14 Points on Randomness

My last post, Game Placebo, got a lot of good feedback – more positive and constructive than usual, I’d say, which is nice. I also sent the article to DanC, who responded to me directly about it. He then wrote a G+ article about the topic of randomness, which led to a Lost Garden article. I feel happy to – at least partially – have been the inspiration for a Lost Garden article, being that that blog was my primary inspiration to begin writing about games almost a decade ago.

My position is that output randomness should not be a part of ideal game design. Right now I’ll try to break down my reasoning into discrete blocks that should help conversation about it.

Output randomness is randomness that affects a game after the player’s decision that decides the outcome. So, I decide to attack, and then there’s a dice roll to see if it worked or not. That’s output randomness. Input randomness, on the other hand, would be something like map generation or some face-up market cards that are available to all players. Although there can be improper implementations of input randomness that cause it to have similar problems as output randomness, input randomness is not what I’m talking about in this article.


The Points

I’ve put this in a list format.  Please read each point, and let me know which point does not work for you, and why (if any).


  • Point 1:  The act of coming to understand something is of value to human beings.  It is both enriching and entertaining to us, by our very nature.
  • Point 2: Games are valuable to human beings to the point that they allow us to understand them. If a game leads us smoothly to understand its lessons (accessibility / easy-to-learn), yet also has a very long, seemingly endless set of lessons to teach (depth / difficult-to-master), then that is a game that has great value to humans.
  • Point 3:  A basic loop of game functionality can be described as the game presents choices to the player, the player makes decisions, and the game then returns feedback. Through this process, players obtain understanding.
  • Point 4:  Games (as defined here) are different than gambling applications.  The value of gambling is not one of learning to understand. Instead, its value is in inducing compulsive play by randomizing reward schedules.
  • Point 5:  Many modern games that we know today are actually some mix of what I’ve described as “games” and what I’ve described as “gambling applications”.  Most card games would be examples, and many of the top aussie casinos have begun to employ these after seeing the success rates from their North American predecessors.
  • Point 6:  What seems like a machine of understanding can later be discovered to not be a machine of understanding, but rather simply an illusion. A clear example is a 100% luck-based activity such as playing Candy Land. As children, we play Candy Land and believe that we have agency – that we have achieved some kind of mastery when we draw cards that help us. Later, we become able to put together that we personally had no role in the win or loss, and the game loses its value to us.
  • Point 7:  It is then reasonable to assume that there could be games that we enjoy even as adults which are illusory in a similar way. Of course, adults need the illusion to be stronger, so we need more of a framework of strategy to exist. This is why so many games are some combination of game and gambling application, and it’s also why different individuals want games with different levels of randomness.
  • Point 8:  Regardless of your level of mastery in a highly random game such as Poker, there are outcomes – feedback – that delivers zero understanding. For instance, you could know quite well what the odds are of a person having a certain card.  You could know that the odds are 90% certain, and then make a move based on that understanding of the odds.  Despite this, it could simply not be the card you expected.  There is no reason, no lesson that can be derived from this event.  Nothing was understood.
  • Point 9:  Of course, there is understanding to be gained in highly random games such as Poker. I’ve never said that highly random games have no strategy involved, but merely that at a certain point, the strategy gets largely solved and it breaks down into sheer gambling.
  • Point 10:  In such heavily random systems, it’s difficult to know exactly where strategy starts and where gambling begins.  Feedback from a highly random game is unreliable.
  • Point 10a: There is a finite amount of time that a player has to play a game. So, if a game is truly a deep strategy game and is highly random (therefore requiring many, many plays to get reliable feedback), it is not practical, and possibly impossible to achieve mastery in that game.
  • Point 11: Depending on the game, some players won’t get to a point with most games where the system breaks down into pure gambling. However, some will.
  • Point 12:  I am concerned with developing a kit of guidelines for optimal game design – games that are better than what has come before.
  • Point 13:  Plenty of popular, well-liked games exist with no randomness at all. Examples would be Outwitters, Chess, Puerto Rico, and Go.
  • Point 14:  Therefore, based on points 7 through 11, and how they conflict with the value of games as defined in point 2, coupled with the existence of Point 13, leads us to the following conclusion:


Randomness should not be a part of ideal game design.



