by Emma Larkins
Have you ever wondered what it would be like to start with a fresh new sky full of stars and discover the constellations for yourself? Well now you can in my new game from Buffalo Games: Starry Night Sky.
In Starry Night Sky, players move their telescopes from constellation to constellation across a new vision of the night sky, drawing stars from a shared bag, placing stars, and completing Exploration Goals as they go. These actions reward players with points, as does discovering new constellations. The Night Track counts down game time as new constellations are discovered. At the end of the game, players reveal their Myth Cards, scoring additional points if certain constellations were completed over the course of the game. Players are rewarded at the end of the game with a beautiful path of translucent, colored stars tracing their unique journey.
The journey to Starry Night Sky’s creation wasn’t quite as simple or relaxing as the game itself. Join me as we delve into the tumultuous origins of this star-themed game!
During a pitch discussion with Buffalo Games in early May 2021, John Bell (Director of Inventor Relations and Preliminary Design) mentioned an interest in a relaxing night sky, constellation, or star-themed game. I’ve always loved space and science fiction, so I got to work brainstorming ideas — mindmapping, playing around with components, and writing down whatever came to mind in hopes that something would spark.
The first idea that grabbed my attention was a dexterity game based on flicking or rotating star-shaped pieces. I borrowed some star bits from Before There Were Stars… and had fun trying to flick them so they’d spin in a pleasing way. To emphasize this motion, I assembled a flicking arm from cardboard shaped like a ladle with a hook at one end.
I used different materials (plastic and electrical tape) for another flicking arm to compare the motions of the two.
The motion was fun, but it needed mechanisms! I imagined players attempting to land stars on star-shaped spaces or regions of a board, similar to Crokinole. I had to make the spinning of the stars matter; was it engaging enough for them to bump into each other at weird angles and bounce off in interesting ways? I also tried to think of ways you could create your own constellations via flicking, or drop the star pieces onto the board to start the game, as in Dungeon Drop.
Dexterity games are a hard sell unless you come up with something exciting and unique since there are so many of them out there. Even with the special tool and the star theme, there wasn’t enough of a “hook” (pun intended) to capture audiences. By the end of May we’d decided against the dexterity angle, and I had to go back to the drawing board.
A Post-Dexterity World
I had another intense, blue-sky brainstorming session, throwing all sorts of things onto the page to see what stuck. Keeping a focus on theme encouraged me to focus high-level on the vibe I wanted to create. I wanted people to feel a sense of wonder and discovery as they played with these star pieces, to connect with their actions on an emotional level.
I wasn’t ready to give up on the spinning motion of a rotating star just yet.
I built a triangular spinner using cardboard and a marble. Instead of trying to flick stars across the board, players would place them based on where the spinner pointed. Each of the three board segments contained a number of “constellations” (rough collections of star spots matching the colors of my star pieces).
I wanted to give the (competitive) game a collaborative vibe. I decided players would get a small reward for completing constellations, but the majority of points would come from helping (tracked by player-specific pawns, or “helpers”). The pawns didn’t make it to the end of development, but the idea of getting rewarded for helping people would become a core part of the play experience.
At this point I also implemented cascading actions — finishing a constellation rewarded a player with additional star placements. I wanted to create a sense of drama with big moments as opposed to having a flat excitement curve.
I made and playtested a rough prototype. It was fun placing the stars onto the designated spots, but players didn’t have enough control over where stars went, giving it a too-random feel.
Despite the lackluster gameplay, this prototype did confirm one of the core reasons that I wanted this game to exist: the beautiful table presence of translucent stars scattered across the sky, building up in a unique way every time you play.
With a heavy heart, I decided to abandon all ideas of spinning and dexterity and focus on star placement.
All About the Stars
I enjoyed placing star pieces onto matching star spots organized into constellations, and it seemed like an idea worth pursuing. When I got rid of the spinner and the segments, it opened up the idea of players choosing constellations to place in, instead of letting fate decide (although I now needed to find a new way to limit player agency).
I finished my next prototype at the end of June 2021.
I didn’t want to build a full-size board right away — in part because I wanted to have the flexibility to easily reconfigure constellations and the paths between them — so I improvised. I used a pen to draw stars on circular pieces of paper, forming constellations. I also imagined players claiming constellations for victory conditions, sliding the stars gently back onto the board to preserve the table presence, of course!
