While cooking up Gears & Spears' setting-influenced task resolution, I started thinking about a task resolution appropriate for Take, my criminal underworld/heist side side project formerly titled "Pull." Because the player characters are "proper villains" in Take, I was thinking about making a system that actually encourages dastardly behavior within the game mechanics, specifically "cheating."

The problem with making cheating part of the system is that it's, well, part of the system. By my reckoning, "cheating" is an act of subverting the system for selfish gains, but when it's allowed and encouraged, that kind of takes away the fun and danger from it. The other obvious snag is that cheating in some way relies on the mark not realizing they're being played for a sap.

So I'm just curious, has this concept been discussed on the boards before? My search has only turned up threads on cheating as a symptom of system dissatisfaction and so forth, nothing on cheating as part of the system itself. Another related question, would it be easier to implement "cheating" in dice-based, card-based or some other type of system?

The only game I recall that has any mention of cheating in the rules is Illuminati (the original, not the trading card game. Although the trading card game may have had cheating guidelines, I can't remember. I know the original game did.) In Illuminati, legalized cheating is an optional rule, and pretty much anything goes - fudging dice rolls, stealing money or cards from other players, peaking at people's hands if they have to leave the table, etc. Dunno if that info is helpful, but there it is.

It's a horrible variant in Illuminati. It takes a fun game and makes it a marathon of scrutiny and book keeping in which no one can leave to go pee. It was worth a try, but I'll never do it again. On the other hand, one of the flare cards in Cosmic Encounters (filch?) gives the power to cheat (as long as you're not caught) and it seemed to work well.

And I agree that cheating means something else entirely in an RPG. And probably different things depending on the GNS leaning of the group/game.

Chris

There's always something like Diplomacy, where the players feel like they're doing something bad (backstabbing each other, betraying confidences, etc) but it's really all part of the rules. If you haven't played the game ever, I think that's a must. There're variations of Diplomacy where "anything goes" as well, where you can lock players in the bathroom so that they can't turn in their orders, but you have to get support from the other players before you do stuff like that, or someone will just let them out. Kind of like the Lumpley Principle, in that way ...minus one player :)

The problem with making cheating part of the system is that it's, well, part of the system. By my reckoning, "cheating" is an act of subverting the system for selfish gains, but when it's allowed and encouraged, that kind of takes away the fun and danger from it. The other obvious snag is that cheating in some way relies on the mark not realizing they're being played for a sap.

Another related question, would it be easier to implement "cheating" in dice-based, card-based or some other type of system?

I think Jonathan raises a good point. Are you encouraging cheating in order to facilitate an advantage for a character while still pushing cooperation, or are you planning on generating inter-player competition. I'm going along the competition route with Powers That Be, but i don't think cheating would work there.

Jonathan mentions Diplomacy (which is a great game), but that's only going to be useful to you if you are planning on fostering competition since there can only be one winner. Are you looking for something like that, or are you instead thinking of using cheating while keeping the focus on accomplishing the goal as a group?

Thomas

Actually, I can think of one game (not an RPG) which does encourage cheating, and which you could easily use as inspiration for a mechanic for Take or similar games.

I'm talking about Cheat, the card game. It comes under a variety of different names, but the basic idea is that you are theoretically being bound to do an increasingly difficult/improbable thing, and assert that you have done it without actually demonstrating success. It's the distinction between lying about your assertion and telling the truth, as the latter becomes increasingly improbable, that gets interesting.

This is easily adaptable to an RPG, and would be really neat. Here's an example system:

Characters' ability is ranked in dice pools (of moderate size, say 5-7 dice normally) of six-siders.

In a single increment of die-rolling (say, a combat round), everybody rolls this many dice - behind a screen. Someone is chosen to go first, and states a result from his die roll. Results are of the form, N dice showing X, whether "One two" or "Pair of fours" or "Three sixes". Who is chosen to go first, and how he decides his initial call, is pretty open to design decisions (esp. based on theme or other mechanics in the game), with the one constraint that there should be some factor motivating the initial call to be fairly low. Let's say you have to spend N*X resource points to open, so you're motivated to call low in general; "one two" costs you two tokens, "three sixes" as an opener costs you eighteen.

You do not show this result, you are merely asserting that you have it. If the assertion goes unchallenged, then you succeed this round. The next player then asserts a better result - a higher pair if you called a pair, or a triplet of any number might also be legal if you called a pair - and, if unchallenged, is again considered successful at his die roll. The next player must improve on this assertion in turn, and so on.

