AGA11: Emperor Stress Test

Sullla AGA11: Emperor Stress Test
v.66 (Build 31873)
Emperor Difficulty: Persia (Cyrus)
Standard-Sized Ice Age Map Script, all other default options

Note: Cyrus had his original Creative / Expansive trait pairing for this game, which he would retain until the Warlords expansion shuffled all the leader traits around. I also wrote this report on a weekend after playing the game on-site using the internal Firaxis test build which would end up coloring some of the commentary.

I wanted to get a look at the Emperor AI myself, so when Sirian posted a Single-Player Group Test with that goal in mind, I thought I would download the save and take it for a spin. Since the computer I use at work is on a different release than what the Play Session is using, however, I was unable to open the save. Therefore, I decided to take the same settings and play out a game on a different map, which may provide even more useful information. I'm posting this here so that I can upload pictures to go along with my report and so that the Play Session group can also take a look at what I found and comment on it. The version is not the same as what most of you have been using this past week, so keep that in mind though.

I took the first start that came along and found that I had corn and stone in range of the first city - not too bad. Persepolis was also on the coast, which would of course be a big advantage too. The build order for my capital was Warrior/Worker/Settler/Warrior/Worker/Stonehenge. That wonder was built elsewhere in the world in 1960BC, when I had 7 turns remanining on it (and the Oracle fell on the next turn - the AI civs are fast on Emperor!) but I could probably have built it myself if I had tried it before the second worker. I also probably could have gotten it if I was Industrious, so balance seems about right in this area. You can get the early wonder, but you really have to sacrifice to do it.

As far as techs go, I planned to skip an early religion entirely and just make due with whatever spread to me from an AI civ. I can't remember what my first research choice was; I think it might have been Agriculture (?) But my scout popped Mysticism from a hut on about the fifth turn of the game, so I immediately changed my plans and swapped research to Polytheism. Buddhism fell VERY early on, but I lucked out and actually got to Hinduism first and founded the religion in my capital. Sirian also found in his game that he COULD have gotten to Buddhism first, so the religions seem to be balanced as well on the higher difficulties. A Spiritual civ will have a good chance of founding a religion on Emperor, the non-Spiritual ones have much less of a shot, but it will still happen sometimes. This all seems good to me.

As usual for an Ice Age map, the land mass was long and kind of narrow (I really do like this map script, it's very unique). I realized early on that if I got a settler out far to the west, I could seal off my neck of land from the AI civs and then safely backfill it with settlers at my leisure. I actually sent out an unescorted settler into the wilderness to do this, since I couldn't have a warrior ready for a couple more turns, and just relied on the AI civs and their tons of units to clear out the barbs for me. Risky, but it worked. Here was where I stuck Pasargadae:

It's hard to see, but this city will get cows, deer, and fish after border expansion. That expansion will also seal off my peninsula, and Pasargadae gets to be a coastal city as well. This was a major early-game coup for me, making up for the fact that Persepolis did not get Stonehenge. Things were pretty uneventful for a while after that. I managed to put another city southwest of Pasargadae that grabbed a gold resource and further sealed off my borders. After that it was backfilling time as I expanded towards my capital. The AI civs all left me in the dust, of course, their scores enormously ahead of mine and possessing all techs that I could see on the tree. This was very reminiscent of Civ3 Deity. Now I just had to keep plugging along and see if the Civ4 AI would collapse in the late game as the Civ3 AI inevitably would do.

Further expansion of my civ. Napoleon is the runt civ of the planet (he got a lousy tundra start) and I have almost caught him in score! The barb city of Ghuzz has appeared on the only tile on my entire island where it possibly could show up. Fortunately, I wanted a city on that tile anyway, but conquering barb cities on Emperor is a LOT harder than building a settler. This is fine for higher difficulty (there are definitely more barbs and tougher ones on the higher difficulties), but I wouldn't want to see this on Noble. From my experience, it seems that the barbs are scaling by difficulty, so more good news on that front.

The AI seems to be settling aggressively now, or at least Napoleon was doing so in this game (does that vary by leader? It would be neat if it did) Tours can only generously be described as a fishing village, and it's already under a lot of pressure from Pasargadae. Napoleon would put a further city on the hill tile where I have a white arrow. Later on, Qin would settle aggressively near me as well. This is not bad necessarily, but it is something new that I hadn't noticed before. Is this also affected by Emperor? Just another wrinkle I'm trying to learn about this game.