Please read this section before commenting to avoid posing a straw-man argument.

  • Clarification 1:  I am not saying that heavily random games have “no strategy” or even “little strategy”.
  • Clarification 2:  I am not saying that heavily random games are “bad games” in any kind of general sense.  I am instead saying that they are bad games insofar as I have defined the value of games in point #2.
  • Clarification 3:  I also understand that randomness can be a useful shortcut for game designers, particularly those designing party, family, or generally “light” games.  For most players, the negative effects may be negligible and for players who do not play much, the positive may outweigh the negatives. However, I would say that this is “planning for low-depth”, which is only appropriate for some design objectives.  I would also remind you that a non-random game can also make a good party/family/light game.


  • Clarification 5:  By “ideal game design”, I’m not talking about “the design of one singular ideal game”.  I don’t believe in the idea of “an ideal game”.  Instead, I’m looking for ideal guidelines for game design, in general.
  • Clarification 6:  Randomness as a mechanism can work in a game that is ABOUT judging odds.  If the game is about tactics or anything besides judging odds, though, it seems like it would always be inefficient design, at best.

So there you have it.  My points, along with some clarifications.  What points do you object to?  Let me know.  I’ll probably update this with more clarifications as people comment.

  • This is interesting.

    I would identify disagreement with points 9, 10, and 12.

    9 and 10: you are not really clarifying the interaction between randomness and strategy in highly random games like Poker. Randomness is not just the fluff in Poker that motivates weak-minded humans to play it; rather, the randomness is key to the strategy. If you don’t think so, try to design a Poker variant with no randomness. But if I’m right about that, then randomness is justified by your point 2: it gives you strategic decisions you can’t easily have otherwise. (The prevalence of mixed strategies in game theory is a kind of evidence for this point, as well.)

    12: I resist the notion that we should be looking for an “ideal game.” I am not making the practical objection that it is hard to actually achieve the ideal game. Rather, I think most of us employ a portfolio approach to enjoying games, where we can play many different games and derive different goods from them. I don’t think you can produce a strong argument whose conclusion is something like: “Chess is the ideal game; everyone should stop playing Poker.” (Chess and Poker are just examples here.) When I play and enjoy a variety of very different games, I am clearly not approximating some ideal game — my choices are not converging. Am I making some kind of error? I think not. Thinking about “optimal game design” may be personally helpful to you as a designer, since you will design a relatively very small number of games. But for players I think pluralism makes more sense.

    That said, I very much share your suspicion of games that are more like reinforcement schemes with randomized reward levels, and your notion that some feelings of control and accomplishment we have in games are actually illusion.

  • keithburgun

    To clarify on point 12, “guidelines for ideal game design” is not referring to “how to design THE IDEAL GAME”, but rather “these are ideal guidelines for game design in general”. So I don’t believe in the “ideal game” either.

  • Interesting ideas, but this seems like a bit of a dangerous blanket statement.

    You cite Poker a few times, and I understand why – however, it’s extremely important to consider the differences between Video Poker and “Real” Poker. In Video Poker, it’s all a matter of understanding the statistics and playing mathematically. The strategy and chances are relatively straightforward to understand, so the game quickly devolves to pure gambling.

    Real Poker, though, is more about reading behavioral tells (at least when all players understand the rules and basic strategies). Is Person A lying? Would Person B believe me if I was lying? Considering my chances and current situation, will I be able to get away with Plan X? This is where Poker gets truly compelling, even though the setup for each round is entirely based on chance.

    You asked readers to run through your points and respond to them, so here we go – in general, the idea here is that you’ve got good ideas, but I’m not sure if they actually add up to the conclusion you’ve given.

    “The act of coming to understand something is of value…”
    This is super true, and one of the main reasons that games are compelling in the first place. However, a game with or without randomness can be fun and rewarding to study. In Poker, the basics of the strategy involve learning your odds. Once you get a feel for this, you can make better judgement calls about what to do next, and that’s cool.

    “…(accessibility / easy-to-learn)…(depth / difficult-to-master)…”
    This is also a valid statement on its own, but it is not relevant to the discussion of randomness.

