March Madness: Civ4 AI Edition

[T-hawk]

Will we ever get a Cultural win in this competition? The AI definitely will go for culture sometimes.

Not on Deity. The AIs develop too fast, and get an overwhelming military advantage or the tech for space before they'd finish a culture run. The culture-prone leaders keep getting killed by the unitmongers wielding Deity swarms of units.

The sweet spot for AI culture wins is around Emperor, where they're less likely to conquer the peaceful ones or reach space before somebody pulls off culture.

[Gazglum]

Amazing game. Great writeup, loved the historical jokes. I was particularly impressed with Saladin's performance, he's always seemed a bit of a chump to me in the past. Not surprised by Catherine though. I think she's the strongest AI in the game, and I'd probably bet on her to end up at the top of the totem pole when its all over.

[Sullla]

OK, Game Eight, last of the opening round games. This one looks a lot harder to predict than the last one.


Asoka

Asoka is the less famous of the two Indian leaders, also probably the leader most commonly thought to be a member of the opposite sex. I can't even remember how many people on the CivFanatics forums thought that Asoka was a woman, it must be the long hair. Anyway, Asoka has Spiritual and Organized traits, a pretty nice pairing with good economic benefit. He has the Fast Worker unique unit and the Mausoleum unique building, both of them well above average for unique replacements. When played as an AI, Asoka heavily favors religion and culture. He has religion and science flavors for his research, and with Mysticism as a starting tech, Asoka will very commonly found his own faith at the start of the game. Asoka AI also likes to demand that others convert to his religion (8/10 rating), loves to build wonders (8/10), and doesn't hold much interest in units (2/10). Asoka has a low aggression rating at 3.7 out of 10, and he's considered to be a "Good" leader on the peace weight alignment scale. Asoka is another one of those AIs who will do very well if left alone, but has a tendency to get run over by more militaristic leaders when he focuses too much on religions and wonders. Either case could play out here.


Bismarck

Bismarck has been played quite a bit in our Multiplayer events, especially in smaller Play By Email games, when he can be paired in an unrestricted combination with civs like India or Mali. Bismarck has Expansive and Industrious traits, meaning that he's virtually guaranteed an early wonder of his choice if played correctly. The combination of cheap granaries and cheap forges can also be very strong. Bismarck is unfortunately tied in this game to the German civilization, which has two very late unique items in the Panzer and the Assembly Plant. They usually arrive too late to matter. Bismarck the AI only has one research flavor: MILITARY. Unlike some of the other suicidally aggressive AI leaders with this flavor, however, Bismarck is much more balanced in his preferences. His aggression rating is just a touch below average at 5.6 out of 10, and he's right in the middle of the peace weight scale. Bismarck AI also has middle of the road ratings for wonders and units, both of them sitting at 6/10. In other words, Bismarck will heavily emphasize military techs, but he won't be as crazy in declaring war as someone like Shaka or Montezuma. This setup may suit him well. We'll see if Bismarck AI can play a savvy game of realpolitik.


Churchill

Churchill is another candidate for the worst trait pairing in Civ4. He has the thoroughly awful Charismatic and Protective duo, a combination that offers no economic advantages and very little in the way of direct combat power either. It's not exactly a power combo. Churchill is the third English leader that we've seen, bringing once more the powerful Redcoat and Stock Exchange. As an AI personality, Churchill has military and gold flavors. His numbers are fairly average for the most part, although it's worth noting that Churchill disdains wonders (2/10 rating) and likes to spend on espionage (7/10). He has a moderate aggression rating at 4.3 out of 10, and a neutral standing in peace weight. Churchill is fairly bland as an AI leader, there isn't a whole lot that makes him stand out from the crowd. He'll have to play a strong game to overcome the weakness of those pathetic traits.