OK, first big bug associated with this test: AI civs are changing religions WAY too often. I made up this compilation shot to illustrate the problem:

The AI civs are basically changing their religions constantly. They do this with no logic behind it whatsoever; Napoleon founded Judaism, but he leaps into Confucianism, then back to Judaism, then back to Confucianism again, etc. Bismarck couldn't make up his mind at all, going through three different religions, and Qin kept going back and forth between the two religions he founded. So much for natural alliances forming up along religious lines! It was very clear that the AI civs were a collection of lines of code on this point and not actual thinking beings. What I believe is happening is that the AI civs are asking one another to convert religions, and they keep selecting "yes" with no thought as to the consequences. The AI civs are similarly changing civics CONSTANTLY. I mean, they changed their civics almost every turn, it seemed like. I'm sure this is triggered by the same phenomenon, the AI civs asking each other to change civics and them constantly selecting "yes". I hate to think how many turns were being lost to anarchy in all this nonsense. I'm going to enter both of these into the Bug Databse on Monday, and they are serious cause for concern. This needs work, badly, or the AI is going to look like a collection of idiots.

That picture is just to show that water borders are still acting strangely. They aren't nearly as crazy as they were in build 31701 (where they weren't expanding into water at all!) but something funky is still going on that needs to be looked at. This will be in the bug database as well once I have access to it again.

Since I had the Hindu Holy City, I was naturally trying to pop a Great Prophet to build the Hindu Shrine. I had roughly an 80% chance to pop a Prophet in 125AD and got a Scientist instead. Argh! The nice thing about Great People is that they are all pretty good, even if you don't get the one you want. I used him for an Academy and got a Prophet eventually with my next Great Person (although I had to run a Priest for about 60 turns to get him).

One of the biggest differences about Emperor compared to, say Monarch, is the lack of happiness and health. Your cities only get 3 happy faces by default on Emperor, and that's not a lot! This isn't Civ3 where you can simply turn up the luxury slider to compensate, you have to find that happiness somewhere else, and I really felt the crunch in this game. I had several happy resources near me - but they were all Calendar resources (dyes and silks), so I ended up beelining for Calendar to get them online ASAP. I think I SHOULD have gone for Monarchy and Hereditary Rule instead, which I'll be trying in the future. For ages though I had a happiness crunch in my cities, forcing them to stay small in size; things looked very bleak for a while with extremely long tech research times and tiny cities. The combination of extra happiness + courthouses (Code of Laws) + markets (Currency) really turned things around for me. The difference was startling, going from 20-turn techs at 70% science to 5-turn techs at 90% science. I believe that the economic side of Civ4 is in very good shape but it needs to be explained better because it is NOT at all apparent how to repair a sagging economy. The manual needs to be very clear about the importance of building courhouses; I'm coming to believe more and more that they are the MOST important building to get research back on track in the Middle Ages. So again, this all makes sense to me and it works well, but most new players are going to get lost here unless we explain to them how the system works.

This was one of the big changes between pre-expansion Civ4 and the later expansions: the default health and happiness scaled by difficulty level. Originally, the player had an extremely low health and happiness cap on the top difficulties which made Emperor and Deity very, very tough to play. The gameplay became much easier when the expansions swapped over to the same flat health/happy cap on all difficulties and I've always felt that this was a mistake. Making due with a mere 3 happiness in your cities took true skill to pull off.

Not surprisingly, Tours eventually revolted and joined my civ. This was the second revolt in that city, which I believe is the first one that can be successful. Again, this seems to be working well - Napoleon got over-aggressive with this city site and paid the price for it when the city flipped away. The one problem is that I had no option to refuse the city; it simply joined my civ without giving me any say in the matter. Now the player is going to want to get the city in almost all cases, but there will be some situations where adding another city is not desirable, so there needs to be some way to refuse a city that flips to you. I'm going to enter this as a suggestion in the bug database.

And this suggestion of mine did make it into the finished gameplay, hooray! I have actually turned down city flips in some of my cultural pressure games too.

So when the borders of Tarsus expanded, I found to my pleasant surprise that there was another island down to the south of me which was unoccupied by the AI civs. Tarsus was the sort of city I wasn't even sure whether to found or not, but it paid off in a big way here by giving me access to another island that would become my second core. The one surprising thing was that the AI civs were not already on this island; both Napoleon and Qin had access to it from their own cultural borders as well. I was going to put in a note saying "AI is not expanding well across water", but later on in the game they were doing a good job of exactly that. In other words, I don't quite know what to make of this. Maybe they didn't want to expand there because of the barbs running around on the island (?)

