Civ4 AI Survivor Season 7: Wildcard Game 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: Given the Raging Barbarians and high number of players, it would make sense to assume the wildcard game is more random than the normal ones. So far, this has not turned out to be the case, with yet another Wildcard Game here that could only realistically go a few different ways. For Season Seven, there were three leaders who had a realistic shot to win the Wildcard Game: these were Huayna Capac, Saladin, and Napoleon. Everyone else had basically zero chance, and in fact no one else won a game until Tokugawa fluked the very last one. Instead of my expectation that more players + raging barbs = more possibilities, it seems that instead having more players opens up the possibility for unfair start differentials, and that was exactly what happened here, as the same few leaders fought for the win again, and again, and again. So why were the Easterners so strong, while the Westerners had to fight over second-place scraps? Well, they all had one thing in common, and that was: rivers! All three of them stared in fertile river valleys that other leaders could only dream of. But otherwise, all three of them had quite different paths to victory. Hilariously, all five of Napoleons wins hinged on the exact same thing happening: Vienne (Boudica's second city) being captured or razed by the barbarians in the first 50 turns. This happened in every game that Napoleon won, and otherwise didn't happen at all, with shocking consistency. It got to the point that in Game 12, I wrote in my early notes "Boudica loses Vienne T42 - Nappy probably wins," and that's exactly what happened! In these games, Napoleon would go on to claim much of Boudica's land as well, steamroll poor Julius Caesar (we'll get to him later), and eat half the map before you could say 'snowball.'

But when Boudica didn't lose Vienne, Napoleon basically had no chance, and the game instead came down to a duel between Huayna Capac and Saladin, who were able to win seven games apiece. These two fought an extended duel for control of the East in most games, and though it happened, it was very rare for them to advance together. They founded two of the first three religions in basically every game, and were simply unable to get along in most cases (in Games 3 and 15, the two where they advanced together, Huayna adopted Sal's religion instead). This meant that they clashed repeatedly, and each was capable of significantly damaging the other. Sometimes Huayna could take and keep Medina (the Islamic Holy City), and sometimes Saladin would annex Cuzco (the Incan capital); in either case, the loser would be ground out and murdered, and the winner took the gold medal. In general, Saladin won this war if it occurred early on, because Huayna Capac's triple-gold start was very high in commerce, but low in foodhammers, and he could often fall behind in commerce. In contrast, Huayna Capac would typically win if it started later, because Saladin was an absolute dogpile-magnet and would get torn apart in a frenzy if he hadn't already snowballed by the midgame. In either case, this game basically had three outcomes: Napoleon could get half of Boudica's land for free and steamroll the map, and if that didn't happen, Huayna Capac or Saladin would win their duel, and thus the game. What was most interesting about this was that the diplomacy unfolded quite differently across the games, occasionally becoming extremely chaotic, but the emerging powerhouses were almost always some combination of these three regardless.

So the question became: what the heck were the other six leaders doing? Well, this game was characterized by the extreme weakness of a few of them. Julius Caesar was absolutely screwed on this map, and his pathetic existence in a sad little corner massively contributed to Napoleon and Saladin's strength. JC was never even close to winning, and his only good came occurred when he sniped two of Saladin's kills (Game 17). Gilgamesh was almost as bad, having exactly one strong game (Game 1) and flaming out without note across all 19 others. Despite having copper at his capital (and creative trait), he was simply a non-entity in most games; I suspect this was because of a lack of commerce, but I'm not really sure. Finally, although he did manage the only 'non-Big Three Win' in Game 20 (when the real contenders all brutalized each other out of contention), Tokugawa was a very weak leader in most of these games, ravaged so badly by barbarians that he could never recover. Though he was never first to die, Toku found himself executed by stronger leaders in the midgame repeatedly, and until the last three games I honestly wasn't expecting him to place at all.

