Battle Merchants is a strategy game for 2-4 players where four fantasy races (Elves, Dwarves, Orcs, and Hobgoblins) are at war. None of the players really cares which race wins the war, because you're just selling them weapons. The richest player at the end of the game wins.
So where did this game come from? One of my earliest designs was an auction game called Wag the Wolf. It turned out to be too weird and complicated for what it actually was, but it did have this really cool auction mechanism that I was dying to put into another game.
At the same time, I had this idea to have an economic game based on a military-industrial complex. I didn't want to make a preachy game; I just thought it would be gleefully amoral fun to have a game where players perpetuated a war by selling weapons.
I didn't want to set the game in the present-day. That would have been less fun. I decided to set it in the future instead, and have the weapons be militarized robots. Everybody loves robots, right?
The first working title was "Warmachines". I knew that there was a miniatures game called "War Machine", but I couldn't resist the double meaning. And anyway, I knew that the game was going to change like crazy from my first draft. And man, was I right...
After about a year of tweaking, I had an auction game where players built and sold war robots. The game was seven rounds long. Each round, there was a bid phase where players auctioned for turn order, then a tech phase where players upgraded their tech abilities or drew special power cards, then a build phase where players built their bots, and then a sell phase where players sold their bots. Then combat was resolved, and we went onto the next round.
Combat in the earliest version of the game was very abstract. There were five levels of bots. In battle, the bots at the lowest level of each side would be destroyed. Surviving bots earned VP for their manufacturers.
There were four nations at war, very cleverly named North, South, East, and West. Each one had a line of spaces that demanded a specific kind of bot. There was a chip between the spaces of each nation. When a chip was surrounded by two bots, the chip went to the middle of the board. You could not sell to any other spaces until they were unlocked by the chip before them moving to the middle of the board. This meant demand on the board was always changing.
Amazingly, this chips-as-demand mechanism made it almost completely unchanged into the final version of the game. Just about everything else changed.
At the end of the round, players would put all chips from the middle of the board into a bag, and draw a few of them out to determine which regions went to war. That way, you were never quite sure where battles were going to happen. Players kept track of how many battles each nation lost. The more battles a nation lost, the more they'd pay for a new weapon. At the end of the game, the player who had the most surviving bots in the nation that lost the least got a VP bonus.
I'll tell you a little secret about me as a game designer, and I wouldn't be surprised if it applied to other game designers: I'm always convinced that I know what the central, critical mechanism to my game is early on, and I'm always dead wrong. I would hang on to this "draw only a few chips out of the bag" mechanism for much longer than I needed to.
The first big change I made was in response to an obvious bit of feedback: Why was the winner decided with VP? Wouldn't an arms merchant care more about money? After some thought, I took out VP and made most money the deciding factor in winning the game. It's a strange change, because in a way, there's still VP; it's just that each dollar converts to 1 VP at the end of the game. More importantly, every time a player spends money, she's spending VP. This is a big change, but thematically, an important one.
I wanted a little uncertainty and risk in the auction. Wag the Wolf had an almost-worth-it mechanism where money cards had critical information printed on them, so every time you spent a large amount of money, you revealed information related to endgame scoring. This created some risky situations where you had to judge whether it was worth it to let everyone else know this one bit of information if you wanted to win an auction.
Warmachines didn't have this mechanism, so at first, I put all the money in hidden cards. You drew cards as income, but since the cards had different values printed on them, you never knew exactly how much you were going to make. My playtesters hated this, and suggested making the money open, and staggering the money values on the auction track instead. This meant that players outbidding other players had to spend extra money, which seemed slightly more interesting. But speaking of interesting...
The game, now called "Battle Factory," got to a point where there was an interesting auction, and an interesting set of mechanism related to building and selling robots. Playtesters started telling me they were more interested in the second part of the game than the second one. It was pretty clear what was happening: the game was rejecting my beloved auction mechanism like a bad donor organ.
I made the decision to try the game without an auction. I tore apart the phased structure of the turns as well. Now a player had to choose which action to pick in a turn. He could choose to improve tech, build bots, sell a bot, or draw a power card. Players could build three bots in a turn, but only sell one (another thing that made it to the final game). The game lasted over the course of five steps. Each step ended when the third demand chip moved to the middle of the board.
Amazingly, the game played much, much more smoothly. It felt much tighter thematically, although I had a lot to tweak.
As for the auction? I built another game around it. It floated around without a theme for a few years (in fact, its early theme was that it had no theme), until I applied a theme of TV networks buying programming. Things went well for that game until my playtesters started telling me that the auction wasn't working for that game, either. But that's a story for another time...
I wasn't crazy about the game's name, so my friend Brett Myers suggested the brilliant name Pax Robotica. I loved it, and so did my playtesters. Seth Jaffee suggested a more direct method of combat resoltuion. Instead of the lowest bot losing, how about each column of bot fight each other? I hesitated, hemmed, and hawed, but eventually tried it.
Turns out Seth is a smart guy. The direct combat mechanism made battles much more interesting. Now every time you pulled a chip from the bag, the chip had a number saying exactly which pair of bots fought. This seemed a lot more interesting, especially when I ignored my playtesters who asked for more battles. Silly Gil. Will I ever learn?
It was about this time that I realized that five levels of bots was too granular. I dropped it down to three levels.
I'd tell people, "Hey, do you want to play this game of mine? It's about building and selling fighting robots." They'd say, "Sure, I'd love to play your game about fighting robots."
