Game Six Alternate Histories Online Spreadsheet
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. Game Six saw pre-game favorite Mansa Musa suffer First To Die, allowing Darius to run away technologically while the low peace weight leaders focused on other targets. 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 we have since used the same general format. This particular set of alternate histories were run by Myth - many thanks for spending so much time on this task! 
Myth posted the resulting data from the alternate histories and then discusses some of the findings below 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 originally noted, 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:
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. Also, in accordance with recent changes, the Apostolic Palace was turned OFF for these tests.)
Myth: Whether it be a leader snowballing a domination victory off of consistently weak neighbors, or a leader being left alone consistently enough to get to a giant technological lead, there is almost always a pattern that emerges in the Alternate Histories (though not always!). On this map, the only thing resembling this pattern was the inability of Hatshepsut to get out from the Gilgamesh and Mao sandwich she found herself cursed with.
Poor Hatshepsut was killed a whopping 75% of the time, usually due to her peace weight situation. The poor girl found herself with a peace weight of 9 in a neighborhood where her closest neighbors were 1 (Mao) and Gilgamesh (2). This inevitably led to a clash with at least one of these leaders, with the earliest contender being Mao. This is what played out in the actual Game 6 and it repeated itself over and over again while running these games. Egypt found themselves in a 1v2 in almost every single game, leading them to either be gobbled up by China or partitioned between Mao and Gilgamesh, catapulting their neighbors into strong positions in the vast majority of games and giving her a 45% First To Die rate.
With that being said, for the few games Hatty was left alone (or beat Mao in the 1v1), she outperformed the two financial leaders in the West. Her economies in Games 1, 6, and 13 were consistently much better than either Mansa's or Darius's. She even managed to win twice, both through culture, and was a SINGLE TURN away from another cultural win in Game 1 (losing out in a last minute Diplomatic Vote to Mansa). I think this shows that Hatty got really unlucky with the map setup here. If she wasn't next to two leaders who consistently invaded her civ, she likely would have been able to win at least 25% of the time.
As mentioned above, Mao was the big winner from Egypt's diplomatic situation. Mao and Hatty fought in every single game, a full 100% of the runs, and it usually was extremely early. To her credit, Hatshepsut usually was able to hold at first, and only really collapsed when Gilgamesh (or, bizarrely, Mansa in 3 games) stabbed her from the other side. She even was able to solo conquer China in Game 13 when everyone else left both leaders alone. However, these times were rare, and China usually ballooned to a large size in game after game, leading to Mao's 50% Top-Two rate.
Gilgamesh was the other winner of all of this, but Sumeria was a lot less consistent. This may be because he often focused on Sitting Bull on his other side more than Egypt, and Sitting Bull was consistently just harder to crack (Protective at work?). Sitting Bull also didn't help this by declaring on Gilgamesh far more often than any other leader in the game. This led to Gilgamesh getting caught up in wars and falling behind in technology to Mao or the financial duo in the west, which led to a lot of late eliminations for him as his neighbors took him out after dealing with everyone else.
As for the other leaders, Brennus and a surprising Mansa Musa were pretty much tied for 3rd on this map. Brennus makes sense - he was low PW [peace weight] and therefore had a much better diplomatic situation with the two eastern leaders. Often, what decided his games depended upon who founded the other of the first religions and who converted to his. When Hatty or Mansa founded it, Brennus would do well; when Mao or Gilgamesh founded it, he would falter. Mansa, on the other hand, depended on being left alone; his corner position and meatshield to his east helped greatly in this. However, his land quality was horrendous, and he really couldn't get his economy going to the same degree as he usually does.
Darius, the winner of the actual Game 6, was unable to repeat his Actual Game Magic in the majority of games. He tied Mansa for the second lowest elimination rate, but could only translate that into a top position 25% of the time. For his start, that's almost sad! He constantly clashed with Brennus to his east (fighting him in 100% of games) and oddly enough started the conflict the majority of the time. This probably is because in the code the AI can attack with JUST Immortals, one of the only units that fall in this category, allowing him to plot earlier than Brennus.