I made paths between constellations using pieces from a retired copy of Takenoko. Players traveled along these paths, which would allow them to place stars in some areas but not others until later in the game. I wanted options to open up organically for players over the course of play so they weren’t overwhelmed at the beginning.
“Helpers” stayed. Now, players received points for starting a constellation, regardless of who finished the constellation. Players marked their contributions with meeples.
“Spillovers” happened when players completed a constellation, with the completing player placing stars in adjacent constellations, an action that could cascade multiple times.
Finally, I created “star pools” to provide a somewhat random selection of stars for placement. Stars were added to the pools each turn from a bag (inspired somewhat by Azul), cleared periodically, and shared by all players.
(Fans of the Winston draft may notice some slight similarities!)
Pandemic pitching meant creating a pitch video, capturing this version for posterity.
How would players receive stars? Was bag-pulling the way to go? I built a star shaker out of a can, cardboard, and (pink unicorn-themed) duct tape to test another option.
The shaker would prove helpful if I decided to make stars different shapes (something that can be problematic in a bag puller, although it works great in Steam Up: A Feast of Dim Sum). Although I liked the toyetic appeal of the star shaker, unfortunately I couldn’t make it reliably disperse one star at a time. And the (imagined) benefit wasn’t great enough to overcome the cost of prototyping and manufacturing, given the uncertain outcome.
Prototyping and Playtesting Online
Because of the pandemic, I didn’t have the option to playtest 3+ player counts in person. In early 2020, I’d taken the playtest events that I was running in the Seattle Tabletop Game Designers group online, and in July 2021 I made use of this group to playtest Starry Night Sky using digital tools, including Discord, Tabletop Simulator, and Tabletopia.
I built the first “board” version of Starry Night Sky — which, incidentally, was the first “board game” I’d designed that included a board! — completely digitally. It took time to learn the new tools, for example, finding a way to visibly distinguish the star pieces from the game board so that players could see and move them. But there were also positives; it was a lot faster to iterate online because I didn’t have to print and cut out a new board every time I made changes.
At this point in development, I thematically envisioned players as tiny gods actively creating and placing stars and other celestial phenomena in the sky.
In this version of the game, one random star was added to each star pool at the beginning of a player’s turn. Players would take all of the stars from one star pool and place them either in a constellation adjacent to the center of the board or in a constellation that had already been started (via spillovers). Large or small pawns denoted the amount of stars placed (a.k.a., degree of helpfulness). Players received rewards such as black holes (to cover unwanted star spots) or wild stars.
Completing constellations awarded you points and ownership of the removable constellation disk. Helpers also received points. Stars would “spill over” into adjacent constellations, sometimes cascading into multiple completed constellations on a turn.
Players liked matching stars to star spots and cascading across multiple constellations. They disliked the lack of strategy around placing stars in one constellation vs. another. There were too many options, and it was hard to see how choosing a particular one would impact the outcome of the game.
In addition, the shared star pools gave players little control. Especially at higher player counts, the distribution of stars in pools changed too much from turn to turn, making it hard to plan for the future.
The First Dark Night of the Soul
It’s important when discussing game design to talk about the times you feel frustrated, stuck, and uncertain.
I like to use the term “dark night of the soul” from the Hero’s Journey (popularized by Joseph Campbell). The dark night of the soul represents the moment(s) when you feel like you can’t go on and should just give up. It feels like everything you try is pointless and nothing you do will fix the insurmountable problems you face.
The above design problems might seem subtle, but the biggest issue was that I hadn’t really captured the “fun” of the game. I needed to go back to the drawing board in a big way, sit down, and hash through a bunch of ideas to find the next version of the game.
Three New Challenges
I wasn’t about to give up just yet, so I wrote down thoughts and sketched images and generally tried to get the creative juices flowing. The three main design directions I decided to pursue were:
a.) Giving players a sense of direction by limiting their movement using a new piece called a telescope. (Maybe the telescope would stay at the center of the board and rotate in different directions, somewhat similar to the way the original spinner version of the game had worked?)
b.) Giving players a sense of purpose by providing interesting goals. Making it clear exactly where their points and bonuses would come from: completing constellations? helping out with constellations? what about long term/strategic goals to balance out the very tactical ones?
c.) Making star collection more interesting by giving players more control over the way they collected stars. (Maybe there could be star pools they had to visit with their telescopes?)
