One of the recurring features of past seasons of AI Survivor have been our "alternate histories", running additional iterations on the same maps to see if the same events would play out again. This game was a dominant Mansa Musa victory on Livestream as the Malinese leader ruthlessly crushed all opposition. Was that something which would unfold in each game? This was a topic that called for more investigation with alternate history scenarios. Following the conclusion of previous seasons of AI Survivor, I had gone back and investigated some of the completed games and found that they tended to play out in the same patterns over and over again. While there was definitely some variation from game to game, and occasionally an unlikely outcome took place, for the most part the games were fairly predictable based on the personality of the AI leaders and the terrain of each particular map. Would we see the same patterns play out again and again on this particular map?
The original inspiration to run these alternate histories came from Wyatan. He decided to rerun the Season Four games 20 times each and publish the results. The objective in his words was twofold:
- See how random the prediction game actually is. There's a natural tendency when your predictions come true to go "See! Told you!", and on the contrary to dismiss the result as a mere fluke when things don't go the way you expected them to (pleading guilty there, Your Honour). Hopefully, with 20 iterations, we'll get a sense of how flukey the actual result was, and of how actually predictable each game was.
- Get a more accurate idea of each leader's performance. Over 6 seasons, we'll have a 75 game sample. That might seem a lot, but it's actually a very small sample, with each leader appearing 5-10 times only. With this much larger sample, we'll be able able to better gauge each leader's performance, in the specific context of each game. So if an AI is given a dud start, or really tough neighbours, it won't perform well. Which will only be an indication about the balance of that map, and not really about that AI's general performance. But conversely, by running the game 20 times, we'll get dumb luck out of the equation.
Wyatan did a fantastic job of putting together data for the Season Four games and I decided to use the same general format. First I'll post the resulting data and then discuss some of the findings in more detail. Keep in mind that everything we discuss in these alternate histories is map-specific: it pertains to these leaders with these starting positions in this game. As Wyatan mentioned, an AI leader could be a powerful figure on this particular map while still being a weak leader in more general terms. Now on to the results as run by TheOneAndOnlyAtesh:
Game One | Game Two | Game Three | Game Four | Game Five
Game Six | Game Seven | Game Eight | Game Nine | Game Ten
Game Eleven | Game Twelve | Game Thirteen | Game Fourteen | Game Fifteen
Game Sixteen | Game Seventeen | Game Eighteen | Game Nineteen | Game Twenty
(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.)
TheOneAndOnlyAtesh: On one hand, this was a Mansa Moneybags masterclass, but on the other hand, this was an incredibly un-Mansa like showing for the fan favorite. The typical Mansa playbook is to peacefully build, race ahead in tech, and then cash in on a quick Cultural victory. However, Mansa was much more aggressive on this map and only ever opted for Culture once, and that one attempt in of itself was extremely bizarre (he for some reason switched on the culture slider at Optics tech).
Mansa cared little for culture in this map. He was relatively uninterested in the first three religions - Suleiman, Hammurabi, and Lincoln all went for those more often - and he also eschewed other cultural pursuits like later religions and the Sistine Chapel in favor of pure commerce paths like Currency and Economics and military beelines like Guilds and Rifling. Instead, Mansa wisely preferred to steamroll through the tech tree and/or through his rivals. In game after game after game after game after game, Mansa expanded like crazy, exploded into a military tech lead, and then Thanos snapped another leader's lands into fine additions to his own. Most often, that unlucky sap was Alex, but no leader save for Hammurabi was fully safe from Mansa's wrath here. (Something I would later come to realize is that Hammurabi's pure Culture research flavor may have caused him to hoard all the cultural prizes, making Mansa more likely to spend his beakers elsewhere).
Map Dynamics
In most maps, the bad guys would bully the good guys, but the tables were completely turned here - this was an utterly dominant showing by the high peaceweights despite the even good to evil split. Even if Mansa faltered, Hammurabi and/or Lincoln filled the void admirably, imposing their wills on the three evil leaders. The high peaceweight trio altogether won 19 games, secured 37/40 of the possible playoff spots, and more than 85% of the total kills.
The reason for this disparity was simple: land quality. Lincoln had by far the best starting position, with an excellent capital and ample room for 10+ good quality cities; Hammurabi's corner was rich in resources and commerce and scaled well; Mansa's land was fine, perhaps not as good as his peers, but good enough for Mansa Musa.