You 'fail your roll' in one of two ways. Challenging someone - and having them prove they were correct - or being challenged, and being unable to prove you had told the truth. Being caught in the lie, or being fooled. It strikes me that putting it that way, this would be two different types of 'failed roll' in something like Take - 'getting caught' and 'being fooled' being two big motifs in the source material.

I know it's not cheating in the literal sense of the word (bending the rules), but it's a set of rules which is certainly reminiscent of cheating despite providing a concrete structure.

- Eric

What you suggest is similar to the d6 bluffing game "Liar's Dice" which is tons of fun and easy to learn. As you said this isn't really "cheating" it's more like playing poker by the rules. Can you read probabilities and intent? I think what is actually being sought here is a mechanical system (or overt social contract) by which breaking rules (at least some of them) is encouraged assuming that you are not caught.

It does make an interesting mechanic though...

Thomas

Si.

Actually, it occurs to me that an element of this could be implemented into just about any mechanic, by loosening some of the constraints. Overlay it on your favorite mechanic - dice, cards, doesn't matter, though it'll work better with one where there's a 'degree of success' in place rather than binary success or failure - like so:

Players roll, but don't show their dice. Instead they announce a result and continue. It is acknowledged that this may be a lie. Play the game normally other than this.

Anyone may challenge a roll in which case it must be shown. If the challenger is correct, then the roller 'Got Caught' which has a negative consequence in the mechanics. In something heist-related, it strikes me that a Nar-form consequence might be best: Your character will get caught. Similarly, if the challenger is wrong, then his character 'Got Fooled' with comparable, but possibly distinct, consequences. The parallel Nar-form consequence to the above is obvious, especially as in most good heist material there are layers of secrecy. A Got Fooled result might be the OK for the GM to outright falsify the results of your senses so as to produce the effect of your being duped, for example.

Makes a pretty simple overlay on any ruleset, but with good consequences for the desired mood. IMO this is actually superior to my previous suggestion to use Cheat/Liar's Dice as the mechanic proper, in that it preserves more of the feel of "cheating in an RPG" as compared to in a dice game.

You could also extend it to any form of cheating, not just lying about your dice, so long as it's something which can be formally and clearly asserted, revealed if necessary, and lied about. Lying about your stats, for example, would work as long as it was explicit (rolling the wrong number of dice in a dice pool system, but rolling them openly in front of everyone) and not implicit (adding the wrong modifier which includes that stat and several other factors - this would make cross-checking too laborious). Laborious is bad - as above, I too find cheaters' Illuminati to be a tragically flawed game to actually play. Restricting the forms of cheating to certain modes only, which are easy to remember (if they contain persistent information like stats) or easy to estimate probabilities on (if they refer to dice results etc), is a must.

One last cheat form also comes out of Cosmic Encounter... the idea of cheaters' cards. One of the cooler optional sets of cards (IMO) for that game is a set of cards which allow cheating, so long as one isn't caught. The Filch flare cited above was the prototype for this set, but there are several possibilities... and what this could mean is that one could define several vectors of cheating (May lie about stats; may lie about die rolls; may forget to spend resources, etc) and give access to only one at a time, based on something fairly low-probability (one card for each form of cheating in a deck; one rare dice result which permits this, in dice), until caught. That is, the mechanics make it legal to cheat, in this certain way, because you hold this card/result, and as long as you don't get caught. As in Cosmic, the logical penalty if caught is obvious: in addition to some penalty, you also lose the privilege which was allowing you to (legally) cheat.

- Eric

The word cheat is being tossed around here recklessly. And that's because games do it frequently, and in the Illuminati example (BTW, there is a "cheating" version in the card game, and even one card that excuses "cheating"). I'd propose that Cheating should refer to actually doing something that the game says is illegal.

What that means is that if a game says that you can cheat, it's actually saying that it's legal to do certain things that in another game would be cheating. These all come down to the same thing. Being able to do something as long as you're not discovered doing it. In the case of bluffing, this means that there's usually some detriment for the player guessing that you're doing something wrong (to prevent constant guessing), but is otherwise the same.

This is a simple mechanic that's in many games in the same fashion. As long as we think of it as cheating, the issue will be clouded, IMO.

Mike

The skill development system in a game posted to RPG.Net's Roleplaying Open was a lot like cheating. You wrote in new skills when the GM wasn't looking.

I think you all know what I'm talking about.
If you don't, search for d02: Know No Limit at RPG.Net. ;^)

Mike: And yet central to the point was that this precise misapprehension of things as 'cheating' might be of great utility to the designer. These acts are perceived as akin to cheating even though they are not; this perception is useful if cheating is a motif of the game.

This is contrasted with a game which encourages actual breaking of the rules, which paradox nobody here has any good suggestions on yet. That doesn't mean we're not (at least, I'm not) thinking about it, though.