For the first challenge (giving players a sense of direction), I originally liked the idea of telescopes staying fixed in the middle of the board, imagining the center of the board as Earth or some other planet from which players could survey the sky. As thematic as that was, it didn’t really solve the problem. The next idea I came up with was for the telescope (or telescopes) to act like workers, moving from constellation to constellation and acting as a focus for player star placement.
However, I didn’t want to turn Starry Night Sky into a worker-placement game with all the mechanical weight that typically accompanies that genre. I envisioned Starry Night Sky as a worker-movement game, with a limit on player actions being provided by the one star pool per turn rule. You could get more telescopes over time, but that would only increase how far out you could get into the stars.
Using telescopes had the added benefit of focusing the thematic direction of the game. No longer were players celestial beings or forces of nature nudging the stars themselves; instead it was to be a game about discovery and viewing through the telescope’s lens.
For the second challenge (giving players purpose), I considered leaderboards and spaces around each constellation to track contributions: discovering, helping, and completing.
This was more complicated than I wanted it to be, but it was the best I could come up with at the time. Even bad solutions are worth testing if they get you closer to your goal!
The “discovery bonus” for the first player to put a star in a constellation would make a big impact later on, encouraging players to explore.
Sometimes you just have to doodle for inspiration! These doodles might seem unproductive, but they’re great for staying focused in designer’s mind when you don’t have any solutions.
In the bottom left corner you can see an image that set up the solution to the third challenge (interesting star collection), something that stayed more or less the same throughout the rest of the game’s development: individual star pools.
I figured stars could be added randomly to each of a player’s pools on their turn, just like they’d been added to the shared pools. I would come up with pool-manipulating powers later if it was too hard to match pools to the board.
Some other ideas I was playing around with at the time (August 2021):
• Space lines! (I’m not sure I knew even at the time what “space lines” were. Something aesthetic that you would see in a space game, probably.)
• Also mused a “shooting star” mechanism to bring dexterity back into the game???
• See the introduction of Myth cards! “2-4 constellation combos… Need to help or complete all on the card.” This additional scoring element had a lot of potential!
Fleshing out Star Pools, Constellations, and Myth Cards
The three circles of star pools became individual player boards, meaning that players now had home bases for their individual star work.
I was still floating around the idea of a narrative element to the game. I wanted to hint at a world in which larger stories existed, and give players the tools to build those stories if they were so inclined.
After going back and forth between using existing constellations vs. making my own, I quickly settled on creating a new universe of constellations, partly because I didn’t want people getting upset with me for toying with celestial compositions, and partly because I wanted to imagine a kinder, cozier world than many ancient constellations suggest.
I was also working on my design philosophy for creating Myth cards (2-3 constellations per card; three decks of Myth cards: easy, medium and hard, basing difficulty on constellation proximity and how close they were to the center of the board).
Next Digital Prototype
By the end of August 2021, I’d placed the constellations in three concentric rings for more consistent and measurable travel times.
I used The Noun Project (which I am enthusiastically subscribed to; paying artists, yay!) to source prototype art. It’s important not to spend too much time on prototype art to keep development flexible. This felt like the right time to add art to the game as it was easier and more interesting for players to fill in “Cat constellation” than “Middle central constellation #3.”
Here’s where I’d landed with the rules at this point:
1.) You may move a telescope on your turn.
2.) Place all the stars from a star pool into a constellation where you have a telescope.
3.) Place a helper token into that constellation.
4.) If you’ve finished a constellation, move your telescope and place a spillover star and another helper in the new constellation.
5.) Other players with helpers in the constellation move their telescopes or get additional telescopes.
6.) If two or more people participated in the constellation, all helpers get new Myth cards (incentivizing players to not work on constellations by themselves).
7.) Add three random stars to your star pools.
8.) Get points at the end of the game from Myth cards.
9.) Reveal a Myth card when it’s complete and get a new one.
10.) Game ends when someone completes their fifth Myth card.
