Playoff Game Two 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. Playoff Game Two turned into an early Cultural victory by Louis with Mehmed unexpectedly tagging along in a strong second place. 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
(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.)
I was looking to run alternate histories for one of the three playoff games to fill the one-week gap between the playoff round and the Championship game. Out of the three games that we had seen, Playoff Game Two looked to be the most interesting due to its apparent open-ended nature. This had been an incredibly hard game to predict with all six leaders sharing low peace weights and none of them starting with Mysticism tech to make it hard to ascertain where the religions would appear. We ended up watching a game where Louis and De Gaulle were contending for the top spot on the leaderboard, only for De Gaulle to fall apart in a 1 vs 2 war followed by Louis coasting to a Cultural victory. I wanted to know how this map would play out if we ran it another 20 times - what was the most common result on this otherwise unpredictable setup?
Well, after going through the full 20 alternate histories, I didn't end up with much more certainty than before I started! The one thing that could be definitively stated was that Louis was indeed the top leader on this map and by a wide margin. He won 12 times in the 20 map replays and it was almost always via the same Cultural path, removing the possibility for anyone else to win through Spaceship or Domination by wrapping things up with a faster Cultural ending. Louis founded one of the two early religions in nearly every game and he used this to great effect in terms of making religious allies. The French leader usually was able to avoid two front wars and typically found himself on the postitive side of the diplomatic game. Many of these games stalemated with endless warring between the various AI leaders and in those situations Louis would usually take home the victory through his dogged pursuit of culture. In that respect, what we watched in the Livestream game was commonplace for this map.
Unfortunately for providing a clean narrative, just about everything else about Playoff Game Two was completely up in the air and failed to follow consistent patterns. Kublai Khan, Mehmed, and De Gaulle each scored very similarly across the 20 map replays and Hannibal wasn't far behind them either. Only Tokugawa was truly hopeless as a leader on this map with the four other leaders all having roughly equivalent odds to land a Top Two finish. We would need a sample size larger than 20 games to get an accurate accounting of their true standing on this map; all that I can say is that this was a total jumble of a game that failed to follow any clear trends (outside of all those Louis Cultural wins).
The biggest surprise for me was the strength of Mehmed in the alternate histories which proved that his Runner Up finish in the actual Playoff Game Two was not a fluke. Before the game started, I had tabbed Mehmed as the weakest leader on the board due to his jungle-heavy start and thought that he would be snapped up by De Gaulle or Louis. This was emphatically not the case; Mehmed handled himself quite well in the expansion phase despite all those jungle tiles and punched above his weight in warfare thanks to a Military tech flavor along with copper and ivory resources near his capital. Mehmed was also frequently the beneficiary of positive diplomacy: he would usually pick up Louis' religion and therefore go to "Pleased" with the northern French leader, thus avoiding conflict with the strongest leader on the map since Mehmed refuses to declare war at "Pleased" relations. This channeled Mehmed's aggression south towards De Gaulle in what proved to be a more fruitful direction.
Mehmed's strength turned out to be De Gaulle's bane in many of the map replays, much as we saw in the actual Playoff Game Two. With Mehmed locked into peace with Louis, he had nowhere else to go except straight into De Gaulle and thus the two of them fought in game after game. Now in theory De Gaulle should have emerged victorious in that clash most of the time since he had the far superior starting position, however Kublai Khan would often intervene as well to create the 1 vs 2 situation that we saw on Livestream, inevitably causing a De Gaulle collapse. Unfortunately for De Gaulle, even if he founded his own religion and converted Kublai Khan, that still wasn't enough to ensure safety since the Mongol leader will plot war at "Pleased". And frankly De Gaulle sometimes just lost the straight up 1 vs 1 against Mehmed due to idiotic tech choices and over-production of wonders, including one game where he threw away a certain victory by going all the way to Industrialism tech (!) without researching Rifling. The map replays convincingly showed that De Gaulle is a poor leader for AI Survivor purposes, wasting a strong position with poor research and build choices. I'm convinced that a leader like Hannibal would have crushed the field from this start and De Gaulle couldn't even manage to keep pace with Mehmed from a far inferior position.
As for the other leaders, Kublai Khan needed to get out in front of the rest of the field with an early conquest in order to do well. This took place in both of his victories and allowed him to snowball over the rest of the board to a first place finish. However, Kublai Khan tended to suffer from a poor economy and there were several games where he swelled up to a big size but couldn't come close to any of the victory conditions before Louis' ticking Cultural victory finished. This was responsible for Kublai having the most second place finishes in the field, if not by much of a margin. In other games, his wars were unprofitable and he simply stalled out in the middle of the leaderboard without accomplishing much.
