Season Seven Fantasy Retrospective

There turned out to be very little drama in the competition as Amicalola's squad ran away with the fantasy scoring with nearly double the points of second place cutHrAXys. Powered by Season Seven Champion Louis along with runs to the championship game from Darius and Mehmed, Amicalola shattered the previous winning tally of 37 points from my fantasy team in Season Six despite having one additional competitor to bid against! This was another difference from the previous fantasy contest where the scoring range ran from 37 points to 11 points; as with the draft process, there were both higher highs and lower lows in this round of AI Survivor fantasy. I found it interesting that Slashin' and Henrik wound up with similar totals despite the former having more than double the total number of leaders of the latter. There didn't seem to be much of a pattern of having more or fewer leaders resulting in better scoring for this particular fantasy season.

After the season concluded and the community started running the alternate histories, I wanted to go back and revist the fantasy contest to take a fresh look at which draft picks turned out to be the best choices in light of running those replays of each map. For example, Saladin was a huge winner for sunhuge at a miniscule price while his expensive bid on Gilgamesh turned into a dud. Were these the most likely outcomes for those picks or lucky/unlucky breaks of fortune? We're therefore going to take a look at each of the opening round games in the specific context of the fantasy draft using the additional data from the alternate histories. Note that this analysis can *ONLY* be done on the opening round games because naturally the playoff matches were dependent on what happened in the first eight games. If a different leader had been victorious in Game One then there would have been different leaders in Playoff Game One, and so on. Similarly, this analysis will naturally undervalue heavy bidding on big favorites because it only counts the scoring from the opening round games. Dropping a heavy investment into Huayna Capac or Justinian is all about getting value from more than a single game and it can absolutely be worthwhile to overbid with the expectation of paying for future playoff performances. So with those caveats in mind, let's take a look at the opening round matches starting with Game One:

The first game of the season turned out to be an outlier result in the alternate histories, with Slashin's minimal bid on Hammurabi hitting the jackpot and translating into 6 fantasy points thanks to a first place finish and a kill. This was one of the best values of the whole fantasy contest even if the alternate histories showed it to be a rather unlikely outcome. Gandhi also bizarrely finished in second place to score 3 points for Henrik, an astonishing result since Gandhi has no runner up finishes and only one kill across all 20 of the alternate histories! This appeared to be an inefficient spending of 45 gold that wound up getting lucky in the actual game. The fantasy contestants who were robbed in this game were sunhuge and especially cutHrAXys as Shaka had the most first place and second place finishes only to score neither in the Livestream game (though he did pick up 2 points from kills and another 2 points for conveinently dying first in the Wildcard game). With an expected result from Shaka here, cutHrAXys might have been better positioned to challenge Amicalola for the fantasy title. The other noteworthy bid was 45 gold for Stalin from nabaxo which was not a terrible expenditure (and delivered 2 points for First to Die status) but far from the best choice either based on alternate history results.

Game Two had a pair of heavy fantasy investments into reigning champion Mansa Musa and Suryavarman by jmie6 and Henrik respectively, along with Slashin' holding four different leaders widely perceived as weaklings that he grabbed at cheap prices near the end of the auction. The initial read from watching the game on Livestream was that Henrik had been utterly robbed, as a dominant Suryavarman was somehow beaten to space by a shocking win from Hannibal who also scored 3 kills in the process to deliver Slashin' a whopping 8 fantasy points. The alternate histories told a very different story, however: this was a monstrous outlier of a game that in no way resembled how the map was "supposed" to play out. The true juggernaut was actually Suleiman who had 14 (!) victories in 20 games and the highest alternate histories score of the whole opening round at 111 points. It's truly wild that Amicalola scored zero fantasy points from Suleiman and still ran away with the fantasy victory anyway - he really should have done even better in the contest! Suryavarman and Mansa Musa turned out to be roughly equivalent in terms of fantasy scoring potential, both of them lagging far behind Suleiman. Jmie6 was a bit unlucky here but Mansa does tend to be a feast-or-famine leader. As for Slashin's four leaders, while they all had poor fantasy scoring numbers in the alternate histories, every leader on the team does open up the potential for a low-odds lottery result like what Hannibal experienced here. Given the cheap price for these leaders, it was probably worth grabbing them and hoping for a miracle.