Gilgamesh was stuck with an ill-fitting seafood capital and a tundra-infested corner that had room for seven cities at most. To tack on, his neighbors were the Mansa death machine and the Americans with the best land. Meanwhile, Alex's land was sparse in resources, especially strategic ones. When combined with his propensity to treat Mysticism as if monuments would give him leprosy, Alex frequently found himself without metal units to defend himself against the aforementioned death machine. Finally, the Silly Man's corner was choked in low quality jungle, and as had happened in the Livestream, the Ottoman sultan was drawn to a horrific zero-production 2nd city location that he always queued his next Settler in, rendering him essentially irrelevant from Turn 5.
Leader Dynamics (aka the Mansa Story)
The 16 Mansa games had two sub-categories: the fast and slow Mansa games. Like with Justinian in Game 4, Mansa's speed generally depended on his ability to quickly conquer one of his neighbors, and his 2nd city location was quite important. Mansa always considered two spots: a flood plains river spot to his east that secured him elephants (the correct choice), and a jungle spot to his south redeemed by a corn resource and its good long-term prospects. The first spot almost always led to the fast Mansa games, as it afforded him Elephants for conquest, guaranteed him a fast start - he is freakin' Mansa Musa, he will scale long-term anyway - while making Alex an easier target by locking him out of what little fertile land he had. The southern spot was alright, but it did afford Alex more breathing room while also getting in Gilgamesh's face too much. The slow Mansa games would see Mansa get suckered into an early 2v1 against his hostile neighbors. This was where being Mansa Musa was especially important; while most other leaders would have stalled out and become irrelevant, Mansa would eventually econ well enough to overwhelm his enemies and win, even if he was limping to the finish line. It was not a totally uncommon occurrence for Mansa to miss out on key Renaissance prizes like Liberalism or the Taj Mahal and only finally establish a tech lead deep in the Industrial Era in his slower games.
Nevertheless, most Mansa victories followed the general pattern of the Mali knocking out a weak Alex and snowballing, like in the Actual Game. Gilgamesh and Suleiman would then get picked off next, while the remaining high peaceweight leaders sat together and sang Kumbaya in a Next Turn fest. These games were more dynamic than the results suggested, especially regarding Lincoln's performance. Lincoln's central position rendered him vulnerable to dogpiles, and he faced multiple attacks from every leader over the course of these twenty games. However, due to his abundant land and his ability to vulture (no pun intended) off territory from Gilgamesh, he was the best at riding the Mansa wave into a respectable playoff appearance. The performances of Hammurabi and Suleiman also varied, albeit not to the same extent as Lincoln's. In most games, the two fought stalemated wars into irrelevance, but Mansa could betray Lincoln late to ensure that Hammurabi was 2nd place by default (like in the Actual Game), Hammurabi could conquer Suleiman to establish himself as the #2 guy, or Suleiman would partition Alex with Mansa to ensure that he had curried enough favor with the Moneybags machine to survive to the Wildcard game - the Ottomans were never even remotely competitive otherwise.
There were three oddities among the 16 Mansa games. One of them was his Cultural win that I already touched on. Another one was Game 3, where a multitude of inopportune backstabs just as one leader seemed to be about to break through led to a prolonged global stalemate, only broken when Mansa had at last built enough of a tech lead to conquer Sumeria and limp to a Turn 364 spaceship. Note how ridiculously late Gilgamesh's First To Die came, on Turn 309, which would have shattered the "latest First To Die" record as of this writing. (I truly believe a post Turn 300 First To Die is actually more unusual than having no leaders die in a game). Finally, there was Game 10, which veered into the opposite extreme with a stunning Turn 251 MANSA DOMINATION victory. That game saw Gilgamesh benefit the most from a dogpile of a weaker-than-usual Lincoln and then bring Hammurabi to the brink of death. Once the Mansa-Musa-ninator came calling, Gilgamesh had just enough to hold onto a playoff spot.
The four non-Mansa wins all shared one trait: for whatever reason, Mansa completely bungled his conquests and ended up permanently gimping his game. Usually, a combination of the following lead to this:
1. Alex researching Mysticism in a reasonable timeframe, and thus being stronger and more evenly matched - those extra cultural defenses made a significant difference, as a lot of Mansa's fights were pre-Catapults
2. Mansa making questionable expansion choices, either getting squeezed on territory or overextending into the South and making his border cities difficult to defend
3. Mansa getting into an early 2v1
4. Mansa investing too much in early wonders to the detriment of everything else
In three out of the four non-Mansa wins, Mansa remained a contending leader but ended up failing to outscale leaders with 2-3x as much territory. In two of those games, his empire was so small that he was relegated to the Wildcard game! There was also the utterly bonkers Game 19, where Mansa utterly fumbled his attack against Alex and was run over by Gilgamesh to become First To Die. That game was a 1 in 100 type of game that nevertheless made this entire set worthwhile for me.
