Most of you probably don't know this, but I have a lot of experience with Civ4 Always War. 
I have not played or written about any Always War games since Civ4 was released, but I have gone through no fewer than four different AW scenarios in the darkness of pre-release testing (three of them with Civ4 and one with Warlords). I even took part in the first-ever Always War test of the Civ4 AI back in version .62! Obviously a lot has changed since then, but I do think there are some lessons that I've picked up which can still apply to the current game. For example, one of the critical things to remember about Always War gameplay (and this applies to Multiplayer games as well) is that you don't necessarily have to conquer a civilization in order to defuse it as a threat. It's enough to park a unit or two outside an AI capital in many cases, which will prevent them from ever improving any tiles or founding any additional cities. I have no idea what the lay of the land looks like going into this game, but if there are any AI civs nearby, I plan to cripple them early on but not necessarily to worry about taking them out. That can wait until later, while I expand and build up my own civ.
Secondly, just because you CAN push forward doesn't always mean that it's wise to do so. With a combination of maintenance costs and impossible-to-remove war weariness, Civ4 makes this a particularly acute problem for Always War. Instead of a slow and agonizing push through a sea of AI longbows with maces and cats, it's often more efficient to sit back and just defend until you have superior technology on hand (*cough* cavalry *cough*) which can then tear through the backwards AIs in rapid fashion. This also has the advantage of virtually eliminating war weariness from the equasion, since it will not accrue while fighting in your own territory. My gameplan here will be based around this concept, since the last thing I want to do is be struggling with unhappiness that I can never get rid of, short of conquering civilizations.
Finally, don't try to be a hero at the start of the game. While Sirian's not likely to put an AI right on top of our heads here, it is a possibility (and there is an additional AI on this map, so it's slightly overloaded with civs). Much like Epic Four, I've seen more Always War games lost right at the outset from players getting too complacent than anything else. You can't muck around with religions or workers out of the gate, it's time to stand up some real units ASAP! 
With that in mind, let's look at how I opened the game.
The starting tile looked excellent, with cows/wheat/floodplains/hills in range. I saw absolutely no reason to move and so founded on the starting hill tile. First build was a warrior, since we began with zero military units and I wasn't about to take any chances! First research was Archery, to get some real units on hand. The starting scout went to the east and found a coast, hinting that we were on the eastern edge of the pangaea. A nice place to be! And with that, we're off:
I timed the growth of Berlin to size 2 to coincide with the production of a warrior. In a normal game, that would be a good recipe for then going on to a worker, but oh no, not here. This is an Always War game, so I head right onto archer next in 3700BC. This is what I would call a "safe" start, taking absolutely no chances with possible AIs in range. Someone certainly could get ahead of me here by pumping out an earlier worker, but I simply refused to take that kind of risk. I was running max food, however, so I don't think I was really hurting myself all that much (you lose out on the growth curve by delaying the worker, but it does complete faster when the capital is at a larger size, so it's not quite as bad as it sounds).
After Archery finished, I went onto Agriculture next for our wheat resource, followed by Animal Husbandry (naturally, having both prerequisites lowered the cost of AH). Buddhism FIDL in 3640BC, triggering the "Spain" alert. I suppose there are worse civs to be matched up against than Expansionist/Spiritual Izzy. Stone spotted in the west (although in the middle of a desert-ish region) which would be a significant priority for a future Pyramids build.
By the time I finished my Agriculture research in 3430BC, I was starting to wonder where the AIs were. Umm, I wasn't expecting quite THIS much of a safety cushion at the start! Those players who try to metagame these things, expecting that Sirian won't put an AI on top of the player at the start, will probably get ahead of me here, but them's the breaks. In retrospect, I guess I didn't need to go right onto an archer after the warrior.
Finally I get a war declaration on Gandhi out of the blue in 3370BC! 
But, err, where's the unit? I don't see anything, do you? (It turned out that the scout was two tiles east of me, and had moved past me in the interturn.) Well, the game is now officially afoot. It won't end until everyone else is dead, so we might as well get set in for the long haul.
Hinduism FIDL in 3340BC (yes, it was Gandhi). I wonder if anyone tried for that and missed - could be a rough start for someone if that happens! This was great news for me, as Gandhi was clearly nearby and had happily provided a religion for me to come and take over. Down the road, I plan to assimilate Hinduism and make use of it myself, if you know what I mean. 
Berlin finishes its archer in 3280BC and now finally begins its worker (it also hit size 3 on this turn, not by coincidence). The new archer does NOT fortify on defense, but heads out to cause trouble for Gandhi. I fully intend to fortify this guy right outside Delhi and make sure Gandhi never gets any tile improvements constructed. I will cripple him in his cradle or die trying! And a turn or two later, Gandhi foolishly exposes his scout to that same archer:
First blood! Down he goes! 
Too bad there's no honorable mention for first unit killed, because this has to be one of the earliest!
