by lightinthelab May 31, 2018

Hello! I’m Dr Adam Symonds, games designer for Light In The Lab!

After spending two hours of my life wrestling with the farcically bad “Subbuteo Angling” game recently, I was asked by one of my playtesters- “What’s the worst card you’ve ever seen in a game?”

Interesting question! It got me thinking, and I immediately conjured up a huge list which needed narrowing down. First, I limited myself to considering only cards from otherwise good games- otherwise I could easily identify any number of obviously terrible cards from obviously terrible games. Second, I made my decision based on how the card impacts gameplay, not any aesthetic merits, for similar reasons. The quality of an ugly, poorly laid-out card is pretty clear- I wanted to pick a card that was not, at first glance, obviously bad.

After spending some time sifting through my collection, there was one candidate which stood out clearly from the pack as actively terrible, rather than merely passively bad. Ladies and gentlemen, I give you “The Vault”, a crisis card from the (otherwise excellent) Dead of Winter: The Long Night expansion:

I’ll begin my analysis in a moment. But first, a few important things to note:

Firstly, does this single card, in an otherwise excellent game, deserve so many words devoted to it? No, absolutely not! At the end of the day, if you don’t like the card, simply house rule it or remove it completely. (Though this is often a poor solution which can cause unforeseen consequences, when a card is this bad, it’s easily an improvement). There are many, many worse effects in many, many worse games more deserving of ire (a good selection of which can be found in “Subbuteo Angling”).

However, I thought it would be useful to perform a very deep dive, a detailed analysis, complete with asides and background supporting theory, of this one tiny part of a much larger game. I hope to try and illustrate the ramifications these elements can have if not designed with extreme rigour. Given my background in academia, this is something of a speciality!

Secondly, attempting to push the envelope when designing games and their content is laudable, and something I feel should be done more often. Plaid Hat Games deserve credit for trying to pull off something so gutsy- in an expansion pack of all places- even if it failed so spectacularly.

Thirdly, for those of you who haven’t played Dead of Winter- and you should- many of the details on the card- such as the ‘3&4’, ‘tools equal to non-exiled players’, and optional ‘morale gain’- are basic qualities of a crisis card. These mechanics work fine, and won’t be mentioned again. (In short, the players face one crisis card each round and collectively expend resources to overcome it, else face the ‘fail’ condition.)

Finally, I’m well aware (and the exasperation of my proofreaders confirms) that this is somewhat lengthier than the typical short, snappy articles with titles like this. Unfortunately, many of these concepts are quite nuanced and need to be explained in sufficient detail- shortening would only make them less clear and harder to understand. Chin up- I could have made this twice as long if I’d indulged in a few more tangents!

Okay, so, I’ll start with the more minor criticisms, and move onwards and upwards towards the main point.

1) The Vault fails to evoke its intended theme.

The theme of Dead of Winter is a combination of a modern post-apocalyptic survival, and homage to classic zombie horror tropes. We see in the art that the vault door has been opened, leaking what we presume to be blood, and warnings against opening this vault have been written on the walls, also in blood.

So… did the people writing the warnings before the vault was open write their messages in blood? Why? In the game, this vault is situated in an advanced scientific complex- were no better materials available? Then, did the vault just coincidentally happen to contain (among presumably other things) blood? Alternatively, were these warnings written after the vault was opened, using the blood ‘created’? But then what would be the point? The vault is already open, its contents released.

Worldbuilding is very important to create a coherent view into your game and allow players to understand where they sit in relation to it. The ‘crunch’- the intersection of your fluff (story, theme) and your mechanics (how the game actually works) actually helps players understand how to play your game. If you use the same motif in multiple places- such as a particular image, symbol, or colour- we create the expectation in the minds of our players that those things are linked.

As a basic example, we understand that the white chess pieces are controlled by one player and the black pieces by another player, even though they have different shapes, sizes, rules, and occupy all sorts of different places on the board. This sort of pattern recognition and association-building is a powerful tool of the human psyche, which we can take advantage of to structure our games in a way that players can learn more easily.

