Civ 4 AI Survivor: Season One Alternate Histories

[Eauxps I. Fourgott]

Next up is Game 5, originally a success story for Lincoln and Augustus Caesar, who conquered the two low peaceweights on the map while Washington did absolutely nothing. Original writeup: https://sullla.com/Civ4/survivor5.html

Original map:


Anonymized "teaser" results:

[Eauxps I. Fourgott]

Game 5




Running the alternate histories for this game quickly revealed that the map had a clear overall narrative, one that was seen in the real game as well: the high peaceweight leaders working together to beat the hapless warmongers. They had the advantage of numbers, 4 to 2, and while Hammurabi didn’t fare so well on this map, Augustus, Lincoln, and Washington all had good starting positions that allowed them to peacefully become strong, then knock out a Kublai and Ragnar who had frequently accomplished little in their early warring. It was rare for one of the warmongers to get a real edge against any of these three, or for both to work together against one of them, so the fact that these guys were a bit less likely to come to each other’s aid than more aggressive leaders didn’t even matter – they could just take their opponents out on their own.

As a result, we often saw games with similar dynamics to the real one, as the map fell to Augustus, Lincoln, and Washington and they then proceeded to a peaceful conclusion. Each high peaceweight leader played his own role as well: Augustus was the main muscle, doing the most fighting and generally performing the best as a result. Lincoln was the group’s biggest techer, often sitting back in his corner and peacefully vying for the spaceship win. Washington was a bit of a mixture, sometimes sitting back peacefully and sometimes going to war, but he didn’t play either role as well and was the most likely to betray the group. Finally, Hammurabi served as the meat shield, delaying Ragnar and especially Kublai and buying the others the time they needed to get out in front while accomplishing little himself.

With that being said, while this was the map’s main dynamic, it was by no means set in stone and the game could spin off in other directions. Indeed, this was the most open map I’ve run alternate histories for so far, as every single leader won at least two games out of the set! While Augustus and Lincoln were generally the big dogs, Washington could occasionally leap in front as well. Ragnar and Kublai occasionally managed to get the upper hand and go on conquering sprees, and when left alone Hammurabi could be a legitimate contender for the win as well. There were clearly better and worse leaders for the map, but it was also one where anything really could happen.

As far as other general trends, this map tended to be pretty light on wars, thanks both to the peaceful leaders’ dominance and the fact that a lot of wars tended to be fought to an elimination. Nearly half the games had single-digit war counters, and the average number of declarations would have been below 10 if not for a wacky outlier game that had 21(!) wars. The Apostolic Palace didn’t seem to have as big of an impact here as it has in some other contests, with few AP dogpiles and not a lot of obvious cases of AP peace prolonging leaders’ demises. This wasn’t the fastest map, as nobody was really great at teching, so we saw a lot of finishes in the Turns 320-360 range, with only a single pre-300 ending. Finally, the space race was by far the dominant victory condition on the map; there were a handful of Domination endings, but these leaders almost never ran the slider, a Cultural victory never took place, and the UN was never an important factor either. This game was dominated by a group that liked to just sit back and tech.

Now for a closer look at the individual leaders:

Augustus Caesar of Rome
Wars declared: 49
Wars declared upon: 27
Survival percentage: 85%
Finishes: 7 firsts, 5 seconds (45 points)
Kills: 28
Overall score: 73 points

Augustus was the map’s clear best performer, winning a third of the games, advancing to the playoffs over half the time, scoring nearly half the map’s kills, and only dying three times in twenty games. It turned out that Augustus had a rather favorable start here, and one that in particular synergized perfectly with his strengths. There was a decent amount of land, just enough so that he could reliably use his Imperialistic trait to expand to the largest size in most games; only a single hostile neighbor in Ragnar, so he almost never had to deal with backstabs; a friendly Hammurabi frequently putting cultural pressure on Ragnar to keep him weak; and thus many early wars where Augustus could use his Praetorians to full effect and lay the smackdown on the Vikings. Augustus and Ragnar’s conflict from the real game was repeated in virtually every game (with Caesar himself frequently starting things off), and it almost always went in Augustus’s favor. Even without the Viking lands, and especially with them, Augustus was able to run a competitive economy and vie for the score lead, being one of the map’s top competitors in many games. He’d also often follow up a successful conquest of Ragnar by going and roughing up Kublai as well, cementing the high peaceweights’ success while himself pulling further ahead.

Augustus’s wins would usually begin with him taking most or all of the Viking lands to get out in front, and going from there. Most involved a conquest of Kublai as well to secure his lead. There was also one game where he reversed his usual strategy, going north and conquering Kublai first before dealing with Ragnar afterwards. Augustus would usually follow these successes up with a peaceful space race which he’d win with his extra territory, but there were also two games where he instead attacked Washington and won by Domination. His less successful outings came when he wasn’t able to profit off of a conquest of Ragnar in the same way, and got outscaled by someone else; he’d often still be in the mix and still got five more second-place finishes this way, but he couldn’t quite compete for the win without extra land. Augustus only suffered three eliminations in the entire set of alternate histories: Game 4 where he suffered an unlikely mid-game dogpile by Kublai and Washington, and games 16 and 19, where he failed to stop an unusually strong Ragnar, who later turned and devoured Augustus as part of the middle of his snowball to victory. These were the outlier results, though, and normally Augustus was strong enough to take on any foe and stalemate at worst.

Overall, this was a pretty impressive performance. Augustus’s willingness and ability to go to war really paid off here, allowing him to get the land necessary to contend where he might otherwise have faltered, and he scored at least one point in all but two of the alternate games. More broadly, this performance goes to support the idea that Augustus is a mid-tier leader who is capable of taking advantage of a favorable start. [note: I wrote this before his Season 8 run, which further bolstered this idea, and my support of him in the playoffs in particular was informed partially by this set.] While he didn’t have the stiffest competition here and will no doubt look much worse in his playoff game’s alternate histories, he still did a pretty good job of capitalizing on his advantages, and showed that he has his own ways to pull ahead.