For our final leader, Sitting Bull was doomed on this map. Outside of a very unlikely and fluky second place in Game 15, the Native American leader just sat around and either died at the hands of his neighbors or did nothing. He often declared on Gilgamesh and subsequently died (great warmongering, SB!) or, more often, got attacked by Gilgamesh and died eventually. He was definitely good at hanging around most of the time, but his 75% elimination and 25% first to die rates shows that he didn't do much else.
Finally, it's important to note how late these games went: there was only one pre-300 finish (Hatty's cultural win on 299 in Game 13), and the vast majority of games went past turn 330 with an average of 341. There were also tons and tons of wars: 10+ in every single game! This undoubtedly helped our warmongers here, who aren't known for their economic prowess.
Now for a look at the individual leaders:
Mao Zedong of China
Wars Declared: 54
Wars Declared Upon: 30
Survival Percentage: 65%
Finishes: 7 Firsts, 3 Seconds (41 points)
Kills: 19
Overall Score: 60 points
As mentioned above, Mao was by far the best leader on this map. I would argue that it wasn't necessarily his own doing and more so the ease of access to Egypt's collapse that led him to this position. He had plenty of games where he completely squandered what should have been an easy win due to the poor handling of his economy. 3 of his 7 wins also came in the last 4 games, so I'm unsure if his 35% win rate would be replicated in a 100-sample test. However, there is no denying that he was the strongest leader on this map.
Gilgamesh of Sumeria
Wars Declared: 63
Wars Declared Upon: 39
Survival Percentage: 45%
Finishes: 3 Firsts, 4 Seconds (23 points)
Kills: 19
Overall Score: 42 points
It always interested me how Gilgamesh would do in these Alternate Histories, seeing as this game was his breakout performance that led to him eventually entering Pool One as of Season 7. For what I saw, his second place performance was definitely not a fluke, though he seemed stronger in the Actual Game than most of his showings. Gilgamesh's biggest issues were over-aggression and economy - late-game he usually would attack someone and get backstabbed or get dogpiled by neighbors with much better technology. This led to a 55% elimination rate, much higher than usual for being second place on a particular map. His impressive 19 kills (tied with Mao for the highest) carried most of his score here, showing that he often could conquer neighbors but was unable to place as often after doing so.
Brennus of the Celts
Wars Declared: 46
Wars Declared Upon: 59
Survival Percentage: 45%
Finishes: 2 Firsts, 4 Seconds (18 points)
Kills: 14
Overall Score: 32 points
Brennus played extremely well in these reruns, but he was stuck in a center position with borders touching literally every single leader at one point or another. He also had the misfortune to often split religions with Mao or Gilgamesh, leading to "Pleased" relations with them, and allowing him to be attacked after they had gobbled Egypt or Native America. To his credit, he never lost a 1v1 unless it was very late in the game. He clashed with Darius in every single game - they never once found themselves at peace for longer than the first 100 turns. This differed from the actual game: Mansa was NOT his most likely first target! Usually, he found himself declared upon by Darius between turn 50-70, though they usually made fairly early peace when Darius was the aggressor for whatever reason. Afterwards, he would often attack Sitting Bull or Mansa, but also quite often found himself in wars with Gilgamesh and Mao. These were actually his best targets, as rolling them up while they were off fighting someone else gave the Celts an extremely strong position that wasn't as precarious as if he went west. In the end, however, he definitely struggled a bit more than his low peace weight counterparts due to being in the middle of this crazy continent.
Mansa Musa of Mali
Wars Declared: 27
Wars Declared Upon: 26
Survival Percentage: 60%
Finishes: 3 Firsts, 4 Seconds (23 points)
Kills: 8
Overall Score: 31 points
Mansa, undeniably one of the best leaders in AI Survivor, graded out as the best of the high peace weights due to his isolation. He wound up in the fewest number of wars of anyone in the game (53) and often found himself completely left alone until turn 150+. This sounds like a perfect setup for Mansa to snowball ahead and destroy everyone, but it was not to be: his land quality was absolutely atrocious. He had few early settling opportunities that weren't directly north or south, and Sitting Bull and Darius also often settled towards him, leaving him with even less room. And even of those cities, none of those were very high quality. He had a great pigs + double gold location to his southeast, but never once did he settle in a way that he could get all 3 resources in one city due to the oddities of the AI's settling coding (thanks Aetryn!). So what we were left with was a pretty sad Mali, who had everything fall into place and still didn't do great through the quirks of the map.