At first glance, the connection between opening a vault and revealing secrets seems clear. However, think carefully about tying the mechanical effect- gaining zombies or revealing secrets- to the fluff effect- whether or not to open a sinister vault:

• Which of these two options corresponds to ‘opening the vault’?
• Which is the mechanical effect of ‘leaving the vault alone’?
• Does the vault contain zombies- as indicated by the art? Or secrets?
• Are secrets somehow being used to keep the vault closed? The art shows that this incredibly sturdy door won’t (or didn’t) open by itself!
• Is it zombies AND secrets? But then where is the ‘ignore it’ option?
• Why are we only letting zombies OR secrets out?
• And so on…

The terrible crunch here simply does not help players to understand the card or feel a connection to the game. Unfortunately, this card needs these things more than most others, because…

2) The Vault produces a unique effect with no reference points.

We want the effects in our game to be understandable to all players, including, ideally, players coming at our game for the first time. A good way to do this is to build on and reinforce pre-existing patterns already in our game. We can design our content in a way that introduces new concepts gradually and by showing them in relation to things our players already understand. This reduces the mental overhead on our players from tracking different effects- players find it less stressful and more fun, although most aren’t consciously aware of exactly why.

This idea is known as ‘templating’, and the best place I’ve seen it done is in Magic: The Gathering. Magic has a HUGE number of different effects- there are about fifteen thousand unique cards at this point. Templating breaks information and instructions down into smaller, neater chunks so we can more easily understand them, assists players in performing simple tasks when simple triggers occur, and helps us learn in a generalisable, modular manner.

For example, say I’ve learned what to do when a card says ‘When this enters the battlefield, draw a card’ and, separately, ‘When this dies, gain 4 life’. If I now see a completely new card which says ‘When this enters the battlefield, gain 4 life’, I can easily understand how to execute this novel instruction, even if I’ve never seen this effect before, because these cards have been carefully designed and written to work in this way.

If these effects were written haphazardly, or even just with a little less rigour- say one effect tells us to ‘gain 4 life’, and another to ‘increase our life by 4’- we immediately lose this synergy. Again, the human brain is amazingly good at recognising patterns- to the point where not finding a pattern where we expect to see one can cause measurable psychological stress, on top of the stress from trying to handle these two things separately.

So, back to The Vault. “Reveal a non-game-related secret they have never revealed to their fellow players before” is extremely difficult to interpret off the bat. It can’t be compared to or contrasted against any other card, or indeed any other effect in the entire game. This raises a relatively large number of awkward questions:

• What counts as a ‘secret’? If I put on green socks this morning, concealed them under baggy trousers, and no one has gone to great efforts to see them, then my sock colour is a secret. Most groups will disallow something this lame- but the decision of what to allow and what to disallow is very tricky.
• Is ‘fellow players’ a strict set? In other words, must the secret be something which is not known to any of them? Or is it something which is not known to all of them?
• Are there any restrictions on what players can do with that information after the game?

Note that almost no cards go into this level of detail on exactly how they work. Is this because their effects are intrinsically more complex than those of The Vault? Not at all- they’ve simply (consciously or otherwise) used templating to establish how some, if not all of their components are handled in advance.

In addition, it’s tempting to say that some of the above questions can be addressed through Dead of Winter’s basic rules- for example, there’s nothing in the rule book which prohibits telling people about the game afterwards, so you can share any revealed secrets. However, the fact that this is such a unique effect, in a manual without an FAQ section, means it would be extremely optimistic to argue that this was intended.

Now, in many cases unique effects are cool. Sometimes you’ll come up with something which is so great that it’s absolutely worth including in your game, even if it’s noticeably more awkward to handle than other cards of it’s type. Here, we can look at the card’s expected impact on, and value to the game, to see if it’s worth keeping.

3) The Vault has an awful expected value.

I find it helpful to envisage best-, average-, and worst-case-scenarios for the content I’m adding to my games. For almost all content, it’s possible to construct a situation where it detracts from the experience of a game being played. The question is then “Given this, is it worth including this effect in my game?”.

