World's Fair 1893 started as something entirely different. The whole process from end to end, inception to initial production, was about two years. It all started with a simple concept I call "stickiness". The idea is that on your turn you're placing a piece somewhere to take what's there (like in, say, worker placement), but then the piece stays there to score for other things. I love tradeoffs in games, and this became an interesting way to make each choice have implications for several things. To me, the fun part of playing games is making those tough decisions, deciding what to pursue and what to sacrifice. I had first explored this mechanism in Gold West, but I wanted to create a game built around that concept as the starting point. I also won't claim I invented this mechanism either — if you haven't played the game Québec, it's absolutely wonderful, and I took a lot of inspiration from that design.
Speaking on inspirations, the two biggest inspirations for the game were El Grande and Ra, two of my favorite games. If you've never played, El Grande is the definitive area majority game, with a wonderful tradeoff between placing more caballeros (read: cubes) on the board, activating more caballeros (so you can place them), and taking better/worse actions. Ra, on the other hand, is an auction game in which the winner of the auction will collect sets of tiles that score in different ways. Each player is collecting different things, so everyone evaluates a particular combination of tiles differently, which is what makes the auction interesting. I imagined a game that combined these mechanisms together — you're adding cubes to different areas trying to control them, all the while also collecting tiles that scored in different ways. This coalesced into the first prototype: a brown bag full of tiles with different symbols on them, and a hand-drawn board spanning four sheets of paper.First "rapid prototype" of what would become World's Fair 1893
From this prototype a few different mechanisms were established that form the core of the game and still form the basis of the final game today. Place a cube in an area to take all the tiles there (cards, in the final version, more on that later), then replace three tiles, starting with the region from which you took and proceeding clockwise. I made it three tiles (instead of two or four) because three felt like a good amount of tiles to pick up each turn, so if I wanted players to usually pick up three, I figured each time you should put out three. Sometimes game design is as easy as that; other times it's not.
At this point, the game was more or less abstracted. When I'm exploring mechanisms, I'm always hesitant to weave in a theme too soon. I've always had the approach that I want to start with a game that creates interesting thoughts in players' heads, and doing that requires an unbounded decision space when it comes to changing the mechanisms within the game. Once I figure out how the "game" works, then comes the second step, which is coalescing it around a theme that fits, using that to refine secondary mechanisms, and tie everything together.
So as the concept slowly developed, it came time to find a theme and let the game coalesce around it. I had read a book called The Amber Room — don't bother as it's not that great; if you're going to read anything, read The Devil in the White City instead — and become interested in the amber trade, amber being the precious stone made from fossilized tree sap. As a fan of historical themes, I decided that the idea of players being amber merchants, collecting amber and other goods, and trying to control different key cities of the amber trade (Bern, Venice, etc.) would fit the game fairly well.
In this prototype, each region had a different value for first and second generated by placing a randomized tile on the area. The game had five "goods", which scored only for set collection (collecting multiples of the same good scored more points) and were not linked to any specific area. All the actions that are in the game today were present, but the actions in an area were executed immediately when you placed there. The subtle change to have actions played in your subsequent turn came later, but I'll talk about it now because my memory is not so great. Playing actions on your next turn gave players more options. For example, in the game today, the Daniel Burnham card lets you place an additional supporter in the same area where you place your initial supporter. By allowing players to pick up the card on one turn and use it on their next, it could be used on any one of the five areas; you're not stuck waiting for the action to pop up on the specific area where you need it.Prototype at Gen Con 2014
I brought the prototype to Gen Con 2014 to play with trusted friends, including Adam McIver, who would end up doing the wonderful graphic design on the game. These playtests inspired a number of changes. I changed the game to a modular board and removed the tiles that increased the value of individual regions. In its place, the game had its first major breakthrough. (Let's call it Breakthrough #1.) I realized that the five goods in the game should correspond to the five areas you're trying to control. It made sense to link each of the goods to an area thematically and have each area be worth more if you had more of the associated good. It was a subtle change, but it resulted in each area having a different value to each of the players, which created interesting trade-offs in the game.
I decided to submit the game to Randy Hoyt at Foxtrot Games. I knew ever since I had played a prototype of Lanterns at Gen Con that I wanted to work with them, and the game seemed like it would fit with the weight and style of game they were looking for. A few weeks later Randy emailed to inform me he'd like to sign the game. There's no better email than that. I was thrilled.Prototype submitted to Foxtrot Games
Intermission: The Theme
Underlying all of the changes that are to come was a major thematic overhaul of the game. All along, we knew the game probably needed to be rethemed. We explored a number of different things, all of which worked okay — but the World's Fair theme was a revelation. Sometimes you just know, "This is it; this is the game".
The Chicago World's Fair theme was both a better fit for the mechanisms and weight, and an infinitely more interesting and appealing theme that miraculously had not been explored in detail before in a game. Having lived in Chicago for over five years, I already had a fascination with it, so when Randy starting mentioning World's Fairs as a possible route in which we could go, the idea of focusing on the one from the city I loved so much only made sense. It helps it was the best World's Fair (completely objective and unbiased opinion).
