Designing “Panic”

This weekend I designed a game. I want to tell you how that went.

I was visiting my friend Greg Parsons in Columbus, Ohio. I’m helping him lay the groundwork for his new game publishing company, and part of that effort revolves around choosing and developing good products for his starting catalog. I remember doing the same thing for Cheapass Games way back in 1996, one year before Snake Plissken flew the Gullfire into New York.

The visit coincided with Buckeye Game Fest, the Columbus Area Boardgame Society’s annual convention. There are a lot of game designers in that group, and many of them guys had games to show and talk about. It was a great convention and I wish I could go to one every week. Well, maybe every other week.

On Thursday night, we decided to invent a new game. I sat down with Greg Parsons and his friend and fellow game designer Mick Sullivan. All three of us had partial game mechanics that we’d been playing with, and a little bit of each of them found its way into our new game. After about an hour of rapid iteration, we were playing our new game pretty much as it is now. In fact, it actually took longer to document than it took to create.

The game that we set out to improve was a horse racing game of Greg’s. The strongest points of that game were its secret information and bluffing. Players bet on the outcome of a horse race, which was determined by card drawing, and each player knew a different bit of information about what cards might be drawn into the race. But the volatility of the betting, and the unsatisfying final resolution of the race, made the game not hold together well.

Greg and I had talked about re-theming this game as a stock market game, and even though I was reluctant to work on a theme that had already been done a million times, we nevertheless started with a stock trading concept and said “what can we build that is new?”

We wanted to keep the secret information, but I wanted to make the resolution of the hand (the race) work differently. We decided to start from scratch on the hand resolution, hoping that this would lead us into better territory for the rest of the game.

One of my favorite games is Anaconda, another game of secret information and bluffing. It is a dealer’s choice poker variant. Players start with seven cards, and pass cards around the table to filter their hands. The card-passing mechanic in Anaconda works because the game is high-low. So, cards that are good for one player aren’t necessarily good for the next. We liked the card filtering but couldn’t copy the high-low part; instead, we needed to come up with a new reason why different players might love or hate the same card.

Mick told us about a card filtering game that he had been working on, a space-themed resource management game in which players looked at three cards, revealed one of them, kept another secret, and discarded a third. Where he placed each of these cards was relevant to his strategy, and although the cards were different colors, no player was specifically attached to any one color. Instead, a player’s final hand would dictate which color he was most committed to; if yellow wins the tactical game, and if you have the most yellow cards, you win. This mechanic was similar to a game of mine, Abandon Ship, in which players need “key cards” in their hands to enter specific lifeboats.

In Abandon Ship, which is played with a standard deck, there are four discard piles, one for each suit. At the end of the game, players must play a card from their hand in order to enter the lifeboat of the matching suit, and if there is a contest for a particular lifeboat, whoever holds the most of that suit wins the entire value of the lifeboat. This rule has the interesting side effect that, the more points a boat is worth, the fewer cards it requires to capture.

With all the game mechanics we’d thrown out, Greg, Mick and I set about to create our new game. We started with a standard deck of cards, and proposed the following rule: At the end of the game, there will be a resolution of hands based on the discards. There will be four discard piles in the center of the table, one for each suit, and these cards will dictate the value of cards remaining in players’ hands. This is somewhat like the race mechanic from Greg’s racing game, somewhat like the lifeboats from Abandon Ship, and a little bit like Mick’s space game.

We had a little trouble deciding whether the most common card, or the least common card, was to be the best. If it was the most common, as in Abandon Ship, then it would be interestingly difficult to score that card. But in terms of the story, it seemed more appropriate that a “flooded” market was devalued, and thus the least common card in the middle should be the most valuable in players’ hands. We tried both, but quickly settled on the latter.

At first we used the whole deck, and ignored the ranks of the cards. Thus, all wheat was wheat, similar to the suits in Pit (though not valued differently). We quickly realized that there were too many ties in the endgame, so we went to a numbers-and-suits model, and stripped the deck down to just the numbers 1 through 10. 

With those rules for hand resolution, we then tried to create a framework of play. Based on the flow of Anaconda, our first mechanic was this: start with a hand of nine cards, pass three, then two, then one card to the left, and then set your hand in the order that you wish to discard it. After each discard, uh, you discard the next card. Okay, there was really no point in discarding them one at a time, except for drama. 

