:strip_icc()/pic1282187.png)
/pic1279781.png)
Rebecca loves networking games like Ticket to Ride, and I'm a sucker for games with modular boards, so the question came up as to whether it was possible to make a route-building game with a dynamic board. Making each hex tile a different planet and setting the game in space seemed like a natural fit, and that idea never really changed. The obvious mechanical issue to solve is that if you create static objectives – e.g., connect tiles A and B – you might end up with some boards where this objective is ridiculously hard and other boards where it's almost trivial. So some mechanism must be in place that balances the effort-to-reward ratio for these objectives, which became the contract cards in the final version.
We came up with two ways of tackling this problem in the initial design. First, each contract card would list three planets instead of two. This lets the game award different point values for connecting one, two, or three locations in the network, which provides much better granularity than a simple binary test of whether two locations are connected. Second, the contracts would be selected from some kind of a draft board each turn. Whichever contracts were not selected would acquire resources on them as an incentive to take them in the future, such that eventually even the worst of contracts would be worth taking.
We also recognized early that we wanted each planet tile to have links out of it, with each link listing half of the resources required to connect to the adjacent tile. And because all planets need to be reachable and it's possible for a tile's random placement and orientation not to have any links into it where both sides match up, we knew that unmarked edges of each tile would need an implicit high cost.
Last, the game needed some mechanism that gave players resources so they could build these routes. Thematically we liked having planets in your network provide resources so that players might have to decide about building to get better resource acquisition versus building to complete contracts. It seemed obvious to make the collect action provide resources from one planet. Bigger networks then provide more options but not necessarily more resources, which should prevent snowball victory problems. We also made moons give one less resource per opponent present, so that connecting to a remote planet could be a resource benefit in some cases.
And that was the initial game. There were no action cards, no guilds, and no bonus points for completing the most contracts of a given color. Oh, and you started with only one initial planet instead of two, which made initial bootstrapping brutal in a particularly puzzlish way.
:strip_icc()/pic1281518.jpg)
A very early prototype of the planet tiles
Rebecca: That was the initial concept for the game, but much of our road trip was spent determining the link costs and production distributions. I had just received my Ph.D. in Algebraic Combinatorics (i.e., the math where you count things) and found this to be a very interesting question. Once you decide how many links you want on a tile (5, 4 or 3) you've implicitly determined a classification of planets (Primes, Colonies, Outposts). We wanted there to be some balance and symmetry in the resource structure without everything being identical.
We started by generating all the unique link configurations for a tile. For instance, there are four different ways to put three links on a hex that are distinct under rotation. I still have our old notebook where we drew out all the configurations and calculations. Having only four Outposts would not work well from a gameplay standpoint, so we put a lot of thought into how many tiles and planets of each type we needed and how to obtain that distribution in a mathematically elegant way. From the start, we intended purple and red to be slightly rarer in the galaxy, but to get the numbers to look pretty we found we needed to make blue a bit more plentiful as well.
As you probably can tell, I found this part of the design interesting. In short, the distribution of planet and moon pairings, link layouts, size, and colors were all chosen with gameplay, color balance, and mathematical elegance in mind. We did something similar with the contracts, although those underwent a few more revisions.
Initial playtests were promising, in that the basic networking mechanism was interesting and the market mechanism really did balance the power level of good contracts versus bad ones. There were some issues, too, of course. The major issue was that there was only one point source: completing a contract. If someone consistently completed her contracts for three points and you fell behind for one turn, it was virtually impossible to catch up. Furthermore, all points information was public, so you had a sinking feeling that the only way you could win was for them to make a mistake. What's more, the game had only one strategy as well: complete all three planets on each of your contracts.
A lesser issue was that turns could be really long and intense. On each player's turn, she acquired resources, then expanded her network, and then had to take a new contract and pay bribes. That's more than double the thought that goes into a current turn, so the entire pacing felt much slower. So on and off for the next five years, all the remaining tweaks were focused on addressing these two issues: the game having only one strategy and turns taking too long.
:strip_icc()/pic1282188.png)
The first list of planet names that weren't just letters
Early on in testing, we added the idea of awarding a distribution bonus for the most contracts of a certain color (plus another for most colors). Since your final three contracts aren't cashed in until the end of game, everyone's final score is unknown until after the last turn. We also wanted to give players some way of acquiring more contracts while being less able to complete them well, so there would be a strategy based on quantity (bolstered by distribution points) in addition to the standard quality strategy (scoring the maximum points from each contract).