This is the notion of expected value, or e(x), from game theory, albeit viewed one level higher as a designer. Just as players want to make good moves to maximise their utility within a game, we want to make good content for our games so the people playing them will maximise their utility by enjoying it.

Let’s consider this for The Vault. (I’ll consider the possibilities only from the point where the crisis is failed- otherwise the text effect of the card could be literally anything and it wouldn’t resolve):

Best Case Scenario:

• Some players reveal secrets which are thematically appropriate and entertaining to the group- perhaps an unusual but harmless phobia, or an atypical reaction to a survival situation.
• After some brief but interesting discussion of this new information, the game continues promptly.

Average Case Scenario:

• At least one player is at least somewhat uncomfortable revealing a secret.
• Players have different ideas as to what sort of secrets are permissible (as explored above).
• Players have different valuations of the impact different numbers of zombies will have (as explored below).
• None of these issues have clear resolutions.
• Most of these issues must be resolved before gameplay can progress past the resolution of the effect.
• All of these issues remove focus from the game itself.

Worst Case Scenario:

• A player reveals a secret which causes the game to immediately end. (Not as a deliberate ‘table flip’, but simply by misjudging the intricate social dynamics of the group, and bringing something up that causes at least one player to suddenly have more pressing concerns than finishing the game.)

This is possibly the worst, worst-case-scenario that can exist for any effect in a game. Now, a bad worst-case isn’t necessary the death of an effect- I’ll include effects with bad worst cases if I think they’re unlikely to come up, and are counterbalanced by something worthwhile in the average or even best cases. To be fair, I don’t think this worst case will come up very often at all. However, the average case is still really bad- most groups will have quite a bit of difficulty handling this effect. And the best case, while slightly more interesting than a typical crisis card, does nowhere near enough work to justify this effect’s inclusion.

Thankfully, the saving grace of effects like these is that players usually have their own ways of dealing with these. Most groups are able to interpret and police these effects, even if they come up mid-way through a game, before they get out of hand.

4) The Vault is actively difficult to interpret and police due to hidden asymmetry.

In Dead of Winter, the players are cooperating to survive a bitter Canadian winter, which is also filled with hordes of zombies. However, there may be a traitor amongst the group, who needs the group to fail its collective objective to win. (The situation is actually slightly more complex, as depending on secret objectives and board state some non-traitor players may wish to hinder certain cooperative actions in situations where the “traitor” may or may not. For simplicity, I’ll use “traitor” to refer to a player who is playing non-cooperatively at the given moment, for whatever reason).

Different groups have a different ‘culture’, a different set of meta-rules governing how they approach games (and other situations). It is not the case that any culture is ‘right’ or ‘wrong’, though we may observe them and critique their rules based on any metric we desire. For example, one group of players may generally allow moves to be taken back and replayed if a mistake is noticed immediately, whereas even asking for such a concession could be seen as ‘bad sport’ in another group.

Now, when a traitorous players looks at The Vault, a lot of its value is inverted for them compared to a non-traitorous. They’re incentivised to get more zombies in play. In this case, this is achieved if players choose not to reveal secrets.

Now, how could a traitor reduces the odds of players revealing secrets? Three methods immediately spring to mind:

1) Create social pressure on players not to reveal secrets. (“Hey, everyone, shut up so Bob can tell us a secret! Haha, come on Bob, speak up! I bet it’s something really dirty!”)

2) Invoke the spirit of the game to dismiss trivial secrets permitted by the effect in the hopes players will be unwilling to reveal something deeper. (“No, you can’t just say what colour your socks are. Come on, that’s lame. Tell us something GOOD!”)

3) Most simply, they could pretend to be extremely uncomfortable about the whole idea. (“Look, guys, seriously, I don’t want to, okay? Stop pushing me and just add the damn zombies.”)