Then there were three mediocre performers, all of whom garnered a few second places, but none of whom were ever able to realistically compete for first. Boudica was the best of the three, and also the most variable in performance. She lost Vienne in a quarter of the games (by far the most of anyone), and also Tolosa (her third city) in quite a few others. But in the games where that didn't happen, Boudica would consistently claim the massive northern share of the map, and was certainly capable of cracking a few skulls. Shaka was probably the worst of the three, and although he tended to expand very well, he was crippled by a very jungle-heavy start compared to the other leaders. He would usually be the first to four, five, and six cities - but at least half of them would be unhealthy at Size 1. He was able to snowball in a couple of games (by far his best were Games 17 and 20), but usually he would annoy too many other leaders, and get dogpiled in the midgame like Tokugawa. Finally, Peter was the quiet achiever of the group, and performed very similarly to the livestream game. He was consistently quite strong coming out of the landgrab, particularly in technology, but it never seemed to quite click. Just as Peter was beginning to snowball an empire that rivalled the game leader's, something would go wrong - a dogpile, a tough 1v1, or cross-map events that made him irrelevant - and Peter would be relegated to second place again. Peter's result in the livestream reflected his performance perfectly, and it was a tough series of games to watch for him.

So, in essence, that was the story. Huayna Capac and Saladin would duke it out for supremacy in the East, and whoever won their duel would go on to win the game... except when Napoleon got a lucky break (which, to be fair, happened a quarter of the time!) and stomped all over both of them. No one else had any real chance to win, and since Napoleon and Huayna each flamed out in spectacular fashion on livestream, I'm glad that Saladin was at least there to pick up the pieces.

Now for a look at the individual leaders:

Huayna Capac of the Incans
Wars Declared: 41
Wars Declared Upon: 30
Survival Percentage: 65%
Finishes: 7 Firsts, 4 Seconds (43 points)
Kills: 17
Overall Score: 60 points

Huayna Capac graded out as the strongest performer on this map, and that was probably justified, as he was the most consistent overall. Huayna had a very interesting triple-gold but low-food starting position that I'd pegged as disgustingly broken and unfair to everyone else. I'd thought Huayna would go on to win basically 100% of the games on this map, but thankfully, I was pretty obviously wrong. What Huayna did have was a very boom-or-bust start, as he was highly competitive in about half the games, and totally irrelevant in the other half. Huayna had two types of bad games: the first was games where he simply didn't expand properly due to his wacky start position (e.g. Game 17), and the second was games where, like on livestream, he started a stupid early war before he was ready (e.g. Games 4 and 18). However, unlike the livestream game, this was almost never against Boudica, and instead Huayna Capac found himself ramming into Saladin over and over again. As I've covered above, the winner of this duel would basically go on to win the game, excluding Napoleon's lucky romps. In fact, Huayna managed to suicide into Saladin twice(!) as his only war, essentially gifting Saladin two victories (Games 4 and 18). However, in the other half of games, Huayna Capac looked nigh-unstoppable; he would settle about as many cities as everyone else, while researching at 3x the beaker count, and could easily smash Gilgamesh, Napoleon, or Saladin 1v1 (or in a dogpile, which happened a lot to all three). This was representative of a trend whereby Huayna Capac could win by sticking to himself, or by joining dogpiles, or even by winning 1v1 wars; he was the only leader who really could do it all. He was very rarely dogpiled himself, with his low peaceweight and high warmonger respect clearly coming in handy. Finally, it is worth noting that while Huayna graded out as the top leader regardless, he was also probably quite unlucky with his second places. In Games 2, 15, and 20, Huayna Capac was very close to winning instead of second. As such, this list of games probably slightly underrates his performance, and he should have won maybe 8-9 games instead. I was definitely wrong about Huayna Capac having a busted start, and it was probably about average overall, but Huayna Capac is just such a disgusting leader, handed so many unfair advantages, that it was enough to take the overall first place regardless.