In Anaconda, the play is very similar to that, but there is a betting round after each revelation. We didn’t have betting, so planning out the order of the discards was not relevant. So, we decided to reveal each card one at a time, so that we could choose the next card based on what was revealed in the previous round. This seemed to work, and we felt like we had the second half of the game, but not the first.

Because the theme of this game was stock trading, specifically “insider trading,” I threw another possible mechanic into the mix. Many years ago I wanted to create a game where players are stock traders. They can get “inside information” by paying to look at cards that no one else can see yet. The theoretical balance in this mechanic was that the player who pays for the inside information must still behave deceptively, or else he will risk giving away what he has paid for. 

I described the mechanic of paying to look ahead, and we decided to create a preliminary round based on that idea. We dealt three cards face down in the middle of the table. These cards would be part of the discard pile at the end of the game. Players would still pass cards 3, 2, and 1, but before each pass, players could bid (using a strange payment method, described below) to look at one of the face down cards. We would then pass three cards, reveal the first card to everyone, and bid on the second. Then we would pass two, reveal, bid, pass one, reveal the last card, and go to the discarding phase. 

This seemed like a decent play pattern, but we needed to figure out exactly how to bid. In keeping with the metaphor of being super-rich, we decided that we would not actually spend our bid; we would just reveal a card from our hand, and it would then stay locked in our hand. Because you now can’t get rid of it, and because it si face up, this card it gives other players critical information about your final hand, and possibly information about what you have seen in the face down cards.

Whether this form of “bid” was effective was a pure guess. We had to test it to find out. My hope was that players would bid somewhere between 1 and 10 points to get early information, and (with some tweaking to the size of the information pool) this turned out to be true. Hooray for landing on a new mechanic with blind luck! In fact, over the course of the game we bid as little as 2 points and as much as 15, so we made it clear that you could pay your bid with more than one card. This felt much better than artificially capping the bid at 10, the size of the largest card.

Another nod to the “rich people” theme was this: We decided that along with whatever points we scored in each round, which was on the order of 50 points, we would also get one million points guaranteed, because that’s what it’s like to be rich. So our final scores after five rounds were in the range of 5,000,250. This is funny on its own, but it’s also a subversive commentary on games that make the scores look close by forcing some minimum number of points on everyone. 

That was Thursday night. After the first night’s play session, I typed up a rulebook and shared it with the group. It’s found in full at the end of this article, complete with rules that we later changed. I’m including it as-is, as an example of a game in progress. Below it are the corrections that bring it up to today's version.

We played and tested the game for about two hours on Thursday. Then on Friday and Saturday we played many rounds of the game with whoever would sit in. We changed the working title from “Insiders” to “Panic,” and Greg made a handsome set of resource cards on Friday morning, including Oil, Pork Bellies, Coffee Beans, Corn, Wheat, and Gold. 

Our first concern on Friday was whether the game scaled up, because we had only played with three people. In our first 5-player game, we realized that the three cards in the middle were less information than we wanted, because they represented such a small fraction of the total discards: 3 out of 23. We tried using three stacks of two cards, and quickly settled on three stacks of three cards. 

I was concerned about the size of the stub (that is, the cards left over after the deal). I wanted a few, but not too many, cards to be left out of the game, so that it was harder for players to gauge exactly where every card is, but so that most of the deck still got used. This worked out nicely in playtest; if we used two card stacks for two and three players, and three card stacks for four and five players, the stubs and ratios worked out like this:

• 2 Players: 2-card stacks, stacks are 6/14 of market (43%), stub is 6 cards.
• 3 Players: 2-card stacks, stacks are 6/19 of market (31%), stub is 7 cards.
• 4 Players: 3-card stacks, stacks are 9/24 of market (38%), stub is 5 cards.
• 5 Players: 3-card stacks, stacks are 9/29 of market (31%), stub is 6 cards.

In each game there are N+1 suits for N players, so for example there are 50 cards in the 4-player deck. The other numbers in the game don’t change: the same number of rounds, the same size of starting hands, and the same number of discards. All in all, those numbers turned out quite well.