    This seems to be a reiteration of point one, but with more focus on PROPER FEEDBACK. The quality/clarity of the feedback given to the player for their actions affects how smooth or harsh the game’s learning curve is much more than the inclusion or omission of randomness. In a game like Candyland, where the player’s input has no effect on the game’s outcome, no amount of feedback will help because there is nothing for the players to do with any of the knowledge that might be gained.

    Think back to traditional logic: If we start with the statement “Games with purely random outcomes are of lower quality,” then it is NOT feasible to infer the inverse statement, “Games with purely player-based outcomes are of higher quality.” To get a new valid statement based on the original, you’d need the contrapositive, or “Games of high quality do not have purely randomized outcomes.”

    “…the game presents choices to the player, the player makes decisions, and the game then returns feedback…”
    This is a reiteration of point two.

    “[gambling’s] value is in inducing compulsive play by randomizing reward schedules.”
    Random elements are not necessarily only based on reward schedules. You’ve described a Skinner Box (like a slot machine), which is agreeably a manipulative use of random elements. Plenty of other uses of random chance (like dungeon layouts in a roguelike) are not inherently manipulative in this way.

    “Many modern games are some mix of ‘games’ and ‘gambling applications.'”
    So what? Again, the way that randomness is utilized dictates the quality of the product, not the fact that it was or wasn’t used at all.

    “…we become able to put together that we personally had no role in the win or loss…”
    Another reiteration of point two.

    “…there could be games that we enjoy even as adults which are illusory in a similar way…”
    If we still enjoy these games as adults, it is relatively safe to assume that the “illusion” is not actually an illusion, but rather a compelling game that happens to incorporate random elements. Again, Poker strategy after you’ve started paying attention to everyone’s faces.

    “There is no reason, no lesson that can be derived from [a random dice roll].”
    Not from a singular dice roll, but observing many examples of a dice roll can reveal the nature of the outcomes that might happen. Royal Flushes are less common than Two-Pairs, but are also more valuable. Once a player has seen enough results to have a solid feel for their luck, they start to consider which other mechanics they can apply these chances to, and their strategy begins to take shape.

    “…at a certain point, the strategy gets largely solved and it breaks down into sheer gambling.”
    Again, this is only true for a game like Video Poker. Human behavior is not solvable and can be studied ad nauseam (and this is a big part of where games like Poker get their depth from).

    “…it’s difficult to know exactly where strategy starts and where gambling begins.”
    A well designed game should always make it clear whether the outcome was a result of chance or skill, and understanding this distinction is part of learning the game. If a player decides to take a risky chance, they should be fully aware of the risk before they begin.

    “…most players probably won’t get to a point…where the system breaks down into pure gambling.”
    This just depends on the depth of the game.

    “I am concerned with developing a kit of guidelines for optimal game design…”
    That’s great! Like I said, though, I don’t think the inclusion of randomness should be a part of those guidelines – just don’t make games like Candyland.

    “Plenty of popular, well-liked games exist with no randomness at all…”
    This is true, but again, be careful with your logic. To apply this statement to your point, you’d have to be inferring the logical converse (which is not a safe move).

    This is a combination of previous blanket statements and logical fallacies.

    Pretending to sound smart is a great way to make everybody around you dumber.

  • Jr

    I think you make some interesting points. What I’m left wondering is what role you think randomness ought to play in games.

    Specifically, I am wondering how you think tabletop RPGs ought to handle things. Should there still be randomness, but it looks more like a bell curve? (Like, 3d6, instead of 1d20.) I am curious to know where you think there ought to be randomness, and where that should be removed, in an abstract sense.

    Like, should the randomness be allowed for things like setting (that impacts everyone), or for individual conflict resolution? I’d like to know what you think.

  • alasta

    “not counting the unavoidable things such as “who starts the game” with Chess.”

    Why is that unavoidable? You could for instance design chess around simultaneous turns, so you play at the same time (which I’m guessing is easier to do in videogames.) You are talking about the ideal aren’t you?

  • I feel like you’re somehow oversimplify. The point should be about finding a way (if possible) to choose between “bad” randomness and “good” one. I dont think it’s easy, or even possible.

