Last time, I talked about the success of providing players with visibility into their Reserve, and that it's time to start adding some real historical flavor into the game, an idea that I explored a little bit as one of my goals for the game. So I sat down and scribbled out some ideas for what a custom deck might look like. For a task like this, there's no way that I was going to get the mix right out of the gate. The only way to move forward is to just try something out and adjust based on how the game feels. My task, then, was to create a strawman and beat it up.

My first thought was that there should be three major suits, representing infantry, cavalry, and artillery. Those are the major components of armies of the time period, so it seemed like a reasonable thing to represent. I figured a few things would also be nice to capture: infantry is generally the weakest but most numerous card type, cavalry is next up the ladder with fewer but more powerful cards, and finally artillery should have the strongest cards but should be the least frequent. I got out some decks of cards and started stripping them and Frankensteining something together. I pulled out 2-10 of Clubs from three decks for the infantry (figuring that no face cards made sense for the weakest suit), 5-Q of Diamonds from two decks for cavalry, and 8-A of Hearts from one deck for artillery. I followed the Bridge heirarchy of suits here because it's internalized for me, so it's easy to remember the types of each suit. An observant reader might notice that the range of cards narrowed with each type, and my thinking there was that I wanted the range to be larger for infantry to capture all the crappy infantry units that might be present, but also sometimes asymmetry in itself can be interesting, so why not?

That left me with 27 infantry cards (2-10 x 3), 16 cavalry (5-Q x 2), and 7 infantry (8-A), for a total of 50 cards. I wanted the deck to be a litle larger than a standard 52-card pack, and I wanted there to be some cards representing commanders and scouts. So I pulled out the six jokers from the three decks that I was using, and then the 2 2s and 2 3s of Spades for the scouts. That was 60 cards, which was a nice round number, which seemed like a fine place to stop. The deck seemed promising: asymmetric suits, a little bit of special flavor, every card representing something from 18th/19th century warfare. As a starting point, it seemed totally reasonable.

I just needed some rules for what the commanders and scouts would do. Commanders seemed simple enough: I could make them always trump, but always the lowest trump. That would make them always reasonably useful, but not overwhelming. Good enough for now. The scouts were a little trickier. My first thought was to allow you to swap the Battle card, which would change the trump, representing choosing the site of the battle better. Seemed scouty? But I avoided that for two reasons. One, it was too much like the 3s from Fox in the Forest, and I didn't want to repeat too many ideas. Two, what happens to the size of the Battle? My next thought was to make scouts always lose a trick, but you could swap a card from your hard with the Reserve. That would be useful both to beef up your current hand but also if you had something good that wasn't useful this hand, you could save it for later.

The final element I wanted to change was how Battle sites were generated. The asymmetry in the deck would result in wonky Battle sites (what would a King Battle site even mean?), and I wanted Battle sites to be roughly equally common across the three suits. So I created a separate deck for Battle site cards using a different color back of cards. 2-8 of each of my suits provided 21 cards, and then I had an idea. I took 2-10 of the other suit for Battle sites, figuring those could represent a sort of no-trump. My idea was that for these Battles, the trick is won strictly by the highest ranking card, but you still have to follow suit. It would make the higher ranks of Cavalry and Artillery relevant, and would change things up from standard play. That gave me a 30 card Terrain deck, and I could test things out from there.

The game has now evolved enough that I wanted to provide an updated summary of the two-player rules for folks to follow along with the design discussions. As before, you can't really play with this summary, as it elides important details, but provides a framework to understand the decisions I'm describing.

Napoleon, Blown Apart is played with two decks, a 60 card Troop deck and 30 card Terrain deck. Deal 8 Terrain cards face up as Battle sites. Deal each player 6 Troop cards for their Reserve. For each Battle, deal each player 2 cards, and then the player with Initiative picks one Battle site. The number on the site is how many tricks will be played in the Battle, and the suit is the trump suit. For Battle sites of the extra suit, play is at no-trump. In all tricks, for a tied top card, the first card played wins. Deal each player cards to fill their hand up to the correct number of tricks, then each player in turn may discard cards from their hand to take an equal number from their Reserve, freely chosen. Tricks are played with normal trick-taking rules, where you must follow the lead suit, and the highest trump (if any) wins otherwise the highest card of the lead suit. For no-trump, you still must follow the lead, but the trick is won by the highest ranking card regardless of suit. Commanders are always trump, even in no-trump Battles, but are considered the lowest trump. A Spy always loses, but permits the player to exchange a card from hand with a card from their Reserve. If a player wins all the tricks in a Battle, they take the Terrain card to record their Rout. If any player ever has seven points in Terrain cards from Routs, they instantly win. Players otherwise score one point for each trick they take more than their opponent, with a bonus point for each Commander taken. If no player wins on Routs by the end of eight Battles, high score wins, but if the point leader loses a Rout in the final Battle, deal a new Terrain card and play another Battle until someone wins.