Along with tweaking the size of the starting stacks, we also dropped the game length to 3 rounds instead of 5. We timed the game only once, and it was just under 40 minutes for a 3-round game with 5 players. That comes out to 13 minutes per round, so another 2 rounds would make the game take an hour, which is certainly too long. There were some new players in each game, so I expect that the actual playing time will drop to about 10 minutes per round. 30 minutes seems perfect, and in playtest the time seemed to go quickly, which is always a good sign.

I was excited to see good play patterns emerge in the weekend’s tests. Players understood the importance of bidding, and yet didn’t lose ground if they never won a bid. Going negative was possible but rare. Players understood the mechanics of locking cards into their final hands, which I considered to be one of the most confusing rules. The other challenge, remembering the point values of the different commodities in the middle, also proved to be quite easy. After one round, players remembered the entire sequence of play, and could focus on strategy (which turns out to be fairly hard to grasp). The only real hassle in the game is sorting the discard piles, which we do after every discard round, but I’m beginning to get a sense for how to simplify that.

In playing the game myself, I enjoyed the sense of panic that I felt, and that I saw others feel, when they realized halfway through that they needed to start discarding big cards in bad suits. And of course, the surest indication that a game is good is that players want to play again, which they certainly did! At the end of the weekend we sold our test deck to one of the testers, and the three of us happily autographed the 10 Coffee card. 

As of this writing this game is just four days old, and there is a lot of development work yet ahead of us. We certainly need a way to distinguish it from the millions of stock market games already out there, and come up with a distinctive look and title. But considering the fun that we have all had playing it, I’m pretty sure we have a winner on our hands.

James Ernest, Monday October 17 2011

ORIGINAL RULES

Here is the rulebook I wrote up on Thursday night. I called the game “Insiders.” You can play from these rules if you also adopt the changes at the end.

• Insiders
• The card game of billionaire stock trading
• First Design Brief
• October 13, 2011
• James Ernest, Greg Parsons, and Mick Sullivan

Summary: Insiders is a bidding and bluffing card game about rich people playing the stock market. Players take the roles of billionaires who know more about the stock market than you ever will. Because the players are rich, everyone wins; so the object is to win more than everybody else.

Players: Intended for 2-5, initially tested with 3.

Components: You need the Insiders deck and a way to keep score. The deck contains one full commodity suit (Beans, wheat, oil, gold, uranium, and corn) for each player, plus one more. For example, in a four-player game, use a total of five suits. Each suit has ten cards valued 1 through 10. (As a proxy, for three players, we use ranks 1 through 10 from a standard card deck.) 

How to Win: Players can win and lose points on every hand, based on how they play their cards. In addition, each player scores one million points at the beginning of each hand. This is what it’s like to be a billionaire. The winner is the first player to break five million points. This usually happens to everyone on hand 5. If several players break five million points at the same time, the player with the highest score wins.

Setup: Choose one player to be dealer / scorekeeper. This player shuffles the deck and deals a hand of nine cards to every player. In addition, the dealer puts three cards face down in the middle of the table.

How Scoring will Work: Cards represent different commodities, which the players will buy and sell over the course of seven hands. At the end of the last hand, cards will be located in one of two places: in the center of the table, or “Marketplace,” or in some player’s hand.

The Marketplace: Cards in the Marketplace are undesirable. These represent the commodities that no one wants. Thus, the more of a commodity that is in the middle, the less it is worth to the players. 

Your Hand: At the end of the round, cards that are left in your hand are worth points based on how rare or common they are in the Marketplace. Those that are the most common in the marketplace will be worth negative points; those that are the rarest will be worth positive points. The exact scoring depends on the number of players, and is covered at the end of these rules.

Play consists of seven hands in two phases. Yes, that’s complicated. Bear with us. In Phase One, players will gain secret information about the marketplace and pass cards to each other. This happens three times. In Phase Two, players will manage their portfolios by discarding unwanted cards into the Marketplace. This happens four times. At the end, players will have a hand of five cards, and will score based on the final composition of the Marketplace.

Phase One: (Do this three times).

Each hand in Phase One happens in two steps. First, players bid to see the face of one of the face-down cards in the Marketplace. Then, players pass cards around the table.

The Bid: Players bid for the right to peek at the first card on the Marketplace. If this is the first hand of the game, the dealer bids first. If it is the first hand of a subsequent round, the player with the lowest score bids first. If it is the second or third hand of any round, the player who won the last bid bids first.