Religion played a larger than expected role in diplomacy, even if it was rarely able to smoothen over first impressions. For example, Suleiman's ability to salvage a Wildcard spot generally depended on if he religiously aligned himself with the good guys. Moreover, Lincoln had a tendency to be in a minority religion, leaving him vulnerable to backstabs (especially since Mansa can plot at Pleased). With Mansa ignoring culture and this being a high peaceweight fiesta, Spaceship was the runaway victory condition. There were as many Diplomatic as Domination wins (three apiece)!
How Typical Was The Actual Game?
9/10. The Actual Game was a cookie cutter fast Mansa game, and even his late betrayal of Lincoln turned out to not be out of left field.
Mansa Musa of Mali
Wars Declared: 30
Wars Declared Upon: 34
Survival Percentage: 95%
Finishes: 16 Firsts, 1 Second (82 points)
Kills: 23
Overall Score: 105 points
Scoring more than 100 points is a rare yet impressive feat, and it would have been more had Mansa not choked in the final two games. This was the most aggressive I have ever seen Mansa: his 23 kills would have made any warmonger happy, let alone a peacenik like Mansa. He should have had more kills, by the way, as Suleiman and Hammurabi did snipe some kills after Mansa had done all the heavy lifting.
Mansa's aggression was also controlled and smart. I can recall just one foolish Mansa attack, where in Game 2, when Alex had only Chariots defending his cities, Mansa sicced his forces on… a Protective Gilgamesh with Vultures and Creative culture defenses. Otherwise, there were still hints of Mansa's weaknesses, namely his propensity to struggle early if he got too bogged down by wonders or was unable to establish military superiority. The removal of Deity starting techs definitely hampered Mansa, as he never had quite as good of an economic game as in the Livestream (although he did come pretty close). Mansa benefited a lot from Alex's extreme weakness, and a slightly stronger Alex could have meant a drastically worse Mansa, as Game 19 showed. Nevertheless, I truly believe that this is the absolute best version of Mansa Musa, and the other leaders should consider themselves lucky that Mansa is usually not the warmonger he was here. Going for culture is a glass cannon strategy, and considering Mansa's high peaceweight, he might be better off focusing purely on commerce and military like he did in this setup. The results say everything.
Best Performance: Either his stunning Turn 251 Domination win in Game 10 or his almost as amazing Turn 286 Spaceship win in Game 13.
Worst Performance: Other than the obvious Game 19 and Game 2, Mansa was quite bad in Game 20, requiring a bailout from other leaders in order to deal with Alex. That may have been the first time I have ever seen Mansa backdoor into 2nd place.
Gandhi Award: Mansa sparked his snowball in Game 5, not through military conquest, but through settling a bunch of cities in open spaces while other leaders were murdering each other.
Lincoln of America
Wars Declared: 14
Wars Declared Upon: 40
Survival Percentage: 70%
Finishes: 1 First, 10 Seconds (25 points)
Kills: 16
Overall Score: 41 points
Although Lincoln's performance was ostensibly good, it was apparent why he is an AI Survivor bottomfeeder. Admittedly, Lincoln was actually quite excellent at most aspects of playing Civ IV. He expanded well, managed his economy effectively, and was a surprisingly competent fighter when pressed. So why do I think so lowly of Honest Abe's AI Survivor capabilities? Well, there was one major caveat: Lincoln's ultra-pacifistic personality - which could not have been more historically inaccurate - proved to be extremely detrimental in these games. Lincoln did just about everything he could to not fight, launching a paltry 14 wars despite his strong position and an abundance of weak targets. A stat I never thought I would see: Lincoln had more kills than offensive wars. That is simultaneously hilarious and sad. He had to wait for others to attack him, and this was a dicey proposition, as dogpiles were the primary catalyst for his eliminations. Moreover, I witnessed him do everything he could to NOT research military techs over the course of these replays. In a select few extreme cases, he was deep into the Renaissance Era in the tech tree, yet still lacked Gunpowder tech. Not Rifling. Gunpowder.