All right, that's dense, but should be good enough to give an idea of how the game is currently played. And enough talking, how did it go? I got things shuffled up and played the game against myself, and it didn't feel great. The suits felt strange, the commanders seemed too strong, there was a mismatch between the frequency of cards in the Troop deck and the frequency of things in the Terrain deck that felt off, and everything just seemed really awkward. I hadn't really changed that many rules, I'd mostly changed the composition of the deck, and it just threw everything off. But it was my first game, and I'd played the game quite a bit with a standard deck, so perhaps this was just a clash between the current game and my memories of the way things used to work. This actually happens fairly enough to me during game development, where I get some dissonance between versions.

I stopped that day and tried the next day, to let the new version marinate in my brain a little bit. And still, it felt bad. Another day, another couple games, and nothing was changing. Overall, I played the game with the new deck five times, and at the end, I still liked it less than the version I had been playing with a standard deck. The new deck was a failure, and it wasn't just mis-tuned. If it was in the right direction, but the suit counts were a little off or whatever, I'd still have had fun. But this just was a step back, so something needed a lot of adjusting.

That was actually an encouraging sign! It meant, among other things, that the game was originally fun enough that I could meaningfully detect a step backwards. For a game I had been exclusively playing solo against myself, that means that the previous version was actually pretty darn good. More than anything else about this step, this was an incredibly reassuring thing to realize. I'm still on the right track with the game, broadly, even if this deck is a mis-step. Dead ends happen!

The next thing that I realized is that part of the problem here was a mismatch between the diagetic meaning of the suits and the results in play. In particular, in an infantry battle, you would have infantry able to trump both cavalry and artillery, which feels very wrong. And given that infantry cards are the most common cards, you end up in this situation pretty often. But more broadly, the properties of the different suits are irrelevant 2/3s of the time (roughly) as they only come into play in the no-trump battles. And even there, they might not work properly. My intuition of how 18th century troops should work, born from reading a bunch of books about the Seven Years War and the Napoleonic wars, was utterly useless in thinking about the game. While I wasn't designing a simulation here, it shouldn't work against my intuition.

Beyond the problems of the game's trappings working against the mechanisms, the game was just a more predictable thing. While there was room for clever play in the previous version, the current one felt more rote. I think that's mostly a consequence of roughly half the deck being infantry, but it ended up feeling pretty same-y in practice.

It was time to figure something else out. There were, roughly speaking, two paths forward. The first was that I could change the way that Battle definitions worked. While I liked the sizing of the battles, the way that trump was defined could be changed around. Perhaps trump suits were always fixed, and there was something else about the Battles that changed conditions. I wanted there to still be a decision of consequence alongside the size of the battle, something that the first two dealt cards would condition, but there were potentially other options there that I could modify. Perhaps something relating to weather, terrain, or other factors. I could look through Napoleonic battles to pull out some of the factors and reflect them to the cards.

The second path was potentially more interesting. I love, obviously, trick-taking games. I've designed more than one, I met my wife playing Bridge, I've been playing trick-taking games longer than any other type of game. It's a warm hug for me, an incredibly comfortable genre. But does it need to be the center here? Roughly speaking, this game was a representation of a campaign, with the Battles being, uh, battles, and the individual tricks representing clashes within that batle. But the overall structure of campaign and battles could be preserved while replacing how those clashes are represented in the game. The more interesting path forward here was to swap out the basic trick-taking of the game for something else that would make use of number/suit cards.

I wanted to take that path. Partially because I don't want to be just known as a trick-taking guy, but also to stretch myself a little bit and get out of my comfort zone. So that's what I was going to explore. The first step here was to get out my trusty copy of David Parlett's Penguin Book of Card Games and read about a bunch of card games for inspiration. It worked for Fox, I assumed it would work here.

Next time, we'll explore what I came up with and see how I can push this game towards history. At least a little bit.