Players skipped over completed constellations, making it easier to get to more distant constellations as the game progressed. The “hopping over” mechanism has grown to become one of the most satisfying parts of the game. In the beginning, you don’t have that many choices, which is good — you don’t feel overwhelmed! — but as the game progresses and you get a feel for the actions, suddenly you find yourself able to move great distances and achieve things that were previously impossible.
Player Motivations and Other Lingering Problems
“What if a constellation isn’t on my Myth card? Why would I go there?”
I had hoped that players would have enough motivation from wanting to get spillovers into constellations that were on their Myth cards, but this turned out to be pretty hit or miss.
A lot of the core design ideas were in the game at this time: working on constellations together, opening up movement over time, getting random stars from a bag every turn, and placing stars from your individual star pools — but many things still needed tweaking, such as:
a.) Placing a random star into each of your star pools on your turn. Without any control over your star pools, it could be really difficult to set up for future turns.
b.) Having to place all the stars in a pool on your turn into a constellation. Combined with placing random stars in star pools, this made it very difficult to complete the constellations you wanted to complete.
c.) Not having to move on your turn. People could camp out in a constellation and drop one star at a time, which slowed the game down.
d.) Players had to participate in a constellation to score Myth cards. This necessitated the use of helper tokens, which clogged up the board with information, and also meant dead cards in hand if players didn’t get to a constellation in time.
e.) Ensuring that someone always has a turn one move when the stars are assigned randomly to star pools. No one wants to play a game in which you can’t make a move on your first turn!
Second Full Physical Prototype
I came a long way since the early dexterity, spinner, and paper circle physical prototypes with a fully printed board, also in late August 2021.
Making a board for a board game is hard, especially because most boards for games are larger than the typical U.S. printing size of 8.5 x 11 inches! My solution was to print and tape together together nine sheets of paper — not the prettiest or easiest to transport, but it got the job done.
The most important thing this prototype proved was the visual appeal of the game. Seeing the art starting to emerge, the paths players would take, and the stars slowly spilling across the sky was an immediate hit with playtesters. Not only were the visuals appealing, but they meshed with my vision of the vibe of the game being a chill, relaxing journey.
Players enjoyed the puzzle of the game, meaning this version had much more strategy than previous versions. The semi-co-op nature of the game that I had intended was shining through. People enjoyed the limitations and strategy created by the individual star pools. People enjoyed the constellations I had chosen, liked that they were real, modern objects.
Sometimes you just need a little help…
Downsides With the Upsides
Players enjoyed moving and placing stars, but if they spent too much time completing their own Myth cards and didn’t participate in constellations other players were working on, they wouldn’t get any more Myth cards and they’d run out of steam and direction.
Players weren’t very interested in the special powers I had implemented in order to make the game function.
The anatomy of the Myth cards — sometimes you just got lucky with a constellation appearing on multiple Myth cards, or sometimes constellations were right next to each other — meant that some players had a much easier time completing things.
Visually, it was hard to tell when a constellation was complete.
More Digital Versions
In this early September 2021 version of the game, if you completed a constellation that other people helped with, you’d get to spillover, and these spillovers cascaded as long as you didn’t run out of stars. The player who completed the constellation got to move their telescope and place another star (which could result in another completion and thus more spillovers). In the meantime, everyone who had helped with that constellation could move their telescope to an adjacent constellation or place a new telescope if they didn’t have one in that constellation.
I had one very memorable playtest with this version. A friend named Andrew who loves “comboing off” triggered multiple spillovers in a row. Because everyone was making decisions on Andrew’s turn, everything became difficult to track. Did it make more sense to completely resolve one constellation before moving on to the next, or was it better to let the active player keep going? There wasn’t a clear answer. Some people were excited about this version, but others found it incomprehensible.
(Reminds me of the Infinite Potato Problem I had with Abandon All Artichokes.)
After that I changed the spillover rule so that players who helped would only get a star of their choice.
I had the idea at this point that players would still get three random stars from the bag, but they would get to choose which star pool to put the stars into. Instantly, the star pool concept became something players had a lot more control over (though with some luck still), and they could make plans for future moves.
Messing with Myth Cards
I tested public Myth cards, in addition to private ones. My goal was to give plenty of things for people to work towards.
The ending condition for the game became ten or more Myth cards being completed across all players as well as the public Myths. It wasn’t the easiest think to keep track of, but it worked okay.
