Bret Devereaux

One thing that strikes me is how it seems like a deliberate shift from a more difficult design problem to an easier one. Older TD games were about defense-in-depth via mazing, but the TaB-likes generally focus on hard-stop perimeter defenses.

Likewise, the aim of the early Stronghold games (back when they worked) was also complex defense in depth. Likewise the sieges in 2011's Shogun II, built around multiple zones and choke-points (a design CA keeps trying; we'll see if Total Warhammer III finally nails it).

I think the core problem is that DiD primarily acts against fatigue, morale and communication delays/lack of information. Games still haven't mastered the first two, and the last is a complete non factor in all but the most niche simulationist games.

Morale especially!!! "We are fighting off the siege of our capital city where we and our families live" should have much higher morale than "The player is 3 cities away from full map completion!"

In their defense, the sieges in Total War games on normal difficulty actually do factor in the morale-motivation mismatch. But no one plays on lower difficulties and on VHard, morale is a non-factor for the AI due to huge bonuses, until the 'Army Losses' penalty kicks in.

It's actually quite sad and frustrating: so much of CA's sometimes brilliant design work actually breaks when played on Hard/VHard in their games, but for experienced players, normal is comically easy, creating a situation where the game becomes LESS strategically rich over time.

I guess you would really need the AI to have better macro economy management, I'm not sure if this is hard, or if CA tried this and it wasn't fun. I play on Medium because I don't have much time and I like winning!