Monthly Archives: March 2021

Making an Impressive Impress-the-Judge Game

Posted on by Jeremy Geist

The impress-the-judge game is one of the pillars of party gaming, a valuable social lubricant for parties of nerds and non-nerds alike since the early 2000s. However, more serious designers have largely turned up their noses toward the genre; this is in no small part due to 500-pound gorilla Cards Against Humanity forcing out competitors and earning the genre the undeserving reputation of being filled with offensive “punch-down” humor. This is a huge waste, as a well-designed impress-the-judge game can bring together new players and be a great time for longtime friends alike. I like to consider myself a scholar of the subgenre, and in this article I would like to go in depth on how to make a great impress-the-judge game.

(Before I begin the article in earnest, I’ll define the impress-the-judge genre for those not super aware of game design topics. The basic concept of the impress-the-judge game is that one player each round is the “judge”, who sets forth a prompt, and each other player has to come up with a funny answer that the judge will pick. Usually, both prompts and answers are premade cards, but several excellent games have gone with other options.)

Blind Judgment

As I mentioned in my previous article on the subject, all party games are comedy Legos, designed to help tables of people find common ground through laughter. This can range from complete strangers at a convention to families who have known each other for ages – no amount of familiarity or inherent comedy skill can’t be enhanced by a good party game. Some kinds of party games are better suited to some groups than others; games that rely on your knowledge of other people, like Wavelength and Hive Mind, are better for people who mostly know each other beforehand, while trivia games like Wits and Wagers are good icebreakers for strangers but don’t add as much to preëstablished groups.

Impress-the-judge are one of the most group-agnostic types of party games. All of the humor is contained on the cards and only requires a modicum of effort for players to rehydrate, which means that no matter how poor a group is at making jokes naturally, it should be easy for them to regularly get a laugh out of everyone most of the time. This also means that the impress-the-judge game is the most reliant party game genre on the designer’s sense of humor and ability to write funny cards.

All impress-the-judge games strike a balance between rewarding writing skills (picking the best card) and acting skills (being able to sell the judge on the game you made). Apples to Apples and Cards Against Humanity are 100% reliant on writing skills, with anonymous submission systems that heavily disincentivize players from promoting their own answer. This has its upsides – the surprise when the shyest member of your group has a slam-dunk answer – and downsides – the despair of having your great card tossed aside because the judge didn’t get the reference and you couldn’t explain it.

Other games rely more on acting skill, though none fully reward acting skill without writing skill, as that is short-form improv and not a board game. Snake Oil is probably the widest-spread of these, though Shut Up & Sit Down darling Funemployed ventures even further into the acting side of the genre, as you have to use every card in your hand, so the ability to select the best one for the situation is taken away from you.

Your first decision when making an impress-the-judge game is where on this spectrum you want your game to fall. Your next decision is how, and if, you want to fix a few of the genre’s glaring problems.

The Three Sinkholes of Impress-the-Judge Games

The impress-the-judge game doesn’t have “pillars” of good game design holding it up; instead, it’s an excellent and time-tested gameplay base with a few large but patchable flaws. Deciding whether to fix these sinkholes at the risk of adding poisonous complexity to your game will inform much of your design work.

Downtime

The enemy of all party games is downtime. You want your players to be engaged with, or at least entertained by, the game at all times, and unfortunately, the base impress-the-judge game has two areas that frequently produce downtime for most players.

The first is in the answer selection process. Generally, each player has as much time as they want to select or create a funny answer. This means that, with the exception of the slowest player in your group, all the other players are going to spend a lot of this phase waiting around for something different to happen. There aren’t any jokes or funny moments here, as everyone wants to keep their answers a secret so as to spoil the surprise (and win the round).

The most common solution to having too much downtime here is to incentivize players to work quickly. In Stinker, there is no preset judge, with the judge instead being the player who fails to assemble an answer before everyone else. This makes players work much more quickly, and also increases the general quality of each answer because the player having the most trouble with a good answer becomes the judge by default. I took a different tack when designing Stand Back, Citizen!. Each answer in SBC! is comprised of two different types of cards, with all of the cards of one of these types dealt face-up into the table; when the round starts, you have to pick a good card from the center and create an answer before an opponent can snatch it from you. Under this system, there will still be one player left trying to figure out what card to play, but their squirming as they’re faced with a single subpar option makes the time a source of comedy instead of frustration.