Players got Major (helped with all constellations) or Minor (helped with some constellations) Myth card points for either completed public Myth cards or their own private Myth cards, but only if their helpers were in those constellations. This necessitated keeping track of everyone’s contributions, which definitely clogged up the board.
Completing a constellation on your own now gave you the ability to cycle your Myth cards, which potentially gave you some more control. However, I found that people often didn’t want to give up on the Myth cards they had (sunk cost fallacy).
More star pool math!
Showing to Phil Walker-Harding and Buffalo Games
In early September 2021, I got to playtest with Phil Walker-Harding. I was full of nerves playtesting with the person who was a big inspiration for Abandon All Artichokes! Luckily he enjoyed the game. Although he did like the cascading spillover chain reactions (“I’m the disease in Pandemic!”), this later got cut way back.
“Lots of competitive games force you to co-operate, this one feels right.”
John from Buffalo Games and Phil both agreed they wanted more “bingo” moments. Could there be points for completing a constellation? Emptying a star pool? Being a helper? They struggled with all points being focused around Myth cards.
During the test we came up with ideas for different types of completion:
• “Bountiful Completion” – completing a constellation with help from someone else.
• If no one helped with a constellation, you got the “Solitary Completion” bonus – cycling a Myth card, with no spillover.
In this version I moved to tracking points with point tokens instead of having to do tons of mental math to add up Myth card points.
Transportable Physical Prototype
Heading to Gen Con in September 2021, I wanted to have a physical version to show to people.
This was the most board-like board I’d prototyped to date. I figured (without doing research) that cutting the paper to size and gluing it to cardboard with Modge-Podge would be my best option. It was quite an undertaking and ended up being a bit of a mess, but it was playable.
Later in August, the awesome Estefania Rodriguez and Arnel Baluyot were kind enough to playtest and give great feedback — and they weren’t even fazed that I forgot the player boards and had to draw new ones on the spot! Their positive direction helped cement the idea that I was on the right path.
Sometimes a game like Fantastic Factories has such great components that you just have to borrow them for prototyping. I used them to demonstrate what it would look like to have spaces around constellations on the board for placing helper tokens.
Moving Turns, Sky Quadrants, Star Distribution
At the beginning of October 2021, I finally realized I had to require players to move on their turn. This ensured that people wouldn’t stay in one spot and drop stars, and thus would push the game towards conclusion.
I attempted to limit the number of quadrants players could visit based on player count (two quadrants for two players, three for three players, and four for four) to keep players close enough together to help each other. This meant that players would have to separate the Myth card deck each game based on player count, something I didn’t love.
I also did the math on the constellations to make sure that star combinations (such as blue-blue-yellow) appeared an equal number of times, that there weren’t too many stars of the same color in each board region, and that each color star was represented the same number of times.
This was not the first constellation brainstorm, nor would it be the last! It turns out it’s really tricky building a night sky with just the right feel.
Signing the Contract
By the end of October 2021, Buffalo Games and I decided to move ahead together, and I signed the contract.
John started setting up playtests with people on the Buffalo Games team as well as with Bobby West.
Honing Myth Cards with Science
The biggest issue at this point in development was the balance of Myth cards. Up until this point I’d eyeballed which constellations should go together on each one, and the result was wild score imbalances with little chance of catching up.
Balancing Myth Card variables — number of constellations, number of stars in each constellation, types of stars in each constellation, distance between constellations, distance of constellations from the center starting area — would have been an incredible challenge to accomplish manually. Luckily my husband Phil volunteered to write an algorithm that created constellation combinations, along with Major and Minor points. The algorithm worked by generating difficulty factors for each combination and cutting off any combinations above a certain difficulty cap.
Here’s a glimpse of the first batch of Phil algorithm Myth cards:
This didn’t solve all Myth Card issues. Players could still luck into drawing Myth cards that had already been completed, triggering them to draw a new Myth card, and so on and so forth.
More Granular Scoring, WAY More Components
To balance out the points issue, at the beginning of November 2021 I decided to bring back an old friend in a slightly different form factor: individual cards for each of the 37 different constellations. You got the card for completing that constellation, earning one point per constellation card at the end of the game. This version luckily never got to the physical stage; no one was a fan of searching through a deck to find the constellation they needed.
