Sullla AGA14: SPGT30 (Fourth Game)
v.71 (Build 33655)
Emperor Difficulty: Random - Victoria (again!)
Small Fractal, Epic Game Speed
Note: Victoria had her original Financial / Expansive trait pairing for this game, which she would retain until the Warlords expansion shuffled all the leader traits around. We were doing a particular push to investigate diplomacy at the time and this was my fourth game on the topic, though unfortunately I don't have the reports saved from the first three since they were played on-site at Firaxis.
One particular area of the game that hasn't been getting as much attention lately is the Epic game speed, so I resolved to carry out my fourth diplomatic investigation on that slower speed and see if I encountered any problems. To offset the speed, I went with a Small map and picked Fractal for lack of a reason to do otherwise. Random civ, drew Victoria for the second game in a row! Ordinarily I would have simply rolled a new game, but again the starting position was too good to pass up:
I don't know what it is with Victoria, somebody must like her to keep doling out these great starts. 
I was quite familiar with the Expansive/Financial combo as a result of the last game, so at least I would be on solid ground there. After my experience in the previous game, however, I resolved that I would avoid building the Colossus to avoid that crutch and play a "real" Emperor game.
Started out building a Work Boat (45 turns?! oh, this is Epic alright!), followed by Worker/Warrior/Settler. My starting Warrior I used to explore the area around the capital, and I was VERY careful to keep him in the forests as much as possible. In a Quick game you can pop out another warrior at the snap of a finger, but in an Epic game you've got to think long and hard about each move that you make. At 10 turns for another warrior I could NOT afford to get the one I had eaten by a barb lion! In case you can't tell, I really like the pace of the Epic games. I wouldn't want to play every game on Epic - it really can be slow at times - but the game becomes so very strategic when each tech research takes longer. With techs popping at 2-3 turns on Quick the path you follow doesn't matter that much, but it's a different story entirely when MYSTICISM takes you 10 turns!
It's a shame that the Multi-Player community will probably almost never play anything other than Quick. Even Normal speed games require a great deal more planning, and here on Epic... well, strategy would count a lot more than a fast mouse button, let me tell you. 
As far as this game went, when I didn't meet anyone within the first 20 turns I correctly guess that I was on an island, and a pretty fair-sized one at that. There were a fair amount of barbs running around, but I made sure to get to Archery early on and guarded all my cities with archers. My second city (York) grabbed horses, and I was able to train a couple of chariots that easily controlled the incoming barb warriors. Third city (Nottingham) went into the backline "safe" territory and was basically a glorified fishing village that could use sheep and two plains hill tiles for production. Fourth city (Hastings) was a dotmap filler that had decent but not great land. I expanded slowly, knowing there was no competiton for land, keeping my science rate up at 90% or higher the whole time. London built the Oracle sometime in the BC years and grabbed Code of Laws with it, founding Confucianism in York. Since that religion again spread to my third city before Hinduism, I again swapped over to the newer religion (second game in a row the exact same thing happened). Here was the situation in 120AD:
I'm researching Horseback Riding in order to train some Horse Archers to take out that barb city. Yet by the time I was ready to attack the barbs, I actually had gotten all the way up to Civil Service and ended up using maces instead! Oh well.
I'm more convinced that ever that the best strategy in Civ4 is a cautious expansion that controls maintenance costs rather than a Civ3-style settler flood that drives you into bankruptcy. Every time I've tried to landgrab too quickly, I've found myself falling horribly behind the tech leaders and having a devil of a time keeping up. But here in this game again, I threw a couple cottages into every city and kept pace with the AI even without benefit of the Colossus. Take a look at the Demographics from 120AD:
I'm among the leaders in every category except soldiers, which of course I'm not building a lot of since I'm alone on my continent. After taking those two pictures above, very little of note happen for the next thousand years. I peacefully built up my civ and didn't found any more cities - since I had to take out the barb city to add any more, and I had more important stuff to build before units to kill the barbs. In 950AD I finally researched Optics and began training some Caravels to go meet the other civs. The game was now officially afoot!
This is the original Demographics screen that Civ4 launched with; the famous Demographics screen with the bar graphs didn't get added until a patch about two months after release. I believe it actually was developed by a community modder and it was so superior to the official version that Firaxis asked if it could be included in the full game. This happened so soon after release that most players don't even know that there was a far inferior original version of the Demographics which was more or less stolen directly from Civ3. (Click here for a comparison screenshot.)