The second likely area for downtime is during judging. This is much more of an issue for games with an anonymous judging system, as arguing in favor of your answer, or against an opponent’s answer, or even for an opponent’s answer if you know you don’t have a chance but really liked Uncle Jeff’s pitch, is great fun and takes a lot of the dreariness out of the process. The game that came the closest to solving this is the Quiplash series, which inverts the process by having only two answers for each prompt and making all of the other players judges. This was helped in part by its status as a digital game, but I could see this method working for tabletop games as well.

Judge Bias

Because impress-the-judge games rely on the subjective judgments of the other players, there is naturally an issue of a judge picking a particular answer either because it appeals to them specifically or because they’re biased towards one of the players at the table (for example, because they are dating). The early anonymous judging games were anonymous because of this second issue; I would argue that this is somewhat overblown and more of a problem with the gaming group than with the design of the game itself. It is true, however, that losing because someone else played a card that matched the judge’s hobbies or interests is one of the worst feelings in impress-the-judge games.

There haven’t been a whole lot of advances in this area, in part because judge bias is one of the flaws that players are most likely to accept as built in to the genre as a whole; if you wanted your contribution to not be measured subjectively, you could play one of thousands of strategy games, or even a more objective party game like Codenames. Stinker, perhaps unintentionally, went some way towards solving this by randomizing the judge after people lock in their answers; this might be a place where further advancement could be found.

The Novelty Wearing Off

Party games are second only to legacy games in their reliance on novelty. After all, comedy is based out of the unexpected, so if you’ve seen one funny answer already it’s not going to be nearly as impactful the next time. When you have a set of cards with all the jokes pre-written on them, the novelty will wear out sooner rather than later.

The easiest way to solve this is to vary up the “answer” components so players can write either anything (as in Quiplash or Say Anything, where players can even draw) or almost anything (as in Stinker, which uses letter tiles). However, this stumbles into a secondary trap, which Roger van Oech described in his book A Whack on the Side of the Head. Without some way of forcing them to create new answers, players are most likely to stick to the first things they can think of, which inevitably become relatively conservative answers that everyone’s heard before. So if you’re making an impress-the-judge game with totally open answers, you have to focus on creating lots and lots of excellent prompts.

If you’re using cards, consider using a combination of cards to massively increase the permutations that players can achieve. All of the answer cards in Snake Oil are a single word, and players combine two of them to create a product to pitch to the judge; because they’re built to change depending on the other word you pair them with, this creates a much larger variety of responses than could be generated through single cards. I used a similar method when creating Stand Back, Citizen!, though that uses an asymmetrical combination of fill-in-the-blank “power” templates and silly nouns – there’s half as many permutations, but they’re more directed and more likely to be a bullseye than cards that can be put in either order.

Order in the Court

Besides the actual prompts and answers, you should also think about the less glamorous nuts and bolts of your game; in particular, thinking about the scoring system and end condition.

You might think that because impress-the-judge games are party games for casual audiences, that whatever scoring system you create doesn’t matter because people aren’t going to pay attention to them. While it is true that party games are more about the journey than the destination, your scoring system is your best way to incentivize your players into being funny and shouldn’t be discarded offhand. Stinker is one of the best examples here: If your answer gets chosen, you get one point for every letter tile you used in it, so you’re encouraged to make longer and more elaborate answers (balanced with not becoming the judge because you spent too long moving tiles around).

A scoring system more complex than “judge picks the winner” can also introduce new skills to the game. In Say Anything, you get points if the judge picks your answer as usual, but you can also earn points by guessing which answer the judge is going to pick. This turns Say Anything into a hybrid of the impress-the-judge game and the venerable “know your friends” subgenre and helps it stand out.

The end of the game is also very important, because it’s critical that a party game not overstay its welcome. Apples to Apples had a huge issue with this, because the game ended when one player won a certain number of times – not only is this obviously problematic with a large player count, it meant that the game could drag on forever as everyone collected N-1 green apple cards, or end swiftly as some cut-up got all of them in a row. Cards Against Humanity has no end condition at all, which will frequently mean it goes on too long, but I suppose was insightful into how party games usually just sort of end anticlimactically in most of the situations in which they’re played.

I generally like going until everyone’s been the judge once. It means that the game is often too short at 4 or 5 players, but it’s better to undershoot the ideal playtime than overshoot, as it’s psychologically easier to just start a second round than cancel a too-long game in progress. It also gives everyone a chance to judge and many chances to create answers. If your game has more mechanical trappings and is expected to go a little longer, like Say Anything, going around the table twice can also make sense. Go with what works for your game and your players.