    You talked about Chess and Poker. Let’s talk about the epitome of strategic game, the one game with no randomness at all: Rock- Paper – Scissor. A set of rule, no advantage to anyone (not even who begin) and no luck involved. But so boring.

    With statement like “Randomness should not be a part of ideal game design.” Then you end up designing RPS.

  • keithburgun

    An RPG might not be trying to be a “game” as I’m defining the word. The type of machine I’m referring to – the contest of decision-making – is a special kind of system.

    An RPG may have the design goal of simulating a fantasy, in which case agency may not be as important as the simulation.

    I address a few of your questions in my “Clarifications” section.

  • Nahil

    Hm, I agree with your conclusion, but not with what led you there. Points 1 and 2 just aren’t sufficient because they don’t fully describe what’s special about games. Yes, they give you a system to understand and understanding is good, but the other extremely important element is that games are about *doing*. You don’t just take a test on what you’ve come to understand, you apply it in the environment provided to you. Games are interactive and they require action from the player to even be considered games. Output randomness isn’t bad for games just because they muck up one’s ability to understand them, it’s bad because it takes away meaning from a player’s action, the driving force of the experience. The player tries to do something and then the system steps in and takes control. Why would the system take control of a player’s actions? That’s the question that needs to be tackled. In a medium driven by user input, when is it ok for a game to take control of the player’s actions?

  • keithburgun

    Well I mention that the process by which understanding happens is through decision-making – which is the Doing you’re talking about.

  • Nahil

    Right, but I’m saying that the argument makes more sense when specifically tackling the fact that games are about player action. When you talk about ideal games needing to be able to be properly understood, the main problem I see is that one could make a game that has plenty for the player to understand even if there is randomness in the game that the player cannot understand. Just because the game contains randomness does not mean that it doesn’t provide a deep system to learn and master. Randomness doesn’t provide anything to understand in a system, but it doesnt inherently take anything away from the system either. It just sort of sits there taking up space. What’s more important is how randomness affects player feedback and the amount of control players have over their actions. This is something that’s mentioned in the article and that you’ve addressed many times in the past, but I think it should really be the driving force in arguing against randomness.

    Actually, I have a small example that isn’t specifically about randomness, but addresses the effects of putting dead weight between a player and his input. Say you’re playing co-op metal slug with a friend. This game happens to have major input lag. You and your friend start jumping around and shooting things and a strange thing happens. Neither of you can tell who is who during the frantic action because the characters aren’t accurately responding to your input! Now you don’t feel like you have any control and the game becomes frustrating and tedious. That’s what randomness does to tactical games, but it’s just harder to noticed and easier to be tricked by the placebo effect.

  • Dasick

    What Keith is referring to in this post is randomness as a mechanic. Evaluating odds is it’s own game, a game that has been solved by mathematics. The real value of randomness not discussed here is randomness to provide context (mostly set-up randomness, but not always) – random dungeons in Roguelikes, random card draws in a bidding game etc. It’s very valuable to have context randomness, and equally important to avoid randomness as a mechanic.

    The big problem with randomness in poker is that you can have a hand that has a 10% chance of winning and then lose. Similarly, if I take a risk that only has 1% of winning, I can still win. Sure, there is the Mind Game aspect, but any multiplayer game has that, no matter how primitive. Good game design is about refining and enhancing the core aspect of a game, to be as efficient and elegant as possible. In a single round of poker, you can’t know if the winner is lucky or a better poker player. All the bluffing, body reading and mind games can be, and often are, overridden by a random card draw. Does that seem right to you?

  • Dasick

    What about situations where priority matters? Say, we both move a figure to a certain square… what then? Or if I move a piece to capture an enemy piece that also moves on that turn?

    In Arimaa the solution is quite smart actually. One person starts, but that person also sets up her figures first (played with chess pieces, but freeform setup). The second player can effectively counter-play the first player’s set-up.

    In Clairvoyance (http://www.youtube.com/watch?v=hNR4tzxrd4c), initiative (ie who gets priority) is something you can capture by spending a turn.

    There are solutions that don’t involve randomness. They’re just hard to come up with and test.