Rebecca: Part of the early reasoning behind distribution bonuses was to reward players who diversified their networks. It also supported what I think of as an Outpost-based strategy. Since the contract color is the color of the Prime, one approach is to get to one Prime, then take contracts only of that color. Since Primes are easy to get to and about a fifth of the contract deck are in each color, this is a relatively easy-to-implement strategy. However, there are only six Outposts, so being at two Outposts means you'll have at least one planet on a third of the contracts across all colors. This means you can potentially earn more points on distribution.
The first idea was to try a scoring system in which completing three planets was worth 3 points, two planets was worth 2 points, and one planet gave you a token you could use to draw another contract later (while keeping the first contract for distribution points). At least there was some semblance of a quality versus quantity strategy, though it still felt bad to cash in contracts for only two of the three planets.
Rebecca: It took us awhile to completely address this issue. Players need to be able to get contracts at different rates so that they can find a strategic balance between doing well with a few contracts and doing the minimum on a large number of contracts. It's the classic quality versus quantity trade-off.
The major breakthrough we had was separating the different parts of the turn onto action cards. On each player's turn, she would choose to collect, build, or contract, then flip that action face down. All three actions refreshed once all three were face down. This helped offload the computation complexity of each player's turn.
However, it still had the same basic problem that players acquired contracts at roughly the same rate as one another. We couldn't make each action available on every turn, or players wouldn't take contract early in the game and would take nothing but contract in the late game. There was also the issue that players need to spend more time acquiring resources than spending them. Since a collect will net a player four or five resources, and even the expansion of a single route costs six resources on average, the game needed more resource acquisition actions.
We needed a wide variety of resource production, so we gradually went from three actions (collect, expand, contract) to six, with every action providing some way of acquiring resources. For example, we gave the expand action an initial acquire and a higher bribe (net +3 resources) instead of no acquire and a bribe of 0 (net +0 resources). The gamble action was added for two reasons. First, it's important players get more than one contract every five actions or so. And second, it adds hidden information to the game, so players can only estimate how many contracts of a color are needed to win distribution points. Without any hidden information, a player with a good memory or a piece of paper can work out exactly how much is needed at the cost of everyone else's time, and that's just no fun.
Rebecca: Splitting the game turn into action cards and adding actions is part of what I think really makes the game work. Gamble went through many revisions, and I like how it now plays. Early in the game I'm often looking for direction in how to build my network or need just one more resource of a color, so I'll gamble for one contract. Later in the game I'll want to take advantage of my network or flesh out my color distribution, so I'm willing to forgo the two resources to gamble for three.
Our publisher Chris Cieslik at Asmadi Games had the insightful idea of setting the bribe cost on the contract action to be the number of unflipped actions you had, then to have contract refresh them. This idea solved three problems at once: It prevents players from doing nothing but selecting contract in the late game, it forces players to take contract in the early game, and it guarantees players will have different numbers of contracts at the end of the game.
Rebecca: This is what I think pushes the game up the Euro-strategy scale. Everyone has the same actions and same number of turns, but your action choices determine both the cost and accessibility of your later actions. I've seen players do well playing a methodical game of using every action before contracting with a bribe of 0. I've also seen great games in which a player minimizes the number of times he takes a connect action and carefully manages resources in order to maximize the number of contracts he gets. It's fascinating how different strategies approach the action reset question.
This was roughly four years into design. At this point, the basic game system was in place. We made a few other small but important tweaks in that time. Colonies start with a bonus point, the resource trade rate is 3:1 instead of 2:1, and the contract deck only includes contracts on which all three planets have different colors. (If you don't include that last one, players end up with almost no solid resource production because they are too specialized. Then they don't do well or have fun, but aren't sure why. Eventually we realized it was the contract distribution.)
Rebecca: Contracts having three different colored planets was done early in the game design, but even then the contract deck had all possible Prime-Colony-Outpost combinations that met those criteria. We found that deck size to be too large and too variable, so we added additional constraints (like the Outpost moon color never matching the Prime) to trim the deck while keeping the underlying symmetric balance intact.