Obviously this wasn’t a convenient or simple solution, but sometimes you have to make something you’re not happy with to get to the truth.
Constellation cards were positive because it gave people more scoring opportunities. In addition, it opened a clear new way to track the end of the game: the number of constellations completed.
It was also a fun callback to the very early version in which you would take the constellation itself off of the board and keep it.
The constellation cards soon vanished, along with their scoring condition.
Second Dark Night of the Soul
The best way to get out of this state when taking a long break isn’t an option is to generate ideas and test as quickly as possible.
I tried many different things in rapid-fire succession:
• 1 free star, 2 for helping (ran out of stars)
• 2 free stars 2 for helping (too many stars in the beginning)
• The bonus for helping being a Myth card (felt like too big a bonus)
• Giving people just constellation cards instead of Myth cards (they could all be too far away, felt bad if someone completed your constellation card)
• Starting with one Myth card instead of three (didn’t solve the problem of people going off on their own)
• Public-only Myth cards (felt weird at two players)
• Starting with six stars and getting a star of your choice on your turn instead of random stars (too much analysis paralysis)
I felt very lost at this point, but little did I know I was on the brink of some big discoveries!
Star Pool Cap, Double-Sided Tokens, Myth Card Streamline
I implemented the “four star per pool” cap. This was a great change as it encouraged people to empty their pools often so they wouldn’t get capped.
I cut the Major/Minor Myth thing. Now you just had to participate to get the points on your Myth cards.
Point tokens were replaced with a point track.
The game was timed with a time tracker that advanced whenever a constellation was completed. Cut a whole deck of cards, woo!
Took the suggestion of using double-sided player tokens to track completed constellations and make them easier to see.
A lot of testing went into determining Myth card variables for the algorithm, how many constellations, how hard they should be to complete, how many points they would be worth. The main theme was making them easier to complete by having only two constellations that weren’t too far apart.
I streamlined the bonuses: If someone helped with a constellation, you as completer got to spill over into a new constellation. If no one helped, no bonus. If you helped with a constellation, you got a point. Felt a lot tighter this way.
Balance Issues and Runaway Leader
I was way behind in points in the playtest pictured below. I found myself not completing constellations to avoid giving other players points. Players were disillusioned at the beginning when their Myth cards were too hard. It was hard to catch up if a player fell behind. Small constellations connected by one path felt like a trap being hard to get to and awarding few points. Players had no motivation on their last turn if they couldn’t complete a Myth card.
Completing a constellation felt like a lot of work, and it was easy to miss a step. In order to complete a constellation, players would have to:
• (completer) add or flip their helper token and receive points
• (helpers) remove their tokens (to reduce clutter) and score points
• (completer) spill over
• (all players) reveal and score Myth Cards immediately when complete
• this could all repeat many times
I landed on the best resource gathering method to date: Start the game with one random star in each of your three star pools; receive three random stars per turn, then place one into each star pool.
Players no longer had to complete a constellation for a spillover: now they got a spillover from landing in a constellation with a helper token.
Completers also would get 1 point per star in a constellation, and helpers got that amount minus 1. This was the first time I experimented with tying points to stars because I’d gotten the feedback that people wanted to be rewarded for their “effort” (how many stars they’d placed). It felt bad when they placed a lot of stars, then someone got more points for fewer stars.
Had the idea that Myth cards just have to be completed; you don’t have to have a helper in them (although you get bonus points if you do, kind of like major and minor).
Also testing out players getting abilities every time they completed a constellation, which would be tracked on their individual player boards.
We started working on official art at the end of November with the amazingly talented Nim Ben-Reuven.
I wanted to make sure that the names of the constellations were done in such a way that it was easy to see which constellation was which. It was also important that all of the components (stars, telescopes, tokens) would fit in the space allotted to each constellation.
The Buffalo Games team did an amazing job with creating the unique telescope player movers, one of the most popular pieces in the game.
Lots of creative ideas that I can’t take credit for, like art on the back of the individual player boards.
Third Dark Night of the Soul
There were still some fundamental problems with Myth Cards, but they were essential to motivating players; without them, people didn’t care where they went and had little motivation.
“I keep testing and changing things over and over again, and nothing is fixing it.”
