Board Game Design Day: White, Brown, and Pink: The Flavors of Tabletop Game Randomness

[Music] good afternoon everybody oh that's loud that's better yeah I hope everybody enjoyed the lunch break I'm excited about giving this presentation before I get started just a couple of announcements just make sure that if you have a cell phone or anything else that makes noise that you put it into the quiet mode at this point also if you have an opportunity to do the survey afterwards it is really useful I did the presentation last year and reviewing those was extremely informative for me especially those that that left comments so really appreciate it if I ever even go do that I'd also really excited to get onto that deck of cards for next year but we'll see if we can do that so I'm here to talk about randomness in games before we get started just a quick just a quick couple of things about my my background so for those of you that are not familiar with my work I'm a game designer I've got about ten published titles at this point ranging from a space cadets the expanse is actually my latest title we are currently working on an expansion for that if anybody has a good title for an expansion for the expanse let me know we're in a little bit of a quandary over that yes expand see on I don't know this this there's some good pun title there that will come up with I've also am on the the dice tower at doing the game tech segment for that podcast and do the loot ology podcast which is a biweekly show that goes into the nuts and bolts of game design and I teach a game board game design at the NYU Game Center in New York City so I wanted to start with this quote from the designer Greg Kosta Keon in his really excellent book uncertainty in games which was published in 2013 which I would recommend and he talked about uncertainty as being an essential part of games as he says in designing games a degree of uncertainty is essential and that if there is no uncertainty then it is not a game you can have other forms of entertainment that don't have any uncertainty a book or a movie if you're watching it for the second time it's going to do exactly the same thing but a game if you know exactly what the outcome is going to be then there is no reason to play so he looks at it and breaks down four major areas of uncertainty in games in ways that you can introduce uncertainty hidden information would examples would be in poker things like that where you've got cards or something that your opponent doesn't know that you have there is skill or performance-based uncertainty that is like darts or basketball or anything where you have to flick a disc or something like that there is opponent uncertainty even imperfect information games like chess you don't know what plans our opponent has or what move your opponent is going to make and you have to make plans on how you're going to react to that and then the last one is the traditional randomizers which are dice and spinners card draws tile flip stuff like that and it's that green box the randomizers that I want to focus on today each of these deserves their own lecture and presentation but I really want to focus on the randomizers in the way that those can impact the game so as I said there's a couple of different types of this die roll card drawers you also have classic spinners one of my goals in life is to design a spinner based game I'm actually working on a game where you build your own spinner like a giant roulette wheel with different probabilities for different things but and then the Venerable cube tower where you throw cubes and then they come out these are all ways that you can create uncertainty in your game but how does that get applied so let's take a look at some basic this is just sort of a really basic representation of the flow of a game there is a state of the game then the player makes a decision and that ends up with a result now there's two main places that the can fit into this picture the first is what I refer to as input randomness and that is where the game state is randomized so first the game state is randomized then the players make a decision based on that and then the results of the decision are going to be fixed when a player makes a decision they know what the outcome of that decision is going to be the uncertainty in this case is on the input side what is the game state now on the flip side you can also have output randomness in output randomness the game state is fixed and you're just making a decision around the current game state but the result of your decision is going to be random you make the decision and then you see what the outcome is you don't know how those two things are going to be correlated so some examples of output randomness this is typically the type of randomness that is used in games and certainly was you know going back millennia you have risk of course in risk when you decide you're going to attack a territory then you roll the dice and decide figure out what the results are gonna be roulette you place your bets and then see it in Dungeons and Dragons every action that you take you roll a die to see what the effect of that is it's not necessarily you know super random things there's also more sophisticated games wargames in particular have a lot of output randomness I've shown here a typical combat results table from a 70s 80s war game where you decide you're going to do an attack and then you roll a die or dice and see what happens so those are all some examples of alpha randomness input randomness tends to be a more modern phenomenon in terms of games with some exceptions but one looking at here is Dominion is a great example in Dominion in one mode of play there is a randomization for the kingdom cards that you select at the beginning of the game but once you select those decks that creates that game state that players then react to Dominion actually interestingly it has a couple of different levels of input random versus output randomness one is the overall game state but also each time you draw a hand of five cards that sort of input