Saladin of Arabia
Wars Declared: 22
Wars Declared Upon: 70
Survival Percentage: 55%
Finishes: 7 Firsts, 1 Second (37 points)
Kills: 14
Overall Score: 51 points

Saladin was the other south-easterner, and was competing with Huayna Capac for top score until he fell short in Game 20. This was surprising to me, because for most of the series, he'd been lagging behind HC and Napoleon in points, until the final six games where Saladin pulled his finger out and suddenly won 4/6 times. Saladin was a strange bird in these games. He wasn't very aggressive, instead being attacked a ridiculous 70 times, but he still managed to score a lot of kills. He was bless-cursed by the same issue as Zara Yaqob back in Game Three earlier this year, whereby he had a floodplains heavy region that was amazing by T150, but that massively lacked production until that point. This meant that in early wars, Saladin could sometimes find himself caught offguard, especially by Napoleon; he had enough strong land that he was never first to die, but Saladin was crippled by T120 in many of his losses. In fairness, he did live in a very tough neighbourhood (other than JC, who was normally the world's most useless ally), but Saladin's early weakness was no doubt also partially responsible for Napoleon/Huayna's early strength. Saladin was benefited massively by founding a religion, which was typically the most widespread in the game thanks to his massive river network. This meant that his peaceweight was only a major liability, instead of an absolutely crippling one, and though the defensive war count looks pretty grim (70 defensive wars is a lot!), I'd wager it would have been a lot worse without Saladin's Islamic brothers helping him out in game after game. Saladin was also the most boring of the Big Three by far, because once he got ahead, he never really put the foot down on his opponents' necks like HC and Napoleon. Instead, he would grind out long, steady spaceship wins, just like we saw on livestream. Sometimes a Napoleon snowball would happen, or sometimes Saladin would get dogpiled, but otherwise he felt a bit like the 'default' ticking clock victory, as his floodplains cottages carried him to (slow) victory. Though it took a while to get going, Saladin's start really was absurd once things clicked, and it was hard to feel like his success was anything other than an average leader getting carried by amazing land. Very 'Season Six Churchill' vibes here.

Napoleon of France
Wars Declared: 46
Wars Declared Upon: 54
Survival Percentage: 45%
Finishes: 5 Firsts, 1 Second (27 points)
Kills: 18
Overall Score: 45 points

Although he fell off a lot in the second half, Napoleon was the only other leader with a realistic chance to win the game. Unlike Huayna Capac and Saladin though, who were strong powers in basically every game, Napoleon's starting location was pretty weak, surrounded by jungle like Shaka's. Instead, his performances revolved almost entirely around those of his neighbours. As I've already covered, Boudica's performance was massive relative to Napoleon's, as he won games where she lost Vienne to the barbs, and otherwise didn't win at all. There was a lot of land to their north, and only one leader could claim it; that leader would typically go on to become a powerhouse, while the other would become an also-ran. Julius Caesar's performances were a lot more consistent than Boudica's but no less relevant. The Romans were totally screwed on this map, and it was nearly always Napoleon who got to take advantage. Napoleon would normally be a middle-tier leader coming out of the landgrab, and a lot of his long-term success hinged upon whether he could solo-conquer Caesar (they pretty much always fought) before Boudica, Huayna, or Saladin slammed into him from the other side. Julius was pretty much a free meal in half of these games (including all of the ones where Napoleon placed), but often he could fight Napoleon in a 1v2, and that was usually enough to wreck Napoleon's chances. The other noteworthy thing about Napoleon's performances is how much he fought... which was a lot! I think this is the first time I've seen a leader crack 100 total wars (which Nappy managed exactly), so this might be a new record. Napoleon fought so much partly because he's a total nutcase, and partly because he tended to be stronger than his neighbours. This meant that he'd either snowball out and eat multiple opponents (e.g. Games 1 and 9), or be dogpiled by many weaker leaders (e.g. Games 5 and 16). In either case, the French were in for a lot of fighting. Overall, this was a strong performance, even if it was off the back of pathetic neighbours, rather than personal strength. Between this and Game Eight, Napoleon was probably pretty unlucky not to make the Playoffs. This is the third time he's been in this position, and Napoleon must be competing with Justinian and Hatshepsut at this point for the Unluckiest Leader Award (TM), at least in recent years.