The alternate histories for Game Three resulted in nearly equivalent scoring between Justinian and De Gaulle. Given that Justinian commanded a king's ransom from cutHrAXys while De Gaulle was purchased by Amicalola for mere pennies on the dollar, however, this turned into a massive value purchase for Amicalola which was further rewarded when the French leader landed one of his commonplace runner up finishes. cutHrAXys was unlucky to take only a single point from Justinian and this was another place where better RNG in the Livestream game might have closed the gap between our top two fantasy competitors. Amongst the other leaders, the big winner was nabaxo who scored a hefty 8 points from Alexander at a medium price of 41 gold. This was about the best possible result from Alex who was inconsistent in the alternate histories and often fared far worse. Zara Yaqob was the other leader purchased at a middling cost and he conversely did much better in some of the alternate histories, with Slashin' paying a more or less fair price for the Ethiopian leader based on his alternate histories performance. Victoria and Ragnar were both seen as being poor leaders in the auction and that was verified in the Livestream game and again in the alternate histories where they were nothing but First to Die bait. For 6-7 gold this was absolutely worth taking a flier on which Slashin' managed to hit for the 2 fantasy points.

Game Four was the sunhuge game as he grabbed four of the seven leaders taking part in this match, with Myth further taking another pair. Sunhuge mostly bid on leaders who were seen as weaker performers for his fantasy team, with the one exception of Gilgamesh who required a weighty 70 gold price tag. Gilgamesh was the giant favorite in the community going into the game only to fail to secure either of the top two spots and score only a single point, a development which unfortunately sank sunhuge's fantasy squad. This was a bit unlucky since Gilgamesh did have roughly 50% odds for a top two finish, however he was overshadowed by the true master of this map in the form of Qin Shi Huang. The Chinese leader won 13 times in the alternate histories for a total of 95 points which was an insanely good value for cutHrAXys at a purchase cost of only 20 gold. But Qin didn't win the actual Game Four either, as the victory was instead snapped up by a surprising Joao to deliver Myth's only fantasy points of the whole competition. The alternate histories showed that this was a plausible outcome if far from the most common result. The other four leaders scored most of their points from being First to Die, including Sitting Bull who scored the entirety of his alternate history points from dying first. That was a smart bid from sunhuge at only 20 gold cost which did not end up hitting as Wang Kon wound up First to Die instead. In perhaps the biggest surprise of all, Mao Zedong was the worst leader to purchase from a fantasy perspective: not good enough to score actual points, too strong to be First to Die. Myth's middle of the road bid of 30 gold was a flop in that regard.

This was another strange game in terms of the Livestream match as Roosevelt failed to exhibit the same strength that he enjoyed in the alternate histories - sorry, Myth! What should have been another excellent value purchase at 20 gold wound up scoring no actual points. The alternate histories indicated that this scenario was pretty wide open for everyone other than Boudica and Asoka, with the other four leaders all having good odds to score some kind of fantasy points. cutHrAXys was rewarded with the big 8 point score from Kublai's victory on Livestream, however statistically Kublai was a worse purchase than Roosevelt, Churchill, and Cyrus given the much lower sums that they required. In retrospect, the worst bid was 37 gold for Asoka from Slashin' who turned out to be nothing but First to Die bait and couldn't even manage that feat in the actual Game Five. The alternate histories showed that Boudica was consistently the weakest leader on this map but at least she was a lot cheaper for Henrik to acquire.

Game Six was one match where the bidding almost perfectly lined up with the outcomes in the alternate histories, with one glaring exception. Amicalola and Slashin' both spent major sums purchasing Darius and Willem respectively, and each of those leaders backed up their investment with nine wins apiece during the map replays. This turned into one of the biggest swings of RNG in the whole fantasy competition, with Darius winning the real Game Six and going on to score 16 points for Amicalola while Willem only produced 2 points for being First to Die. Since Darius and Willem appear to have had virtually identical odds to win, flipping those results could have created an alternate world where Slashin' ran away with the fantasy contest instead. The other five leaders were all effective non-entities who had no real chance to win and didn't produce any fantasy points aside from random kills and First to Dies. They were perfectly fine as cheap options that didn't require much of an investment... except that poor jmie6 dropped 65 gold on Julius Caesar in one of the most expensive bid of the auction. Much of the AI Survivor community was equally fooled by Caesar who was a big favorite to win this game in the picking contest. The Roman leader let everyone down though by playing a completely mediocre game which the alternate histories confirmed was the normal course of events for this scenario. Unfortunately the big purchase of Caesar was a major factor in sinking jmie6's fantasy team over the course of this season.