Was I even on the right path? Should I drop everything and change the game completely? For example, switch to pick up and deliver, getting tokens from the board that wanted to go elsewhere?
Idea: “Goals should be achievable, but not too easy.”
Myth Cards and constellation cards merged! Now each “Myth Card” was just a constellation, worth more points the farther out you went/the harder they were to get to. No more multi-constellation Myth cards, just individual “bounties” for constellations based on the cards in your hand. You’d get points no matter who completed your cards, and bonus points if you did the work.
The problem with this version was there was no longer an incentive to help out with constellations.
At the beginning of December, I had another breakthrough. What if, instead of Myth Cards, players had “goal cards” that could be achieved in one or two turns? Things like “Put at least one star in Wolf, Hedgehog, or Friends” or “Place exactly four stars this turn”. Things that needed some work/set-up but not as much as having multiple constellations completed.
I played around with these being worth either points or special bonus actions: getting a random star, getting a star of your choice, scoring 1 point, moving an extra space.
This felt good, but I still wanted players to have long-term goals (in addition to short-term goals).
What if players had three Myth cards in hand that they revealed at the end of the game and they didn’t get more throughout the game like they had been doing? And what if these Myth cards were…
a.) roughly equal difficulty
b.) only made of two constellations
c.) low points compared to the goal cards
d.) no overlaps; exactly 12 cards, two unique constellations on each (I then bumped up to 36 constellations and 18 Myth cards, so the same set wouldn’t be present for every four player game), and
e.) endgame scoring
I also wanted to preserve the fun of simply placing stars onto the board, not focus too much on extrinsic motivators. I wrote in my notebook that placing one star per constellation leading out to the constellation on your Myth card would be bad because you get one point per star placed. (When did I commit to this? Uncertain from rules…)
Helper tokens finally went away for good. It turned out that I didn’t need to track who had helped in constellations now that there was no minor bonus and you scored your Myth cards no matter who completed them. This really cleaned up the board and made everything easier to see.
This is not to say that they shouldn’t have been there in the first place. Without the focus on helping, I never would have achieved the final vision of the game as a collective, competitive experience.
But it was just as important to know when it was time to let go.
I fleshed out the idea of the game length timer track. Each time a constellation was completed, players would move the Moon token down the track.
I had to decide whether people should get bonus actions from the Goal Cards or from the game length/time tracker.
Time tracker bonuses got better over time, were shared by players, and were tracked on the player board below the star pools.
As soon as I had this idea, I saw flaws in it. Since it was a public track, it felt weird fighting over the bonuses. (You might want a particular one that was taken from you.)
What if the bonus was all points instead? This could create a feel-bad moment if you left empty spots in a constellation because whoever completed it would get the spillover AND the completion bonus.
What if the tokens were “started” bonuses instead of “completed” bonuses? The person who first went to a new constellation would get the bonus, and the time tracker would be the number of constellations started, not completed. Players get more and more points for starting constellations over time.
This subtle shift was a huge breakthrough and balanced the feel good of starting and completing constellations.
Constellations Get Their Flavor
It was time to commit to my sky filled with invented constellations so that Nim could finish the art! I wanted a little extra spark and potentially the hook for story, so I decided to add adjectives to each of the objects we placed in the sky.
It was surprisingly hard to find a good, flavorful distribution of flora, fauna, objects, and human roles, but I’m happy with where we ended up.
Myth Cards — Getting Closer
Phil iterated on the Myth card algorithm many times over the course of my development of the game. After I’d figured out that there wasn’t a good way to balance cards with overlapping constellations, he made a list of unique two-constellation combinations of roughly equal difficulty.
I did a lot of playtesting with different values and tweaked the combinations manually until the cards felt right. With my Google Sheets spreadsheet and InDesign data merge, it was easy to make quick changes on the fly — and by testing digitally I didn’t have to print them out every time!
Time Officially Tracked
During playtesting we came up with the great idea to fill the time tracker with Moon tokens that would be taken from the track and placed in a constellation when that constellation was first discovered. Then the player flipped the token after completing a constellation. This solved a bunch of problems at once: tracking game length, giving points on constellation discovery, and marking completed constellations to show players which constellations they could jump over.