Luckily for Honest Abe, his land, suicidal enemies, and natural diplomatic bond with Mansa ensured that he was the runaway pick for second place. To his credit, he actually fought well when he had to, and he certainly had the means to keep the military momentum going into crushing wins if he just had the willingness to do so. Unfortunately, the Civ IV iteration of Lincoln was saddled with the cowardly George McClellan rather than Ulysses S. Grant or William T. Sherman as his commanding officer. As a result, Lincoln failed to take any sort of initiative to win, preferring to sit back and watch Mansa go to space, and with his land, he should have had more than one win even with Mansa in this game. It is no wonder that Lincoln has done little of note beyond Opening Round fodder since winning in his AI Survivor debut: in the majority of situations, his stubborn refusal to fight becomes a fatal flaw for the Great Emancipator.
Best Performance: I really thought his Game 2 win was rather lucky, as it saw a weak Suleiman suicide into Lincoln. His Game 16 near-win was more impressive to me, as he nearly outteched Mansa and lost the space race by one turn.
Worst Performance: His near First To Die finish in Game 10 was quite embarrassing, to say the least.
Shaka Award:
That's a gigantic army for 740 AD. What wasted potential from this leader.
Hammurabi of Babylon
Wars Declared: 28
Wars Declared Upon: 13
Survival Percentage: 95%*
Finishes: 2 Firsts, 7 Seconds (24 points)
Kills: 8
Overall Score: 32 points
*This survival rate is fool's gold; Hammurabi was saved by Mansa multiple times, either through the UN or through military liberation.
Hammurabi was clearly born in the wrong universe. His Aggressive/Organized trait combination is a warmonger's dream, combining the stronger units from Aggressive with the ability to maintain and quickly develop one's conquests with Organized. Unfortunately, Hammurabi preferred to play peacefully, and he was ill-equipped to do so save for a good-but-not-great unique building - even his defensive unique unit is just an inferior version of Mansa's Skirmishers.
In most games, Hammurabi stubbornly stuck to his favored "sit-back-and-develop" strategy, to generally mediocre results. He was excellent at surviving due to his corner position, his easy-mode neighbor situation, and his ability to befriend Mansa (the two never fought), but there was still much left to be desired. His wars were usually of little note, either pitiful attempts at conquering the Ottomans or a meaningless cross-map wars against a dying civ. Hammurabi was in many ways the troll of the map, either twiddling his thumbs on his way to an undeserved playoff appearance (like in the Actual Game) or stealing kill points from others who had done all the heavy lifting. In the unusual Game 19 where Mansa died, Hammurabi's small size also came back to bite him, as despite his tech lead, the combined efforts of the much larger Gilgamesh and Alex would overwhelm him to bring the evil leaders their one turn in the limelight.
In the rare cases where Hammurabi actually utilized his warmonger traits, he was able to succeed in this friendly diplomatic environment. His primary avenue to success was to murder Suleiman - his three best games, including his two wins, all saw the Ottomans as First To Die - and with the entire Eastern slice of the map to himself, his Organized trait made him a viable economic contender, occasionally helping him outscale Mansa. He was certainly more impressive than Lincoln. Nevertheless, Hammurabi still proved to be a tragic case of a leader who continuously tries to jam his square peg into round holes to no avail.
Best Performance: His Game 20 teching performance was legitimately strong, and he was on pace to win around Turn 310 when he was prematurely elected World Leader.
Worst Performance: Nearly suiciding into a much stronger Lincoln in Game 8, requiring Mansa to swoop in and save him.
Wang Kon Award: Somehow losing a war against five city Suleiman while Suleiman was stuck in a 2v1 yet still making the playoffs in Game 12.
Chivalry Is Dead Award: Hammurabi's Game 10 teching was uncharacteristically terrible - he did not have Knights until Turn 230, and had to be saved by Mansa to survive.
Gilgamesh of Sumeria
Wars Declared: 29
Wars Declared Upon: 20
Survival Percentage: 20%
Finishes: 1 First, 1 Second (7 points)
Kills: 2
Overall Score: 9 points
I must say, this was the most impressive single-digit Alternate Histories performance I have ever witnessed. Gilgamesh was the only viable evil leader for two main reasons. First, he was by far the most competent of the low peaceweights in this field, scratching and clawing his way towards what little success he could get. He was the only evil leader to win, the only one to make the playoffs in a Mansa game, and the only one who could maintain a semblance of an economy and accomplish anything militarily. Secondly, if one or both of Mansa and Lincoln failed, he was best placed to take advantage. His second place showing saw him run over Lincoln relatively early and use that to hold on when Mansa came calling, and the wild Game 19 of course saw him attack Mansa at the perfect time to become the game leader and eventual winner. His teching was pretty bad - he would have had a post-Turn 400 Spaceship win had Alex not suicided into him - but it did the job and was still impressive considering his horrible land and situation.