I am not suggesting that any given player would deliberately set out to make others feel uncomfortable or unwelcome for in-game advantage (although, unfortunately, it is a behaviour I’ve observed). The psychology of why players “cheat” is a very complex question, but, in essence, it happens very naturally amongst almost all humans at a point where there’s an incentive to do so, and they don’t think they can be caught.

However, the human brain is an amazingly advanced machine, capable of finding and exploiting flaws in systems with ruthless efficiency and without conscious thought. The lack of clarity around this effect, as we’ve explored, increases the number of “social moves” here which can be taken without chance of detection, and further increases the risk of this negative behaviour occurring.

By the time an invested player notices and corrects their behaviour, it may be too late. But, at least all the loyal players will be pulling together to tackle this card as a team, right? Well…

5) The Vault attempts to assign real-world value to in-game actions

Although this is, in my opinion, a remarkably fruitful avenue for further investigation and development, it must be handled with extreme care. Dares, bets, and wagers can add spice to even the most trivial ‘game’, such as spectating a sporting event or competition.

Similarly, playing for some reward or ranking outside of the game itself, no matter how insignificant (such as placing on a high score table or unlocking more content) is a powerful incentive for further play. (Broadly speaking, we understand the threat of a negative outcome to be equivalent to the chance of a positive outcome in mechanism design.)

However, problems arise when players are asked to compare in-game value with real-world value:

• Which of my secrets are worth less than two zombies?
• Are your secrets worth more or less than my secrets?
• What’s the expected value of this secret’s effect on my relationship with these people, vs the expected value of two zombies on this game state?

Another difficulty with this valuation is, unlike almost every other category of game action, we as designers care if the players get it wrong. This is an unusual notion- players playing games less-than-optimally is often our bread and butter, the way by which skills are tested, players meaningfully interact, winners are determined, and consequently fun is had. We do not tweak our games because a player played a bad card, or made a weak move.

However, a player who overvalues their secrets forsakes the opportunity to interact with our cool mechanic unnecessarily. A player who undervalues their secrets reveals them when they should not, with potentially worrisome real-world consequences as discussed.

There’s also often a sense of tribalism about these different valuations. Players who assign their games and actions relatively high levels of real-world worth are lauded as ‘hardcore’, while others are ‘casuals’. Needless to say, this both compounds the players’ analyses and introduces an additional element of feel-bad which would otherwise simply not be present in the game.

Now, for the big one…

6) The card completely violates Dead of Winter’s social contract.

A lot of the design theory that we take for granted these days has only really been around for the last 20 years, at most. Even the most rudimentary notions of game theory, such as Nash equilibria, were only conceptualised about 70 years ago. In the consideration of all the games made by humans, that’s barely a blink. There are always, always more ideas we can try, more theories we can formalise, and fundamentally more we can learn about designing things well. It took this card for me to realise that one particular idea has yet to fully transition into board gaming, being generally more at home in sociology and, in terms of gaming, larp.

“Social contract theory” was originally formulated as a branch of political philosophy dealing with the interaction between rational individuals and the authority of governments. More recently, it’s been reformulated to deal with questions of entering into agreements with broader communities and activities.

At its simplest, social contract theory says that individuals and communities form agreements, which are often de facto unspoken, about the rights and responsibilities of both when they interact. In the modern world, our ‘community’ is not a simple thing to define- in fact each of us is, in varying degrees, active in a variety of different communities at a time.

For instance, I might be eating a pleasant meal out in a fancy restaurant with my family while texting an old friend, in which case I’m actively participating in several “communities”, each with their own rules and ways of doing things:

• My friend. (“Take pictures of anything funny, don’t talk about exes”)
• My family. (“Feign interest in your little cousin, don’t take the last piece of garlic bread”)
• The restaurant. (“Don’t bring outside food, tip 10%”)
• The country. (“Don’t steal, you won’t be stolen from”)
• Common decency to my fellow man. (The golden rule of ethics, “Do unto others as you would have them do unto you”)

There’s some overlap, some separation, and some communities care about things which others don’t. However, we can easily distinguish between these communities, and study how they operate individually.