  • Dasick

    That’s just lazy design 😛

    Design is about *balancing* two contradictory things:

    Minimizing complexity.
    Maximizing utility (in games, that’s interesting decisions to make).

    Good design has high utility per complexity ratio.

    Kieth is talking about removing a lot of un-needed and harmful complexity in the face of randomness-based mechanics, but that’s only part of the picture. If you *only* focus on removing complexity, you might end up with little utility, and that’s as bad, if not worse (personally, I don’t mind games with bloated UIs and Encyclopedia-sized manuals so long as they deliver lots and lots of interesting decisions to make in return)

  • I would not call Rock Paper Scissors the epitome of a strategic game at all. The big problem with it, and in fact the reason why it is so boring, is because you have close to no context for what your opponent is going to throw. The inability of the human mind to truly simulate randomness is the only thing that keeps the game remotely playable, because there is in fact an optimal strategy of percentages of each element to throw, so long as they are randomly dispersed. It’s solved in kind of a strange, unplayable way because you still have psychological tendancies that will try to override your intended strategy.

    However let’s make RPS more interesting – gamble money with RPS and have unequal payouts for winning with each element, e.g. winning with rock wins you $10, scissors gets you $3, and paper gets you $1, each provided that you win that round. Now the game is providing you with some context for what your opponent actually wants to do. Clearly if he is behind in money (or perhaps just always), he wants to throw rock all the time because it has the highest payout. Knowing that, you can throw paper and ride on his assumption that why would anyone throw paper for a measly $1? Once you train this behavior and have him throwing scissors out of fear, you can start throwing rock for YOUR $10. At this point, you have a game very similar to a console fighting game where the entire crux of the game comes down to what you think he thinks you think he thinks you will do, and player’s skill at manipulating your opponent’s expectations can shine through with much higher clarity.

    I also think it’s very noteworthy that this improved version of RPS looks a lot less like randomness than the traditional version, but of course neither of them contain true randomness (i.e. elements that exist outside of player control). It’s just that the second has clearer reasons for individual victories and a better picture of player agency.

  • keithburgun

    >Just because the game contains randomness does not mean that it doesn’t provide a deep system to learn and master.

    Totally! I tried to make this clear several times. Again, I’m talking about *ideal* *guidelines* for game design, not practical RULES.

    > but it doesnt inherently take anything away from the system either.

    Well – this is something I didn’t really address in the article – but it tends to be a major source of ambiguity to a large degree. So yeah, I see what you mean.

    And I like the MS metaphor.

  • Nahil

    I understand that this is about ideal game design guidelines, but I’m saying that based on Points 1 and 2, there isn’t an extremely strong argument saying randomness should be absent in an ideal game. In a game with small random elements, players can come to understand and learn a deep, rich system without randomness getting in the way of the understanding. If a player gets a great move and it fails because of randomness, he still can perfectly understand the way things played out, the problem is that he wasn’t properly rewarded for his understanding. He/she understood that there was a 10% chance of failure or whatever, but it failed. The problem isn’t that randomness limits possibilities for understanding, it’s that it messes with the system’s ability to acknowledge understanding. A player shows their level of understanding to the system through an action, but the system takes away meaning and weight from their action by not allowing it to properly reflect the ability of the player. So what I’m arguing is that randomness isn’t an enemy to understanding but instead an enemy to the expression of understanding, the player’s actions within the system.

  • Dasick

    In the old Spacewar game you could use Warp, which would randomly teleport you with an increasing chance of being insta-gibbed. Is that an alright use of randomness or not, and why?

  • keithburgun

    Is that a joke? =]

    I mean, given that it was pretty much the first electronic game, I understand it, but no, that’s not OK!

    Warp’s main function is to save your ass when you’re about to collide with a projectile or the sun. So let me phrase it another way: is it OK to have it be that in Chess, when your king gets checkmated, you roll a die, and if it comes up a 5 or 6 then you don’t lose, instead you move your king to some random spot on the board?

    Keep in mind, I love Spacewar, I think it’s better than most digital games that came after it, but that is not OK.

  • keithburgun

    >The problem isn’t that randomness limits possibilities for understanding, it’s that it messes with the system’s ability to acknowledge understanding.