Lincoln of America
Wars declared: 19
Wars declared upon: 28
Survival percentage: 75%
Finishes: 4 firsts, 8 seconds (36 points)
Kills: 2
Overall score: 38 points

Lincoln was the winner of the real Game Five, and the alternate histories showed this to not be an unreasonable result, but they also clearly indicated that he was much more likely to come in second. Fortunately for my analysis, it was very easy to see what was different in the real game. Lincoln was able to score two kills on his road to victory in the real game… and then went on to score two kills in all twenty alternate history games combined. In other words, his real performance was unusually aggressive and he normally tended to play the role of the pacifist instead, declaring less than one war per game on average despite his excellent survival rate. Lincoln did in fact win both games in which he scored a kill – getting that bit of extra territory usually gave him the edge he needed to come out on top – but he simply wasn’t going and taking territory most of the time. Only three alternate histories saw him take a significant amount of territory; most of the time he just didn’t want to fight.

Instead, Lincoln teched, and it turned out he was in the best position on the map for this. I’m not sure why exactly he ended up doing such a good job of it – he doesn’t have any particularly economic traits and there’s nothing special about his land to look at – but once he’d established his cities and blanketed the land with cottages, he was able to consistently produce a strong research rate that rivaled Augustus and Washington’s, even when they’d expanded off of conquests. Quite a few of these games saw those three leaders left at the end, and Lincoln still sitting on his original six or seven cities, but still right in the thick of the space race! While most of the time this wasn’t quite enough, he did manage to win a Game 1 where his only military activity was to capture a single barb city, and most impressively won Game 8 without firing a single shot – he was involved in zero wars that game! (Lincoln also successfully avoided war entirely in two other games, both of which saw him take second place.) He also was only barely beaten to a couple more victories by a handful of turns. Most of the time, though, he’d make a game effort but come up a bit short. A lot of those cases saw him still have grown strong enough through peace that he’d still come in second place, and then there were a handful of cases where he was well and truly outgrown and consigned to the wildcard. There was only a single game where the high peaceweights came out on top and Lincoln didn’t live to tell about it, an unusual match where Augustus chose to attack and conquer him in the lategame. Otherwise, he clashed with Washington in a couple of indecisive conflicts (and one unusual one where he got the upper hand), and aside from that was great friends with his compatriots.

As for Lincoln’s other deaths, they came in the handful of games where the low peaceweights succeeded – Lincoln would inevitably be caught up in their snowballs. His fighting across these alternate histories was merely average, sometimes holding off foes impressively, sometimes folding, but rarely getting a decisive edge unless he had the tech lead. Overall, this seemed to be a good look at Lincoln getting to play according to his desired gameplan: sit back, tech, and wait for the competition to implode. That’s usually not a very reliable way to play the game, but he got a map here where it was actually a pretty successful strategy.

Washington of America
Wars declared: 38
Wars declared upon: 31
Survival percentage: 55%
Finishes: 2 firsts, 5 seconds (20 points)
Kills: 14
Overall score: 34 points

Much like in the real game, Washington was the tertiary high peaceweight leader in importance, although it was for different reasons in the alternate histories; his completely peaceful game was never repeated, as he declared at least one war in all twenty games while getting attacked in 80% of them. Instead, Washington just struggled to be as effective as his cohorts, generally not teching as effectively and not as well-positioned for conquest. The clear takeaway from this group of games was that, despite a spacious yet sheltered starting position that left him in the lead after many landgrab phases, Washington was ultimately not able to compete unless he was able to successfully conquer some land. All seven of his top-two finishes saw him manage this at some point, usually working with Augustus and Lincoln and edging one of them out for second place, although there was one unusual game where he instead worked with Kublai to dismantle Augustus and came in second that way. Washington’s wins both came from games in this mold as well; in one, he simply had better tech than normal and narrowly won the space race, while the other was an unusual situation where he and Augustus fought after finishing off the low peaceweights, and Washington came out on top. He was also a bit unlucky to miss out on a third win, when he was first to finish the tech tree but had a ship part sabotaged at the worst possible moment to allow Augustus to edge him out, but those three games were it as far as Washington’s bids for the win.

The other notable feature of Washington’s runs in these games is that he acted as more of a Benedict Arnold, by far the most likely of the high peaceweights to turn and attack another. Most often, this was Hammurabi, as those two leaders founded the first two religions in most games and would be rivals; whenever this happened it was very bad news for Hammur, and quite a few of Washington’s successful games started with the annexation of Babylon. Washington also got tangled up with Lincoln a few times, which was usually inconclusive, and with Augustus more often, which tended to go worse for him. A couple of games also saw Augustus attack Washington on his own, to be fair. Washington’s eliminations generally came either from the low peaceweights taking over or from his getting on the bad side of Augustus; he had a real problem in that regard that Lincoln and Hammurabi did not. Overall, Washington was clearly the second-best fighter in the high peaceweight cohort, nearly tripling Lincoln and Hammurabi’s kill total and scoring as many kills as the two warmongers combined. He had decent odds to take home a top-two spot, but also struggled to get it done a lot. Ultimately, that ends up making this a case of “damned with faint praise”, as given his privileged diplomatic situation and generous starting location, it certainly seems like Washington should have been able to accomplish more than he did. While this performance wasn’t an embarassment, it does overall support the idea that Washington is not an effective leader for AI Survivor purposes.

Kublai Khan of Mongolia
Wars declared: 54
Wars declared upon: 30
Survival percentage: 30%
Finishes: 2 firsts, 2 seconds (14 points)
Kills: 8
Overall score: 22 points

The last three leaders all were about the same in terms of effectiveness: usually at the bottom of the totem pole, able to occasionally come out on top when things broke their way, but otherwise getting trod on by the others. Kublai ended up with the best score of this trio thanks to doing very slightly better (8 points compared to 1 for each of the other two) when he didn’t come out on top, but it was definitely a rough outing for him. Kublai faced an uphill battle on this map as a low peaceweight warmonger surrounded by high peaceweight leaders – and a generally-useless ally in Ragnar. (At least he didn’t have to worry about Ragnar turning tail; the only time the Vikings attacked him in this entire set was at the end of a successful game when they were the only two civs left.) It was pretty much a given in each game that he’d end up falling behind in tech to at least one of the peaceniks, making his long-term position untenable and leading to his ouster sooner or later.

