caesar was only able to pull out one win, an atypical game that saw him solo-kill Zara and then plough into Elizabeth, who'd been slowed down by repeated attacks from Qin and Cyrus. Otherwise, he had a few strong second places, but was unable to repeat the Real Game Magic. It seemed that Caesar's biggest weakness was his economy. Caesar was next to the weakest leader in the game, had a green/isolated start, and a friendly diplomatic environment, but he often floundered about with macemen against not only English redcoats (or infantry!), but also Persian and Ethiopian ones. Apart from being strong in both of Cyrus' wins, Caesar was rarely a close competitor, and pretty much everything had to go right for him to have a shot. It just turns out that's what happened in Real Game Three after all. A disappointing performance from the Roman juggernaut.

Brennus of the Celts
Wars Declared: 38
Wars Declared Upon: 51
Survival Percentage: 55%
Finishes: 3 Firsts, 0 Seconds (15 points)
Kills: 14
Overall Score: 29 points

Brennus put up a pretty strong set of results in the first half of these games, winning the second-most games and having the highest score of everyone. His wins all followed a similar pattern - profiting the most off dogpiles of Qin, Elizabeth, or Pericles, and snowballing those into victories. However, the second half made this look like a total outlier, with Brennus completely unable to make anything of his outsider starting position. The big problem for Brennus was Cyrus, as he almost always fought with the Persian leader, and usually lost those wars; it was in the games where Brennus and Cyrus worked together against a third party that Brennus performed well.

Brennus, interestingly, was attacked the second-most in this game, with 51 defensive wars against only 38 offensive ones. Over time, it became clear that this trend was due to religious spreads: Elizabeth and Brennus usually split the first two religions, and Elizabeth's was far more likely to spread across the map. This left Brennus as by far the most likely target for the low peaceweight leaders to turn on, and was reflected in his war count and poor survival rate compared to the similar scorers of Caesar, Cyrus, and Zara. Unlike the rest of the 'chasing Elizabeth' pack, Brennus was never able to score second for the same reason - if another leader was stronger, Brennus was often next on the chopping block. Overall, Brennus was screwed by his starting position and religious spread, and was pretty unlikely to advance out of this one.

Zara Yaqob of Ethiopia
Wars Declared: 21
Wars Declared Upon: 27
Survival Percentage: 65%
Finishes: 2 Firsts, 7 Seconds (24 points)
Kills: 4
Overall Score: 28 points

Zara was a bizarre leader to watch on this map, with a tough jungle-belt location that was slow to get going but had a lot of potential. Sometimes he would come out of the landgrab phase very strong, cripple Julius Caesar, and then defend Elizabeth to the death - Zara was responsible for saving her in a few games. Other times he would expand terribly, get rolled up by Caesar, and look like one of the worst leaders in AI Survivor. It was hard to know what the difference was; sometimes barbarian cities would notably ruin either Zara or Caesar, but sometimes it just seemed purely random. Zara and Caesar pretty much always fought, and the winner of their battle usually decided second place between himself (when Zara won), Caesar (when Zara lost), and Cyrus (when they stalemated). His play was characterized by a bizarre pacifism that did not reflect his strength at all - sometimes, Zara would be by far the strongest leader in the game, but he just wouldn't do anything! It was like Gandhi had taken over Zara's game, and this was responsible for a few losses. In fairness, some of Zara's wars were complete disasters, and he had multiple deaths with only one war fought, so perhaps his AI was actually making a smart decision. Still, it was weird to see a supposed 'middling aggression' aggression leader put up many strong performances, while only equaling Qin in kills.

Zara's two wins came from atypical games. The first was a backdoor diplomatic when Elizabeth built the UN while about 30 turns away from her usual cultural victory, and the second came after Pericles had solo-conquered Julius Caesar (letting Zara expand far more than usual), while Elizabeth was dragged down by wars, and Zara was smart enough to attack her before she won. A lot had to go right for Zara to win these games, and most of the time he was a puzzling leader with a schizophrenic performance.