Players bid in point values, 1, 2, 3, etc. The winner of the bid must reveal one or more cards from his hand that have a total value equal to or higher than his bid. For example, if the bidder won with a bid of 5 points, he must show 5 or more points worth of cards. These cards become locked, and can not be discarded or passed away. They will stay in this player’s hand until scoring. Note: A player is not allowed to make a bid that he can’t pay, and is not allowed to reveal more than five cards, total, as a result of bidding, because his final hand will contain only five cards.

The winning bidder may also peek at one of the cards in the Marketplace. He does this before he pays his bid, so that he can use the information he gains to inform him of which cards are best to lock into his hand. Note that other players may learn something from the card he plays!

The Pass: After the bidder looks at the card and pays his bid, everyone must pass cards to the left (or right). On odd numbered game rounds, players always pass to the left. On even numbered game rounds, players always pass to the right. 

The number of cards you must pass is 3, 2, and 1. 3 cards on the first hand, 2 on the second, and 1 card on the third round. This is easy to remember because it is also the number of face-down cards in the Marketplace. The pass is simultaneous, so you can’t look at the cards you receive until after you have handed your cards away. 

Strategically, you want to pass cards in a way that makes your hand better and other hands worse. Unfortunately, until you have played a few games, this is hard to judge. But that’s the idea. 

After everyone has passed their cards, turn over the Marketplace card that the winning bidder peeked at. Now that card becomes public knowledge.

You will do this step a total of three times, with three bids followed by three ever-shrinking card passes. After that, you move on to Phase Two of the round. 

Phase Two: (Do this four times)

Simultaneously, every player discards one card from his hand. Sort these cards into the Marketplace so that it is clear which commodities have the most points, least points, and so on. You will do this four times, revealing and sorting the discards each time, at which point players should have five cards left (including any cards they have locked through bidding).

Scoring the Round: Cards in your hand are now worth positive or negative points, based on the number of players. 

Two Players: There are three commodities in a two-player game. The most common one is worth negative points, meaning that cards of that type count as negative values in your hand. The rarest one is worth positive points, and the middle one is worthless. If there is a tie for high, then both are worth positive points. If there is a tie for low, then both are worth zero. If there is a perfect tie, deal cards off the deck until the tie is broken. This never happens.

Three Players: When there are three players, there are four commodities. The most common (worst) is worth double negative points. That is, every card of that type in your hand counts against you at 2x. The second rarest is worth negative points (1x), the second rarest is worth positive points (1x) and the rarest is worth double points. If there is a tie at any rank, both commodities are worth the value of the better rank. If there is a three-way tie anywhere, deal cards from the deck until it is broken. This never happens.

Four Players: The point values in a four-player game are spread across five commodities, and they are -2x, -1x, 0, 1x, and 2x. The same tiebreaking rules apply.

Five Players: The point values in a five-player game are spread across six commodities, and run from -3x to +3x. The same tiebreaking rules apply.

Scoring: Each player scores the total point value of his hand, plus one million points, each round. The object is to break five million points, which nearly everyone will do on the fifth round. When this happens to more than one player at the same time, which it always does, the player with the highest score wins.

“Insiders” was designed in 2011 at the Buckeye Game Fest in Columbus, OH by James Ernest, Greg Parsons, and Mick Sullivan. This is the first draft. More will come soon.

Notes on these rules: We updated the game since this first draft, so if you want to play from these rules, here are the changes:

1. Play three rounds, not five. (Or play as long as you like, really)
2. The cards revealed in Phase 1 are sets of cards, not single cards. With 2 or 3 players, these are sets of 2 cards. With 4 or 5 players, they are sets of 3 cards. 
3. To be clear, the point values for different game sizes are as follows: Two Players: -1, 0, 1. Three players: -2, -1, 1, 2. Four players: -2, -1, 0, 1, 2. Five players: -3, -2, -1, 1, 2, 3. We have not tested with six players, but there would be seven suits and the points would be -3, -2, -1, 0, 1, 2, 3. 
4. The tiebreaking rules are different: Any slot that is tied with the one above it slides forward into the higher slot. So if there are six commodities, and three of them are tied for 2nd place, then the final values are -3, -2, (empty), (empty), 2 (which contains the three tied commodities) and 3. There is no random card drawing to break ties. If all six commodities were somehow tied, they would all be worth 3x points. As we like to say above, this never happens.