Conclusion

The actual biggest flaw with impress-the-judge games is that they’re tough to publish after both the wild success of Cards Against Humanity and the deluge of terrible imitators in the early 2010s. However, only sticking to games that are easy to publish means you’re constantly playing catchup with whatever trend is rocking BGG (or, more realistically, rocked BGG two years ago). At worst, designing a good impress-the-judge game means you’ll make your friend group very happy; at best, you can make thousands of people all over the world very happy. As long as your game deserves to exist and advances the medium, there’s absolutely nothing wrong with trying to compete in a saturated category.

Balance, On Average

Posted on by Jeremy Geist

When you think about game balance, you probably think about making all possible player options be as equitable as possible; about maintaining a fragile balance between each player’s starting position so everyone has an equal chance of winning every game. But what if I were to tell you that there was a different way to balance your games? What if that way combined a sense of fair play with all the excitement, desffpair, and storytelling opportunities that came with unequal positions? Well, good news! I’m not speaking hypothetically. Welcome to an article about making your game balanced…on average.

Dr. Ro-Galt-Nik’s Mean Bean Machine

When I was a kid, my parents would regularly take me to see the Museum of Science and Industry in Chicago. During one of my visits, I stumbled across a giant version of this thing, which was one of the coolest things I had ever seen:

This Galton Board, also known as a “bean machine”, is a neat visual representation of the natural law of regression toward the mean. Even if you have a function with a wide range of outcomes, if you run that function often enough, it will eventually deliver the distribution of results you expect, with a hill centered on the mathematical mean outcome. To use plainer language, everything will even out as expected, provided you try enough times.

What this means for us as game designers is that there is another option besides trying to make each player’s starting position equal every time: We can also have a wide variety of unequal starting positions and redistribute these starting positions so many times it regresses toward the mean.

I’m Pretty Sure The Genius Did This Once

That last section was pretty abstract, so to make it more practical, I’ll use an example. Let’s look at a game that’s very much in need of redesigning: I’ll call it “Small War”.

Small War is played between two players using part of a standard pack of playing cards: The 2 through 10 of hearts and the 2 through 10 of clubs. These cards are shuffled together and then dealt evenly to both players. Then, each player selects a card and puts it face-down. Once they’ve both chosen a card, they turn them face-up; the player who played the higher card “captures” the other card and forms a pair. Whoever has the most pairs at the end of the game wins.

So obviously, Small War has some problems. The most obvious one is that, because the cards are randomly distributed across both players, it can create a heavily imbalanced scenario where one player might have both 2s, 3s, and 4s and the other has both 8s, 9s, and 10s. How can we fix this so every game of Small War is equitable to both players?

A conventionally balanced version of Small War would simply distribute one copy of each card, 2-10, to each player every game. This turns the game from one based on luck to a tense mindgame where each player tries to figure out when their opponent is going to try to pass their small cards through or try to win the round with a big card. This is cool and everything, but the meta would crystallize quickly after a few rounds and the game would become stale for all but the most hardcore players.

However, there is another way: Keep the gameplay exactly the same but make the game best out of 9 rounds (1 round being what we’d call a “game” in the original). Individual rounds might be extremely unfair, but there’s so many rounds that there’s a good chance of both players’ luck of the draw cancelling out each others’ over time.  Plus, this wider range of opening hand adds more variety, prevents the game from boiling down to set opening and closing strategies, and creates amazing stories where you get dealt an awful hand but manage to overcome your opponent afterwards.

This second case is what Balance On Average (BOA) design is all about.

Defining The BOA

BOA games have a few key features.

  • A BOA game is made up of a number of individual rounds, ending after a set number or when a player reaches a particular score.
  • Each round begins with players being given a random assortment of components to make do with. The distribution of components is random and each player’s respective strength can be unequal, sometimes brutally so.
  • The results of each individual round don’t have anything to do with the rounds after it, except in a “metagame” sense where you might adjust your play based on your score or the amount of time left in the game.

You might be wondering, why even make a game like this? Well, compared to a conventionally balanced game, BOA games have a number of benefits.