Cyrus

Cyrus finds himself in the role of another leader who had his traits changed in Civ4's expanions, and not for the better. Cyrus now has Charismatic and Imperialistic traits, a weak pairing that doesn't offer too much in the way of economy. It does let him get off to a decently fast start, but there are other leaders who fulfill the same role and do it better. Cyrus brings along the Persian civilization's Immortal unique unit and Apothecary unique building. Neither one has ever impressed me very much, not unless you're abusing AI archers on Marathon game speed with immortals. Cyrus AI has military and growth flavors for his research. He's a surprisingly aggressive leader (rating 7/10), although this is tempered somewhat by Cyrus' love of wonders (8/10). This mixture doesn't really make that much sense, and we saw earlier how Gilgamesh AI struggled with a similar setup. Cyrus has a low peace weight, one point above leaders like Catherine and Stalin from our last game. He will be predisposed to dislike someone with high peace weight, such as Asoka. I honestly don't have a whole lot more to say about Cyrus, he's somewhat aggressive but not enough to be interesting like Ragnar or Temujin.


Hannibal

Hannibal has one of the weaker Financial pairings, mixing the game's best trait along with Charismatic. It's... OK, I guess? None of the three extra traits added in the expansions have ever been that interesting, although I think we need to thank Firaxis for not doing the typical expansion thing and making all the new traits super overpowered. Hannibal is the only leader of the Carthaginians in Civ4, who have the Numidian Cavalry unique unit and the Cothon unique building. These ones are fairly average, not too weak or too strong. The Hannibal AI was apparently programmed from a Roman point of view, as Hannibal is clearly considered to be a villain with this setup. (I guess the winners do write the histories in this case.) Hannibal AI has military and gold flavors. He doesn't bother with wonders much (2/10), his unit emphasis is slightly above average (6/10), and Hannibal is notably above average in aggression rating (7/10). Hannibal AI gets rated as an "Evil" leader via peace weight, which is kind of unfair to him historically. In any case, he'll be a likely figure to stir up some trouble in this game, particularly if he starts near one of the leaders on the opposite end of the alignment scale.


Mao Zedong

Mao is one of the best leaders in the game to be saddled with the Protective trait. He pairs it alongside Expansive, so at least Mao gets one excellent trait to play around with. His Chinese civilization has the Cho-Ko-Nu unique unit and the Pavilion unique building. The AI probably doesn't build enough crossbows to make this too useful, although they will get the free Protective promotions if they show up. As an AI, Mao has the unusual distinction of being the leader who cares the least about religion in the whole game. He attaches almost no benefit or malus to shared/separate religion, and Mao has a rating of 0/10 for "Demand You Convert to His Religion." He simply doesn't care. Mao has growth and production research flavors, a very unusual combination. I'm curious to see what kind of route through the tech tree that produces. Mao AI doesn't focus too much on wonders either (2/10), but he does have a soft spot for espionage (8/10). He is slightly below average in aggression rating (5.7 out of 10), and Mao is considered to be an "Evil" leader by peace weight. It's actually a very low number, only one tick above Montezuma (and De Gaulle, heh) for the lowest spot possible. Someone at Firaxis was not a fan of the Cultural Revolution, and probably with good reason. In any case, Mao is made rather unique by his disdain for religion, his unusual research flavors, and his extremely low peace weight. What that means in the context of this game though, I have no idea.


Wang Kon

Wang Kon is the other Protective leader with a second good trait, in this case pairing it with the Financial trait. This is almost enough to make Wang Kon a viable leader in our events. Almost, but not quite. (Side note: the way that Firaxis gave the Protective trait to a bunch of Asian leaders is arguably racist, and at the very least in rather bad taste. We've got Churchill and Charlemagne as Protective leaders, and then Gilgamesh, Mao, Qin, Saladin, Sitting Bull, Tokugawa, and Wang Kon. This deserved a little bit more tact.) Wang Kon is the only Korean leader, and has the Hwacha and Seowon unique items. These both work rather well in the hands of the AI, especially the hwacha since the AI loves to build catapults in vast numbers. Wang Kon AI has gold and science flavors, a clear economic emphasis for his research. He is very average in all of his ratings, although Wang's aggression rating is higher than you might expect at 6.1 out of 10. In every other way he's a peaceful economic guy, but not that one. Wang Kon has a high peace rating as "Good" leader, same as Asoka. He could develop and become a major force in this game if he's able to avoid some of the more testy leaders at the other end of the peace weight scale.