Darius of Persia
Wars Declared: 36
Wars Declared Upon: 31
Survival Percentage: 60%
Finishes: 3 Firsts, 2 Seconds (19 points)
Kills: 6
Overall Score: 25 points
Darius, our actual game 6 winner, wound up only grading out only slightly lower than the Brennus/Mansa duo, and I can honestly say he did more than I expected. Just look at those wars: he declared more than he was declared upon??? Who is this??? 
All kidding aside, Darius was pretty proactive early on in this game. He declared on Brennus way more often than he was declared on in the first 100 turns, and honestly might be the reason why Mansa did as well as he did on this map. Taking the aggression/attention of the biggest threat to Mali was such a nice thing for Darius to do! However, he often didn't do much with the land he was given. He had a great starting position with an amazing city spot with double gems to his northeast (fairly, though, these were under the jungle) and in most games did absolutely nothing with it. He was proactive in starting wars, but often failed to win them without someone else helping. He often made peace at the first opportunity and sat there as someone else ballooned to an unassailable position. His only truly impressive game was Game 8, where he managed to conquer and conquer until hitting domination.
Hatshepsut of Egypt
Wars Declared: 18
Wars Declared Upon: 45
Survival Percentage: 25%
Finishes: 2 Firsts, 2 Seconds (14 points)
Kills: 2
Overall Score: 16 points
Oh Hatshepsut, how much you pulled off with this terrible start is darn impressive. 2 First Places (should've been 3!) and 2 Second Places with those neighbors is much better than I ever expected going into this. That being said, it wasn't all great from her. Her war tactics were... less than stellar... and she didn't do herself any favors by settling strangely in many games. Her culture was insane, but she almost never lived to reap the rewards of it. There's not much else to say that hasn't already been said: she was doomed from her inception onto this hateful continent.
Sitting Bull of the Native Americans
Wars Declared: 23
Wars Declared Upon: 37
Survival Percentage: 25%
Finishes: 1 Second (2 points)
Kills: 3
Overall Score: 5 points
And then we have Sitting Duck Bull. I can't blame his low score completely on his personality: he had no room! In game after game he was squeezed horrendously from both sides. Having a creative leader to his east especially made him unable to do much in that direction. That isn't to say the map was the only issue for Sitting Bull. The poor guy found himself locked in an eternal struggle with Gilgamesh and he could not defend himself to save his life (literally). He often lost cities with equal military power, or started the war only to be eliminated in the 1v1 after. Of his 3 kills, only 1 was really deserved, the other 2 being last hits on a target weakened greatly by someone else. His singular second place finish was so insanely unlikely that if I repeated this map another 80 times I don't know if he'd repeat it more than once or twice. The combination of an unfortunate map position and a terrible AI personality for Survivor led to a leader with nearly no chance.
This game was a lot closer than a lot of games we've seen in AI Survivor. Mao Zedong was obviously the best leader, but the next 4 were pretty similar in strength. To repeat, the true story was the peace weight situation in the east leading to strong empires on that side of the continent, and almost every game came down to some kind of leader in the east taking a top two. However, every leader except Sitting Bull had at least 4 top two finishes - that's a pretty balanced game!
In terms of comparing this game to the actual Game 6 of Season 4, we saw both a typical and atypical game. First, and most glaring, Mansa Musa was only First To Die 10% of the time, and survived 60% of the time! The pre-game favorite had very good odds to move on to at least the Wildcard, something that unfortunately did not develop in the actual game. Of course, the pre-game favorite for First To Die was definitely the correct pick, though she also wasn't FTD 50% of the time. Second, Darius was not most likely to win, though he often survived to move out of the game in one way or another. The truly robbed leader was Mao, but at least he survived to do well in the Wildcard and make it to the playoffs anyway. Finally, Gilgamesh definitely deserved his second place, but that could've gone in so many different directions as well. In the end, this game was an utter mess and without Mao's streak toward the end I would've struggled mightily with writing this.
I hope everyone enjoyed looking back at this game from Season 4!