Yep, my civ still looks about the same. York was lagging horribly behind in growth because it had no tiles that would provide more than two food - and it had to drop every shield it possessed in order to work those, so I had to swap it between shields (no growth) and growth (no shields) to get anything done. The other cities are doing fine. I had finally gotten around to attacking Zhou, and it would be razed in 1050 (hey, it didn't fit the dotmap!)
Met the first AI civ (Hatshepsut) in 1100. I signed Open Borders with her, then swung a MASSIVE tech trade (I gained three techs, she gained five) that swung our relations to +3 despite the fact that I was getting a significant minus from having a different religion. The next turn I met Montezuma, who (surprise, surprise) was unfriendly and refused to sign any deals. The Aztecs turned out to be a runt civ, behind in territory and tech compared to the others. On the same turn I ran into Frederick, who was also friendly and signed Open Borders. Then finally I met Peter a couple turns later, who was also willing to start trading. One unusual thing quickly jumped out at me - ALL of these civs were Jewish when I met them! Never seen that before.
Monty soon jumped back to Buddhism (which he had founded very early in the game), but that actually hurt the Aztecs more because their relations worsened that much further with the other AI civs. No doubt as a result of those lowered relations, Hatshepsut declared war on Monty in 1305, and on the next turn asked me to join in. Since I had no deals with the Aztecs and was in absolutely no danger, I thought "why not?" and agreed to do so. This improved my relations with Egypt at the cost of a relations hit with the Aztecs - but since the Aztecs hate everyone anyway, that was no big loss.
Here's a shot of one thing that still appears problematic. Frederick definitely has Feudalism, yet his defense in his capital city is still made up of archers and not longbows? The AI civs are just not upgrading their units well enough. If they need to turn down their science for a turn or two just for upgrades, they need to DO IT in order to provide some real defense. All these archers are doing is costing Germany money in unit costs. At the very least, if the AI finds itself in a war it needs to stop whatever it's researching and upgrade those units. Right now it seems like a human could exploit this.
At this point, the critical tech for me was Astronomy. My landmass was pretty good, but I desperately needed access to the resources of the other continent and couldn't trade for them until I got Astronomy. I ended up painstakingly researching it over 25 turns (25! gotta love Epic speed) but it was well worth it when I did and could start importing some more happiness resources. That allowed my cities to grow to larger sizes without having to divert research into the culture slider, which I was reluctant to do. As my tech research was running neck-and-neck with the AI civs at this point, I didn't want to have to cut science at all if I could avoid it.
By 1500AD I had scouted out the whole world and began to investigate the diplomatic situation. The Aztecs were small and backwards, hating everyone, a complete non-entity for this game. (Since this happens way too often with the xenophobic civs, Soren needs to change them around a bit to make them more competitve. In this game Monty was still researching Paper and Liberalism when the rest of the world was building the spaceship.) At the other end of the spectrum was Egypt, who was the dominant civ on the other continent and was leading in score. Frederick's Germany was running a three-city challenge (he had apparently lost some cities to Peter in an early war) but doing surprisingly well, only a handful of techs behind the leaders. And Peter's Russia was crammed onto the end of the continent, but was actually the AI tech leader, slightly ahead of Egypt and running even with my civ. The fact that the smaller civs could compete with Egypt (and even lead the tech race) was another indication that Civ4 is doing something right when it comes to combatting the problem of bigger civs always being better. 
Those are my relations with the civs in 1510. Basically, everyone except Monty likes me (and Egypt really likes me). There are two problems with this, however. Although Egypt is my best friend, Egypt is also for sure going to be my opponent in any UN election that might come to pass. No one else is ever going to challenge Egypt on population (the Egyptian core cities were full of irrigated floodplains and contained several 20+ cities even before Biology). The second problem is that while everyone liked me, they liked Hatshepsut even MORE, thanks to the shared religion of Judaism. Frederick was "Friendly" (the best attitude) with Hatshepsut, and Peter probably wasn't too far off either (he was "Pleased" but likely had more pluses with Egypt than with me).