Finally, Hannibal and Tokugawa simply lacked enough space to be truly competitive on this map. Hannibal could never quite get his economy going despite a nice capital city because he was always boxed in by Kublai Khan and Tokugawa, leading to frequent wars with both of them. Kublai interestingly went east with his initial settler in many of these games, not west as we watched on Livestream, which only squeezed Hannibal that much further. He was able to hang around for some second place finishes but the only Carthaginian win was a Diplomatic ending that he probably didn't deserve. Meanwhile, Tokugawa had even less territory available while also having terrible economic traits and a personality that caused him to hate everyone else diplomatically. This was not a recipe for success and he wound up with the worst results by a clear margin. Kublai, Mehmed, De Gaulle, and Hannibal all were close competitors with somewhere between 15 and 22 points from Top Two finishes while Tokugawa managed a grand total of 2 points from a single distant Runner Up spot. He was First to Die bait and nothing more in this setup.
Now for a look at the individual leaders:
Louis of France
Wars Declared: 24
Wars Declared Upon: 28
Survival Percentage: 80%
Finishes: 12 Firsts, 1 Second (62 points)
Overall Score: 71 points
On a map full of uncertainty, Louis was the one great constant figure as he consistently made successful bids for Cultural victory. Louis had a strong river-heavy starting position and he was typically able to use a self-founded religion to make allies and avoid lopsided conflicts. It helped that Mehmed most frequently fought with De Gaulle over to his west while Tokugawa tended to get into scrapes with Hannibal off to his east, leaving Louis to sit back and build lots of wonders in peace across a bunch of games. On a map where there were no truly dominant economic leaders, Louis was often left uncontested as he raced to a fast Cultural ending long before anyone else could come close to an alternate victory condition. That's not to say that Louis was unbeatable in this scenario, as sometimes the diplomacy went against him and he was dogpiled, including a First to Die result in Game #14. That was an outlier result though, and the most common outcome on this map was a lot of fruitless warring from everyone else leading into a Louis Cultural victory somewhere in the Turn 270-300 range.
Kublai Khan of the Mongols
Wars Declared: 40
Wars Declared Upon: 24
Survival Percentage: 75%
Finishes: 2 Firsts, 6 Seconds (22 points)
Overall Score: 33 points
Kublai Khan was unsurprisingly the most aggressive leader on this map, narrowly starting the most wars thanks in part to his ability to declare war at "Pleased" relations. He had the most success going west and taking over De Gaulle's wonder-filled and floodplains-heavy territory which helped to compensate for Kublai's typically mediocre economy. He also attacked Hannibal to the east roughly as often although even a successful conquest there didn't do that much to advance the Mongol position. The biggest problem for Kublai was the existence of Louis on this map, as Louis would consistently win by Culture before Kublai could get into a position to win via Domination or Spaceship. With the northern French leader cutting the total game time down to less than 300 turns, there simply wasn't enough time for Kublai to win by his preferred victory types. The Mongols were able to win in Game #7 because Louis lost one of his three Legendary cities to cripple the Cultural attempt, and then again in Game #14 because Louis was First to Die in an early dogpile. Absent these unusual events though, Kublai was typically relegated to the second place spot since he couldn't out-race Louis to the finish line.
Mehmed of the Ottomans
Wars Declared: 38
Wars Declared Upon: 23
Survival Percentage: 60%
Finishes: 3 Firsts, 2 Seconds (19 points)
Overall Score: 30 points
Mehmed was the king of making something out of nothing on this map, taking a weak starting location that some community members said shouldn't even have been allowed for an AI Survivor game and turning it into a solid performer across the alternate histories. His performance was inevitably tied to his life-or-death struggle with De Gaulle that manifested in basically every game. In most games only one of them could survive to the finish and astonishingly Mehmed came out on top more often that not. He was a bit of a feast-or-famine leader with only a 60% survival rate but three victories, more than anyone else other than Louis. Granted, some of that was small sample size bias, but it wasn't a fluke that Mehmed tied Kublai Khan for the most kills with 11. Mehmed was actively trying to advance his position by hunting for advantageous wars and taking pieces out of De Gaulle's hide. The ivory resource near the Ottoman capital seemed to make a big difference as Mehmed was consistently able to attack with elephants which De Gaulle struggled to counter. Whatever the reason, Mehmed overperformed in a huge way and his strength was my own personal biggest misread of how the map would play out.