    Well, I think those are one in the same, because understanding is a process that I describe as the entire feedback loop of presenting choices, making decisions, and getting feedback. In theory, a game that produces bullshit feedback even 1% of the time is less preferable to one that produces bullshit feedback 0% of the time.

  • Nahil

    Right, I think the bullshit feedback is what makes randomness not ideal, but I don’t think your audience is getting that from the article. I’d say randomness is specifically bad for the feedback part of understanding, and should be talked about that way rather than just understanding as a whole because most of the people reading the article can point out ways that randomness doesn’t affect the non-feedback parts of understanding. I just think this is more about feedback for your actions than anything, and most of the arguments against what you wrote don’t really have anything to do with feedback, which suggests that you didn’t put enough emphasis on it. The Fantasy Strike dudes in particular are talking about how randomness doesn’t really affect understanding substantially, and I think it’s because they’re not specifically thinking about actions and feedback, the stuff randomness really screws with.

  • Arbitrary Value

    I disagree on two grounds.

    First, randomness doesn’t require dice rolls by the computer. Any game without complete information will inherently have a non-deterministic outcome even if the rules of the game are entirely deterministic; the trivial example is rock-paper-scissors. Randomness also comes in despite deterministic rules whenever skill is required to execute a strategy. I am aware that you frown upon that mechanic, but it is very widespread.

    Second, and building on the last point of the previous paragraph, randomness can have a useful role for smoothing out the learning curve of a game. If there is absolutely no randomness, the better player will win 100% of the time. This is, as you put it, an outcome that delivers zero understanding for the losing player. He will keep losing 100% of the time even if he improves a little, for as long as he remains the worse player.

    Thus, it can be desirable to create a game where, for example, a new player will lose to an experienced player only 90% of the time. This randomness need not come from a dice roll – consider instead a scenario in Chivarly: Medieval Warfare where a Man at Arms (a fast, melee class) is charging at an Archer. The Archer’s ability to fight in melee is relatively limited, but he has a crossbow that can kill the Man at Arms in one shot if he does not miss with it (he does not have time to reload). The more skilled the archer, the more often he will make the shot, but even an unskilled archer will hit sometimes, not due to an in-game dice roll but simply because even a shaky hand will click true sometimes. This way, the archer can improve gradually, from making the shot 10% of the time to making it 90% of the time.

  • Dasick

    The real solution to that problem (giving each player an equal chance at winning) isn’t rubberbanding because it reduces the impact of the player’s agency(can we agree that it’s bad?). The real solution are handicaps. There is no shame in playing and winning against a handicapped opponent. There can even be a ‘bonus’ for the handicapped player, like extra ranking or winning gives ‘honor’ points that unlock special skins in the shop that cannot be obtained anywhere else.

    In a ROFL-stomp as you described, the loosing player is still getting strong feedback for her actions, and the margin of the stomp does get reduced. (Personally, I find myself improving rapidly when facing overwhelming odds)

  • Arbitrary Value

    First, I am not proposing that each player should have an equal chance at winning but rather that the better player should not invariably win. It would actually be rather difficult to have the better player win invariably, at least when the two players are not widely separated in skill. I imagine even chess masters have bad days on which they might lose to someone inferior but still skilled. (Hence why the winner in a tournament is not usually determined by a singe match.)

    Second, I think it is possible to have built-in, self-correcting handicaps. Consider, for example, the various weapons available to the Pyro in Team Fortress 2. He has one flamethrower which does significantly more damage than the other available flamethrowers; that one is a good choice for newer players. More experienced players will benefit more from the flamethrower which allows them to push back enemies with an “air blast” and perform combos more quickly. In the right hands, these abilities are more effective than simple higher damage, but not so much more so that the newer player with the high-damage weapon is helpless against a more experienced player with the special abilities.

  • Dasick

    In most competitive games that’s actually a good solution but it doesn’t involve randomness and it shouldn’t involve execution as well – only the understanding of the system. So I guess we agree then?

  • Thanks, it’s a lot clearer, now, with the minimizing/ maximizing approach.