On the same turn, my scout finds not only the location of Delhi, but our second opponent in this brouhaha as well:
Tokugawa, not really someone I wanted to see. Aggressive civs are nasty foes in the early stages of the game, and he will build a lot of units too. Fortunately, locating Delhi this early on means that my archer will be able to zero in and apply the choke in textbook fashion. Depending on how close Japan turns out to be, there was the possibility of going after Toku as well (of course, as everyone who played this game knows, he was too far away to do much in that regard).
When I discovered Animal Husbandry in 3070BC (and man, these Noble techs sure are cheaper than I'm used to!), naturally the first thing to do was scope out the location of the horses. Hmm... one in the tundra south, another one a good distance to the west. I have to say, I like the scenario design here from Sirian. The capital is a very strong location, but the player is going to have to work a little bit to get the early resources. Things would be disappointing if there were horses right next to Berlin, and the player could simply beeline to Horseback Riding and crush everything with horse archers.
In other news, my scout was now up-close and personal at Delhi:
I fully expected the warrior in Delhi NOT to attack out of the city and kill my scout, and the AI did exactly that. (I'm not sure I've ever seen the AI attack out of a city if it only has one unit there guarding it.) This was the start of a long run of situations where the AI civs could have chosen to kill my scout with one of their units, but decided not to. I won't be mentioning the starting scout much, but he managed to reveal a huge amount of the pangaea before getting killed. If you watch the minimap, however, you'll have an idea of the vast and extremely valuable map info he brought back to me.
My archer heading for Delhi would end up killing another Japanese scout en route, but had to stop and heal up again after winning that victory. Meanwhile, I finished a worker at Berlin and built another archer, while researching The Wheel and starting Bronze Working. There was an Indian warrior poking around my capital, but I had the situation well under control. Finally, I was able to pounce on him on this turn:
My archer kills Gandhi's warrior, and then my warrior kills Alex's scout. Woohoo, another bloodbath! You can see that I've already met Alex and Roosevelt, and more war declarations are still to come. There were seven AI opponents; Peter, Isabella, and Saladin were the missing faces I hadn't contacted yet. Notice that Berlin is already on a settler; I plan to use this second city to grab either horses or copper, depending on where the locations turn out to be. Sending out a settler with only one archer for escort could be dangerous, but I wanted to get moving on the resources now, while I still had a grace period before the AI civs started showing up in force. And yes, this contradicted my earlier "safe" gameplan, but I never said all my actions make perfect sense. 
Meanwhile, the archer I produced in 3280BC finally made it to Delhi, and just in time to cripple Gandhi forever:
Gandhi has one tile improvement, a mine on his gold resource, but that's the only one he will ever get. I pillaged it with my archer and then plopped my archer on the forest tile next to the city. For the rest of the game, he will never accomplish anything with his workers, just sit in his capital basically waiting for death to come. Now I can't actually TAKE the city anytime soon, but the beautiful thing is that I don't have to. Gandhi has been completely neutralized by the parking of this single archer. I will get around to finishing him off eventually, but for all intents and purposes he's basically done already.
A couple turns later, my scout finds the Arabian capital on a peninsula west of India:
One warrior defending his entire civ. My kingdom for an axeman! 
Of course, we have to do this the hard way, and so I won't be back for several thousand years. Good to know where Sal is located, of course.
Once I discover Bronze Working in 2320BC and can see the copper locations, the info my scout has brought back is enough to draw up a dotmap:
Civ4's staggered expansion period often makes dotmaps considerably less important than they were in Civ3, but with all the space here I thought it would be justified. There is a huge patch of desert to the west of this screenshot, so I have no intentions of settling past what you see here. I am going to draw an imaginary line in the desert and make no move past it, while instead concentrating on building up everything to the east of it (more on this later). I already consider Delhi to be a part of my civ (it's only a matter of time...) so it was included in the dotmap.
Where to put the second city? I'm sure there's many possible locations, but with my obsession for efficient land use, I chose the purple dot you see just to the west of Berlin. The more obvious spot would be one tile north on the desert hill, but in fact that location only gains three useless desert tiles (the river health bonus is irrelevant for an Expansionist civ on Noble). The spot I picked also allowed me to reclaim an otherwise useless desert tile. Working off of that spot, yellow dot in the north and green dot in the west then became the next obvious locations. Yellow dot would lock down stone and have fish for food, while green dot had horses, cattle, and rice, although the land otherwise was poor. These three locations I judged to be most important, due to the resources they obtained.
Everything else was more "long term" than immediate priority. The capital's otherwise fantastic location wasted some water tiles to the east, so I planned to squeeze in two cities there to work the seas. The red dot in the north looked particularly good, with clams there (and it turned out to have a fish resource too, outstanding!) Blue dot in the south was founded ON the horses, the only way to get the whale tile in range in a reasonable time frame. Light blue dot is another coastal city development project, but at least it has sheep to help out a bit. Brown dot has serious overlap, but has to go there to get the wheat tile in range. Lavender dot is a weak location, but grabs silks. I would eventually put another city in the unseen southern fog, because there was marble down there by the crabs. The only place I wasn't sure what to do was in the north by the gems, but I just resolved to work that out at a later date.