His issues did not stop at his cold land. He was always atop the hit list of a much more advanced Mansa or Lincoln, and he was effectively in the wrong side of a 3v1 due to the ineptitude of his fellow low peaceweights. Despite these insurmountable obstacles, he still toughed it out and managed two miracle playoff finishes, demonstrating why he is generally regarded as an above average leader for AI Survivor purposes.
Best Performance: Game 19, obviously, but also Game 10 was probably good enough to win if there was any leader other than Mansa there.
Worst Performance: Getting shredded early by Lincoln in a war he started in Game 13.
Apostolic Palace Cheese Award: Being denied an easy kill on Hammurabi by the UN in Game 10.
Suleiman of the Ottomans
Wars Declared: 24
Wars Declared Upon: 27
Survival Percentage: 35%
Finishes: 0 Firsts, 1 Second (2 points)
Kills: 3
Overall Score: 5 points
Season Three was played on a Tropical map, and an unfortunate side effect was that many games saw a leader doomed from Turn 0 due to having too much crappy jungle. In this game, that leader was Suleiman, who settled this exact city and queued a 30-turn Settler in every single replay:
This was emblematic of the HURR DURR MUST GET RESOURCES mindset ingrained into the Civ IV AI coding, which baited the Silly Man into believing that a location with only Calendar resources locked underneath jungle was a viable one for a second city. Making matters worse was that his third city was usually not much better, as he opted to settle a food poor coastal spot that claimed zero contested land. If Suleiman was not Imperialistic, things would have been even uglier.
By the end of the expansion phase, Suleiman would only have 5-6 weak cities to his name and be a non-entity as a result. There is little else for me to say: he was a boring and irrelevant leader. Sometimes he helped dogpile Alex or Lincoln, other times he launched random wars against Mansa and Ham, but I never cared to pay much attention. I noticed myself multiple times nearly forgetting to record Suleiman's war declarations; he was that bland. I guess he did make the playoffs once, when Gilgamesh had murdered everyone else in the outlier Game 19.
With that said, this set should not have had any bearing on Suleiman's true capabilities. Even Mansa Musa would have failed in this spot due to the quirks of AI programming.
Worst Performance: His Game 20 exit was peak embarrassing Silly Man, dying on Turn 144 to pave the way to a runaway Hammurabi.
Frederick Award: Still getting murdered by Mansa on Game 14 even when Mansa had turned on the Culture slider with just a Cuirassier army.
Alexander of Greece
Wars Declared: 36
Wars Declared Upon: 27
Survival Percentage: 5%
Finishes: 0 Firsts, 0 Seconds (0 points)
Kills: 3
Overall Score: 3 points
Like the Silly Man, Alex was screwed by his land, but unlike the sultan, he was entertaining. Alex was plagued by the same affliction that other ultra-warmongers suffer from: a complete aversion to Mysticism. This was even more apparent in this setup, as Alex's only source of Copper was in a faraway peninsula, and he would always settle his cities in such a way that his iron was in the second ring, like in the livestream. Thus, Alex was a car without an engine, unable to fight the wars that he lives and breathes for. If Mansa was typical Mansa, Alex might have fared a little better, but this was an aggressive Mansa who was more than glad to continuously spank Alex in these replays.
Notice the Iron resources outside of the Greek borders.
Even if Mansa was feeling more forgiving, Alex was still running on borrowed time. Mansa could culture flip cities, he could tech up and run over Alex whenever he wanted, or someone else could bring an end to the Greek farce. All Alex could do was play the role of spoiler, whether it was slowing Mansa down or helping out with dogpiles. This was just not Alex's time to shine.
Best Performance: Believe it or not, Alex did come close to winning the crazy Game 19, as he was within 20 delegates from being elected World Leader of the UN. Perhaps if he had patiently developed his Babylonian conquests, he could have won or at the very least punched a ticket to the playoffs. However, being Alex, he instead threw away his ticket trying to attack a much more advanced Gilgamesh.
Worst Performance: Getting rekt by Mansa on Turn 137 in Game 10.
Ooga Booga Award:
A nice attacking force you got there, Al!
Putting it simply, the best economic leader in Civ IV had easy conquest opportunities which he gladly took advantage of. Combine this with some incredibly mediocre competition, and well, you have the recipe for perhaps one of the most dominant showings ever by a single leader.