Still, I’d have expected better. Kublai did have a fairly spacious starting position that allowed him to get a lot of strong landgrabs, and his only hostile neighbor was the usually-small Hammurabi. He certainly tried to capitalize, with a lot of war declarations against Babylon, usually at an early date, and less often some midgame war declarations against the pacifistic Lincoln. However, where he really seemed to fail was in these wars: he almost never was totally successful, as his targets would hold him off time and time again despite inferior numbers, until Kublai either gave up, saw Babyon snatched up by Washington, or was dogpiled by an Augustus or Washington who had already dealt with Ragnar. After that, he’d be too far behind and eventually knocked out. I want to emphasize here how bad Kublai was at conquering his neighbors: one of these games had him attack a Hammurabi who only had four cities thanks to an early Ragnar attack, but completely fail to take any territory. Two others had him attack, in the midgame with military tech parity, a Lincoln who had half his cities (10 to 5 in one case and 12 to 6 in the other), only to eventually LOSE cities to him! This really felt like Kublai should have been more successful, at least doing a decent job of knocking out Hammurabi to establish a stronger position, but he only managed to kill Hammur twice in the 16-game “normal” set. His real Game 5 performance was in this mold, as while ultimately successful against Babylon, he took forever in his campaign and so was hopelessly behind by the end of it regardless.

As for the flip side – Kublai’s successes – he did manage two wins in the end. One of these saw him succeed against Hammurabi for once, then work with Washington to take down Augustus and assume a dominant position. The second was a proper snowball, conquering Hammurabi and then Lincoln and (with Augustus’s help) Washington en route to a Domination win. Kublai also finished in second place twice, both times to Ragnar in games where everybody else was dead (he notably was distinctly behind Ragnar in both of these instances), and otherwise got two lucky survivals in games where he managed to stay out of the way of the high peaceweights. As for the overall implications of this run, I’ve long been skeptical of Kublai’s true skill level, and after I’d been warming up to him more recently, this series of games renewed that skepticism. He has a habit of drawing friendly fields with buddies that will cover for him and not call him out on his weakness, and here in a less-friendly environment that was nevertheless not terrible, he didn’t make himself look good at all.

Hammurabi of Babylon
Wars declared: 9
Wars declared upon: 48
Survival percentage: 35%
Finishes: 3 firsts, 0 seconds (15 points)
Kills: 3
Overall score: 18 points

Hammurabi drew a tough starting position in the center of the map for this game. Not only did it leave him bordered by literally every other leader (including being Kublai’s only next-door enemy), but it also left him cramped and without much room to expand. Most games saw him out of space at just four or five cities, and while he might could have squeezed out one or two more with a bit more hustle, overall it didn’t feel like he was doing a particularly bad job of expansion. His surroundings just filled up quickly, and the least cramped part of the map (to his southeast) also had a big patch of desert that served as a natural border anyway. While he had good early culture (thanks especially to usually founding one of the opening religions), he still would tend to be near the bottom of the scoreboard after the landgrab.

Then the real problems began. Hammur was constantly beseiged by Kublai and Ragnar in this game, usually being attacked by one if not both before turn 75. All things considered, he did a good job of holding them off, but with his small size he was never able to properly get the upper hand on them, and his best-case scenario was generally holding on long enough to get peace or an ally jumping into the conflict. His allies weren’t the most reliable in this regard, however; sometimes he held on for ages, only to continually receive no support and eventually fall to attrition. While he did manage to avoid wars with Augustus or Lincoln (I don’t think he fought either one a single time in these games), he was also backstabbed with some frequency by religious rival Washington, which never ended well for him. (In fact, he died more times to Washington than to Ragnar and Kublai combined!) Even in the relatively rare circumstances where he made it out of these early wars alive, he’d generally be set too far back by his small size and warring and have no hope of competing for the win.

To Hammurabi’s credit, he did manage to secure three wins in these games. Two of them came from unusual games where he managed to avoid getting attacked early at all (save an irrelevant early war from Washington), then peacefully built up a tech lead in games filled with inconclusive warring and come out on top in the end. The third came in an otherwise-normal game where he managed to capture a pair of barb cities in the west; this allowed him to dominate the ex-Viking territory when Augustus conquered Ragnar, pulling ahead from there. If Hammur could overcome the inherent problems with his starting position, then he was able to perform well in this rather weak field of leaders. He never finished in second place, though – if he was strong enough to contend, he was strong enough to win. He also did very little aggressive warfare in this game, with only a single kill off an AP dogpile in his non-winning games. His very low wars declared and kill counters speak to both his unaggressive nature and to the tough start he was stuck with that usually saw him attacked first. Overall, I think these games offered little to evaluate Hammurabi’s broader merits as a leader, and mostly spoke to his difficult starting position. He had enough economic heft to pull ahead a few times but otherwise was generally set back through no particular fault of his own. Still, he was undeniably impotant on this map, helping to hold back Ragnar and Kublai long enough for his pals to clean up and come out on top. Sometimes the rewards don’t match up with the work performed, and that was certainly the case here.

