When we last left things, I was optimistic with my Durak-derived system of battle resolution, with the back-and-forth of a Durak bout (which I called a skirmish) capturing some of the dynamics I wanted in the game. There were some problems with it, and I hoped that I'd make some changes and it would button things up and get to a fun experience. Well, I'm here to report that I made the changes and they were an improvement, and the game still wasn't as fun as the last trick-taking version.

I played three games with the new setup, and the changes kind of worked. Kind of. There was some more dynamism with the switches from attacker to defender, and there was more thinking about "do I really want this to switch right now, I can't really follow-up on the attack?" And it's nice to have that decision, but it felt kind of bad as well. You shouldn't decide to stay on defense just because if you went on attack you'd lose even more. That decision worked very much against the fiction of the game. It should be a good thing to go on the offensive, and often the decision here was that it would be bad to do so.

I tried tweaking the rules around how you win skirmishes, letting the attacker play any card when the attacker role switched, but that just felt really artificial in the other direction. It just felt wrong. I finally realized something that I had been overlooking in my enthusiasm for trying a combat resolution systemt that wasn't trick-taking: Durak-style gameplay relies on larger hands, and the narrow hands that I was using for the battles was essentially incompatible. The interesting decisions in skirmishes should come from large hands and allow you to kind of predict where things would go, and those decisions just couldn't happen in a low hand count situation. It was a fundamental mismatch, and it was time I recognized that.

So, I punted. I switched back to trick-taking, with the modification that there was a strict hierarchy of suits: infantry was lower than cavalry which was lower than artillery. Scouts would basically work the same way (can discard to draw cards equal to their rank and keep one, or modify the rank of a played card and draw a card), commanders would be the lowest card of the next suit up. This wasn't exactly the same game as the last time I'd done trick-taking. The deck was different, and the trump rules were intended to better match the setting of the game. And you know what? It was excellent. It was fun again!

The lesson here, fundamentally, is that sometimes game design goes down a blind alley. I had an idea a while back, it led me to some exploration, I made a bunch of tweaks and larger changes, and the new system just didn't work. It was a mismatch for my design goals. But, some of the changes I made were a positive thing, I learned stuff about commanders, scouts, and the composition of the deck, and I could switch back to trick-taking in a better spot.

The focus, though, now turns to the Battle sites. If I'm not going to change trump based on the choice of Battle site (which I no longer want to do), there still needed to be something interesting about the different Battle sites, something that might push you to choosing one vs. another. It's time to go to history. My next design task is to go through every one of Napoleon's battle and do some quick analysis on them to see what made them interesting. How big were they? Were there factors in them that made them exceptions to the usual way that 18th/19th century battles went? When I complete that analysis, I can use that to shape the Battle card deck, and then maybe we'll really be somewhere.