malekithe said:
2:1 with cannons failed against longbows??!!?
Longbows get a bonus defending cities, and even more of a bonus if that city is on a hill. When I go to gunpowder, the longbows I have defending cities on hills are the ones I replace last.
When taking a city my general technique is:
1) First stack: obsolete units I was going to delete anyway. These are "catapult fodder" for the cats and mounts and crossbows, et al., that an AI typically throws at the first stack to go adjacent to their city. 4 or 5 obsolete units in the stack means they'll burn cats to kill my units.
2) Second stack: more obsolete units, as the AI typically keeps some cats in reserve and won't expend all of them on the first stack to come along. 4 or 5 more spearmen, archer, axemen, etc.
3) The REAL stack of doom: at least one of each contemporary unit for the time period I'm in, and a generous overflow of cats, cats, cats, and more cats (or cannons, cannons, and more cannons!) In addition I have an extra 3 to 4 macemen, 1 or 2 with City Raider, 1 or 2 with "stars" (for field combat), and 1 that acts primarily as a healer unit and for initial city defense (until a longbow force comes in for permanent garrison). Lastly, I also include 2 or 3 obsolete swordsmen who usually have City Raider II from barbarian conquests and what-not.
Stacks 1 and 2, their main purpose is to tempt the defender into a counterattack and to burn catapults as much as possible. Even old school spearmen will also be able to kill off whatever horse archers and some of the knights they throw out at me. If they do survive the counterattack onslaught, and they often do, they simply guard the high ground or forest overlooking the city so that the "real" stack won't have to fight their way into the attacking position.
When the "real" stack shows up, my force of siege weapons typically wipe out the city's defensive bonuses in one or two turns. That's how many I bring. It's important not to be parked there too long because time is on the side of the defender, and if you're on offense, it's conquer fast or die. For that reason, if your stacks encounter lone intercept units on the road to the city you're going to attack, just avoid them, move around them, don't even acknowledge their existence (but choose well-defended terrain as you move around them).
After the city's defenses are at zero, I employ a Darwinian test of my obsolete swordsmen: they go in as the spearhead of the raid, and if they survive the attack, they're destined for promo to macemen or grenadiers after the city is taken. If they die, they are purged from the array of units costing me money.
The swordsmen attack is to wear down the strength of the defender's most powerful garrison units, which are usually longbowmen. What this means is that when I unleash my catapults into a direct attack on the city, it won't be WHOLLY suicidal, and a good number of them have a chance to withdraw from combat or even win a few, and I start from the least experienced cat, and hang back with my elite, highly-promoted catapults, who only begin THEIR attack when the odds favor their winning. After all the catapults have had their go at it, if the city is still defended, the remaining macemen, mounted units, pikemen, etc., get involved, whatever's the best matchup for the strongest defender unit at the time.
If I'm down to 1 or 2 healthy units and the city has more than that number of defenders at full strength, I stop the attack and leave them fortified, to retain that siege position for a healing session and to bring in reinforcements. It never bodes well if I have to do that though, and under ideal conditions I have multiple sets of city-raid stacks: usually while one is healing, the other is busy attacking somewhere.