Ragnar of the Vikings
Wars declared: 41
Wars declared upon: 46
Survival percentage: 15%
Finishes: 2 firsts, 0 seconds (10 points)
Kills: 6
Overall score: 16 points

Ragnar was generally ineffective on this map, although really he had three distinct classes of games on this map. By far the most common scenario, taking place in about 60-70% of matches, took place when Hammurabi sent his first settler straight towards Ragnar, then founded one of the opening religions. The resulting Holy City culture squeezed Ragnar badly and he wasn’t able to cope with it; these games generally saw him pathetically weak from an early date, never able to accomplish anything. He tried attacking Hammurabi sometimes, but this never went anywhere; he once even lost two cities to a two-city Babylonian empire! He also frequently fought with Augustus, with Rome coming to him when he didn’t attack them, and this never ended well. Augustus had a stronger empire, he had Praetorians, and thus Ragnar was screwed. These games were reponsible for Ragnar having the map’s highest First to Die rate, and Augustus’s successes tended to be fueled off of successful early conquests of the Viking territory. Washington and Lincoln would join the attack at times as well, and that certainly didn’t help his case. In general, Ragnar’s performance in these games was horrible.

A less common variant came when Ragnar wasn’t so squeezed by Holy City culture, whether because another civ founded the opening religions or Hammurabi sent his first settler somewhere else. Ragnar was a bit more credible in these games, expanding decently and being competitive in score and GNP early on, but they still weren’t good performances on the whole. He still fought with Augustus most of the time, and would generally lose or stalemate; his best outcome came once when he conquered two Roman cities, then stagnated (helped, admittedly, by an inopportune Kublai backstab, the only one to happen in these games) and was eventually mostly conquered. Ragnar’s performance in the real game was one of this category; he stalemated in early wars with Augustus, only getting conquered in their third conflict, but still was ultimately doomed. Ragnar wasn’t a total joke under these circumstances, but he still performed poorly and bolstered his reputation as a bad leader for AI Survivor.

The most rare class of performance for Ragnar, showing up in just two games, came when he himself founded one of the game’s first religions, thereby putting Holy City culture on Hammurabi. Ragnar was stronger than usual in those games, and he was able to successfully snowball by first gaining a quick contest of Hammurabi, shrugging off early attacks by Augustus and in one case Washington, then working with Kublai to run over the rest of the map. Ragnar was clearly the better partner in these games and cashed in with a pair of Domination wins as a result. This was how Sullla predicted the original Game 5 to go, but the rest of the alternate histories show that it was very much the exception rather than the rule. Outside of those two games, he survived once when Hammurabi let him off the hook with two cities left, and scored a single kill in a zany result where he, himself crippled by an earlier war with Augustus, marched across the map to finish off a similarly crippled Washington, and that was it. Overall, while Ragnar’s been one of the least successful leaders in the real competition, these results show that he’s probably about the same as most of the other warmongers: a terrifying sight when he gets ahead, and usually comically ineffectual when he can’t. This map turned out to be one where he usually couldn’t.



Conclusion
The alternate histories showed that we’d seen typical dynamics in the real Game 5, although the specific results weren’t the most likely. Augustus and Lincoln were indeed the top two leaders on the map, although it was usually Augustus and not Lincoln winning, Washington was their less successful ally, and the other three were generally slated to be eliminated. It’s even easy to see why the real game diverged: Hammurabi didn’t squeeze Ragnar with an early Holy City, allowing him to hold out for longer and avoid being first to die, while Lincoln was considerably more aggressive than normal, allowing him to jump into the top spot.

One thing this game didn’t really do was provide a lot in the way of new insights into these leaders. Perhaps because it was a pretty straightforward map, or perhaps because the leaders involved are mostly in the lower tiers of competition, I felt that this mostly just reinforced existing impressions without providing new angles on them. Still, it continued to be an enjoyable endeavor to watch and record these. Augustus’s not-really-expected excellence and the double wins eventually recorded by each leader especially helped to keep them sufficiently interesting. Next time is another game where the high peaceweights had the advantage of numbers; we’ll see if it too proves largely straightforward, or if the different map, extra low peaceweight, or increased economic aptitude make it spin in more interesting directions. Until then!

[Amicalola]

Very interesting! I had thought Augustus' position did not look that strong, so it is surprising to me that he performed so well here. Maybe he really is a decent AI who underperformed in the first few years...

[williams482]

I was also underwhelmed by the land available to Augustus here, and surprised that he put up such a dominant showing. I guess having praetorians and being able to run over an inept Ragnar in most games goes a long ways in a weak field.

Hammurabi did surprisingly well for himself from that central starting position.

[Eauxps I. Fourgott]

Next up, Game 6. Original version was a Pericles/Elizabeth game. Peaceweight balance would seem to favor that being a repeated outcome, but are looks deceiving?

Original writeup: https://sullla.com/Civ4/survivor6.html

Map:


Anonymized results:


Full writeup coming Monday.

[Amicalola]

Next up, Game 6. Original version was a Pericles/Elizabeth game. Peaceweight balance would seem to favor that being a repeated outcome, but are looks deceiving?

Original writeup: https://sullla.com/Civ4/survivor6.html

Map:


Anonymized results:


Full writeup coming Monday.

Without looking too closely, I would guess:
1. Elizabeth
2. Pericles
3. Pacal
4. Frederick
5. Darius
6. Julius Caesar
7. De Gaulle

Thanks for running these, as a side note. They are always interesting.

[williams482]

I'm guessing that Freddy was not the guy who snuck in 12 kills but died a lot. That screams JC to me. Other than that though, not a lot of grounds for disagreement. Pericles and Elizabeth both have a ton of space to expand into uncontested, and the 3rd place spot with the relatively low survival rate sure scans as center-start Pacal.

I'm going with:

1. Elizabeth
2. Pericles
3. Pacal
4. Julius Ceasar
5. Darius
6. Frederick
7. De Gaulle

[Eauxps I. Fourgott]

Thanks for the support, Amica! Always good to have some positive feedback.

And Williams correctly guessed the final order!