Now, because of my pre-established awareness of what a restaurant is, I understand the sorts of expectations which that restaurant and its staff are likely to have of me. I don’t need to sign anything or even voice my explicit consent to this before I walk in- by choosing to enter the building, I willingly accept these restrictions for as long as I’m active in that ‘community’. This implicit two-way agreement is the ‘social contract’ itself.

Furthermore, entering these communities and loudly demanding that they do things another way would ‘violate’ the social contract. Even if the way those communities are operating is questionable, or even objectionable- perhaps I’m an ardent vegan and the restaurant serves meat. By entering that community, I’m implicitly accepting some expectations and restrictions on what will and won’t happen and, in return, I have expectations and restrict the things the community can do to me. I won’t kick up a fuss about what’s on the menu or not being able to eat outside food (“It’s a free country! I can do what I want!”), and in return, I can expect the restaurant will ensure people bring me food and drink.

However, problems arise when an entity places an expectation on us which we did not assume or understand. Let’s say this restaurant adds a 50% charge at the end of the meal for people wearing rainbow hoodies. In this instance, I would quite rightly object on the basis that this rule wasn’t made clear to me before I ordered. These unusual expectations must be clearly presented up front, so everyone understands what they’re agreeing to when the interaction begins.

Critically, there’s nothing wrong with these expectations– personally I think more communities should try new ways of doing things! However, if they’re different from what you’d expect, they must be made explicit before the interaction occurs. Then, people can decide if they wish to participate in that community (and enter the restaurant), or not (and go elsewhere). This is an effective model for making sure peoples’ rights are generally respected and interactions are generally harmonious.

These concepts come into their own when applied to larp. For example, suppose I sign up to a larp about surviving horrific monsters in the dead of night. I can reasonably expect, even without investigating further, that the game might be a frightening, or even in some way uncomfortable, experience. I might be in all sorts of situations I wouldn’t normally allow myself to be put in- like fleeing for my life in a dark, muddy wood. Notably in larp, I experience these unpleasant sensations directly, as a person in the world (not just a player in the game). However, we can run games with these themes and mechanics, because players understand and accept their social contract.

Conversely, if I sign up to play a game about genteel afternoon tea in 19th century Britain, my expectations change accordingly. While it is generally permissible to jump out from behind a tree wearing a rubber mask and scream at me in the former example- even though I have not explicitly consented to this extreme behaviour- the same is not okay in the latter. Doing so would violate the social contract between player and game. I came to eat crumpets and gossip about lords and ladies- your frightening roleplay is as out-of-place and unwelcome as if I were to bring takeaway fish and chips into a fancy Italian restaurant.

The Vault jumps out from behind a tree, wearing a rubber mask, and screams at Dead of Winter players.

I understand that the THEMING of Dead of Winter is a scary survival situation in which no one can be trusted, and yet all must try to cooperate. However, the lens through which the audience experiences that fluff is, at first glance, a relatively standard format board game. The same is true of second and third glance- while the (generally excellent) “crossroad” cards contain (optional) mature themes, thisMECHANICALLY boils down to selecting from a set of options and executing the consequences. That is, what the players are DOING is sitting around a table, making cerebral strategic choices based on goals and probabilities.

To go from this sedate, well-understood communal experience to revealing actual secrets on the flip of a single card is as bizarre and incomprehensible as capturing my opponent’s bishop, only to find tiny writing on it saying they have to take their top off.

In conclusion, The Vault is the worst card ever because of how this forceful effect, stripped of all safeguards, runs roughshod over an otherwise excellent game. To make these demands of a player partway through a game, when both themselves and their friends are invested, without any sort of warning, is borderline aggressive. I can conceive of no situation where this card will be returned from the darkest place of shame to which I have banished it- under the plastic insert of my spare Dead of Winter box.

To those of you who are still with us- thanks for reading! I’ll try to write more of these analyses if I have time between designs. If you have any comments on this, or suggestions for what else I should write about, feel free to email me directly at adam@lightinthelab.com for a chat. Perhaps “Subbuteo Angling” will be next…