Actually, I do have one thought. The use of deliberate drift could be arranged so as to produce a feel of cheating, by de facto making things legitimate which were prohibited by the rules proper. Honestly I can't think of a way to implement this in a serious game, but in a Toon/Junta hybrid (gah!), a set of rules like "Everybody's money must be on the table in plain view at all times. Honest." which could be understood as breakable, would probably end up with a feeling of corruption as people generated fairly logical "house rules" which were in fact intentional. Slippery and dangerous, but one possible resolution of the above paradox.

- Eric

Wow, there are a lot of really good suggestions and nice feedback. Thank you so much, everyone!

I'll clarify a few things people have asked about the game itself. First and foremost, Take about taking people's stuff, be it on a small or grand scale. Working smalltime is usually a one-man operation, like pickpockets and con artists. The big jobs, heists, require a complete crew that is skilled, functional and won't rat out on each other. When you've completed a job, the Heat will be on you for a while, depending on how much you stole. This is where the social element of the game comes in, as you try to navigate your way through the criminal underworld, forming alliances, hearing gossip, earning favors and making connections to better and better crews.

Because there is such a cooperative emphasis on the heists, I'd prefer an inter-player metagame conflict to be reduced to a minimum, but I don't want to completely preclude the possibility of a player betraying the crew. Rats are a common element in crime drama, after all. I think this can easily be covered by a player just wanting to be a rat, no mechanics necessary.

From what has been mentioned, I think I'd rather go with the illusion of cheating rather than actual, malevolent metagame backstabbing. Things like bluffing results hoping to fool an opponent seem particularly interesting. Because much of the system in Take is centered around a character's "style" trait, the penalties for being caught in a bluff are fairly easy to implement.

Say the system used some kind of height/width mechanic like GODLIKE. You've rolled some dice, not revealed them, and announced that the highest matching set you have is five 6s. Someone tries to call your bluff, you reveal that you only have three 2s. Should your style rating be reduced by the height (value of highest matched dice) or width (the number of matching dice)? Or would it just be easier to use playing cards and some super-simplified poker mechanic?

Another idea, incorporating the concept of "one person can cheat this way, but not these other ways":

A player can only call a bluff if they possess the "bluff token." The game starts with the GM having the bluff token. A player announces a scarcely believable result, the GM calls "bullsh*t!" and tries to call the bluff. If the GM was right, he can keep the bluff token. If he was wrong, he gives the bluff token to the player. This is a nice, non-numerical penalty for miscalling bluffs. If you're wrong, you can't bluff again until the token passes back into your hands.

Wow, there are a lot of really good suggestions and nice feedback. Thank you so much, everyone!

I'll clarify a few things people have asked about the game itself. First and foremost, Take about taking people's stuff, be it on a small or grand scale. Working smalltime is usually a one-man operation, like pickpockets and con artists. The big jobs, heists, require a complete crew that is skilled, functional and won't rat out on each other. When you've completed a job, the Heat will be on you for a while, depending on how much you stole. This is where the social element of the game comes in, as you try to navigate your way through the criminal underworld, forming alliances, hearing gossip, earning favors and making connections to better and better crews.

Because there is such a cooperative emphasis on the heists, I'd prefer an inter-player metagame conflict to be reduced to a minimum, but I don't want to completely preclude the possibility of a player betraying the crew. Rats are a common element in crime drama, after all. I think this can easily be covered by a player just wanting to be a rat, no mechanics necessary.

From what has been mentioned, I think I'd rather go with the illusion of cheating rather than actual, malevolent metagame backstabbing. Things like bluffing results hoping to fool an opponent seem particularly interesting. Because much of the system in Take is centered around a character's "style" trait, the penalties for being caught in a bluff are fairly easy to implement.

Say the system used some kind of height/width mechanic like GODLIKE. You've rolled some dice, not revealed them, and announced that the highest matching set you have is five 6s. Someone tries to call your bluff, you reveal that you only have three 2s. Should your style rating be reduced by the height (value of highest matched dice) or width (the number of matching dice)? Or would it just be easier to use playing cards and some super-simplified poker mechanic?

Another idea, incorporating the concept of "one person can cheat this way, but not these other ways:

A player can only call a bluff if they possess the "bluff token." The game starts with the GM having the bluff token. A player announces a scarcely believable result, the GM calls "bullsh*t!" and tries to call the bluff. If the GM was right, he can keep the bluff token. If he was wrong, he gives the bluff token to the player. This is a nice, non-numerical penalty for miscalling bluffs. If you're wrong, you can't bluff again until the token passes back into your hands.