Also notice above that the cultural borders in the water are REALLY screwed up outside Arbela.

Lots of peaceful building up to this point. By 1000AD, some of the AI civs actually didn't have techs that I could see on the tree! Pathetic as it sounds, this was a major step forwards for me. I like the fact that it's no longer possible to make up ground on the AI by trading your way out of a tech hole; you actually have to research the techs yourself and get ahead the hard way. No more running 100% cash for the entire game! More on AI diplomacy a bit later on. You can also see I'm in the process of getting my second core up and running. I would place three cities down there myself and capture barb one for a total of four - almost as many as on my starting continent. Much later, after getting Astronomy, I would also found some cities on islands to the east. I like how expansion can actually take place throughout the entire game now, rather than simply in the opening 100 turns in Civ3-style. The maintenance system is a definite winner.

Another bug. I've already researched Civil Service, yet I can't irrigate off of this French farm and spread farms back towards Pasargadae. (It's not just a gray graphic on the display; I actually can't build the farm.) I thought this might be a design feature, not being able to irrigate off a foreign farm, but something like 5 turns later I went back and found that I could, in fact, irrigate off of the tile. This is going into the bug database too.

I captured this barb city, and while it was in resistance, this was what the city screen looked like. Did the game default to Artists instead of Citizens because I was running Caste System at the time? Weird. Obviously this had no effect on gameplay, but I'll post it here anyway so that Soren can take a look at it. I guess this counts as a bug. Just strange.

Time to mention AI Diplomacy. Unfortunately, this is one area where Civ4 still needs a LOT of work to become more logical. First of all, the AI civs still pop up WAY too often with demands to change your religion and civics to match theirs. It's not as bad as in some of the earlier builds (where they pestered you practically every turn), but it's still bad. I've said this before, but I'll say it again: I think the diplomatic system in Civ4 would be improved if AI demands to change religion and civics were simply axed entirely. It's a neat idea, but the implementation is extremely flawed and leads to all kinds of craziness (as I discussed above). My relations with other civs keep getting poisoned as the AI "demands" I change to some idiotic civic or swap to their religion, neither of which I'm ever going to do. Right now, these demands are irritating and pointless, and they're messing up the diplomatic engine. If Soren can't get them to work soon, just cut them and the whole system will be improved.

Towards the end of the medieval era, I started noticing something else bizarre with diplomacy. The AI civs randomly began "refusing to talk" with me. Some of the time, they simply wouldn't talk with me at all about anything - then a couple turns would go by and they would again talk with me. No change in relations, just sometimes they refused to talk for any reason. As the game progressed, more and more civs refused to talk with me despite NO action on my part whatsoever towards them. I have no clue what was going on, it must be some kind of a bug. Bismarck remained PLEASED with me, yet he refused to talk with me for the last thousand years of the game! And - he made a demand on me, which allowed me to get into the diplo screen, where I was able to trade techs with him! We were maintaining Open Borders and trading a resource, yet I could not contact him because he refused to talk with me. My civ's best friend simply stopped talking to me, even though we stayed on friendly relations! This needs to be fixed.

I was able to make a lot of trades with the AI, take a look:

Notice also that the leaderhead graphics on this screen are missing, another bug. Anyway, I was able to trade my excess resources away for some gold per turn, which helped me to run 100% science and help catch up. Furthermore, I was able to get Open Borders with some friendly civs (Bismarck, Kublai Khan) while some other civs were sourpusses and refused to do much trading (Alex, Isabella, Napoleon). There is the makings of a good diplomatic system here, with logical trading partners and alliances based on shared interests and religion. Right now - it's buggy as hell. But I think it can be fixed and fine-tuned with a lot of work and rigorous testing. I am definitely keeping my eye on diplomacy as much as possible, because right now it needs a LOT of work. It needs the following things to take place:

1) An end to pointless demands to change religion/civics.
2) An end to bizarre "refuses to talk" messages. This should only pop up if you've been at war with another civ or something equivalent.
3) AI civs seem to randomly get angry at you for no reason right now. I do nothing and my relations with the AI civs get worse. Why? This needs to be fixed.
4) AI civs need to stop proposing idiotic trades. They still offer 10g for Fission. Yeah, right. I've seen the AI propose decent trades - but far too many are of this variety and make it look stupid.
5) The AI civs keep demanding I "cancel deals" with other civs. This is a stupid demand (the game doesn't even tell me what these "deals" are!) and Civ4 would be better if this was simply axed entirely.