Game 6





Running these alternate histories quickly made it apparent that the dynamics of the real Game 6 were largely representative of the map’s broader trends. This was a somewhat unusual map mainly characterized by two factors: a rather large continent size with lots of land, and a group of AIs dominated by peaceful builder personalities, with Julius Caesar the only true warmonger of the bunch. As a result of these two factors working together, while the map did see combat, decisive wars were comparitively rare. Results with four or more of the leaders surviving were commonplace, and I even got one playthrough in which none of the seven leaders died! (A first for me.) When tallied up, this game actually had less total kills than Game 5’s – and that one not only had long stretches of peace, but also one less AI in the field! Conquest was simply not a reliable path to victory on this map, and as a result I think the range of possibilities was a bit more limited. There were certainly some twists and turns along the way but this might be the least interesting set I’ve run so far.

When it came to individual leaders’ chances of success, the single biggest distinction that emerged was which hemisphere of the map the competitors started in. The southern hemisphere had much more available land, while the coast pressed in much more tightly on the north. Now, a lot of this southern land was tundra and ice – but it was still land, and still worth settling for the Deity AIs. As a result, these three leaders had a much higher performance ceiling than the four northern leaders, and went on to take all but one of the victories. They were just too strong; it was nearly impossible for one of the other leaders to come out in front. (It also helped that all three of these leaders had a distinctly useful trait – Creative to help flesh out territory quickly, or Financial to pull ahead economically.)

Of those three leaders, then, Elizabeth and Pericles emerged as the power duo of the map just like in the real game. Liz was the unquestioned economic leader in most games, resulting in a dominant performance, while Pericles was able to consistently expand out to one of the map’s biggest empires and make himself a contender that way, and the two combined for 70% of the map’s top-two finishes. Pacal had a much tougher time than these two thanks to his odd-one-out peaceweight and central start, but in the games where he could get going, he was a serious threat in his own right. As for the northern leaders, Caesar came out looking the best, with moderately successful warring giving him a decent score, while Darius was too cramped to accomplish much, Frederick was his usual ineffective self, and De Gaulle looked just as bad as, if not worse then, he did n the real game.

Spaceship was the runaway winner for victory type on this map. Despite Liz and Pericles being dominant, neither one went for culture more than a couple of times, this seemingly being a field that split the cultural prizes too evenly, so we only saw two wins by that victory type. The map was too big for more than two Domination victories to surface, and the UN was only important a couple of times. On the other hand, I did see quite a lot of wars declared in most games; the AP had a significant presence in this game, often forcing peace or starting unnecessary dogpiles, which helped tick that counter up. There were a few games where the fighting was decisive and I saw only 6-8 wars despite half the field dying, but more often there would be 15+ wars declared. A lot of those wars didn’t actually do that much, mind, and it could often feel like I was just watching squabbles more than real conflicts of note.

Now for a look at the individual leaders:

Elizabeth of England

Wars declared: 38
Wars declared upon: 36
Survival percentage: 85%
Finishes: 12 firsts, 4 seconds (68 points)
Kills: 16
Overall score: 84 points

Elizabeth was the premiere AI on this map, winning over half the games, failing to advance only four times, and scoring the most kills despite her generally peaceful personality. Liz had a dream setup for this map, with a fairly spacious start from which she could expand to a decent size even in the worst scenarios, and facing zero serious military threats in the early game. She leveraged this well and would basically always become the game’s top techer; sometimes others could keep up for a while, but it was rare for her to not pull ahead sooner or later, and then she’d keep going and going until reaching runaway status. Elizabeth’s shadow loomed over the rest of the map, and for somebody else to win, they’d usually need to either put on an exceptional performance or else figure out how to get rid of her.

As mentioned, a big part of Liz’s success was that she was never really threatened militarily in the early game. The only rival she fought early with any sort of consistency was Darius; they fought semi-frequently, and Liz was always the stronger party in these conflicts. On most of these occassions she’d conquer him to expand her empire a little bit (only one of Darius’s deaths did not come from Elizabeth), but in any case she never came out of those fights weaker. Then later in the game, for the most part, by the time she’d get attacked she’d be strong enough to withstand the assault, if not get the better of her attackers. Liz was good at teching Rifling in a timely manner in these games, and with her tech pace, any would-be aggressors would have to work just to fight with her on an even footing – getting a military tech edge against her was pretty much out of the question. Liz was moderately aggressive in these games, declaring the third-most wars, and while she’d sometimes sign peace early she was also fairly good at pursuing her conquests to completion. Plenty of her mid-game wars saw her advance from a winning position to a runaway one.

Even in the games she didn’t win, Liz was usually still an important player and could take second place easily, as we saw in the real game. She only suffered three deaths in the alternate histories; all three came in the late 200’s and required major efforts by multiple strong civs to finally bring her down. (She also ended one more game barely clinging to life, and that was an outlier result where an unusually strong Pacal was able to grind her down solo.) In most of the games, though, she was clearly the strongest overall leader. Interestingly, she almost never pursued a Cultural victory in this set of games despite being prone to it in other seasons; clearly it just wasn’t the right field of leaders for that. She only picked up a single one of those victories, and otherwise would simply keep teching and growing her lead until she reached space (even her one Domination win came just four turns before her spaceship would have arrived). Overall, this game did a great job of showing what Liz can do when she’s actually given a favorable situation. Going into Season 8, she was clearly more than a bit unlucky to have zero wins in her real AI Survivor career.

Pericles of Greece

Wars declared: 33
Wars declared upon: 16
Survival percentage: 90%
Finishes: 4 firsts, 8 seconds (36 points)
Kills: 11
Overall score: 47 points

Pericles’ win in the real Game 6 turned out to be a fairly unlikely outcome, but he was clearly the second-strongest AI on this particular map and deserved his ticket to the playoffs. Pericles occupied the eastern mirror-image of Liz’s strong starting position; while his actual capital was stuck on the coast, he still had easy access to a lot of land in the southeast and was able to use it to grow to a large size in most games. I think the Creative trait really served him well here, helping solidify his claim to these regions and more easily poach captured cities on his borders. Pericles was often in the running for a top-two spot solely based on his peaceful expansion, and further benefited by being nice and out of the way, which combined with his moderate peaceweight to keep him largely safe from attack except by the hapless De Gaulle. Indeed, oftentimes Pericles didn’t do a whole lot in the rest of the game to secure a top-two spot, but was already in a good enough position to compete for it anyway. His quantity of cities was enough to make up for the quality of the Financial leaders’ research and be competitive in tech in most games.