randomness and then you have to decide how you're gonna play out those cards there may be some output randomness in that like if you're playing a blacksmithing you have to draw cards you're not less so what you're gonna draw but in general you have a pretty good indication of the way that your turn is gonna flow based on the input randomness of the cards that you draw a similarly backgammon is input randomness you kind of roll the dice and then decide what you're going to do based around that make your plans based around that contract bridge to a certain extent has a level of this in that you know as the declarer you see all your cards and your dummies cards and you can kind of build a plan around that there is some unknown factors of course and what your opponent's hold but typically you can have a pretty good idea of a plan and then finally I just threw in a video game here the recently released into the breach which is a great example of input randomness in that it very very deliberately shows you exactly what you're the opponents are going to be doing on their turn so in this case for those of you that haven't played into the breach you can see that there's a those red striped lines there those are the areas that the enemy is going to be attacking on their turn unless something happens to them if they get moved out of the way or they get destroyed you know exactly what's going to happen and your moves are also very have limited randomness you just know if you move here and you shoot here you're going to do this damage you're going to push this kind of the water he's gonna die you know exactly what's going to happen so those are some examples of input randomness now the you know it's it's good to divide it up and to think about it as you're designing as a way of approaching randomness and these decisions that you build into games but you know they're not you know they're not very clearly you know absolutely perfectly delineated in games because you know you think about risk there is a fixed set up to when you start your turn about what the board looks like and then you have to decide what you're going to do and then you roll the dice the outcome but the Renault the input from your turn is the output randomness from somebody else's turn so there's this constant flow this constant cycle but it still can be you know it's it's still I think very useful from a design standpoint there also can be longer-term consequences like in settlers of catan there is also sort of input randomness for your turn where you roll and see what goods are available and then you decide how you're going to deal with it and when you're building a settlement you know if I play you know these four cards I'm gonna be able to put a settlement and this area in settlers of catan but I don't necessarily know how that's going to produce for me down the road so that's kind of a longer-term output randomness so there's a lot of gray areas in this but in general I think this is a useful classification typically input randomness built into a game will lead to strategy a more strategic feel to the game gives the players more sense of control and more rewards more skillful play whereas output randomness is typically going to feel more random it's you know there's there's little things that you can do if it's it becomes more tactical in that you have a shorter planning horizon and is typically associated with lower skill caps now these are broad generalizations there's a lot of examples on both sides for example wargames wargames are filled with output randomness every time you do an attack you typically roll the dice and see what happens and those still feel very you know obviously have a lot of strategy have a lot of tactics there's ways to rise above that but if you take the example of like into the breach the video game example that that would game would have a very very different feel if you didn't know the results of your combats going into them if you just planned your moves and then there was some sort of randomization about what happened it's a very very different game and it's a very less puzzly sort of a game becomes more of a you know just more random so there there's there's a blend there of of what you want to approach interestingly also which is a little bit outside the scope of this but the more the more randomness that you have in your game you can actually make it feel less random the more dice rolls that you have you're going to start to approach an average value that's a whole separate talk but just keep that in mind and war games take a lot of advantage of that there are so many dice rolls that you can pretty much predict over the long run what's going to be happening now there's - well the more that at least two axes of randomness input output is one of them I ran through that one kind of quickly but I want to spend a little bit more time talking about the correlation axis of randomness because this is not one that's considered all that often and I find it really interesting to look at games in this light so what is correlation in terms of this random randomness and the way I'm approaching it the idea is basically that there's going to be a series of events or a series of random numbers that are happening over time time one time two time three times four and if you know what's happening at time X how much does that tell you at time X plus one if I know that this is happening this turn how much does that tell me about what's happening next turn and there's a lot of different answers to that question sometimes the answer is you know nothing you know absolutely nothing about what's happening next and that in a scientific or mathematical terminology is called white noise and what that means is that again there's there's no correlation between the last result and this result if I'm playing risk and I roll the dice the result battle result that I got last time is not impacted at all does not impact in any way the result that I got this time they're completely unrelated to each other just as a fun fact we'll come back to it it's actually