That's all I can think of off the top of my head. I don't think this area of the game is getting enough attention, and we really need AI diplomacy to be at least functional or the long-term life of Civ4 isn't going to be very good. At some point, this will have to be addressed.

I spent a lot of time in August and September working on AI diplomacy because it was one of the few gameplay systems that was still in rough shape. To Soren's credit, most of these issues were hammered out by the time that the game released at the end of October; I still think that Civ4's diplomatic system remains one of the best in the genre.

I felt that things were going pretty well in my Emperor game up to this point. I had two cores of cities, a Forbidden Palace in Pasargadae, and thanks to my resource trades and religious city shrine income, I was able to run 100% science without losing money. The AI civs were still ahead in tech, but I had closed the gap somewhat, and was pulling even with the tech laggards. Qin was the tech leader, which is interesting because he was the second smallest after Napoleon (and that right there shows that Civ4 is doing something right, if the small states can compete with the big boys). If there was one area where I was somewhat lacking, it was in defense, but I was doing my best to counter that by beginning a slow military buildup. Evidently, it wasn't fast enough, because in 1440AD Napoleon declared war and troops started swarming over the border:

I survived the first turn without losing any cities, although two workers were snatched up. The fun was just beginning, however, as Napoleon's REAL attack force poured across my borders on the second and third turns. It was... scary how many units he had. This was not the Civ3 "sneak attack with a single unit", Napoleon had clearly built up a significant army and thrown it at me. Good for the AI, but, umm, bad for me in this situation. I upgraded a ton of units in my border cities, but it still wasn't enough. Tours and Susa fell on the second turn and the floodgates were opened as a huge stream of French units poured into my territory. Here they come!

I could not stop them, naturally. My only hope was to hole up in my remaining cities and wait for them to sign peace. I got about a half-dozen units behind the walls of Pasargadae and prepared to face the worst of the onslaught. And on they came, stripping the land bare of improvements and advancing well into my borders. I saw a lot of good stuff from the AI in this war. The AI units were clearly attacking in groups; some turns Pasargadae wouldn't be attacked at all, and on other turns a group of 4 or 5 units would all attack at once. The AI also used catapults in a semi-intelligent fashion, sometimes using them to bombard defenses and at other times sending them in to cause collateral damage. Napoleon also had a nice mix of units to prevent any one unit of mine from having the edge.

As for what wasn't so good... Well, the fact that the AI was attacking in groups was great to see, but the groups themselves were still acting pretty mindlessly. A group of three chariots is good, but that group of chariots still shouldn't throw itself at muskets behind walls! Napoleon also had ridiculous numbers of un-upgraded units in his attack force. Tons of archers, spears, chariots, etc. If they had all been upgraded to their most recent form, things would have gone sour for me REAL fast. Take a look at the following picture:

This group threw itself at Pasargadae the following turn and was slaughtered. If it had been a group of two longbows, a horse archer, and a pike, then the odds of winning would have been much greater. I'll also point out that you can see the promotions on foreign units by highlighting their flag with the cursor; this is a bug and I will put it in the database on Monday. (You are not supposed to be able to see the enemy's promotions!)

This was later changed so that everyone could see everyone else's promotions, which was pretty much mandatory for Multiplayer purposes.

In the end, Napoleon wasn't able to take Pasargadae; it was simply too heavily defended. He did a ton of damage to my civ, however: slowing down my research, taking two cities, and pillaging a large area back into the stone age. Despite some problems, the AI fought this war much more intelligently than anything I ever saw in Civ3. Napoleon (1) looked at who to attack and picked the least-defended target - me (2) attacked with a REAL force and not a trickle of units (3) used groups of units in semi-intelligent fashion and (4) did as much damage as possible and then made peace. I don't want to sound too critical, because I was quite impressed overall by the AI performance. The 10:1 kill ratios of Civ3 are dead dead DEAD, and the AI will be extremely dangerous on the higher difficulties. As it should be.

The battlefield just after peace was signed. The flag flying over Pasargadae may be tattered, but it was still Persian! (Notice too how the AI was intelligent enough to pillage the iron next to Arbela.) I resolved at this point in time that I would get Tours and Susa back at some point before the game ended. If there's one thing I learned from Civ3 Deity, it's never to give up. Napoleon may have sucker-punched me once and did some damage, but I was far from out of the game.