  • Compared to a regular game, where being put in a bad position early in the game will make you lag behind for 90 minutes, a BOA game can get away with being pointlessly cruel to players because the round will be over in 5 minutes anyway. In fact, cruelty often engenders a desire to play again later and to laugh about it after the fact.
  • There’s a wider variety of game states in a BOA game. In a conventionally balanced game, the designer (and developer sometimes) works hard to ensure that every player starts the round at around a 5/10 and maintains that for most of the game. BOA games can deal out gamestates ranging from a 0 to a 12, greatly increasing the number of possibilities.
  • BOA games can be deeply emotionally impactful. Getting a strong hand and doing well is more likely to stick in your memory if you’ve dealt with medium-to-weak hands for the last 5 rounds; getting a come-from-behind victory is more meaningful if you were actually behind the other players in a way beyond “number of games played”.

The BOA In Action

The majority of BOA games are traditional games hailing from long before the advent of the modern “designer game”. The one Americans are probably the most familiar with is poker. You might have a terrible hand in one round and a godly hand in the next, but besides the metagame of playing more or less aggressively depending on how many chips you have compared to your opponents, the outcome of one round doesn’t affect the performance of your hand in the next.

Other card or card-equivalent games that make use of BOA include bridge, rummy, and mahjong. Much of the strategy of any of these games involves not only capitalizing on good rounds, but knowing when your starting hand is good compared to the other player or team’s and throwing away your hand as safely as possible if you you can’t catch up. For games with an easier learning curve, Uno is the textbook definition of BOA (as the distribution of wilds and special cards like draw 2s unevenly weighs starting hands), though presumably many groups only play one game at a time.

The majority of designer games that use BOA balancing are trick-taking games, likely because trick-taking card games have more fidelity to the classics of the genre than most other kinds of game. Some, like The Bottle Imp, adhere strictly to keeping rounds separate and tallying up scores on a piece of paper; others allow you to mitigate a bad hand (Nyet!) or give players a way to catch up after a nasty round (Tournament at Camelot). You could also charitably describe impress-the-judge games like Stand Back, Citizen! as BOA, since the suitability of your hand to the prompt will change each round.

Some games that look like BOA games but aren’t include “dude smasher” games like Magic, Hearthstone, and Netrunner (as you usually play three rounds max against a single opponent); drafting games like Bunny Kingdom or 7 Wonders (as your actions in earlier rounds dictate what you do in later ones); and fast social deduction games like One Night Ultimate Werewolf (as you’re not playing multiple rounds in one game – you’re playing multiple very short games).

One of the most interesting interpretations of BOA design is Condottiere. Condottiere is a sort of hybrid between BOA and conventional design because “getting a new hand” and “starting a new round” isn’t a 1:1 ratio: Players draw new hands once all but one player are out of cards. This means that players with bad hands can force balance on average by increasing the number of hands per game, while a player with a great hand might be tempted to string their hand across multiple rounds while flirting with the disaster of throwing away most of their good cards. I still haven’t seen this mechanic anywhere else meaningful, and it remains a great source of inspiration for me.

BOA Constrictions

Despite the several advantages of BOA design, games of that nature have a tiny foothold in the tabletop game industry as a whole, for reasons both meaningful and coincidental.

Part of the reason for the modern unpopularity of BOA design might be due to the meteoric rise of video games right around when the properties of a “well-designed” board game were starting to become mainstream around designers. It stood to reason that a balanced board game should resemble a balanced video game, with players selecting an asymmetric faction, strategy, or character near the beginning of the game and having a roughly even chance of winning with it at equal player skills. 

Another, perhaps more legitimate reason is that it’s hard to build a sense of momentum in a BOA game because each round is specifically detached from the next. You can have excitement and progress within a round, but compared to an Agricola or Dominion where a player’s engine and strategy slowly builds up to a splashy finish, BOA games frequently feel more “even”, with emotional impact primarily arising from unusual situations that may not always occur every game.

BOA games also have many more restrictions than conventional games. As discussed, BOA games have to have a long series of short rounds, each of which randomly redistributes starting positions. There’s only so many ways you can achieve this, and many of the good ones have been covered by conventional games. There are additional UX concerns with this system, as you need the redistribution itself to be relatively quick and painless (like shuffling and dealing out a deck of cards, as is common) and the scorekeeping to be something more modern than someone having to grab a pencil and sheet of paper. That said, I believe that BOA games are so overlooked by modern tabletop designers that their true potential remains unexplored – maybe there’s some mechanic hiding in the BOA space waiting to revolutionize the industry.

Conclusion

All game designers can fall victim to a kind of design myopia, where we grow so used to the evaluations and demands of our particular slice of the community that we forget that there are many ways to make a fun, dynamic game, some less mainstream than others. Looking at BOA games has been extremely helpful for me in this regard, and I hope to one day find even more interesting methods of balancing a game and keeping its players happy.