Shinuyama – the ants go marching two by two…

[Baalz]

Nice analysis, and certainly more rigorous than what I did. I was thinking about why my conclusion differed from yours, and I think it's a factor that I didn't think to mention in the original guide - bakemono warriors are going to benefit disproportionally from more advanced deployment tactics. A blob set to hold and attack (with the benefit of an archer screen) is probably how most people would do this, but this is considerably suboptimal for bakemono warriors (well, any swarm troops) in situations where they are fighting lowish protection, human hitpoint troops (most of your test cases). Look at it this way - in a theoretically perfect deployment round one of melee would be each of your guys attacking somebody. That perfect case is never going to be achievable, but the closer you can get to it is a fairly good yardstick for how effective your deployment is. For example, if 80 bakemono warriors were able to each attack 40 barbarians on turn one, there probably wouldn't be any survivors to even rout, giving you a casualty free victory. In the same perfect situation Dai Bakemono are going to have to take a couple rounds of pummeling from the barbarians.

The most obvious factor is who attacks first? Holding and attacking in a blob from mid field seems to generally result in bakemono warriors defending first round of melee. Another factor is what percentage of your troops are on the front line? The percentage is much higher for Dai Bakemono vs bakemono warriors in a blob situation meaning the offense of the warriors is hampered. The warriors are going to perform much better if you can get a bit of flanking going on. How I generally deploy the bakemono warriors for indie expansion is (depending on what I'm attacking of course): split them up into two groups on each side of the battlefield set to hold and attack with a single guy forward of them set to guard commander (preferably a shielded indie troop). What will generally happen is a sandwiching as the indie forces chase the bait into the vice and disintegrate.

I found the warriors to be pretty brutal expansion units, though you're almost certainly right that Dai Bakemono are better blunt instruments and more user friendly.

[DragonRider]

I'll admit that I'm not very good at predicting troop movement to get the first attack. Under normal conditions, I do split up large blobs into multiple smaller squads to improve surface area, but I didn't do it this time because Bakemono Warriors have morale 9 and that means they rout like it's their job. The fewer of them you have in a squad, the more likely they are to do it, and having half of your army pack up and leave in the middle of a fight is a good way to put an end to your expansion plans. This is a problem I frequently had in my informal tests. I'd have a bunch of Dai Bakemono and 15-20 Bakemono Warriors behind them to cover their flanks or fill cracks in their formation, then three of the warriors would die, the rest would flip out and run away, and then the Dai Bakemono also panic and rout and then I would be screwed. Since routing is a very random thing I tried to minimize the odds of that happening for a more controlled environment, but maybe I'll try a few more runs with divided squads and see what happens. (Although this brings up another problem: like an idiot I created the map without max gold and production multipliers, so I had to wait a lot of turns to get enough gold and resources to set up all my armies, so my dominion is all over the place. That means I'm always fighting with 10 morale, instead of sometimes having to deal with 9 or even 8 like in a real game.)

One other potential wrinkle is that I only did endurance tests based on the path of least resistance. It may be that 80 Bakemono Warriors do better than 16 Dai Bakemono when their only job is to mass up and hit those individual provinces with 25 Cataphracts or whatever and attrition is less of a problem. That's another thing I should probably look at.