When Pericles did fight, it was most often with De Gaulle, a conflict that, uh, never favored De Gaulle. Pericles was often able to take at least a couple of French cities; even when he didn’t directly fight De Gaulle, his culture could flip the closest cities once they got conquered by somebody else. Pericles wasn’t too consistent in fighting anybody else; his low declaration counter really tells the story here, as he averaged less than 2 per game despite frequently being strong and rarely being under attack himself. He just wasn’t that interested in going on the warpath most of the time. One thing that was quite uncommon was the real game’s result of him conquering Julius Caesar; this conflict was only repeated a couple of times in the alternate histories, and for the most part Greece and Rome were able to coexist peacefully. When Pericles did fight, his results were pretty average. He’d often have a power edge based simply on his bigger and better empire, and gradually leverage that to win the wars. He didn’t do anything particularly impressive or embarassing in his wars, although there definitely were some games where he delayed military tech longer than he should have. For the most part, though, it was pretty unremarkable warfare.

Pericles’ path to victory in these games was to somehow grow stronger than normal and get an edge over Liz. Once this was through an early conquest of France, once through simple very-wide expansion, and once through a lategame conquest of Caesar after having been stronger-than-normal through the rest of the game. Pericles simply outraced Liz to space in one of these three games, but in the other two he ended up going to war with her and fighting for the top spot – these were two of the most exciting games in the set as he was able to work with other leaders to gradually wear the titan down before getting the upper hand and conquering her completely, after which he was the clear winner. Pericles also picked up a fourth win in these alternate histories, pulling out a Diplomatic victory while clearly in second place, and then cleaned up eight additional second-place finishes from games where he did a solid job, but just wasn’t as strong as the leader (usually Liz). Then there were some games where he just didn’t do quite as well, and played an average peaceful game that saw him simply get outscaled. Pericles only died twice in this entire set, both times at the hand of a runaway game winner (Pacal in one case and Liz in the other), though it should be noted he was a bit lucky here as he was nearly eliminated and saved by the bell in one or two other matches. Still, early-game danger for him was nonexistent here.

We clearly saw an unusually strong Pericles in the real game, explained by his rarely-repeated conquest of Caesar that let him pull out in front. While he was still an important player most of the time, he usually played second fiddle to Liz and only rarely eclipsed her. Overall, this wasn’t an especially impressive set of games from Pericles, nothing that revealed any sort of hidden strength from him. He simply was placed in a very strong starting position and was competent enough to get decent results from it. In a broader picture, Pericles does seem to be a lesser leader as he doesn’t have any particular strength, but he at least isn’t embarrassingly incompetent. I’d say this map comes close to representing a best-case scenario for him.

Pacal II of the Maya

Wars declared: 23
Wars declared upon: 83
Survival percentage: 40%
Finishes: 3 firsts, 2 seconds (19 points)
Kills: 8
Overall score: 27 points

Pacal was the final member of the southern trio, and his start had high potential between the abundant available land and his great economic traits. However, in direct contrast to Liz and Pericles, Pacal’s position was anything but sheltered. Instead, he got stuck in a tricky central position, and furthermore was one of the only low peaceweight leaders in a high peaceweight field. His peaceweight is normally a significant asset to Pacal in these games, but here it was a liability instead, and the overall result was the highest getting-invaded rate of anybody in the game. Pacal would usually face invasions from an early date, often from multiple foes, and would resultingly either get crippled or eliminated early on, or else simply set back far enough that he had no hope of competing with a peaceful Liz to his west. He was most assuredly not playing the same game as her.

Pacal usually started his games well, expanding far and wide and often being one of the score leaders early on; the problems only started once he got attacked. The most usual adversaries for Pacal were Frederick and Julius Caesar; Darius invaded semi-often and disastrous wars with Liz and Pericles sometimes happened as well, with De Gaulle being Pacal’s only true ally on this map – and De Gaulle wasn’t able to be much of an ally. When Pacal just fought Fred, he was usually able to perform fairly well and even get the upper hand as often as not, but most of the time somebody else would dogpile him and it would all fall apart. Wars with Caesar were less successful since Pacal had no counter to Praetorians. While Pacal was rarely eliminated all that swiftly (he was only First to Die four times, one of those times quite late in a weirder game), the peaceful bent of this group tended to mean that even in the better scenarios, he’d be too far behind post-war and be permanently set back as a result. Other times he’d lose one or two cities at a time, slowly getting ground down over a series of wars.

To Pacal’s credit, he proved himself capable of performing well on the rare occasions where he wasn’t crippled by early wars. He was competitive with Liz in tech in games where he kept his empire intact, and was able to outtech her and pull out in front to collect wins on three separate occasions. All three of these games saw him either remain entirely peaceful, or else fight only one opponent at a time without a backstab and thus be allowed to complete a conquest and pull out in front. In addition to his three wins, Pacal suffered several near misses: in one game, he finished his spaceship less than 10 turns behind Liz, despite having spent the early game fighting for his life while she teched in peace; in another, he was competitive with Pericles in a space race until a suicide attack by Caesar distracted him enough for Pericles to pull out in front; and in a third, he was in a tight space race with Caesar, only to risk an all-out attack on his rival and get his butt nuked off as a result. Outside of these games, though, Pacal was a weak and bullied civ. He never scored any second-place finishes outside of the two near misses above, and all of his kills came from these more successful outings. Still, on the whole, this was a pretty impressive outing that I think bolstered Pacal’s image as one of the game’s stronger leaders. He was able to convert a difficult position into the third-best performance on the map, and did a lot better for himself than many of his competitors would have.