The Bluff token seems dangerous (though intriguing). The problem is that if you get the token you can bluff with impunity (that could be solved by multiple tokens i guess). I would suggest that your Style changes by a fixed amount when you bluff. The fact that you are caught in a big lie as opposed to a small one is unimportant. It seems to me that it's just better to have a single penalty. If you get caught in a big lie, it's your fault for attempting to pull it off just as much as if you get caught in a small lie (though small lies would be less likely to be called).

I would suggest that unless a bluff is called the given result (whether true or not) is accepted mechanically. If you are called, but spoke truthfully then it's an auto-success (so if you sucker someone into calling your small lie that would have failed, you end up succeeding anyway thanks to your dupe). If you are caught in a lie you automatically fail (this will keep you from being too ambitious in your lies, especially if you have a probable success anyway). This also does some interesting things because if you have a terrible roll you are more likely to bluff since you have "nothing to lose."

Thomas

Maybe the bluff token could be said as "Anyone can call the person with the bluff token in a lie, but only the holder of the bluff token can call other people's bluffs."

In reading this thread, I thought of two traditional games in which "cheating" is part of play, and I think both teach something about how to implement it in play.

The first is a card game called in polite circles I Doubt It. In this game, on your turn you are required to discard, face down, some number of cards of the value that has been reached, and declare how many you have discarded. Thus, if the player before you discards two sevens, then it's up to you to say how many eights you are discarding. Anyone can call you on it, and check the cards you've discarded. If they are right, you must pick up the entire pile; if they are wrong, they must pick up the entire pile. However, the game doesn't say that you can't discard three cards and claim you've discarded two; and it's part of play that you're expected to discard cards that you claim are a different number from what they are.

The lesson here is that "cheating" may well mean nothing more than that the player is claiming he is doing one thing when he is doing another, and that there are penalties if he is caught doing the other.

The other game is Monopoly. In this classic board game, if you land on someone else's property, you owe them rent. However, you do not have to pay them rent. If the next player rolls the dice before the player on whose property you have landed requests his rent, you don't have to pay. This is in essence "cheating", in that you got away without paying an in-game debt you accrued and rightfully owed; but because you didn't mention it and he didn't notice until "too late", the rules say you got away with it.

The lesson here is that "legitimate cheating" must be "caught" before a specific in-game event or it becomes fair by default. (The same is true in I Doubt It: if the next player plays, it's too late for anyone to call you on your discard.)

So I think that something like this could work in an RPG, if it recognizes these two aspects: cheating is a legimate action which is contrary to the player's declared action, and it can only be penalized if caught before a certain subsequent point in play.

How that works in a particular game is more difficult, but not impossible.

--M. J. Young

I agree that the atmophere of calling something cheating might be valuable. I think Evil Stevie makes a lot of mileage out of it. But mostly as a sort of advertisement. It gives the game some cache. You'll also note that he makes it optional, and that in actual play it sucks.

Because, essentially, in saying the "the player can cheat" you're saying, "the player can do somthing that he can't do." This sort of implicit paradox causes multiple interperetations, which lead to problems in play. As someone said, in the end, you'll have people making house rules.

What I'd do is put cheat in quotation marks, and explain that it's a legitamate part of the game. Because I can't cheat, personally. I can "cheat", meaning bluff.

MJ, that was exactly my point. What you called being caught I called detection. The best rules like this are like those in Bullshit (AKA I Doubt It), where there are repercussions for being caught (having your bluff called), and repercussions for incorrectly identifying an "infraction" (calling when the opponent isn't bluffing). That has the largest number of potential strategies associated a priori.

BTW, Game Theory makes regular references to bluffing as a central idea in creating strategy. Indeed, it's the most complex decision-making matrix in warfare. Other strategies like Mass and such are relatively easy by comparison.

As long as we're at it, here's an article on Game Theory and Bluffing as Game Theory analyzes it: http://www.gametheory.net/News/Items/037.html

Mike

That's an interesting article, Mike, thanks for posting it.

So, in summary, this suggestion...

LordSmerf wrote: I would suggest that unless a bluff is called the given result (whether true or not) is accepted mechanically. If you are called, but spoke truthfully then it's an auto-success (so if you sucker someone into calling your small lie that would have failed, you end up succeeding anyway thanks to your dupe). If you are caught in a lie you automatically fail (this will keep you from being too ambitious in your lies, especially if you have a probable success anyway). This also does some interesting things because if you have a terrible roll you are more likely to bluff since you have "nothing to lose."

Lxndr wrote: Maybe the bluff token could be said as "Anyone can call the person with the bluff token in a lie, but only the holder of the bluff token can call other people's bluffs."