I think that Game Seven might be the most infamous opening round match from this season of AI Survivor, a completely unexpected Montezuma victory that the alternate histories proved was an absurdly unlikely result. This translated into 6 fantasy points for cutHrAXys (and then another 2 points in the playoffs when Monty was predictably First to Die) when the Aztec leader was the worst-performing individual on the map during the alternate histories. He actually moved into the third place spot for fantasy scoring in the alternate histories specifically because he was so terrible that he loaded up on 28 First to Die points! The actual 800 pound gorilla on this map was Hatshepsut who had the third-highest alternate histories score of the opening round at 93 points. Her elimination without scoring a single point for sunhuge was a cruel twist of fate, oddly matched by Suleiman scoring zero points for Amicalola back in Game Two. Hatty was even more dominant in terms of finish points (if not scoring kills) with 17 victories in the alternate histories only to be denied in the actual Livestream game. Elizabeth was the other strong leader on this map who also scored nothing other than First to Die points for Amicalola.

Meanwhile, there were massive bids on Catherie and Pacal for this game from Myth and Henrik which turned out to be complete duds. Catherine disappointed everyone with only a single victory in the alternate histories while Pacal was unable to muster up even one win despite heavy community backing. These were horrible picks given the giant gold sums that the two leaders required to purchase and yet Henrik managed to capitalize against all odds in the Livestream game with Pacal taking the runner up spot along with claiming 3 kills. The Mayan leader had 6 kills across all 20 alternate history games and somehow he had 3 kills in the real Game Six! Myth was nowhere near as lucky and his 80 gold investment into Catherine yielded exactly zero points which essentially torpedoed any chance of his fantasy game being competitive. This was an utterly bizarre game of a sort that we've rarely seen in AI Survivor.

Game Eight had the highest fantasy bid of the whole auction as nabaxo spent a whopping 100 gold on Huayna Capac. This was a high price to pay but not ridiculously so as the Incan leader performed well in the alternate histories and of course had excellent playoff scoring potential as well seen in past seasons. However, Huayna Capac narrowly missed a runner up finish in the actual Game Eight and then crashed out of the Wildcard game (despite performing the best in that game's alternate histories) to leave nabaxo with zero points in an unlucky turn of events. With the benefit of hindsight, the better investment was Napoleon who scored even better than Huayna Capac at a fraction of the price. Slashin' managed to get 3 points out of Napoleon thanks to a series of kills but could have done even better with a bit more luck. The big winner from this game was Amicalola though as Louis delivered enormous value from his medium-range cost by winning Game Eight followed by both of his playoff games en route to the overall championship. This was an excellent purchase given that Louis won about half of the time in the alternate histories while still being a bit lucky - Louis wasn't going to deliver 17 fantasy points every time that the competition was replayed! The other leaders mostly scored few points and could be purchased cheaply as the fantasy competitors correctly determined their low value ahead of time. The one exception was Charlemagne who was far worse on this map than anyone expected and recieved a medium bid of 36 gold from jmie6. His score of 6 points in the alternate histories was the worst of anyone in the opening round; it was quite lucky for jmie6 to secure 2 points from Charlemagne claiming First to Die status in the Livestream game.

Here's the combined fantasy summary from the alternate histories of all eight opening round games:

I included the cost to purchase, the actual points scored across all of Season Seven, and then the alternate history points scored to get a better sense of the true value of each leader in the opening round games. Again, this analysis necessarily does not include any results from the playoff rounds and therefore undervalues teams of AI favorites like the one put together by jmie6 so please do keep that in mind. Nevertheless, I still think that the exercise is interesting as it helps showcase which fantasy teams experienced good luck and bad luck throughout the opening round of games. The winner of this look at the opening round was again Amicalola although by a smaller margin than his runaway performance in the real competition. Amicalola had the top-scoring leader of the alternate histories in Suleiman along with other big performers in Darius, Louis, and De Gaulle. Elizabeth was amazingly another leader that the alternate histories indicated should have been a major contributor who did practically nothing in the Livestream games while scoring 45 alternate fantasy points. Amicalola did a fantastic job at identifying leaders who were situated in the second tier below the most popular favorites, avoiding the 60-70 gold prices that they required while instead spending multiple times in the 30-50 gold range. If there was an effective Moneyball strategy for this season of AI Survivor, Amicalola appeared to have identified it.

The closest competitors to Amicalola in the real fantasy contest were cutHrAXys and Slashin' who both had equally strong results in the alternate histories. cutHrAXys had the huge, unlucky whiff on Justinian which was counterbalanced by an insane pants-on-head outcome from Montezuma which never should have taken place. cutHrAXys also suffered from poor RNG in scoring only minimal points from Qin instead of a giant opening round payout. Mathematically Qin was one of the most efficient purchases of the whole auction at 4.75 alternate history points per gold spent. And Shaka did reasonably well in the actual Livestream games but also could have scored even better according to the alternate histories. This was probably the closest fantasy team to overtaking Amicalola's juggernaut lineup. As for Slashin', the team with the biggest roster was always going to land some fortunate low-odds results which certainly happened with Hannibal and Hammurabi. This offset the bad luck of Willem getting taken out so quickly and Napoleon greatly underperforming his scoring potential. This analysis of looking only at the opening round and aggregating the total alternate histories scoring of every leader also tends to overvalue a team of weaker AI leaders to some degree which should be kept in mind. Nevertheless, Slashin' had so many lottery tickets in play that slightly better results from his more expensive leaders could have put him right into contention for the top position.