Julius Caesar of Rome

Wars declared: 79
Wars declared upon: 20
Survival percentage: 55%
Finishes: 1 first, 3 seconds (11 points)
Kills: 12
Overall score: 23 points

This was a poor map for Caesar – there was virtually no way he could keep up economically with five peaceful AIs all at once – but he did his best anyway, and came out as by far the most effective of the four northern leaders. While a true snowball was almost impossible for him to achieve, he was able to get some isolated military successes to pick up a dozen kills and be in the best position to capitalize if two of the three southern leaders managed to collapse. Caesar was able to play the early game pretty competently, expanding fine if never putting together anything really special. Then when it came time to go to war, he would usually choose either the weak De Gaulle or the dogpile-prone Pacal for his first target, giving himself a winnable battle. (He fought Frederick more rarely and saw less success against him, as most of the time the two seemed to be evenly matched.) This was the source of most of his kills, although it does bear noting that there were quite a few games as well where he’d simply stall in his early wars, failing to accomplish much despite his Praetorians and stagnating as a result. Even when he did get the upper hand and score a conquest, though, he’d usually be last in GNP among the remaining leaders (unless De Gaulle was still hanging around) and fall hopelessly behind the tech leader(s) in short order. Most of the time, there simply wasn’t a path to victory.

Still, Caesar could and did hang around a majority of the time by playing smart. While he picked a whole lot of fights over these games, he was decent at avoiding ones that he couldn’t survive, and was able to usually stay friendly with his rivals despite his lower peaceweight (I’m guessing he rolled higher than normal when the map was generated, although his peaceweight is high for a warmonger regardless). In addition to a bunch of Wildcard performances, this allowed him to collect a trio of second-place finishes, in games where Pacal died early, Liz or Pericles killed the other, and Caesar was left as the next-best option. On the other hand, there were also plenty of games when Caesar found himself up against a stronger foe and was eliminated as a result. Any of Liz, Pericles, Pacal, or even Frederick could get it done – it just needed a game where Caesar wasn’t very strong and one of the others got the chance to fight him one on one (or in a dogpile). Caesar only died before Turn 200 once (and suffered the First to Die fate one additional time in an odd game where nobody died for a long time), but he did suffer eight more eliminations after that point. The specific result from the real game, where Caesar attacked Pericles and was conquered as a result, was repeated once or twice here, but was not in any way a common result.

Caesar did have one triumph in this set, though – one game where everything came together right and he was able to properly snowball. This came in a particularly war-filled game, where he was able to come in at the right moments to dogpile a rival and take a good portion of the spoils – first against De Gaulle, then Frederick, then Elizabeth in a global dogpile – to emerge as one of the top two leaders, in a tight space race against Pacal. When Pacal made a bid for glory and attacked him, Caesar was able to get the upper hand (using plenty of nukes), completely eliminate him, and thus collect his only win. While he never came close to the victory in any other games, the fact that he could win at all from this long of odds is pretty impressive in and of itself. These games suggested that Caesar is at least a decent leader, well above the lowest tiers of competition. Others no doubt could have performed better from this position (not least because their skills would have been more suited to it), but Caesar was able to perform solidly despite the deck being stacked against him.

Darius I of Persia

Wars declared: 32
Wars declared upon: 13
Survival percentage: 65%
Finishes: 0 firsts, 2 seconds (4 points)
Kills: 4
Overall score: 8 points

The bottom three leaders were largely hopeless on this map, and luck played as big of a role as anything in Darius finishing as the best of them. Darius was probably stuck with the worst starting position on the map, crammed up against the northern coast on the narrowest part of the continent, with Elizabeth straight to his south. Look at the map for this game and you’ll see that they were competing for the same land between them and to the east… and Darius had nothing else to vie for, whereas Liz had another half a continent to her south free to settle! To further the disparity, Liz had nice temparate terrain around her capital that could more readily be used, whereas a lot of the land Darius could settle was buried in jungle for a far slower start. Thus, Darius was doomed in this direct comparison – a leader with the same economic hat as him was right next door and guaranteed to get a stronger empire. With Liz next door and as strong as she was, there wasn’t room for Darius to flourish on his own.

Even leaving the direct comparison aside, Darius was awfully cramped. Liz, who was inevitably strong in culture, would push up from the south, while Pacal and Frederick loved to try to expand into the far eastern land, and as a result Darius never had room for more than four or five cities in his main area. Sometimes he could expand southwards, stringing cities down towards the tundra in space between England and Maya, but this wasn’t reliably available and the resulting cities weren’t very strong. Darius also didn’t help himself here, seemingly a very slow expander. This could have been partially because of the jungle around his start – not sure – but there were in any case some particularly telling games where he would simply stop trying to expand at just three or four cities, his inertia that fans are familiar with coming into play once more.

In any case, once the landgrab was over, Darius was inevitably stuck near the bottom of the scoreboard, and had no hope of competing with the southern trio (or usually Julius and/or Frederick) who had more land to work with. He could sometimes keep up all right in tech, but that was about it. The good news about his start was that it was pretty sheltered, and thus he rarely faced invasions; he was able to probe at distant enemies with his own (usually ineffectual) forces at his leisure but wouldn’t have to worry about retaliation. The only real danger to him, responsible for six of his seven deaths, was Elizabeth; sometimes she’d attack him and sometimes he’d boldly try to take her down early, but as a direct result of their available land, Liz was guaranteed to be stronger. The best Darius could hope for was a white peace, and more often Liz would conquer his nation and knock him out. Even when this never happened, though, he was pretty much stuck sliding into the wildcard as happened in the actual game. Darius did collect two second place finishes in the very first two games I ran, both games where he managed to expand his territory some against De Gaulle of all leaders (despite their distance, the two fought quite a bit in these games) and thus gain a competitive position for second place. He was genuinely strong in one, later working with Liz to conquer Pacal, but the other relied on him backdooring second at the end of the game, and in both he was well behind Elizabeth in any case. Overall I think Darius is a capable techer who can be dangerous in the right position, but this was the wrong position. It wasn’t really a map where we could draw a lot of broader conclusions about him; he was just handed a raw deal and struggled to perform accordingly.