Normally this is where I would try to preview the game and outline what I think is going to happen. However, I genuinely have no clue what we're going to see in this game. We have a more or less totally random collection of leaders here with no unifying theme behind them. They could take this game in almost any direction once it gets underway. Rather than prattling on, let's go ahead and get started. It's our final match of the opening round - who's going on to the playoffs?

Our map, spoilered for size:

You all know the drill by now. Make your picks for:

Winner
Runner Up
First to Die
Number of Survivors
Victory Date and Type

Because of the weekend, let's close this game in three days, on Monday evening April 14. That will also give me more time to work on these games, they do involve a lot of effort. Best of luck to everyone with their predictions.

[Sareln]

Winner: Hannibal
Runner Up: Bismark
First to Die: Cyrus
Number of Survivors: 4
Victory Date & Type: Space, 310

[slowcheetah]

Winner- Bismarck
Runnerup- Hannibal
FOTD- Asoka
Number of Civs- 3
Victory Type/Date- Spaceship (T310)

[Brian Shanahan]

Wow, big difference in the map to the previous ones. I'll roll the d6s and d300 in the morning.

[Lord Parkin]

Will we ever get a Cultural win in this competition? The AI definitely will go for culture sometimes.

Not on Deity. The AIs develop too fast, and get an overwhelming military advantage or the tech for space before they'd finish a culture run. The culture-prone leaders keep getting killed by the unitmongers wielding Deity swarms of units.

The sweet spot for AI culture wins is around Emperor, where they're less likely to conquer the peaceful ones or reach space before somebody pulls off culture.

Actually I've tried a few Deity games with a full house (17 leaders + 1 observer), and most are won by culture. Haven't seen it happen in the smaller games yet, but it should be possible.

[Bobchillingworth]

Winner: Hannibal
Runner Up: Wang Kon
First to Die: Churchill
Number of Survivors: 4
Victory Date & Type: Space, 302



Really shooting in the dark here. I'm guessing Churchill is in trouble, since there's a lot of "evil" civs around him. Cyrus and Hannibal could end up being buddies. If they do fight Hannibal will win, Darius didn't do too well with Immortals and Hannibal will surpass him in tech fairly quickly. Wang Kon has one really good trait and a great UU, I'm thinking he'll dominate his peninsula although that chunk of desert nearby isn't doing him any favors. I don't forsee any of the leaders becoming a massive runaway here and rampaging over the map. I suppose Hannibal could, but lacking any discounts and having a mediocre UB for this map is going to hold him back.

[Dhalphir]

Winner: Hannibal
Runner Up: Wang Kon
First to Die: Cyrus
Number of Survivors: 4
Victory Date - t305, Diplomatic

[Lord Parkin]

I take it you've gone off the maps with a central civ?

This one is very difficult to pick, I was hoping the geography would give more clues. Alright, here's my shot in the dark...

Winner: Bismarck
Runner Up: Cyrus
First to Die: Asoka
Number of Survivors: 3
Victory Date & Type: Space, T323

Reasons why:

From my experience Hannibal does well once he gets going, but Cyrus likes to war a lot and will keep him in check during the early years. Cyrus's position with a single neighbour should give him an advantage, and if Churchill or anyone else gets in on Hannibal from the south then Cyrus stands to gain a lot. Churchill is probably a bit too squished to do well in the long term, unless he takes the lion's share of Hannibal's land. Mao should do alright but not brilliantly - he tends not to be aggressive enough to rank near the top. Bismarck tends to do quite well in some games and quite poorly in others, depending on whether he can be enticed to go to war - I'm rooting for him here because of the map, but it's a risky bet. I can't see Asoka surviving long-term, it's just a question of who will finish him off... especially if he and Wang Kon found Buddhism/Hinduism next door to one another. I still haven't seen a game where Wang Kon has survived to the end, but perhaps he'll surprise me.

Full prediction:

#1 - Bismarck
#2 - Cyrus
#3 - Mao Zedong
#4 - Wang Kon
#5 - Churchill
#6 - Hannibal
#7 - Asoka