Qin Shi Huang of China
Wars Declared: 44
Wars Declared Upon: 39
Survival Percentage: 45%
Finishes: 1 First, 1 Second (7 points)
Kills: 4
Overall Score: 11 points

Qin was a big disappointment on this map. He placed twice in the first five games, and I thought he'd just be another warmonger capable of putting up respectable results without beating Elizabeth, but then for the rest of the games he just sort of, collapsed? Qin's only victory occurred in Game 2, when he was the lucky beneficiary of repeated dogpiles of Pericles, Brennus, and Elizabeth, and was then able to kill Cyrus in the endgame. Otherwise, he was a weak leader that was unlikely to do much other than take part in the regular Massacre of Pericles, or flail ineffectively against Elizabeth's superior land and tech. Qin's immediate area was pretty jungle-heavy, and I suspect his terrible traits did him no favours either. His cities would also often be crushed by Elizabeth's culture, losing their second or even first-ring tiles.

To his credit, Qin did try to dig himself out of the hole. He started more wars than he faced, which was surprising considering how weak he was, and his personality - Zara started less than half the number of wars, with a considerably higher aggression rating and a stronger position! These wars were rarely able to achieve much though, and for most of these games Qin was a second-rate power unlikely to advance and with the second-lowest survival rating. Perhaps worse, he was also just quite boring.

Pericles of Greece
Wars Declared: 21
Wars Declared Upon: 43
Survival Percentage: 25%
Finishes: 0 Firsts, 0 Seconds (0 points)
Kills: 2
Overall Score: 2 points

Poor, poor Pericles. The Greeks were relentlessly attacked, and nearly always stuck in 2v1 wars coming out of the early game. However, Pericles also displayed some pretty poor play here, and his deaths were frequently his own fault. Often, Pericles would start his first war, only for someone else to come crashing in from the other side; Caesar, Brennus, and Qin were all frequent targets for Greek aggression. These games usually ended badly, with Pericles being first to die in nearly half of them, and surviving for only a quarter.

But more problematic was Pericles' inability to make something of the games where he didn't get dogpiled. Sure, Pericles had a few strong third places (Games 8, 10, 17, and 20), and was probably unlucky not to get at least one second place, but it was insane how much had to go right for even these mediocre results. The best examples are Games 17 and 20, where Pericles scored his only two kills. In the latter, Caesar and Pericles fought two bloody wars that unusually ended in Caesar's death; Pericles was so far behind though that he never recovered economically. More embarrassing was Game 17, where Pericles attacked Caesar on Turn 49 with phalanx armies against warriors, and essentially conquered the Roman (and later Celtic!) core for free, while still apparently unable to advance. This was inexcusable; sure, Pericles was 40 points behind Zara, but he should have been at least 1000 ahead! Thus, I walked away both feeling sorry for Pericles and his central position, and unimpressed with his performance. He died the most, died first the most, and was completely unable to take advantage of even the craziest outlier scenarios.

Conclusions

Overall, the outcome of this game was clear: an Elizabeth cultural victory was the 'default' outcome, and the other leaders needed to actively cooperate against her to prevent this; this happened in Real Game Three as Qin was able to slow the English leader enough for Julius Caesar to eliminate her, and this strategy provided a victory or two per warmonger, but it seemed nearly like blind luck for who would profit the most. Second place went to Zara, Cyrus, and Caesar about a third of the time each, but none of the three were a runaway favourite for this. First to die was the only consistent thing with Real Game Six, with Pericles as the runaway favourite. Finally, Culture was by far the most likely win condition, as Brennus and Elizabeth each heavily prioritized this in every game.

It is worth noting something that Wyatan brought up in one of the original alternate histories - running these games as two sets of ten each gives very different perspectives on leader performance. Looking at the first half, Brennus actually grades out the strongest by points, with Elizabeth a close second (far from dominant!). Zara wins his duel with Caesar more often than not, and has a series of strong second places as a result, while Caesar looks even weaker than Qin Shi Huang. Meanwhile, the second half shows Caesar winning the western duel more often, and a few strong Cyrus and Pericles performances, only for it to often be irrelevant as Elizabeth steamrolls them all with culture near-invariably. Brennus and Qin do basically nothing, while Zara manages only two weak seconds until the final game. It's interesting how these larger samples can still have pretty big streaks sometimes.

I hope you enjoyed reading this blast from the past, and I'll hopefully be back soon with a look at a different game from this season.