Boudica of the Celts
Wars Declared: 42
Wars Declared Upon: 51
Survival Percentage: 45%
Finishes: 0 Firsts, 5 Seconds (10 points)
Kills: 16
Overall Score: 26 points

There was a massive drop-off from the Big Three down to everyone else, but Boudica squeaked out fourth place by virtue of having the most second-place finishes, and slightly more kills than the other mediocre performers. Boudica's performance was highly variable, and had massive implications for the rest of the map. She didn't struggle with barbarian pillaging as much as Tokugawa, but for some reason her cities were far more likely to be captured; I'd guess Boudica lost a city in about half the games, with no other leader at more than a quarter. As mentioned, this sometimes allowed Napoleon to claim a lot of Boudica's northern territory, which could totally derail the game. Normally though, Boudica did get her fair share, and this was enough to typically see her be relevant without actually threatening to upset the victors. Her central position was very high-risk, in that sometimes she would get attacked mercilessly by all her neighbours, which was simply bad luck; sometimes, like in Game 6, she could turn those wars around for a strong performance, but normally she would just get picked apart. Tokugawa in particular was very likely to attack Boudica if he wasn't previously engaged, and the pattern of 'Boudica attacks Napoleon, then Tokugawa attacks Boudica' was a common one. The one neighbour that Boudica almost never fought was Huayna Capac, with their war in the real game seeming like an outlier (and probably the reason that Boudica was so unusually strong). Really, Boudica's problem was the same as usual, in that her technology rate was awful, and this knocked her out of contention even in the games where she was quite big (Games 7 and 14 especially). Boudica was sort of unlucky not to get one win (Game 7, where Huayna Capac won by culture 5 turns before the Celtic spaceship was due to arrive), but this was clearly an outlier regardless, and Boudica's performance on livestream was much better than her average result here.

Peter of Russia
Wars Declared: 44
Wars Declared Upon: 27
Survival Percentage: 50%
Finishes: 0 Firsts, 4 Seconds (8 points)
Kills: 13
Overall Score: 21 points

Peter was a tragic figure in these games. Nearly every time, he'd come out of the landgrab looking like one of the competitors, with a strong tech rate and decent expansion. He was often able to get on the right side of a 2v1 with either of his neighbours (Shaka and Gilgamesh), and expand that way. But in the midgame, it would nearly always fall apart for the Russian leader, who would collapse or find some other unfortunate way to lose (in Game 16, Peter was the dominant leader with 4 kills when Huayna Capac won an out-of-nowhere cultural victory). The reason for Peter's failures in general though, were twofold. First of all, he had major production issues. A lot of Peter's cities were coastal, and that was very apparent in the demographics where Peter would typically be in the top three in Food and GNP, and the bottom three in Production. It meant that Peter needed a tech lead to fight successful 1v1 wars, in contrast to leaders like Boudica and Napoleon who could simply run over their opponents with weight of numbers. Peter's other issue was diplomacy, and this was exactly what we saw on livestream. Peter would usually found either the third or fourth religion, as his tech lead became a curse. No one else would convert to this religion, which meant that Peter would get dogpiled by 2-3 other Islamic or [Huayna Capac's random religion] leaders before he could really snowball. It all added up for a series of 'almost' games, with Peter never able to quite get it together. It's rare that a leader's performance on livestream matches the Alternate Histories perfectly, but Peter's '1 kill before dying to dogpile' game was pretty damn close.

