Game Seven Alternate Histories 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 Seven was one of the most surprising and exciting games of the opening round, with noted lunatic Montezuma winning a victory by narrowly stopping Hatshepsut's Cultural attempt. 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:
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
Livestream of First Eight Alternate History Games
(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.)
Game Seven was one of my top choices to subject to additional investigation from Season Seven, enough so that I ran the first eight games of the alternate histories on Livestream prior to beginning the playoff round. The match that we had watched in the real Game Seven had presented the community with all sorts of questions that needed to be answered. What would happen when Elizabeth didn't lose one of her core cities to the barbarians? Would Catherine flub her early expansion as badly in repeat performances? Could Frederick really be THAT pathetic on a consistent basis? And, most importantly, how would the duel play out between Montezuma and Hatshepsut, the two leaders who had clashed for the victory? Based on what we had seen, I thought there were pretty good odds that the real Game Seven had been somewhat of an outlier result that didn't reflect how the map normally played out.
That absolutely proved to be the case as the alternate histories immediately revealed that Hatshepsut was the most likely leader to win the game, overwhelmingly so, and any result where she didn't win was a very unlikely outcome indeed. I had been expecting Hatty to do well and was still somewhat shocked when she captured victories in 7 of the first 8 alternate histories played out on Livestream. She would add another nine victories to that total, including TEN in a row at one stretch, finishing with 17 victories in 20 games. Furthermore, Hatshepsut was never eliminated in a single one of these repeat performances which confirmed that she had been extremely unlucky in the real Game Seven. Hatshepsut won by culture every single time and her single-minded focus on that sole victory condition resulted in lots of early finishes. One of the keys to the map was Elizabeth fighting off the various low peace weight leaders long enough for Hatty to claim her Cultural victory, something that happened in game after game after game. In fact, Hatshepsut and Elizabeth were typically the two strongest AI leaders in the game and usually held hands to finish first and second. Only in the unlikely scenario where Elizabeth was knocked out early was there a realistic chance for Hatshepsut to falter.
The other key to the map was the relationship between Pacal and Montezuma. They bizarrely shared the same religion in the real Game Seven and never fought one another, with their long peaceful border allowing the Aztecs to go conquering off in England and later Egypt with great success. This, uh, did not happen in the alternate histories. Montezuma and Pacal typically founded different religions and fought with one another relentlessly, over and over again, with their conflict being one of the defining characteristics of the map. With two of the three low peace weight leaders at one another's throats, the more militaristic civs were unable to work together against the peaceful economic ones as we'd watched on Livestream. Elizabeth in particular fared much better since she was often able to avoid the 1 vs 2 wars which were her undoing in the actual Game Seven. She was eliminated only 25% of the time and had a ton of second place finishes which would have been first place finishes if Hatty wasn't winning by culture so quickly. We saw the single weakest Elizabeth performance out of the 20 map replays; she was exceptionally unlucky and did vastly better in pretty much every alternate history.
The biggest disappointment from a community perspective was Catherine who had been the favorite coming into Game Seven. However, the alternate histories confirmed that she hadn't been unlucky in Game Seven but rather simply didn't perform especially well in this scenario. Catherine continued to refuse to settle her nearby gold resource and she was hampered repeatly by barbarian cities that popped up along her borders and curtailed expansion. Even when Catherine had a solid early game though (which did happen as she sometimes came out of the landgrab phase in top position), she was unable to translate her opening into consistent victories. She basically needed help to take down Elizabeth and start snowballing, something that rarely occurred as Montezuma was too busy fighting Pacal or Frederick or even declaring war on Catherine himself. Elizabeth and Hatshepsut were simply too strong for Catherine to beat either one of them consistently, and even when she rarely conquered Elizabeth, Hatty was still lurking to sneak out her usual Cultural victory. Catherine had one atypical Domination win in Game #5 and that was basically it for the alternate histories.
The map replays also confirmed that Pacal was a nonentity on this map, suffering from a poor starting position and a bad case of neighboring Montezuma. Even when left alone, his economy was always inferior to Elizabeth and he couldn't outrace Hatshepsut's lightning-fast Cultural finishes. This left him in no-man's land with a few random second place finishes and not much else to show for his efforts. Frederick was even more pathetic on this map, with an excellent survival rate since Elizabeth and Hatshepsut wouldn't attack him but nothing else to his name. While he fared a bit better than the utter disaster of the real Game Seven, he was completely unimpressive and badly overshadowed by the other high peace weight leaders. The data has really piled up by this point that Frederick is a poor AI leader, just not good enough at playing the econ game and helpless at anything else.