This isn't just an issue for idle curiousity. Together, Egypt + Germany + Russia could probably marshall enough votes for Hatshepsut to win by diplomacy! At the very least, Hatshepsut would be elected UN Secretary and could make my life miserable with all sorts of awkward resolutions. What I need to do, therefore, is pry Germany and Russia away from Egypt's camp and onto my side - while trying to avoid irritating Egypt too much, since I need my luxury deals with them very badly. This looks like it's going to be fun. 
Still continuing with my Epic game, I headed up the top of the tree in the Renaissance period and went for Democracy so I could build the Statue of Liberty, one of my favorite wonders (especially when combined with Representation). Now I knew that the Statue would be expensive to build here on Epic, but with copper hooked up for my civ it couldn't be too bad, right?
34 turns?! 
Well, you've got to love the Epic speed! The AI civs basically ignored Democracy until I had less than 10 turns to go on the Statue, and I did end up getting the wonder despite the lengthy time of construction. What a vastly superior system to wonder prebuilds and cascades!
One thing that's been getting talked about a lot in regards mostly to Multi-Player is the scoring system for Civ4. I found some more interesting data to report here, from the very different persepective of Single-Player. Take a look at my civilization in 1728:
I'm leading in land area and have more population than anyone other than Egypt's insanely irrigated floodplains civ. From the demographics of the same year, I can also report than I am #1 in GNP (361 to 213), #1 in Manufacturing (167 to 75), and #1 in crops (263 to 154). In short, I'm basically dominating in all categories. But my score is behind not only Hatshepsut, but behind Peter's tiny civ as well?! What's going on here?
The answer appears to lie in terms of military. I have only a handful of cardboard cutouts for my military, while all the AI civs are stacked to the gills with military forces. In fact, I am dead last in military on the demographics, 145K to a world average of 400K (ouch!) So it appears as though your military factors quite heavily into the scoring for your civilization; that's not a bad thing at all, but it's something to be aware of. By any non-military estimation, I was far ahead of these civs that were technically leading in score.
The in-game scoreboard was causing some problems for Multiplayer games since they often ended without any clear resolution (no one dying or achieving a victory condition), with the highest score player being declared the winner. That's a problem if your scoring system isn't functioning very well! A lot of these old screenshots have extremely low scores compared to modern Civ4 since they were operating under a different scoring mechanic. There was a lot of community discussion on this topic and around this point Soren switched over to the current Civ4 scoring, which grants points for population, territory, techs researched, and wonders built. Soren also included a requirement that territory must be controlled for 20 turns before counting for score points, specifically to address Multiplayer games where people were dropping extra cities and launching Great Artist culture bombs on the final turn. None of this refinement to the scoring system ever would have happened without the work of the testing group and especially the MP crew.
Getting back to diplomacy, one of the things that I found was the Free Religion removes the diplomatic penalty you get from having a different religion from another civ. The tradeoff is that you also lose any benefit you might get to relations from having another civ share the same religion as your civ. Soren, this is EXACTLY how it should work, so please don't change this! 
In the above shot, the lack of a religion next to the name of my civ should show that I was running Freedom of Religion, and in fact had been doing so for some time. This wiped out the minuses from having a different religion from the civs on the other continent, and I hoped that it would be enough to get the Germany/Russia civs onto my side.
It wasn't enough though; Germany still prefered Egypt to my England, and Russia seemed to be leaning that way as well. This called for some more direct action therefore... In 1734 I paid Peter 400g to adopt Free Religion - and it definitely helped things out. Peter fell from "Pleased" to "Cautious" with Egypt while keeping the same good relations with me. Success! But Peter swapped his civics back to Organized Religion in 1748 - gah! Stay bribed, darn you! Furthermore, Germany's relations with Egypt were so good that even changing to Freedom of Religion wasn't going to solve things. It was going to take more effort still to get these civs on my side.
The answer was to begin sending a series of Confucian missionaries over to the other continent and begin systematically converting the German/Russia cities to Confucianism. This was a slow process, but the overall slow flow of the Epic game helped make it worthwhile (this is another thing I like about Epic games, much more of a chance to use missionaries due to having more time to move them). In time I converted all of their cities, then asked the two leaders to convert. For a significant amount of money, they agreed to do so. Success!