The combination of Myth cards (long-term goals) and Goal cards (short-term goals) made for just the right balance of incentives. Now it was just a matter of honing them!
I iterated on testing the Myth cards and Goal cards, tweaking goal card abilities and Myth card points. (All worth 2? Worth 4?) Feedback was that the abilities were a lot to keep track of, so I simplified and removed abilities and made Goal cards worth 1 or 2 points (slight variations in difficulty). In addition, I made Goal cards easier to achieve without having to use those bonus abilities.
The Final Countdown
We made the Discovery Goal cards round for fun and to help distinguish them from the Myth cards.
At the end of December 2021 was pretty much when the current rules were locked in:
• Three stars per turn, each in a different star pool of the player’s choice, but at end of turn
• One goal card (Exploration Goal) per turn
• Spillover becomes one bonus move and place per turn, don’t have to complete the constellation, just help out (this is where helping went)
• Pick one star pool, place as many stars as you want (don’t have to place all the stars, more control)
• Four ways to score:
a.) point per star placed
b.) discovering a new constellation (place a discovery marker from time track), with points escalating over time
c.) goals for exploring (1 or 2 points)
d.) end game scoring with Myth cards (variable, but low-scoring. Players could score more points in a turn through stars, discovering, and goal cards, so these are basically “cherry on top” points)
However, at this point you could get a “star of your choice” instead of points for completing a Goal card.
Are We There Yet?
Despite the game being very close to the final design, it wasn’t always hitting with playtesters. It’s not always easy to know when designing a game whether you’ve gone down the wrong path or you’re only a few feet away from victory. I wrote out a list of all the things Starry Night Sky was doing well that the game had struggled with in the past. (This can be a great motivator for personal development as well!)
I made the last three mechanical tweaks, and though they were small, they had a big effect on tying everything together:
1.) Players would draw stars at the beginning of their turn, instead of the end of their turn. Drawing at the end of your turn is often a good choice because it gives you more time to plan your move, but because the board could change a lot from turn to turn it made more sense to draw at the start.
2.) Players can discard a Goal card for a star of their choice. Sometimes you have a streak of bad luck and can’t get the last star you need to finish a constellation on your Myth card. Goal cards are approximately the same amount of points as a star, so it’s not too much of a feel bad, especially if you doubt you can finish a Goal card.
3.) Players can have a maximum of three goal cards. This pushes players to either complete their goals or turn them in for stars. Sometimes players can struggle with taking advantage of mechanisms that might help them. When they know Goal cards aren’t infinite, they’ll pay more attention to completing them.
In February 2022, I got to playtest the game with Nim’s final art. The effect of the game and the art was incredible when everything came together.
Myths and Goals in Perfect Balance
In February and March 2022, I finalized the balance for the Discovery Goals and the Myth cards. Getting these details right was what finally pulled all the systems and incentives together.
The point values had to be close enough that people didn’t feel bad when they got low-point Myths at the beginning of the game.
The point values had to represent the varying difficulties of the constellation combinations.
The point values had to be low enough that they didn’t overshadow the amount of points players received from the core actions of the game: discovering, placing stars, helping.
It’s hard to understate how valuable data merge was at this point to rapidly test the numbers.
My friend Mike M. pointed out that having a two-star constellation on your Myth card was a feel bad because those Myth cards scored less. (I had thought they were “easier”, which would be a good thing.) In reality, you ended up getting fewer points overall because of the point-per-star rule. A great example of how challenging assumptions can lead to better gameplay!
I showed off the pre-production copy I received in April 2022 throughout the rest of the year to friends, family, and attendees at Gen Con 2022…
and at SHUX (after which we made some last minute tweaks to ensure players would get their goal cards before their stars on their turn, make sure they could choose where to put stars based on a Discovery Goal, and weren’t confused by the numbers on the Night Track).
The game is now available for purchase, there are how-to videos available, and reviews are looking good!
Thanks to Helpers and Additional Playtesters!
Phillip S., Ashwin Kamath, Amelie Le-Roche, Amanda Panda, Jason W., Andrew W., the Woodford family, Kari J., Mike M., Jason H., Jay Bell, Joe S., Connor Wake, Sara J., Jim B., Tyler B., Jason Schneider, Scott, K., Jay, Rob M., Devon T., Carol, and more!