A word about Tours: as anyone can see, it's going to flip back to me. There's just too much pressure on the city for anything otherwise to happen. Napoleon has NO chance to hold onto this city; no sooner did it go out of resistance than a Persian revolt took place. I'm somewhat uncertain what to think about this. On the one hand, a city under that much pressure flipping doesn't seem out of line... but on the other, there was nothing Napoleon could do to keep the city. Nothing. This could make for some very bad mojo down the line, less random than the Civ3 culture flip but still very frustrating. I don't have an opinion yet one way or another as to whether this is a good thing, but this is the reality of what's taking place in the game. I'd be interested to hear opinions from some of the other testers on this since I don't have a strong feeling either way.

Susa, the other city I lost to France, revolted three times but never flipped to me. I thought that a city was guaranteed to flip on the third revolt, but perhaps I misunderstood the situation. It never returned to me by peaceful means, so I would have to take things into my own hands eventually...

Another bug, one I saw a while ago but I guess hasn't been fixed yet. This is a captured barb city, yet it is getting "we yearn to return to the Motherland" unhappiness, which should not be happening. What has probably taken place is that a barb city has appeared somewhere on the globe, and since I'm "at war" with the barbarian civ, the game has added this unhappiness in the city. It too is going in the bug database.

With the French war behind me now, my civ continued building peacefully. Astronomy let me settle some islands that were previously unreachable, and since I was running 100% research at a profit, I saw no reason not to do so. This would probably be a good place to discuss the Organized trait. It is definitely not underpowered anymore; I think it definitely holds up to the other traits in terms of strength. Half price courthouses are quite strong, and on maps with a lot of water (like this one), the cheap lighthouses are a big deal too. The main feature of course is the free civics, and that's nothing to laugh at. This trait does NOTHING in the early turns of the game, but the longer it goes on the more powerful the Organized trait becomes. (It's almost like the opposite of the Creative trait, which is strong at the start and then drops off as time passes.) The ability to run whatever civics one pleases and never worry about the cost is great - I really like it. This was also the main reason I was able to run 100% research in this game; I estimate Organized was easily saving me 50+ gold each turn, without which I would have had to drop down to 90% science. This trait is subtle, but it can potentially be very strong. I would advise not changing this trait at all, we've got the balance dead on at the moment.

The Organized trait went through a lot of reworks during the testing period as it was often seen as one of the weakest traits. In this version, it was slightly different from its eventual function as it removed all civic costs completely instead of reducing their upkeep by 50%. It had the same basic gameplay pattern though, much stronger in the late game and much better when playing on higher difficulty.

As for the Civics themselves, I was running Representation/Bureaucracy/Emancipation/Free Market/Organized Religion. I'm not sure I approve of the latest change to Organized Religion; I really can't see many situations where I would ever want to run Theocracy now, whereas before I would sometimes go there for the building construction bonus. Now that bonus is at Organized Religion, AND I can still build missionaries without a monastery? That seems awful strong to me. I kind of liked the Religious civics the way they were before, but that's just me... I'll also mention here briefly that I love Representation; this is my new "favorite civic." Extra happiness in big cities really helps, and the +3 beakers for each specialist is a very big deal. Late in games, I'm running a lot of specialists to prevent my cities from growing into unhappiness, and Representation provides a nice boost. It's not overpowered, but it's the sort of thing that very devoted Civ4 players are going to be using a lot. And in the rare game that you can build the Statue of Liberty and run Mercantilism with Representation - look out.

In this game, I was beaten to the Statue of Liberty by 5 turns in 1695AD. I still had built no wonders up to that point. I got a message in 1705AD that I was #5 (of 7) in tech - there's TWO someones behind me? Yay!

My game might have played out peacefully from that point until the end, but when I got up to Railroads I found that there was no coal anywhere in my territory, or anywhere unsettled on the map. The only nearby source was... under the French city of Marseilles. Now I needed coal not to build railroads (although that would be nice), but to power my factories. I planned on going for a space race victory, and building all those parts without rails or factory power would take much too long for my liking. So war it would be - but war on my terms this time. I had no oil either, but I founded a city on an island to the east that secured a source of it, and from there I beelined on the tree to Combustion (to hook it up) and then along the top of the tree towards tanks. I built myself a mixed stack of infantry, artillery, and tanks and then declared war in 1874AD.