That shot shows the religion present in all of their cities, plus my new relations with Frederick. And yes, I too have gone back to being Confucian and running Pacifism (even though I would have preferred to have been running Free Religion), so that I can enjoy the religious diplomatic benefits of sharing a religion. Now Germany was "Friendly" towards me and "Pleased" with Egypt because they were no longer sharing the same religion! Peter had also gone to "Pleased" with me and "Cautious" with Hatshepsut. I think we've managed to swing some UN votes onto the English side here!
Now the one cautious note in this otherwise good news is that we need to be careful that this is not easily exploitable. For example, AI civs should not even CONSIDER changing religions unless the religion the player suggests is present in most of their cities, preferably their most populated core ones. I can see a player converting a couple junk cities and then getting the AI to convert to their religion, destroying the AI in the process. If the AI does accept a blatantly unfair deal like that, it should just convert back to its former religion almost immediately. In this game Frederick and Peter never converted back to Judaism - because ALL of their cities were Confucian while a couple of their cities were missing Judaism. If a player goes to the effort of converting all or almost all of an AI civ's cities, he or she should be able to get them to convert so long as the AI civ in question doesn't possess the holy city for their religion. As long as we can keep the human from being able to exploit the AI, this will be a REALLY cool feature. This game was awesome in that religion really played a major role in diplomacy, just the way it's supposed to. We've come a long way from the days of the AI randomly swapping religions at random every few turns!
So is this the solution to air power being too strong - have the AI mass SAM units in their cities? I had quite a surprise when I noticed this. Air power is not something I tested in this game, but it still needs a lot of attention. Just thought I'd post this picture so that everyone can see what the AI is actually building in v.71 when it comes to late-game units...
Anyway, I built the UN in 1850 in London (it still took 15 turns with Factory/Coal Plant/Iron Works, heh). The election for Secretary-General came up a couple turns later, me versus Hatshepsut (no surprise there). So had my diplomatic work paid off?
It sure had! Clean sweep of the votes that I was expecting to get, making me the UN Secretary by a pretty wide margin. Oh, and notice that Soren changed the percentage needed to become Secretary from 50% to 40%. Looks like a good move, because in many games no one was getting elected Secretary, thus making the wonder useless. Since the resolutions are an interesting part of the game, I like having it a little bit easier to get them into play.
Four turns later I get the option to pick a resolution, and so I pick Diplomatic Victory. Why drag this out, let's just see if I can win right now. Everyone votes as excepted, and the result is:
Victory! Just like that game over in 1856AD. This is my earliest peaceful victory so far (and in an Epic game too), not bad. The critical question is whether or not the vote threshold for victory is too low, and my feeling is that it's not. 60% seems like such a low percentage, but when you consider that civs can abstain from the vote altogether (and frequently do so), plus the fact that the #1 or #2 most populous civ is always going to be a candidate and vote for themselves - well, it's actually quite hard to get to 60%. Even with two of the other four civs voting for me, I just barely reached the number needed to win. This doesn't seem like it needs to be adjusted upwards to me, especially since I had to work for the votes I did get. UN may need some tweaking, but it's actually in fairly good shape at the moment. I think I'm one of the few people to actually win by the UN, and we could certainly use some more data on it, but that's my impression anyway.
Few people in the testing group spent much time investigating the Diplomatic victory condition. I did more investigating here than just about anyone and I'm pleased that this victory condition ended up working quite well. Certainly the diplomatic victory in Civ4 works better than the diplo victory in any of the other Civilization games and by a pretty wide margin.
And those are my relations with everyone at the end of the game. One thing I'd like to ask for - Soren, is there any way we can get some additional names for the AI attitudes? Right now there are only five of them: Friendly (usually about +10 and up), Pleased (roughly +3 to +9), Cautious (-2 to +2), Annoyed (-3 to -9) and Furious (-10 and lower). It would definitely help to have some more names in there; the numbers are great, but telling attitudes at a glance with words helps a lot. In particular, there's a big difference between a +3 Pleased and a +9 Pleased! Two additional tags (for 7 overall) I think would be right on the nose.
More AI attitudes would have been nice but now that I know more about how the AI is coded (with specific programming associated with each of the five attitudes) I understand that this wasn't a realistic request.
Four SPGT30 games finished so far, can I get another one done this week before going back to school? We'll see...