The big stunner from this look at the alternate history opening round games turned out to be sunhuge's fantasy team. Sunhuge's scoring potential was effectively identical to Slashin's team! More than anyone else in the competition, sunhuge was the victim of bad luck throughout the opening round. Hatshepsut was an absolute monster in the alternate histories with an 85% win rate and failed to produce anything for his fantasy team. Similarly, Gilgamesh did far better in the repeats of Game Four and really could have, should have secured more than a single fantasy point. Isabella was also a sneaky-good purchase at a cheap cost from all the way back in Game One that definitely underperformed her average result. Fortunately for sunhuge, an exceedingly unlikely 11 point score from Saladin boosted this roster to the middle of the pack where it was at least within shouting distance of its expected placement. The other roster that performed far better in the alternate histories was Myth's team which finished in last during the real competition. While Myth struck out badly on Catherine (one of the worst values at 0.26 alternate history points per gold), his team had an excellent value purchase in the form of Roosevelt and several other solid drafts in Pericles, Genghis Khan, and Joao. Unfortunately Joao was the only one of these picks to hit in the actual competition which left Myth with a mere 6 points. This roster was never going to be one of the top contenders but it more properly should have finished around 5th or 6th place.

Inevitably some of the fantasy teams had to come up short and this analysis suggests that Henrik, nabaxo, and jmie6 had the weakest rosters. Henrik was the fantasy competitor with the most fortunate RNG of the opening round, as Pacal managed to produce 7 fantasy points despite a poor showing in the alternate histories. Churchill was surprisingly the best pick here of the rest of the group, both in the alternate histories and in the Livestream games, making it to the Championship game in unlikely fashion at a mere 13 gold cost. With Pacal, Suryavarman, and Gandhi all expensive leaders that performed worse than the community expected, this was the fantasy equivalent of an aging team suffering from injuries. Nabaxo's fantasy team was easy to explain: he hitched his wagon to Huayna Capac and the Incan leader (who still did well in the alternate histories) simply didn't deliver. With Stalin also underperforming, it was only Alexander that kept nabaxo's team afloat and the Greek leader simply wasn't enough. Finally, jmie6 payed giant sums for Mansa Musa and Julius Caesar only to see both of them faceplant in the competition. The alternate histories indicate that this wasn't particularly unlucky either as neither of them did especially well in repeats of the map. Charlemagne was also the single least efficient purchase of the whole auction at 0.17 fantasy points per gold in the alternate histories. Instead it was Ramesses carrying the torch for this fantasy team (in what the alternate histories suggested was an outlier outcome) to the tune of a championship game appearance and 8 points. Again, keep in mind that aggregated leader scores like this will inherently undervalue fantasy teams with smaller rosters of expensive favorites since it doesn't include potential playoff scoring. At the same time, the teams with the smallest rosters didn't do particularly well in the actual fantasy contest either; this could be useful to keep in mind for bidding strategies in future seasons.

Conclusion

Here's one final table showing the gold efficiency of the auction purchases based on these alternate histories from the opening round. A surprising number of these were pretty close to 1:1 with one gold in the auction converting into one fantasy point across the 20 map replays. I thought that this was a useful snapshot to see which bids were most effective at scoring fantasy points across the alternate histories relative to the cost that the more popular leaders required. Now this is again a limited analysis even aside from not being able to include playoff scoring: it's inherently easier to be more gold efficient when it comes to the cheaper leaders who didn't cost much in the auction. You would much rather have an efficiency rating of 1.51 on Louis by paying 41 gold for 62 fantasy points as opposed to an efficiency of 2.29 from 7 gold spent translating into 16 points with Ragnar. It's probably more useful just to look at the most desired leaders that cost 40+ gold to see how they stacked up against one another; Louis, Willem, and Elizabeth seemed to score the best here with a lot of the most popular leaders having relatively poor value. Again, adding playoff scoring would make this look a bit more reasonable to the favorites if there were a way to do that.

Overall, this competition was once again a tremendous amount of fun and helped to make Season Seven of AI Survivor so entertaining to run. I'd like to thank our eight participants in the fantasy contest once again and hope that we can do this again whenever the next season of AI Survivor might be. Amicalola has already said that he'll be back to defend his fantasy title against what will hopefully be seven new competitors from our AI Survivor community. Until then, thanks once again for reading and following along!