Susa of course fell immediately, and it was happy to be back as part of my territory. From there, I needed to capture Marseilles (which is off the bottom of the above picture) and then raze Lyons and Orleans to free up enough cultural space to be sure of holding Marseilles. I was curious to see how strong of a response Napoleon would be able to put up to my attack. Well, the AI was by no means weak and punched right back at me on the interturn:

It's a little hard to read (this red text just does NOT covert into .jpg format very well), but basically the message log shows that Napoleon attacked 18 times on the interturn. That's right, he attacked with EIGHTEEN units. And I was sneak-attacking him! He hit me with more on the following turn too, probably close to a dozen units. I was definitely impressed by the AI here. Even though I had a tech edge (infantry/artillery/tanks against rifles and cavs), I still almost lost Susa. I just did NOT expect the AI to hit back that hard. There's still some bad news here (tons of outdated units running around - I actually had tanks fight spears! ) but the AI really can put up a tough fight. Even with a war on my terms, it gave me fits.

Now there was one other problem with the way Napoleon fought this second war. He threw all of his units at me at once; so many, in fact, that after about 5 turns of war he literally stopped throwing units at me at all! I saw maybe three offensive units in the last 10 turns of the war. This isn't all bad news; if the AI is going to choose between dribbling in forces or throwing them all at the player at once, it should definitely choose to attack with them all at once. But once they were all dead, Napoleon was basically helpless. I think he may not have switched over to military production because I was the one who declared war. In any case, when the player is marching around and burning down core cities left and right, it would be nice if the AI could switch into "enemy at the gates" mode and beef up its defenses to prevent more cities from falling. Of course the problem is how to get the AI to recognize when this is actually the situation and not a trick by the player... Anyway, I saw a lot of good stuff from the AI here, it definitely fights a lot tougher than in Civ3. I'd just like to see it get even better before the project is done.

The aftermath of this war. Interestingly, the real winner from this war wasn't even me but Qin, whose score and borders both increased a lot. He even founded two new cities in the opened up space. We'll probably see a lot of this in Civ4; not that it's necessarily bad, just unusual. I now had the coal I needed, and that was what I had attacked for.

I noticed this problem in the late stages of the game. The AI civs were planting way too many of these junk cities that claimed no resources and would never amount to anything. Just look at that picture above! I don't even want to know what the AI is thinking here. There are more of these scattered around the globe too. These cities obviously were costing the AI maintenance costs and slowing down their research; the AI should definitely add island colonies, but they have to be able to pay for themselves and contribute something to the overall civ. These will never do either, and will only be a drain on the AI's cash reserves. The Civ4 AI has been doing this for quite some time now, and hopefully Soren can find a fix for it soon. The AI looks mighty "Civ3ish" in this kind of situation.

Finally I got down to the business of winning the game. Knowing that the UN victory still has "issues", I made the easy decision to go after the spaceship and see how well the AI fared in the space race. I stuck to the top of the tree and researched Rocketry as soon as possible so that I could get to work on the Apollo Program. I completed the wonder long before any of the AI civs and therefore got an insurmountable lead on the spaceship. Their tech lead had basically disappeared by this point as well, and I skipped a lot of techs I didn't need to be able to forge ahead on the spaceship ones. I thus got to Superconductor first and built the Internet (my first wonder!), which caught me up on everything I had missed anyway.

The AI civs spent too much time researching the bottom part of the modern-era tech tree. I've found in my admittedly limited late-game experience that the AI tends to favor the techs that lead to Broadway/Hollywood/etc. at the cost of the military and spaceship techs elsewhere on the tree. This could probably use a little rebalancing - or simply make the AI less predictable. A predictable AI will be easy to beat no matter what it does. I like the fact that the UN and the Apollo Program are at opposite ends of the tree; this makes the human have to choose one or the other in a tight game, instead of the no-brainer UN from Civ3. That has the potential for good stuff down the road. But the AI needs to be better at building the spaceship, because it wasn't even close in this game.

They don't have anything other than the casings done, and I'm already launching. And that's with their huge research and shield discounts on everything. At the very least, Rocketry has to be a higher priority tech so that the AI can get off its behind and get to work on the spaceship! Once that wonder is built, the AI seems to do OK, but it takes too long to get it built. That's my tentative first suggestion as to what to try, and I'll have to see how things go from there.

So I won my first Emperor game. It definitely wasn't easy, but I think the challenge level was just about right (though this was closer to Civ3's Deity than to its Emperor). I see lots and lots of positive things about Civ4, and I hope that we'll have enough time to clean up some of the bugs and other problems that are holding back an otherwise fantastic game.

This was my first win on Emperor which was definitely "high difficulty" for the testing group. Sirian and Aeson were about the only ones who played and won games on Emperor and no one was winning games on Deity (note that Immortal difficulty wasn't added until the expansions). This was therefore a really satisfying result and a sign that I was continuing to improve at Civ4's gameplay.