Our playtests were going well, but players still wanted a little more in the way of strategic variety. The basic game thought process was still about maximizing the quantity and quality of contracts you completed. Our solution to this was to add guilds. At the beginning, all we knew was that they were an alternate or supplemental point source. Their use would not be required to win the game, but some winning strategies could focus on guilds. How many points they were worth or how players could join was to be determined.
Rebecca: It should be noted that the strategic variety requests came from our hardcore playtesters, who by this time were very familiar with the game.
We tossed around a number of guild scoring systems, but discarded most of them for either being too complicated or rewarding the same plays that improve contracts. It's not a strategic option if it doesn't actually change your strategy! The one thing that worked was awarding players points equal to the number of planets they had infiltrated, so the blue guild is worth four points if you've infiltrated four blue planets by the end of the game.
Rebecca: This is a different strategic direction since all the planets on a contract are different colors, so going to an extra blue planet doesn't increase the value of a contract. Being on three blue planets might help complete contracts for distribution, but it will depend on which blue planets you are at.
This worked well for the most part, but presented two challenges to solve. First, the value of joining a guild is constant, so players have an incentive to join guilds as late in the game as possible (when the opportunity cost of losing resources is at its lowest). So the cost to join needs to start cheap and get expensive as the game goes on. And second, the red and purple guilds were much worse than the others (and the blue was a bit better) because there are few red and purple planets and one extra blue planet. We tried making a combined red/purple guild (too confusing), awarding a bonus point for joining these guilds (too fiddly), and adding Sabean Core as an extra red/purple planet (doesn't help enough). In the end what worked best was creating separate cost tracks to join each guild, and making the red and purple guilds cheaper than the others.
Very early concepts for the planets
Rebecca: We did keep Sabean Core, however, to replace an asteroid tile. The link costs for Sabean Core are slightly higher than the asteroids, but the planet itself potentially adds more purple and red. Since it's not on any contracts, this change didn't affect the contract distribution, while still supporting the red and purple guilds.
Players also needed some way to join guilds, which is where the scheme action came from. Previously we tried an action called Fence, which let you cash in contracts for bonus points, but it didn't create interesting game play. We kept the concept of cashing in contracts for scheme, but made it an alternate source of resource acquisition. The opportunity cost of cashing in a contract earlier is traded for always getting six resources, possibly from a world you don't yet control. Then infiltrating a guild during the scheme action gives the player the option of trading resources in the present for some number of points at end of game. And as a bonus, if they take this option they'll have additional strategic direction.
Rebecca: Scheme (or Fence) went through so many revisions. Being able to convert a difficult contract into six resources (even without an agent there) and possibly some points is very useful to certain strategies. Other strategies use it to fund their Guild bribes. It's one of those actions that's very flexible.
I really like how Acquire, Gamble, Scheme, Contract Board, and 3:1 resource trades are such different ways to get goods. Sometimes when we mention that Syndicate was partly inspired by Ticket To Ride, people think that set collecting might be a game mechanism; however, our good production mechanisms are quite different. One option is to build routes using only the goods of planets/moons in my network or explore a nearby world just for its resources. If I'm short a single good, I might try my luck at gambling or take the more expensive 3:1 trade that doesn't cost me an action. I could instead decide to scheme a contract away to get colors from a planet where I don't have agents, forgoing the potential for additional points. Sometimes the colors on the contract board will determine which contract I choose. The game is set up so that players have choices in how they get goods; they aren't at the mercy of the train deck or die roll to get certain resources into the system.
Last, there was some discussion as to whether the complexity of guilds was suitable for an initial play of the game, or if they should be added into later plays (similar to the action cards in Agricola). We tried both ways, but in the end they stayed. Even with the option of infiltrating guilds, The Phoenix Syndicate is slightly less complicated than Endeavor and Macao, two excellent games that didn't need a "training wheels" mode.
Rebecca: I'd say The Phoenix Syndicate is "less complicated" from a rules or teaching perspective as there are fewer things for a new player to track than in, say, Endeavor. Strategically, it feels to me on par with Macao (one of my favorites), without a deck of unique ability cards. One thing I like about The Phoenix Syndicate is that I feel like I can see how my choices are affecting my strategy.
For more details on game play, you can download the not-quite-finished-as-we're-awaiting-final-artwork rulebook (PDF) or check out the Kickstarter project which ends in early May 2012.
Ted Vessenes
/pic1114287.jpg)
:strip_icc()/pic1288200.jpg)