All told, what a marvelous scenario we've got here from Sirian. There's some breathing room from the AIs, and both copper and horses nearby, but the player has to work for them. If the player techs all the way to Iron Working, there was iron at Berlin, however (a nice saving grace for those struggling). Stone and marble are out there too, but the player will have to contest with Gandhi for the former and do some major expansion to get the latter in the southern wastelands. Very few happiness resources in range pre-Calendar, so clearly some work to do in that regard as well. The Arid terrain is rugged, but smart city locations and worker actions can reap great value from it. All in all, very well crafted. Of course, Sirian will probably post here and say it was all done by chance, but if so, the map generator rolled a rather nice start for us here. 
Back to the narrative... After finishing Bronze Working, I researched Pottery, Mysticism (to open up Stonehenge) and then Writing. I built another worker after the settler, then snuck in a barracks and started Stonehenge (it would be insane not to build that with non-Creative, Industrious Bismarck). I founded Hamburg in the spot I wanted, but while hooking up the copper, Gandhi's final exploring warriors, produced prior to choking his civ, showed up to give me grief:
This is not good. One warrior here is no threat, but two of them could be a serious problem. Hopefully they will go after Hamburg itself (which is in no danger) and ignore my copper mine, which I just completed but haven't hooked up yet.
No such luck. There goes 8 worker turns mining that copper down the tubes. 
This is the risk you run when you go light on defense, having critical stuff get pillaged on you. I actually killed one of the warriors with my archer, as you can see, but I wasn't able to get them both. This is a good lesson on why you should defend frontline cities with two units! (And now Sal has a unit in the area too, good grief!)
I swapped Berlin onto an archer, built that in 3 turns, then used the new bowman to clobber Gandhi's warrior. Things had been tense for a minute or two, but by 1480BC things were back to normal again:
Losing the progress on the copper mine was a setback, but fortunately not a critical one. I took away the lesson to be more careful in the future. Fortunately, Gandhi was basically crippled by now, and that was the last attack of any kind I ever saw out of him. What WAS he doing during this time? Happily founding more religions while doing his OCC thing: 
Delhi has become the HinJew holy city. Wow! When I take that away from him, I'll have a virtual monopoly on world religions. Of course Izzy will still spread Buddhism around, but she'll have to work for it. Denying AI civs access to religion definitely slows them down and hurts them.
Back at home, I build Stonehenge in Berlin in 1360BC, then proceed to go archer -> settler. These turns are quiet times, with few AI units appearing. I continue to build up my civ in peace, focusing on expansion while slowly accumulating some axes to take care of Gandhi. Here's what things looked like shortly after 1000BC:
Munich (green dot) has just been founded, which I intend to be the cornerstone of my defenses for ages to come. I already have a road to it halfway completed, and once its borders pop from Stonehenge, I'll have horses as well. Gonna take a lot of worker turns to improve that site, but it will be pretty good once I do. Notice that between my city placements and outlying archers, I have plenty of warning of any incoming AI units. Just like in Multiplayer, knowing what's coming ahead of time is one of the best ways to stay ahead of the game in Always War. And yes, my cities have been built in a straight line up the river valley towards India. When I take Delhi, I intend to have a road network already in place. I'm sure some people will rush Gandhi and take him out around 2000BC, but I like doing things more methodically. 
As expected, I got the Oracle before anyone else:
I'm sure that a lot of people will choose to go the Civil Service route with the Oracle in this game, and that may very well prove to be a strong strategy, but my goal all along was to take Metal Casting with the Oracle. Double-speed forges in the hands of an Industrious civ are very, very dangerous, and I wanted to get that into play at an early date. Cheap forges became one of my very first builds in pretty much every city, and I got a ton of value out of the early tech here. It's entirely possible that others got even more out of early Bureaucracy civic, but I'm happy with how I played things here.
Much to my surprise, Gandhi did eventually send out a settler with a pair of archers for escort, and with only my one archer on hand, I couldn't do anything about it. I was able to cut off and isolate one straggler, however:
Needless to say, that guy was toast. (City Raider II axe, yay!) Gandhi also continued to prove he was a moron by attacking OUT of Delhi with archers against my Cover archer fortified in the forest. 
Right before I went to attack Delhi, however, Gandhi must have passed the 750 culture mark, because his city went from 40% cultural defenses to 60%. Eep! That turned a romp into a slaughterfest of my soldiers. My axes all died, but they redlined the last defender, and my choking archer was on the spot to clean up the mess:
JUST got the city by the skin of my teeth. Whew! I guess there were some advantages to attacking earlier after all! 
With the capture of Delhi and the virtual destruction of Gandhi (he did have one more city, but it was irrelevant), the first part of the game was complete. I now planned to turtle up behind impregnable walls of defense for a long time while building up an unparalleled research and production machine. Let them come to us! They shall reap that most bitter of harvests know as "war weariness." 