Shaka of the Zulus
Wars Declared: 47
Wars Declared Upon: 36
Survival Percentage: 25%
Finishes: 0 Firsts, 3 Seconds (6 points)
Kills: 12
Overall Score: 18 points

Shaka was the worst of the mediocre leaders, but really it was pretty close between all three of them. For whatever reason, while the other two were semi-consistent across all 20 games, Shaka only picked up in the second half, which was responsible for his lower score. Shaka had a tough start, in a central, jungle-bound location. He did his best to escape, ramming his impis against the walls of Celtia, Sumeria, Russia, and Japan, but this rarely worked. Most of the time, Shaka would get out-teched and then lose a 1v1 as Construction was unlocked, so he needed a dogpile to get going. This was what happened in Game 9, for example. Shaka's other route to success was simply to get lucky in his early wars. This did happen a couple of times, such as Game 13 (where he took Moscow on T75 with the initial strike) and Game 17 (where he somehow managed to 1v2 Peter and Giggles), but it was a rarity. More commonly, Shaka's cities were unhealthy and had no improved tiles until Iron Working, and this crippling drawback was too much for the Zulu leader to overcome. Honestly, considering the rough start he had to deal with, Shaka expanded very well (often the first to 4-6 cities), and did much better than my expectations after watching the first few games. His tech might have been a disaster, but at least Shaka was choosing the right priorities in a tough situation. Even if he couldn't get it together often, he did more than the leaders below him, and this game felt oddly like a positive outcome with the above in mind.

Tokugawa of Japan
Wars Declared: 44
Wars Declared Upon: 37
Survival Percentage: 30%
Finishes: 1 First, 1 Second (7 points)
Kills: 8
Overall Score: 15 points

Behind the Mediocre Trio was another group of three leaders that were, frankly, appalling. Coming into this game, I'd thought Tokugawa would be quite strong, and unlucky to have missed out on the win. Instead, he was one of the weakest leaders, with the livestream game being arguably his best! Yes, Tokugawa won a game (Game 20), but it took a ridiculous string of events for it to happen, including the Big Three all knocking each other out of contention, Shaka killing Peter and Gilgamesh simultaneously, Tokugawa getting all of the spoils in an anti-Boudica coalition, and never getting attacked in the process. Yeah, that wasn't going to happen again anytime soon... and the worst part was the Huayna Capac's (very delayed) spaceship was also en route - he nearly lost anyway! At least the issue for Tokugawa was clear: he got absolutely rolled by barbarians in every game. Interestingly, Toku never actually lost a city (that I saw, at least), which was probably thanks to Protective, but his tile improvements got pillaged at Kyoto until Turn 60 or even 75 in most games. This always set Toku far enough behind the pack that he spent the rest of the game catching up, and normally another leader (Shaka or Boudica, usually) would whack him before he got there. Even in games where he was similarly left alone (such as Game 7), Tokugawa wasn't remotely competitive, and it really did take an extraordinary string of circumstances for him to win Game 20. With that in mind, Tokugawa's livestream result was on the extreme upper end of his potential, and I'm not really sure what the difference was. Don't let his win fool you here - until Game 18, it looked very much like Tokugawa was going to share zero placements with Julius Caesar, which was a pretty remarkable possibility.

Gilgamesh of Sumeria
Wars Declared: 25
Wars Declared Upon: 30
Survival Percentage: 30%
Finishes: 0 Firsts, 1 Second (2 points)
Kills: 4
Overall Score: 6 points

Gilgamesh was another big disappointment here, and unlike Tokugawa and Caesar, I'm not really sure why. Gilgamesh had an excellent capital, with copper on hand to defend against the barbarians. He also had Creative/Protective traits, which are probably literally the ideal combination in the Raging Barbarians game; I didn't see him ever lose a city, at least. He had decent land nearby too, with grain resources and a fertile river valley to the northeast. But for whatever reason, Gilgamesh was consistently very weak in these games, clearly sharing runt-status with Julius Caesar after the landgrab in most games, and totally ineffective at snowballing even when he was competitive. His tech was awful, but so were his demographics, as Gilgamesh seemed to essentially just play a series of very bad games of Civ4. He couldn't beat Peter, Shaka, or Huayna Capac in a 1v1 fight (instead losing to all three), and was always the weaker party in any dogpiles. He didn't fight many wars at all, really, with his total of 55 just barely higher than Julius Caesar's. Even his kills were a total disappointment, as he scored the same amount as JC(!!). If not for a random second place in Game 1 (admittedly, this was a genuine if inexplicably strong performance), they actually would have scored the same number of points. Considering what each leader had to work with, that is seriously embarrassing for Gilgamesh. For whatever reason, the Sumerians were totally screwed on this map, and had basically no chance. At least that fits with what we got on livestream.