I really credit Randy for all of the incredible thematic details in the game. Though a lot of the development happened under the "amber" theme, for sake of ease of understanding, I've re-couched all of the terminology in the diary below to match the final game today. What were goods became exhibits, the amber in the game became the Midway tickets, and the actions became the influential figures of the fair. If you're interested in more of the thematic aspects, we discussed them in detail on a podcast of The League of Nonsensical Gamers.
This was just the beginning of the game's development, and we needed to figure out a way we could reasonably playtest with me in New York and Randy in Texas. Randy imported the files onto Roll20, an online platform that would allow us to playtest in our disparate locations. I love the future. We played regularly almost every Friday.
The first thing that became apparent was that the player who collected the most exhibits would more often than not go on to win the game. The exhibits simply scored triangularly the more of a single type you collected. We also realized that we could probably do more than just reduce the number of points they're worth. After all, the more exhibits you collected, the more they were worth, so once you reached a certain point it made sense to just collect as many as possible — and if you weren't collecting a certain type of exhibits, there wasn't much incentive to start (with the exception of denying your opponent, which isn't all that fun), which meant you were better off just trying to earn points from controlling areas.
Through lots of experimentation, we arrived at two huge breakthroughs. These would end up being two of the most mechanically important aspects of the game.
• Breakthrough #2: The exhibits must be approved in order to score. In other words, you needed to control the area that the exhibit was associated with in order to make it worth anything. This was critical because it integrated what were previously disparate strategic avenues: area majority and set collection. With this change, the set collection didn't mean anything if you didn't also spend energy on area majority, and the area majority wouldn't earn you much if you didn't also have exhibits to approve.
• Breakthrough #3: The exhibits should score for diverse sets. This rule is critical because of the one above. Because you must control areas to approve exhibits, and you must approve a diverse array of exhibits to score meaningful points, this means (does mental connecting of the dots) you must control different areas over the course of the game. I love this rule because it alleviates the problem of some area majority games in which you can accumulate an insurmountable lead in an area, then reap the benefits all game. The rule that you have to remove half of your supporters each turn helps with this, but even so, the game now makes you want to control different areas. The end of each round should feel like an accomplishment when you approve exhibits, but it also quickly changes your focus as you must re-evaluate which new areas you need to try to control next round. This (hopefully) keeps the game play from feeling repetitive, as the goals you're trying to accomplish continually evolve.
Speaking of scoring rounds, the game always had three scoring rounds, but they weren't always triggered the same way. The Midway tickets were originally just a way to score points. Instead there were separate "trigger" tiles, twelve of them, that when taken would be placed on a track; once certain thresholds were reached, a scoring round would be triggered. This worked okay, but the scoring rounds were highly variable, in that some rounds would be extremely short and others very long. We needed a way to decrease the variance of trigger tiles.
Enter Breakthrough #4. The Midway ticket cards are the triggers. It may not seem like a major change, but it had a twofold improvement on the game: First, it decreased the variance of the scoring rounds because there could be a ton more of them in the deck, and second, it gave the Midway ticket cards a more important role, mechanically and thematically. The Midway was what made money for the fair and was really the only part that was profitable. The Ferris wheel was the biggest attraction on the Midway and became the centerpiece of the fair, so it was fitting to make it also the centerpiece of the game.Prototype at Gen Con 2015
Okay, one last thing. You may have noticed the game was once tiles, and now it's cards. I originally designed World's Fair 1893 as a tile-drafting game, mostly due to the inspiration of Ra. The game had a fixed number of tiles in the bag, which meant a fixed number of turns for each player. There were ten starting tiles and 108 in a draw bag. The first two rounds ended when players triggered them, but the third round ended when the tiles ran out. The timing of that third round felt quite different from the other two, flat and predictable by comparison.
Randy and I brainstormed solutions, and he pushed to have us try a deck of cards, rather than tiles. The deck could be shuffled, and the game could continue until the third scoring round is triggered just like the other two rounds. This worked only because of Breakthroughs #2 and #4. Actions were already being discarded when used, but now exhibits could be approved and discarded (#2) and Midways could be cashed in and discarded (#4) at the end of each round, making a discard pile that could be reshuffled into a deck when the draw pile ran out. This was a subtle change from a gameplay standpoint, but it allowed a lot of flexibility in terms of the flow of the game. There's also a broader point about game design there. The idea to use a draw deck made sense only after a couple of other subtle changes were made to the rules. (Those rules didn't originally change anything about the components, just how they were treated during the game.) It was a reminder to constantly assess a developing game in its current state as new improvements can open the door to even more improvements that wouldn't have made sense prior to the previous improvement.
That said, having a shuffled deck also provided a number of challenges. The timing of the reshuffle is critical. If the reshuffle happens too early before a scoring round, there can be a disproportionate number of action cards in the deck compared to Midways and exhibits. There were three variables we had to play with: the number of Ferris wheel spots in a scoring round (i.e., the length of the round), the size of the deck, and the number of Midway cards — and it had to work with two, three, and four players. It took a lot of math and a whole lot of testing, but we managed to figure it out, so (hopefully) everyone who plays the game can take it entirely for granted.
And that's the long, short story of the development of World's Fair 1893.
—J. Alex KevernFinal production game in play
To submit news, a designer diary, outrageous rumors, or other material, contact us at firstname.lastname@example.org.
31 May 2016
- [+] Dice rolls