See the difference in those two sentences? I saw that when people sat down to play my game, they were a little confused. Where were the fighting robots? Why was combat so abstract, even with the new change? Couldn't the game be only about fighting robots?
It took me a few years, but I finally realized that I needed to make a change. The theme was not working. Players were getting an economic game where they were expecting a wargame.
I have a personal vow that I will never design a game about city-building, castle building, or bloodless colonization. I have that vow because for one playtest, this was a city-building game. It was awful.
I don't remember where I got the idea for setting the game in a fantasy world instead of a futuristic world, but the more I thought about it, the more it seemed to work.
So: retheme! The game was now called Sword Merchants. The four different kinds of bots were now four different kinds of weapons. The four different nations were now four races. I printed out all-new components and gave it a spin. After a few playtests, it was clear that players weren't as confused as they were before. They knew the kind of game they were getting into. Success, finally?
At this point, I had dropped my other game design projects to focus on Sword Merchants exclusively. A bunch of critical tweaks happened.
I'd mentioned that nations (now races) spent more money for bots/weapons based on how many battles this lost, and that the player who had the most surviving bots in the "best" nation (the one that lost the least) got an endgame bonus. This part of the game was quite opaque. It was very difficult to divine whether a sale you were making was actually going to help you here or not. The mechanism, while thematically sound, was too many steps from what you actually did.
So I tried something else. When you sold to a race, you'd get a bonus from the race opposing it. Thematically, they were jealous of the other race, and would give you more money to even out the race. Mechanically, this encouraged you to spread out across the board. If I offered a tile from the same race you sold to, I figured players would just always sell to the same race (what we game designers call a "positive feedback mechanism" - yes, we are so down on ourselves, we view positive feedback as a potentially bad thing).
This worked okay, but was clearly confusing. One playtester suggested that if I was worried about players always selling to the same race, why not limit the reward tiles to one per race per season? I think my jaw dropped a few feet when I heard this suggestion. It was so ridiculously simple, and yet it forced players to sell to different races. That was a huge leap forward.
It was about here that I put the Bonus Craft powers into the game also. If you reached a certain level of Craft (formerly called Tech) in a weapon type, you would earn a special power. There were four weapons and four types of action, so I linked each weapon to an action boost. One craft mastery would allow you to sell two weapons in a turn, another let you collect extra power cards, another would let you collect extra Craft cards, and the final one would let you forge and sell in the same turn.
I showed this to Michael Keller, designer of City Hall and Captains of Industry, a couple of times, and at one point he had enough and rolled his eyes. "Gil, one of these four bonuses is completely broken."
"Really? This one, right?"
"No, not that one."
"Okay, this one?"
"Gil, are you sure you're a game designer?"
I was down to two bonus powers, and my credibility was at stake. "Okay, it's gotta be this one."
I'd never seen Michael so exasperated. I had selected all the bonus craft powers, except the one that let you forge and sell in the same turn. "Gil, it's that last one you didn't pick. That craft completely breaks the tempo of the game. It's much more powerful than the others."
I didn't believe him, so I asked him to focus on it in our playtest and beat me with it. He did, doubling my score.
Another big improvement was when my playtesters convinced me that they wanted more battles in the game. Remember, I had pulled demand chips going into a bag, and only a limited number of chips coming out to determine where battles happened. Around this time, someone suggested: why not cut the bag out of the game and just have everybody fight?
So much of game design is a striving for simplicity. This was another example. I had a bunch of rules around this bag, and it turns out I never needed them. It was so much easier, thematic, and fun to have all the weapons fight at the end of the game.
Remember how I had three levels of bots/weapons? One tester pointed out that no one ever bought the middle level. He suggested playing without it.
We did, and I was shocked again. Another needless part of the game gone! There were now two levels of weapon. I used to refer to them by number, but now I could give them names: "Standard" and "Vorpal".
Also, remember the bonus craft dilemma I had, where one of the powers was much stronger than the others? I tried to nerf it by forcing a player to spend money every time she used the power, but a playtester gave me a better idea: Just make all four craft bonuses that power, and link them to that one weapon. So Sword mastery would mean you could forge and sell a Sword in the same turn.
It hurt losing that asymmetry from the game, but it was for the greater good. It was much easier to balance, and I could still keep crazy powers in the power card deck.
Funny, at Protospiel 2012, I had no intention of pitching any games. But one playtester enjoyed the game so much, he insisted I show it to James from Minion Games. Thankfully, James also enjoyed the game, and in a few months, I had a contract to sign.
I'm really happy to have Minion putting this game out. James is a very good publisher who knows how to take care of his fans.
It's been a long journey from Warmachine to Battle Merchants. I started designing this game in August of 2007 using an auction mechanism I'd started working on in 2003.
I have one other published game, my word game Prolix. That was a difficult game to design. It has a real-time interrupting mechanism that took a few years to tweak. Also, being a word game, I couldn't fix problems by adding additional rules; I'd wind up with a 60-page rulebook for a family game! And changing mechanisms for a relatively simple game is difficult because of emergent complexity. Everything I fixed over here broke something over there. When I fixed that thing over there, that would unbalance something else down the line.
A few times, I wondered if I should just shelve Prolix and work on something else. I would tell people that my next game would be a crunchy economic game with plenty of rules. Thankfully, Prolix turned out really well; I'm extremely proud of it.
It's been six years that I've been working on Battle Merchants, but I never got the helpless feeling I had with Prolix. I knew there was something here, and it was just a matter of time until I hit upon the correct rules balance. I'm delighted with how this game turned out, and I hope you all enjoy it as well!