The most logical thematically was that you'd stop at various towns and villages along the way and spend time rebuilding and reconfiguring your invention, and this was how things worked in early versions. However, this paradigm had two basic issues I never found elegant solutions for. The first was how to deal with players breaking down between towns. Would you go straight back to the previous town, or have to drag yourself to the next? Both were pretty dreary options. The other issue was that in a racing game, spending turns not getting to race feels like work. Even if the decisions are interesting, it's still like “eat your peas and next turn you can have dessert.”
The other idea for distributing parts I never got to work properly is having players' destroyed parts create card-draw tokens which can be picked up by players behind them. I still think this is really thematic and adds a cool catch-up dynamic, it just maybe belongs in a different game. If the draw tokens were claimed by being landed on, then a player directly behind someone playing very recklessly would get a huge windfall, so to be fair they probably needed to stay on the board, benefiting every player behind the one who spawned them. But this made the rate of card-draw very swingy and nearly impossible to balance because I couldn't (or rather didn't want to) directly control the rate at which players lost parts. The mechanic also provided no cards to the player in first, so it didn't fully address my original need.
It became pretty clear that, for things to stay exciting and strategic, everyone had to somehow be getting new parts constantly.
I knew I wanted players to be able to remove dice from their inventions in case they couldn't manage to get new parts that synergized, but I also knew I wanted it to be more efficient to add new parts so as to encourage players to keep building and changing their invention. Having new part acquisition and dice removal hinge on the same resource would let players choose how heavily they wanted to lean on one or the other, and a player with a big awesome invention would have tons of dice to remove, thereby pulling their resources away from the n00bs with small inventions and few dice who thus needed new parts the most. So players used “time” every round to bid on the right to select new cards first, and any time they had left over was used to remove dice.
The auction worked great, and it did exactly what I wanted it to. The problem was that it slowed the pace of the game to a crawl, and felt especially out of place because the rest of the game was resolved simultaneously. This was the hardest change for me, but eventually I was convinced (largely by Gavan) that it had to go.
What would take its place that could do the things I needed in a hastier manner? It turned out that the solution was more simultaneity: a card draft! Originally part cards could be discarded during the draft to either receive or remove dice, presenting a difficult choice between adding parts, renewing parts, and powering them. The final change was having cards be discardable for cogs, which could be spent to remove dice, rather than directly removing dice through discarding. Both functioned nearly the same, and the primary reason (aside from adding a bit more planning) was that playtesters frequently didn't realize that they could discard cards during the draft to remove dice, whereas it’s harder to miss a whole separate phase devoted to dice removal.
The lesson I learned is that occasionally a purposefully clunkier and less elegant turn-structure improves everything for everybody.
The vast majority of boardgames out there are turn based, and resolving stuff in turn order has some notable advantages: it's more familiar, it's easier to learn a game when you are watching other players interact with the mechanics, and when you do something cool everyone gets to (has to) watch you. The problem is that if I made the current Steampunk Rally mechanics turn-based, the game would be maybe three hours long with crushing amounts of down-time.
The damnatory term “multiplayer solitaire” has been leveled against many euro-style games over the years, but I see this not as an indictment but rather an opportunity. If a certain portion of the game does not involve other players, why must they wait for me? Steampunk Rally has interactive elements (counter-drafting, offensive boost cards, watching in fear as an opponent creeps past you), but when it comes to operating your lovingly-crafted invention by placing dice on it, it's all between you and the machine. We experimented with several turn-based round structures, and eventually Gavan conceded that these didn't do a whole lot except slow the game down. Through the development process I learned the value of traditional turn order as well, and I'd suggest resolving the racing phase in turns for the first couple rounds if you're learning the game for the first time. But by resolving simultaneously, Steampunk Rally crams a ton of game into a short playtime with virtually no downtime, where everything resolves at the speed of the slowest player. And I think, whether or not you prefer it to more traditional turn resolution, it will feel different from any other worker-placement game you've ever played.
But even as action cards, these powers needed to be constrained. It wouldn't feel good to resolve your turn carefully budgeting sustainable damage only to have someone play a weapon card right before resolving damage and incur catastrophic losses. Weapons were meant to add another challenge for players to deal with, not random screwage. So I made these action cards only playable before the racing phase, right after the draft, but this introduced other issues. If players could play them in any order, there were frequently timing issues, so the opportunity to play action cards was resolved in turn-order, which felt incredibly clunky. And since action cards mostly affected multiple opponents (to limit kingmaking), the game changed radically with different player counts since this affected the rate at which these cards were played.
I experimented with several variant ways of incorporating these interactive effects, at one point even having public events that players could claim or trigger by blind-bidding cogs. (As I said, I like auctions.) It turned out the solution was much simpler: I just had to reduce the number of cards that affected other players, replacing them with Gavan's suggestion of cards that give personal bonuses which can be played at any time (and changing the name to “boost” cards to reflect this change in focus). These cards add lots of fun planning since they are essentially mini-goals you can strive for to try and maximize their effectiveness, and a reduced frequency of offensive cards sped up flow of play, evened out the wonkiness of different player-counts, and actually made the attack cards feel a lot more special and impactful.
Sometimes it turns out the mechanics are working fine, it's just the math that's off.
As always, I'm also extremely impressed with our video guys Tom Sarsons and Joe McDaid. As I write this, I'm merely preparing to spray-colour a very important wig, but I know that by the time this is posted, Joe will have edited together whatever disjointed shenanigans we manage to capture this weekend into some breed of cinematic masterpiece. [Update: Was I wrong??]
I also want to give a huge shoutout to Adrian Vaughan, our voice of Tesla. I based the script for our video off of the amazing trailer for the indie videogame Crawl, and having their voice actor agree to do ours feels like writing a Star Trek fan fic and then having Leornard Nimoy agree to narrate it. This man deserves to be famous.
Above all though, I want to thank Gavan Brown for funneling some of his manic energy into a project so dear to my heart. After seven years, the feeling of seeing this game come to life with such verve and passion behind its production, and having it meet with such cheerful enthusiasm by the gaming community, can't be adequately described in words. (The fact that I still love playing it after all this time is just icing on the cake.) And so instead I'll just say that I know with his talent and dedication, Roxley's going to have a breakout hit sooner or later, and regardless of how this one ultimately pans out I thank him for letting me along on his caffeine-powered ornithopter and feel like a hero-scientist for a while.
[Update: This one seems to be panning out pretty well seeing as we funded in a week! You people rock so hard!!]
Several years ago, an intelligent, informed and well-intentioned man advised against my developing a game with a steampunk theme as he felt it was probably a flash-in-the-pan fad that would be over in a couple years. Much thanks to all of you for making this wonderful subculture bigger than ever, or at least letting it stick around long enough for me to finish this darn thing.