called white noise it's related to a frequency spectrum so when you have noise it generates different different frequencies in it and white noise has all the different frequencies from low frequencies to high frequencies low correlations to high correlations small jumps and big jumps so in a light spectrum sense that would be white and so that is why it is called white noise and this is as you can imagine the most common way of treating randomness in games but it's not the only way next option is called brown noise and brown noise you know a lot the results at time step one are very very highly correlated to what's going to be happening at time step two by the way white noise is called white because of the color of light but brown noise is not called Brown because of the color of the spectrum it's called Brown off of Brownian motion in a spectrum standpoint is called red noise which you'll sometimes see as a term but Brown noise is a little bit more of a common term a really simple way to generate Brown noise is to say you're starting at a certain value you're starting at a value of eight and you flip a coin and on a heads it goes to nine and on a tail as it goes to seven okay so you have a really good idea of what the next result is gonna be but you can over the long run predict exactly where it's going to be this is also sometimes called a random walk type thing a type of randomness because it's it's tied to to where you start from and it will spread out from there but it spreads out in a controlled and slower way and then the last type of noise is cutting it in between these two white noise and brown noise it's called pink noise and pink noise which goes back to our spectrum again you have a very high chance of a small change but a much smaller chance of a big change so over time that's going to start to you know gradually drop down and you could get really really big jumps but you have most of time it's gonna be focused on those small areas now pink noise actually is sort of interesting because it's ties a lot into what humans like and the way that humans experience the world is pink noise is actually the best representation of randomness and the randomness that we see there's been studies of us and Clark did two studies in 1975 where they took music and they plotted the frequency of the notes over time and how the frequency from one note to the next and it's exactly corresponding to pink noise they did the same with human speech human speech in terms of its tone in terms of its volume in terms of the pacing of it also is a pink noise type curve financial systems follow this you get a lot of small changes in the stock market and every once in a while you get big movements and crashes and big jumps and you also have what are called so-called Black Swan events like the giant stock market crash in 2008 which are supposed to be you know exceptional events so systems tend to be you know you have an idea of where the Dow Jones is gonna end up every day but every once in a while you get these big giant jumps so a pink noise in a very real way represents a clearer represent a closer representation to reality and it's also been found that systems that model pink noise musical systems particularly in voices and speech and things like that that people respond to them better people like them even if you just sit I was gonna play some noise for you but I'm not gonna I'm not gonna burden you with that but if you listen to samples of white noise audio white noise and brown noise and pink noise people universally like the sound of the pink noise better even though it just sounds like noise it just sounds like a fuzzy noise but it's a better fuzzy noise than white noise so a couple of game examples that I wanted to jump into and talk about how these three different types of noises are represented in some board games so this is a game from 1987 called Shaco and company about building a chocolate company and in this game as a business simulation there's at the end of each turn or beginning I forget if it's beginning or end of you sure but you flip over a card that says what the market conditions are and it basically tells you how much chocolate people want and how much they're willing to pay for it but there is zero correlation from one term to the next you have no idea if the market is going up or the market is going down so you can invest in all this infrastructure to build your chocolate and process all your chocolate and then you flip over a card and there's only ten bars of chocolate that are wanted in the world and you're in trouble it makes it very chaotic feeling it makes it very very difficult to plan for the future and also there are some special abilities that's game that will let you peek at cards coming up and that makes those extremely powerful but it as a business simulation it doesn't really feel right it just feels like you're just taking a guess at what's gonna be happening as a counterpart counterpoint to that the game crude or make multi which was originally released in 1974 in that game you roll two dice there's an economic situation so there's like prosperity or there's expansion you roll two dice in that game to see what areas on the board are going to produce actually it's really similar Settlers of Catan it's really interesting this is a sort of a precursor to that system but on doubles the economy changes to a new state and when you get doubles I don't know if you guys can see on on the thing here but at the bottom of those cards it tells you where it goes to next so like if you're in prosperity it may go to rapid growth most of the time if you world doubles is going to go to downturn or it may go to depression so that is sort of a combination of white noise and brown noise you have a pretty good idea of what's going to be happening but kind of the trigger of that is more random but it makes the the economy feel a lot more alive this was refined even more in the re-release in 2012 of crude the rules were changed in that point instead of the economic changes being triggered by doubles on that roll it was