Frederick of Germany

Wars declared: 45
Wars declared upon: 38
Survival percentage: 65%
Finishes: 0 firsts, 1 second (2 points)
Kills: 5
Overall score: 7 points

Frederick was completely uncompetitive on this map despite enjoying a favorable peaceweight situation, a decent start, and good conquest opportunities nearby. His games in the alternate histories tended to go along similar lines as in the real game: he expanded to a decent size, performed decently in tech but could never sustainably compete with whoever was leading the particular game, didn’t accomplish a whole lot else, and generally survived to reach the Wildcard without coming particularly close to a top two finish. Fred did fight Pacal and Julius Caesar a fair amount, with results all over the spectrum: sometimes they would get the better of him and conquer him (they were responsible for his demise in all but one of his deaths; the last one was an outlier result where Liz attacked him), sometimes he would get the better of his rival, often with assistance (and thus he picked up five random kills across the set), but most often they would stalemate. What never happened was Fred emerging from these wars as a major power or competitor for the win. He did finish in second place a single time; in this game, he conquered a random barb city in a choice location and was noticeably stronger as a result, then worked with Liz to conquer Pacal in the midgame to pull into second place. Aside from this result, though, Fred was just the tagalong kid for the high peaceweights.

The other significant thing to note about Frederick’s performance here was that he displayed heavy amounts of the classic economic leader’s tendency to ignore Rifling research for ages. Normally, it’s a sign of an extraodinarily bad case of ignoring Rifling when a leader techs Assembly Line first – for Fred, it was pretty much his habit! Teching Industrialism before Rifling also happened on multiple occassions, and I’m guessing this is primarily due to the German unique stuff being at those two techs (though Fred’s Production flavor undoubtedly didn’t help either). In any case, I did see at least one game where he went straight from musketmen to panzers. These delays cost him at least one game, where he should have been a generation ahead of Caesar and Pacal in military tech with infantry, but instead was a generation behind, fielding muskets against rifles, and was killed from a solid position as a result. That might be something to keep in mind for any future games where it looks like he might be relatively strong. Overall, though, this was a very telling performance, one that strongly reinforced the narrative of Frederick as one of AI Survivor’s very worst leaders. He can’t hold out in violent fields and he couldn’t keep up in this peaceful one; it seems that victory for him requires particularly narrow and favorable circumstances, and otherwise he’s destined to be an also-ran.

De Gaulle of France

Wars declared: 23
Wars declared upon: 67
Survival percentage: 15%
Finishes: 0 firsts, 0 seconds (0 points)
Kills: 2
Overall score: 2 points

De Gaulle performed exactly as his reputation would suggest in these games: terribly. It wasn’t just that he had a rough start, although this didn’t help either; his very low peaceweight made him quite unpopular in this field, leading to plenty of dogpiles and the very lopsided war counter, and starting next to Caesar and his Praetorians did him no favors. Still, he wasn’t always dogpiled or attacked by Rome right away. There were plenty of times when he had time to get established, and chose his first conflict himself (almost always against Rome or Greece, of course). And he was no more successful in these instances. He could sometimes wrest away a small border city or force a stalemate, but almost never accomplished more than that. His landgrab and especially cultural development weren’t great, either, resulting in him routinely occupying last place on the scoreboard even before fighting. In any games where De Gaulle survived a decent while, he’d end up horrendously behind in tech, hanging around waiting to be knocked off by somebody at some point or other. More often, though, he was knocked out early, by a solo-operating Caesar or Pericles or by multiple leaders working together. De Gaulle was the heavy favorite for First to Die in these games, and his performance in the real game was completely typical.

De Gaulle had exactly one comparatively successful game, one where he managed to backstab Caesar at the perfect moment to conquer him with Pacal and had the highest power rating on the map coming out of that conquest. Despite then backstabbing an already-at-war Pericles, though, he wasn’t able to conquer more than a couple of extra cities, failing to convert his power lead into anything useful and thus settling into a third-place position behind Pacal and Elizabeth. Aside from that, there were two games where he managed to limp to the end alive, and he scored one more kill in an opportunistic strike against a single-city Frederick, and that was it. Overall these games strongly reinforced the idea of De Gaulle being one of AI Survivor’s most incompetent leaders. His start was bad, but it wasn’t that bad, and he should have been able to make something out of it in some games. Instead, he was just a punchline in this contest, time and time again.

Conclusion

As is becoming the norm for these, the real Game 6 hewed fairly close to the alternate histories while differing in some of the particulars. Pericles and Elizabeth were indeed the two dominant leaders on this map, although it was Liz who normally came out on top. De Gaulle was the usual pick for First to Die, and Darius, Frederick, and Pacal all had mostly-normal games as well. The most unusual aspects were Pericles’ strength, Julius Caesar’s early death, and the Diplomatic ending, all of which were seen in alternate histories, but only a couple of times.

Up next, the game of Russian dominance: Catherine and Stalin’s debuts in Game 7!

[Magic Science]

Thanks for the comprehensive report. I love how deep the Civ IV rabbit hole goes.

While he picked a whole lot of fights over these games, he was decent at avoiding ones that he couldn’t survive, and was able to usually stay friendly with his rivals despite his lower peaceweight (I’m guessing he rolled higher than normal when the map was generated, although his peaceweight is high for a warmonger regardless).

Wow, I never knew that peaceweight was randomized. Is there any way to see its exact value for a leader in a particular game?

[Eauxps I. Fourgott]

Wow, I never knew that peaceweight was randomized. Is there any way to see its exact value for a leader in a particular game?

Not that I know of, but yes, when a new game is started, a random modifier from 0 to 2 is added to each leader's peaceweight.

For these earlier entries in the AH series, that same roll applies to all 20 replays (but is different from the one that applied to the actual game). More recently, I figured out about this and for newer AHs I load the game from the WorldBuilder save each time, which causes fresh PW rolls.