De Gaulle of France
Wars Declared: 20
Wars Declared Upon: 45
Survival Percentage: 45%
Finishes: 2 Firsts, 5 Seconds (20 points)
Overall Score: 28 points
De Gaulle was the opposite of Mehmed, the leader who did the most to throw away an excellent starting location with poor decisions. De Gaulle was in some ways the victim of bad diplomacy, frequently stuck in a 1 vs 2 situation against Mehmed and Kublai Khan that the AI always struggles to deal with in Civ4. He was also attacked in bizarre fashion by Hannibal in multiple different games; I have no idea why the Carthaginian leader wanted to charge across the map to sabotage both of their games but it happened several different times across the alternate histories. De Gaulle was invaded 45 times which was far more than any of the other leaders - not what I would have expected given his relatively sheltered start. However, De Gaulle also failed to help himself by consistently refusing to research military techs and over-building wonders that he didn't need. There were many games where he could have crushed Mehmed by using his superior economy to hit knights or rifles way ahead of time, and instead De Gaulle would be faffing about with Physics or Constitution or whatever technologies instead. He both lacked the killer instinct of a true warmonger while also failing to have the economic traits of Mansa Musa or Willem or Darius. De Gaulle was therefore left in a worst of both worlds situation, effectively acting like a budget Willem who rarely managed to accomplish anything. Unquestionably the most disappointing leader in the alternate histories for this game.
Hannibal of Carthage
Wars Declared: 32
Wars Declared Upon: 34
Survival Percentage: 50%
Finishes: 1 First, 5 Seconds (15 points)
Overall Score: 21 points
Unlike the result of his opening round game, Hannibal was unable to dig himself out of the cramped starting position in which he found himself. He made every attempt to fight his way out of the corner of the map, with Carthage involved in more wars than any of the other nations, however it just wasn't sufficient for Hannibal to translate into victories. He clashed constantly with both Tokugawa and Kublai Khan across the alternate histories, with Hannibal able to hold his own against either one individually only to falter when trapped in the dreaded 1 vs 2 situation. This was the reason why Hannibal ended up as the leader most likely to be First to Die, suffering that ignominious fate eight times in all. On the other hand, when Hannibal was able to survive he had pretty decent odds to slip into a Top Two finish with five different Runner Up spots. Hannibal was never really in contention to win outright though, with his sole victory coming as a Diplomatic win shortly before Louis was about to take another Cultural ending. While he was competitive on this map, Hannibal seems to have been a tier lower than the Kublai/Mehmed/De Gaulle trio in terms of overall strength.
Tokugawa of Japan
Wars Declared: 31
Wars Declared Upon: 31
Survival Percentage: 35%
Finishes: 0 Firsta, 1 Second (2 points)
Overall Score: 8 points
Tokugawa also had a cramped starting position but lacked the Financial trait and better overall AI personality enjoyed by Hannibal. He spent most of the alternate histories warring unsuccessfully with all of his neighbors without ever achieving much of note. Tokugawa's isolationist AI personality meant that he had difficulty making friends and he was prone to lashing out at anyone in range. Toku was unquestionably a tough fighter thanks to all of the promotions on his units, however his terrible economy and isolated diplomatic status usually resulted in him getting eliminated at some point in the Turn 150-250 range. At no point in time was he ever remotely competitive to claim a victory himself, only serving as a spoiler candidate to ruin the games of other leaders who actually had a shot. He was definitely the weakest AI leader on this map and the best thing that can be said about Tokugawa is that he was slightly less likely to be First to Die than Hannibal.
I feel as though the writeup for this game was kind of a mess and I wish that I'd been able to tell a more coherent story based on the alternate histories. As best I can tell though, this game really was a total crapshoot for everything other than Louis winning by Cultural victory. Four of the other five leaders all had roughly the same odds to claim the Runner Up spot, and if Louis didn't manage to win via Culture, it was again completely open as far as who would claim the vacated victory. Basically Louis was strong and Tokugawa was weak, then who the heck knew what would happen with the other four leaders across the various alternate histories. I was impressed by Mehmed and disappointed by De Gaulle, otherwise it was hard to know what to make of this game. Very much a weird game overall!
Thanks as always for reading, I hope you enjoyed this look back at Playoff Game Two!