changed instead that there was an accumulator that was created when you roll those two dice you've got you took the difference between those two dice he rolled a 2 and a 4 it was 2 and you added that up on a track and when the track reached 8 then it moved into a new economy and then you again roll the die to see where it was which economy it would go to the great thing about that system that's that's a real pink noise system because you can have big changes somebody could roll a 1 and a 6 somebody could roll doubles for a while in a row and it's not going to change at all but you can kind of see coming with the first system increment Balti you knew ultimately what most likely economy was but you couldn't see it coming here you can you can say okay it's at six it's at seven it's getting close to that egg point it may change next turn it may not change next turn do I want to deal with it now do I want to wait so you've got big swings little swings and some idea of where things are going so a much more robust and interesting system for players to look at another genre I want to compare to dinosaur games the first one is Evo in 2001 this is the environment track from Evo so one of the things that's really important to your dinosaurs is whether it's temperate or desert or whatever and so the you guys can see my mouse you can't see my mouse okay so you have a marker on this track and at the end of each turn you roll a die to see where it moves and you can see that on a three four five or six it moves forward to the next one on a two it stays where it is and on a one it actually moves backwards so again as a player I have a really good idea where it's gonna be this is classical Brown noise so it's kind of a random walk but it's even you know more directed than a random walk but it has a very staid feel to it you you it's not gonna move that much over the course of the game again its counterpoint the game Toronto X from 1990 which really should be reprinted God please somebody reprint this game has these environment tokens that are put onto a onto each separate environment that represent you know the fruits or Suns or whatever the way that this works though is when players add environment tokens they put them on face down so you know what you're putting in but the other players don't know what you're putting in and when the number of face down tokens exceeds the number of face up tokens the environment changes you throw away all the face-up tokens you flip everything else up and the majority becomes the new winner and depending on how that goes that can actually cascade into multiple changes like immediately it can also leave the in a really stable state if everyone put in the same kind of feature or it can leave it in a really unstable state if there's a lot of different things that are in there so it's it's very much a pink noise process and it's a really really dynamic and interesting part of Taranto X because of those cascading features you also get that cascading feel a little bit in pandemic when you get kind of these super critical states and small differences can make big changes on the board but that's it's not as much of a pink noise process as there is here so generating pink noise is actually pretty challenging in a boardgame environment it's really easy in a video game environment from a board game it can be challenging there's a couple of different ways to do that one is by using the dice Delta like I talked about which was used in crude you can have I'm going to talk about exploding dice in a second but you can also have something like this heat of battle table from advanced squad leader which I call extending the extremes in advanced squad leader crazy stuff can happen anytime you roll a 2 or a 12 this is the chart that what happens if you roll a 2 so if you roll a snake Eyez on two dice you then go to this second table that can result in people going berserk and could result in people promote heroes arising from a squad that was just about to just be destroyed there's all different kinds of things that can happen and there's this very small chance of gating into this other world this other table of things that can happen so that can be really interesting exploding dice which as far as I know we're originally used in the game the Uncharted see the miniatures game is really interesting because in this game you just roll lots of dice when you're doing an attack a 1 2 or 3 is a mess of 4 or 5 is 1 hit but a 6 is 2 hits and you get to re-roll those dice to die all the sixes so it creates these moments it creates this tension that you you have a pretty good idea of what your distribution is going to look like but you've got a tail now that extends infinitely out and I've seen some crazy stuff that happens because you're rolling like 20 dice and so you can get tons of 6 and it just keeps going and going and it can lead to these really dramatic and exciting moments so it leaves out the possibility for real extremes while keeping things generally focused in the middle I'm just gonna throw out one other thing just because I've got five minutes left here so violent noise as far as I know has never been used to in a board game violet noise is the exact opposite of pink noise there's a very very small chance of something small happening but really big chances of big things happening I'm not sure how that would be used but I'm just throwing it out there for the sake of completeness on the different colors of noises so just a couple of design tips just to kind of wrap it up when you're doing making decisions for the players in terms of how they're approaching things look at input versus output randomness and how that's going to affect planning how that's going to fix strategy how that's going to affect analysis paralysis of players that have to consider every last thing and lastly people like pink noise so white noise is easy to do Brown noise is easy to do try to come up with systems that will generate pink noise that will give this opportunity for you know you know mainly what's gonna happen but it has the chance