Julius Caesar of Rome
Wars Declared: 37
Wars Declared Upon: 13
Survival Percentage: 25%
Finishes: 0 Firsts, 0 Seconds (0 points)
Kills: 4
Overall Score: 4 points

Oh. My. God. This has to be contending for the worst start AI Survivor has ever seen. It's competing with Hannibal's stinker in Season 6, but I honestly think this one might have been worse. Julius Caesar had one food resource available outside his capital, and otherwise had a tiny stretch of foodless jungle to claim for himself. He had no luxuries (the gold was never settled), and no rivers. He was surrounded by two of the top leaders on the map (and I do think both would have been strong even if Julius wasn't there), and could never settle more than six cities peacefully (again, 4/6 had zero food resources). In retrospect, I really think no leader could have possibly won from here. To his credit, Julius Caesar did try to break out of his hellzone, with nearly three times as many wars declared as faced. But even these wars tell the story - Julius Caesar was shockingly weak in every game compared to every other leader, even the other runts like Gilgamesh and Tokugawa. If anyone ever attacked him, the Roman game was literally over, with zero games facing more than 1 war - Caesar simply died before anyone else had a chance to join in! Sometimes, he would get on the right side of a 2v1 vs. Napoleon, but he was never able to actually take any cities, and simply served as a distraction for the Celtic or Arabian armies. The closest thing Caesar had to a 'good game' was Game 17, where he killstole from Saladin twice and ended up with over ten cities. Obviously, he wasn't even close to coming second though. I honestly think you could run this map 100 times and Julius Caesar would be lucky to place at all, and to be honest, we definitely shouldn't be making any conclusions of his AI strength from this one. This was pretty much a Kobayashi Maru scenario for the Romans, and try as he might, Caesar was never going to make it out.

Sullla Note: Just to provide some additional context, we had 7.5% of the contest entrants pick Julius Caesar for first place and 20% pick him for second place, in a nine-AI field! The hopelessness of this starting position apparently wasn't clear before the Wildcard game took place.

Conclusions

Overall, this game on livestream went somewhat as it 'should have,' although there were definitely some oddities. Saladin winning was very within the norm, as once Huayna Capac self-destructed and Napoleon failed to get an early snowball, he was pretty much the only plausible option left. I'm certainly glad Tokugawa and Boudica each failed, as those would each have been massive outliers. The victory date and number of wars on livestream were also well within the normal range. On the other hand, the actual path that Saladin took to get to his win was very unusual. Boudica and especially Tokugawa were much, much stronger than usual, which had knock-on effects for the rest of the game. This was probably thanks to Huayna Capac's bizarre war against Boudica, which was very unlikely, and then led to each of the northern leaders claiming more land than usual. Similarly, Huayna Capac dying first was certainly plausible, but a lot less likely the poor Julius Caesar, who was probably lucky to die first less than half of the time. It was an interesting series of games to watch, with the Big Three competing for dominance, but everyone else on the map having a lot of relevance for each individual game. Even if Boudica and Peter weren't competing for the overall trophy, they certainly could and did throw wrenches into the main competitor's plans! Normally I'd also talk about Fantasy, but the makeup of the Wildcard Game is so random that it honestly seems pointless. Huayna Capac and Saladin were both good bids in the context of this game, for example, but Saladin was so unlikely to make it here that it's hard to quantify the implications. Suffice to say, there are a lot of directions Fantasy could have gone here, and the winners definitely have to be both lucky and good.

I hope you enjoyed this Alternate History - only two more left for the season now!

Cheers ~ Amicalola