http://www.4shared.com/file/IBPJx2-F/4p_-_Loris_River_replay.html
This battle took place on the Loris River map, which is actually built for 4 players (with two teams, one in each corner). I tried to do it for 2-player, just me and MysteryExclamation ("ME").
When I started, I realized this map was a bit big. I sent off two spearmen to grab things in the northern part of the map, with... some success. My main interest, however, was to rush down to ME's base and start harassing her immediately. With luck, she would pick a fight with me with the front-line close to her base. Since it's a square map with us in either corner, some villages are quite out of the way.
I had a good success with my Horseman-Cavalry tag-team. I had picked the two of them because the Loyalist Cavalry has 20% resistance to blade, and 30% to impact - only orcish archers and goblin spearmen can hit their 20% weakness to pierce. Plus they were fast, which was the point of this whole endeavor. I killed a bunch of people, and did pretty well in the south. I was gradually driven back, but on that front I got a Lancer, and had killed a good deal of her units.
On the topside, my idea of grabbing everything with the spear-men seemed kind of silly. The map is wide, but it's not incredibly long. I got into a continuous fight with ME's units up there, doing so-so. Nothing special.
The Lancer I got in the southern campaign was... fun. It naturally deals 12-3 damage, which is not bad, but then it charges, so whenever it attacks both in and the enemy deal double damage. So it's basically 24-3. During the day, that goes up to 30-3 damage. That's right, he can deal up to 90 damage in a turn, and can one-hit anything with less than 30 life.
He killed a unit almost every turn. Watch his spree in the replay. He just wanders around sticking his lance in anything that comes his way. Bad-ass. I prefer Lancers to Knights for multiplayer because while a Lancer is more likely to die, he will probably kill more units in his short life than a Knight will an entire game.
Of course, this also means that other units get less experience, right? In a campaign, this is fairly important. However in multi-player, killing a few units means that your level 1 units will both survive better, and have a better chance of ganging up on the remaining enemy units. I think it's fair, though having your level 1s deal the finishing blow is generally preferable.
After a fairly even battle on both fronts, I started to win a bit in the southern front, and eager to grab villages (I was below-half of the villages most of the game) I went and grabbed a bunch.
...silly me. As it turns out ME had a decent-sized assault-force ready, and picked off my units the same way I like to pick off her units. I was still massively ahead in terms of kills, but this had narrowed things a lot.
My plan at this point was to lose (or at least only mildly threaten) the lower area, and focus on wiping out the northern area, where I was on the verge of winning (I had done a bit of imbalance, sending more troops there when it looked like I was winning on the southern area). The next day, I was going to wipe over that side of the map, then cut downwards between ME's keep and her troops in the field.
This would have been a feint, of course. The Lancer (who had been sent back for repairs at this point) would be riding a route that kept him out of enemy sight-range. He's got a speed of 10, so he can do that pretty well. One the battle was drawn to between ME's leader and the rest of her troops, I would have the Lancer run in out of sight range and lance her leader.
Troll has 55 health and a 20% resistance to pierce. A day-time blow with the lancer would deal 24 damage. More importantly, her troll leader has 40% defence in a castle. I had a 64.8% chance to hit at least twice, with a 21% chance to kill her leader out-right. If I only hit twice, then I would use another horseman (84% chance to hit once and kill) or any other unit I could get there - possibly fencers (82% chance to kill) to deal the last 7 damage.
Otherwise, it would just be a big battle where I could reclaim the villages of the south back while my units did their best to take equal amounts of northerners with them. The units to support them would be streaming in through the lower continent, and taking the villages as they go (slowing the first re-enforcements to create a proper "wave", not just a stream of attackers).
By the time the reinforcements came in, I would control the vast majority of the villages on the map, and be able to send in enough troops to at least keep MysteryExclamation from re-expanding, and then eventually crush her. Unless she fought to keep the southern continent, in which case I would have a very good chance of ganking her leader in the meantime.
Of course, then ME went to get fitted for a corset, and since she had forgotten to keep her computer plugged in, it ran out of battery and died. Which killed the game. Thus the replay ends half-way.
Lame.
Other general notes: Starting with 100 gold meant that I didn't have many units at the start. Neither of us did, which made Loris River very, very big. This emptiness is what made my cavalry tactics very potent. However, starting with 200 gold (as if we each were two players) might be overall better. It would certainly make for an epic fight.
I also did an over-all bad job of xp-management during this fight. I let a lot of experienced units die. (Plus the pseudo-loss from Soddreoc killing everyone. He kills 9 units in the game, 3 to become a lancer, and 6 afterwards. I'm not sad about his performance.)
There's an interesting problem with having to claim land on a map like this. I often lose units to going and grabbing villages. I should be more carefully about this.
Anyway, enjoy the replay. If you have questions about particular decisions on my (Aelaris's) part, feel free to ask.