for these outlier events that keep things really interesting and spice them up so that's all I've got for today thank you and I'd be happy to have any questions I think we've got we've got about five minutes left for questions so I think there's a microphone in the middle if anybody hasn't he otherwise I will just keep talking so your initial quote said unser was about uncertainty yeah and randomness is one way to achieve uncertainty in games that I've designed that have a lot of uncertainty being generated by player behavior but no randomness I've ended up with situations where the play testers will tell me your game is way too random because there it's too unpredictable to them what's going to happen based on other people's behaviors if they're not the designer and they don't have they don't see the table well enough to predict can you think of ways that one might use randomness to provide more predictability to unpredictable player behavior which seems kind of intuitive but I think that one way of looking at that is I I'm almost kind of reminded of a bluffing aspect of a way to do that so you're you're the you kind of have an AI more of an idea of what you don't want it to be absolutely certain what the players are doing and I'm gonna lean back on what Tom Lehman talked about in the last talk about race for the galaxy and the way that players are making choices and the first level heuristic is just pick what's best for you and then after that start to leech off of what other people are doing so I could kind of see that where you you don't want your choices to be too obvious of what people want to do if all choices are equally valid for people then that's a design related issue that you need to fix there needs to be things that are in general better for people but not to the level of certainty you know this is absolutely what I have to do I have no other options so I would just try to constrain people's choices and that'll make it feel a little bit better thank you sure yeah in the beginning of your talk you talked about lots of different kinds of randomness other than randomizers like other players and so on is there how do you see being able to apply this concept of different colors of noise to that because I'm trying to wrap my head around how would you get pink noise out of other human behaviors as opposed to randomizers that's an interesting question I'm not sure if there's a direct analog really as much I think with your skill based stuff you know there's always going to be a certain distribution of of player performance you know in flicking games or shooting three-pointers or what-have-you so I I'm not sure it's it's as easy to control it in terms of uncertainty with hidden information stuff like that I think you can tune it to a certain extent by by allowing information come out partially like if you look at social deduction games like we're go for you know Avalon or things like that there's ways that information leaks out so then you've got you don't have a perfect you're basing it on a little bit of information which may not be suspect which I think kind of lays some of that groundwork that you're talking about yeah hi my name is Totti and thank you for the talk it was I loved it and I have a maybe a comment then I'd like you know to for you to elaborate on if so basically I think it's interesting that you have these like you said about Pima and a pink randomness right or pink noise or not that that's kind of the Holy Grail right that's what you wanna that's what you wanna have that's what people and I like the most and you say that it's it's a trick - it's a trick to make it in a board game and environment it's easy in in a digital game or a computer game I have been designing a memos and they're it's extremely difficult to prevent it and being a game killing event which is like a world killing event right extremely difficult so I I think it was very interesting that you were in Mandelbrot and chaos theory and and and Black Swan right this is these are chaos theory is my events and so yeah yeah this is just my comment that I want you to maybe love it thank you yeah I you know I think that you don't want to have and as a designer you to keep the game on track I mean and that's part of the the the fun and challenge and frustration about being a game designer is you kind of set up this machine but then it's got a run on its own and the players are providing the input for it and you can't set up a situation where the players can push it all the way off the edge of the cliff because they will if they have an opportunity to do that so so you can extend the extremes but still kind of try it keep it constrained in some fashion you have to look for you know closed feedback loops that can move things off really quickly and stuff like that so I think that there's as a designer you know I also do software development alright and as a computer programmer you know 99% of the things I screw up were at the edges of my loops or you know I forget to check for nulls or at the corner cases and you know in game design it's the same thing you got to look at your corner cases what if somebody just does this move over and over and over again or if all the players try to do that or somebody tries to tank the game or somebody tries to end the game on the second turn you know you've got to look for all of those extreme events and so you you you want to create good randomness but you want to create it in a bounded sandbox that the players can operate in you can't let them just completely run amok because you're not going to get the experience that you want out of it so okay I think that's time if anyone else has any other questions I'll be outside afterwards at the meetup area so thank you all and please remember to do the surveys if you have the opportunity you

Channel: GDC

Views: 29,061

Rating: 4.9648352 out of 5

Keywords: gdc, talk, panel, game, games, gaming, development, hd, design, tabletop games, board games, video games, dice, randomness

Id: qXn3tGBztVc

Length: 32min 1sec (1921 seconds)

Published: Wed Nov 07 2018