If you’re playing a grand strategy game and have a major loss, do you do what I do? Assuming available resources, the plan is often to max out the build queue, build an even bigger army, and then go after the force that just defeated us. Essentially this was the Roman Republic’s reaction following the losses at Trebia and Lake Trasimene. Rome raised a new army centered around eight legions, more than she had ever deployed in her history. By way of contrast, a Roman field army typically consisted of two legions under the command of a Consul. In extraordinary circumstances, two consular armies could combine to present four legions and accompanying support. Hannibal got eight and and the proceeded to wipe them out at Cannae, including one of the consuls.
Cannae is one of the “Epic Battles” in Field of Glory II. Playing it, as I recently did taking Hannibal’s forces, gives some insight into Field of Glory II scenario design.
The introductory text to the battle’s scenario explains Hannibal’s strategy. He knows he is outnumbered but also knows he is facing a newly-recruited army. Because, you are told, the Roman’s know the weakness in her green troops, they deploy them in a deeper-than-usual formation. Hannibal puts his strongest forces on his wings, particularly beefing up his left-flank cavalry. He also deploys his infantry in an inverted crescent formation, calculated to draw the Roman’s inward, allowing his lines to extend beyond the Roman flanks, despite his inferiority in numbers.
As the game begins and the battle lines come together, this is exactly what happens in Field of Glory II. The initial meeting of infantry is at the center of the line, with the advantage going to the Carthaginian veterans. This draws the Roman wings inward allowing the Carthaginian wings to flank them. The slaughter of the Roman army ensues*.
It would seem next-to-impossible that a generic AI would react in just this way, reproducing Rome’s loss. Particularly given the depth of the deployed forces, it would seem smarter to feed the rear lines into the weakening center while using superior numbers to try to outflank the Carthaginians. Note, I haven’t tried this strategy, nor am I speculating that it would work; it is just what I’d expect the game to attempt to do given the situation in the early turns. My point is that, in order to create a satisfying Battle of Cannae, the AI has to be nudged in the historically-accurate direction. Not that there is anything wrong with that.
What it highlights is the qualities that will go into a well-crafted historical scenario in this game system and that it means more than a well-researched order of battle. It also has strong implications for randomly generated battles for play against the AI and the single player campaigns which are made of those random battles. I commented on this before, specifically with respect to Hannibal, but it may apply across the board; a hard-coded campaign of chained historical battles is probably the only way to reproduce history.
*I note that the victory screen did not reflect that complete destruction of the Roman forces that history records. Field of Glory II presents an option to continue fighting after you’ve won “on points.” The way it is phrased, it implies that such continuation can allow for higher casualties but I don’t know if it works out that way. One of these days I’d like to experiment – is it possible to achieve the complete destruction of the enemy by refusing to allow the game to end until all enemies are removed from the board?