Finally, there were some individuals in the community wondering if we had underestimated Montezuma as an AI in the wake of his victory in the real Game Seven. Maybe he was better than we had all thought? The alternate histories emphatically undercut that line of argument, decisively showing once again the Monty is among the worst AI leaders in the game. Montezuma was utterly incapable of repeating his Domination victory, with zero first place finishes and only one game where he achieved anything at all (the final alternate history in Game #20). Instead, he engaged in the exact same self-destructive behavior that we've come to expect over the years, launching wars against anyone and everyone that rarely led to any success. Monty even managed the rare coup of having all five other AI leaders at war with the Aztecs simultaneously in one of the alternate histories which I didn't even know was possible without the Apostolic Palace. Long story short, Montezuma was First to Die an incredible 70% of the time (!!!) which is all that you need to know about his performance. We basically watched the 1 in 100 toss of the dice where everything lined up perfectly for Monty - not something sustainable over any length of time.
Now for a look at the individual leaders:
Hatshepsut of Egypt
Wars Declared: 20
Wars Declared Upon: 13
Survival Percentage: 100%
Finishes: 17 Firsts, 1 Second (87 points)
Overall Score: 93 points
The alternate histories decisively proved that Hatshepsut was the leader who "should" have won Game Seven. She nearly did win the game that we watched despite all sorts of unlucky breaks elsewhere on the map, between Elizabeth faltering early and Montezuma somehow choosing his wars in an efficient manner. Hatshepsut would typically have a solid but not spectacular landgrab, avoid much in the way of aggression from anyone else, and then turn on the culture slider at an early date to win well before Turn 300. This wouldn't have worked most of the time, however her isolated starting position and the presence of Elizabeth as a friendly bodyguard combined to result in a win something like 80-90% of the time. I want to stress again here how successful Hatshepsut was at avoiding conflict: she only faced 33 total wars (and a mere 13 defensive wars!) across 20 map replays. Compare that to Montezuma who found himself in 89 total wars and it's clear that they weren't even playing the same game most of the time. Hatshepsut basically did her best Gandhi impression and it worked nearly all of the time.
Elizabeth of England
Wars Declared: 31
Wars Declared Upon: 44
Survival Percentage: 75%
Finishes: 2 Firsts, 10 Seconds (30 points)
Overall Score: 41 points
Elizabeth was probably the strongest overall leader on this map in terms of raw food + production + tech, and she would have won the most victories if the Cultural path somehow didn't exist as an option. The fate of Elizabeth was tied together with Hatshepsut with these two high peace weight leaders mutually reinforcing one another's positions. Unfortunately for Elizabeth, despite being incredibly strong in many of these games and on pace to win multiple pre-Turn 300 Spaceship victories, it was nearly impossible for Elizabeth to complete the tech tree and win by space before Hatty could complete an earlier cultural finish. Elizabeth wasn't helped either by having a much higher degree of difficulty; she was attacked 44 times (more than three times as often as Hatshepsut) and found herself involved in heavy warring in most games. The English queen got the better of most of these fights though, as she was only eliminated 25% of the time and scored the most kills of anyone in the alternate histories. Yes, Elizabeth finished with the most kills from this group! That was another sign of her strength on this map and the overall peaceful nature of the alternate histories - the total number of kills was low for everyone. Elizabeth was genuinely unlucky in the real Game Seven and had vastly better odds to make the playoffs than Pacal or (especially) Montezuma.
Catherine of Russia
Wars Declared: 49
Wars Declared Upon: 27
Survival Percentage: 50%
Finishes: 1 First, 3 Seconds (11 points)
Overall Score: 19 points
Hatshepsut and Elizabeth were the top two leaders in the alternate histories by a wide margin. (Don't be fooled by the much lower score of Elizabeth, as she would have won by spaceship in most of these games if Hatty hadn't been even faster to win by culture.) Then there was a huge drop off to the remaining four leaders who were essentially also-rans that almost never had any chance at claiming a victory. Catherine was the only other leader to score a single victory, in the atypical Game #5, snowballing off an Elizabeth elimination and managing to reach Domination by stopping Hatshepsut's Cultural victory attempt before it could get three cities to 50k. The Russians were competitive in many of these alternate histories despite some early game follies, however it was nearly impossible for Cathy to run over enough of the map to claim Domination before Hatshepsut could complete her cultural ending. I was reminded of Cathy's playoff game from Season Six where she found herself in a similar position with Mansa Musa although she was even worse off in this game. The fact that her low peace weight "allies" couldn't stop bashing each other over the head was a real problem as well. Cathy was solid enough in this game but that just wasn't good enough against Hatty's culture and Elizabeth's teching.
Pacal of the Mayans
Wars Declared: 21
Wars Declared Upon: 34
Survival Percentage: 80%
Finishes: 0 Firsts, 4 Seconds (8 points)
Overall Score: 14 points
None of the remaining three leaders were able to claim a single victory in the alternate histories or come particularly close to winning. Pacal was thoroughly unimpressive as he followed one of those "sit in the corner and do nothing" strategies that we've seen him follow from time to time. His greatest problem was bordering Montezuma as the Aztecs attacked relentlessly and prevented Pacal from ever getting out to a dominant economic position. Pacal typically won that fight eventually (with help from Elizabeth and Frederick) but he was never able to achieve a true snowball and start running over the rest of the field. Thus he had a great survival rate at 80% but achieved very little in the way of kills or conquered territory. If you scroll through the ending screenshots above, you'll see that Pacal's blue borders sit unchanging in the same spot across seemingly every game. Even when he wasn't attacked, Pacal simply couldn't keep pace with the two eastern high peace weight leaders. Elizabeth would always beat him to space if she hadn't been conquered and Hatshepsut was even faster with her cultural stuff. This was a dull, uninteresting performance from Pacal which didn't reflect the strength he's shown in other settings. It was definitely a subpar scenario for the Mayans.
Montezuma of the Aztecs
Wars Declared: 37
Wars Declared Upon: 52
Survival Percentage: 20%
Finishes: 0 Firsts, 1 Second (2 points)
Overall Score: 6 points
In median terms, Montezuma was clearly the worst leader on this map as he was First to Die an incredible 14 times in 20 games. No one else was First to Die more than twice while the Aztec leader suffered that fate fully 70% of the time. It was amazing how Montezuma managed to anger literally everyone else in the game with his neverending warring; he displayed none of the patience from the real Game Seven and instead struck out wildly in every direction at once. As expected, this was a total failure of a strategy and resulted in Monty suffering even more invasions than Elizabeth. In a situation where Montezuma needed to work with Pacal and Catherine to take down the high peace weight AI leaders, the Aztec monarch instead attacked both of them incessantly, especially Pacal who rarely managed to avoid an invasion. It all blew up in Monty's face as he died repeatedly without achieving anything. Montezuma did outscore Frederick under our scoring system since he actually managed a pair of kills and a second place finish in Game #20 as Aztec aggression had one of its rare sucesses. Monty was therefore more of a feast-or-famine leader who could very rarely find success in his wild gambles. However, it was vastly more likely that he would flail around with suicidal aggression that would later come home to roost. I'm struggling to put into words how much of an outlier result the actual Game Seven was for Montezuma, basically the idiot who repeatedly goes all-in with nothing but straight draws in poker who happened to hit the nuts on the river.
Frederick of Germany
Wars Declared: 25
Wars Declared Upon: 13
Survival Percentage: 85%
Finishes: 0 Firsts, 1 Second (2 points)
Overall Score: 5 points
Frederick was an equally terrible leader with the big difference of surviving to the finish in nearly every game. This wasn't due to his own efforts, of course, but rather caused by high peace weight Hatshepsut and Elizabeth being the strongest leaders on the map. Neither one would attack him and thus their power kept Frederick from dying off. The German leader did basically nothing with their protection, of course, only managing a single runner up finish despite surviving in 17 out of 20 games. He was the true archetype of the "sit in the corner and do nothing" AI leader that we've seen from characters like Hammurabi and Roosevelt in the past. Frederick also faced a mere 13 attacks over the course of the alternate histories and proved to be nearly as sheltered from fighting as Hatshepsut. Unlike her, he was never remotely close to winning either Spaceship or Cultural victories and largely existed at the sufferance of his allies. Frederick was kind of the anti-Montezuma: instead of being inept due to over-aggression, he was inept due to over-pacifism while lacking the leader traits or starting position to keep pace with better economic competitors.
Oftentimes I'll talk here about whether or not the game we watched on Livestream was representative of the "normal" outcomes coming from the alternate histories. Game Seven doesn't need discussion on that point though: it's obvious to everyone that we saw an extreme outlier result with Hatshepsut and Elizabeth both getting eliminated while Pacal and Montezuma (!!!) took the top spots for the match. We legitimately might have needed to re-run the map 50 or 100 times to see another Montezuma victory, it was that unlikely of an event. Amusingly, there was heavy betting on this match in the fantasy contest with the bids going to Pacal and Catherine. Neither one turned out to be a good choice at all though Pacal's fortunate runner up finish resulted in him scoring decent fantasy points. I'm glad that we were able to watch such an entertaining Game Seven as a group, as it was one of the best matches of the opening round from a viewer perspective even though it turned out to be a wildly unusual result. I do feel bad though for Hatshepsut and Elizabeth - they were tough luck losers on this map, especially Hatshepsut who remained the only leader with zero official points. This was her big chance and she was hit with terrible RNG at the worst possible moment!
Thanks as always for reading, I hope you enjoyed this look back at Game Seven!