    So the point for Keith was to prove that actually randomness is just complexity. Or that you can always have a better Complexity/”utility” ratio without randomness. That’s a theory I can believe. But those 14 points didnt try to prove that. But again, I dont know if randomness as a bad thing (which it obviously is sometimes) can be proven.

  • Very interesting deconstruction, thanks.
    But for someone so articulate about logical fallacies, you’re pretty unfair to use an “ad persona” implying Keith just wanted to just “sound smart”.
    I think he just had a statement (random is bad) prior to actually proving it. So he didnt sounded smart but he sounded vague and with an agenda.

  • Yeah you changed RPS into a poker-like game. When I take RPS as an example I had the feeling that somehow it had a “hidden” randomness. You addressed that feeling by pointing out that the player has no way to anticipate the outcome , of a given match.

    (I know AI can beat player statistically after some games, and that there is human champions of RPS whit meaningful win strike)

    So my point was that with no luck involved you could build a random game. Dasick above nailed the problem : It’s not luck Vs strategy , it’s more about Complexity Vs Meaningful decision. I dont know the role of Luck in that trade-off. But I hoped Keith could somehow figure it more precisely than by hitting randomness on the wrist.

  • You’re right – that last bit is rude.

    It’s rude because Keith’s writing, based on previous articles as well as this one, makes me angry. It always gives me the same impression – he’s starved for things to write about on his blog so he Kerouacs all of his thoughts about an arbitrary topic onto one gigantic scroll and hopes for the best. That’s all well and good (and probably the purpose of most blogs, anyway), but Keith’s writing is disguised as useful scholarly advice (“I wrote a book! Here’s what I think about similar topics…”). With the incredible level of convolution and poor assumptions being made, the result seems dangerous for readers who might be looking to get better at the craft.

  • Jeremy Watson

    1 – Longevity: I would think that the ideal game design would include longevity. If to use “Poker” as an example was deterministic in one game. No one would play anymore except for games where players are evenly matched. I guess you might still call that the ideal game design but very few people would play it.

    Poker player consistently “blow up” because of bad beats. “I did everything right and you played like an idiot!” They do this on purpose. It keep the bad players coming back. Sure they’ll win a few hands but the game continues and the pros win in the long run.

    2 – Risk: I would ask is risk a required element to ideal design? It is difficult to present risky options without random elements because the only way to pull them off would be to rely on a mistake from your opponent. If risk is a needed element, how would you pull it off without relying on mistakes?

    3 – New Players: I would also ask is new player introduction a required element to ideal design? I think I know your answer on this one already.

    4 – Humans: The players… The players themselves are a random element. So unless your ideal game design is one player or against a computer, there absolutely has to be random accountability. Spending years balancing each choice so there isn’t an exponential run-away (or not CIV I’m looking at you) victory condition or shortening the game to a level you wouldn’t be happy with so that you can get games over with isn’t what I’d call ideal.

    5 – The Lie: What if a game was designed so that all the randomness was actually a lie? Here is a simple example: You have two dice [2][5], reroll one or two of them and you must avoid getting a total of 5, 7 or 11. What is the best move? Would you feel cheated if/when you found out?

  • dpeg

    I don’t know if it matters much for the discussion, but in my world, there is a crucial difference between how I play with. I am an avid Go player, and it’s the pinnacle of game design for me (clarity, depth, etc.).
    On the other hand, when playing with several persons, there is always indeterministic behaviour — this is a reason why I feel many boardgames don’t work *really* well (gang up on the leader, kingmaker etc.). There is a good reason why among card games Bridge and Tichu are so much better than the rest (that I know…): they are partnership games, and reduce the matter to a two-party affair. However, as opposed to abstract boardgames, there is mixed indeterminacy from card randomness and multiple players. (My point is that these are different types of randomness, and can mix well.)
    Finally, for single player games, I appreciate randomness (or “procedurally generated content”, as it is called nowadays) because I am interested in variation. When I play a game of Go, I can rely on my partner to provide an interesting that game which will differ from last week’s. When I play a game of Dungeon Crawl, I am happy that the game produces completely different scenarios for me every tme — it would be unplayable for most of us otherwise.

    Bottom line: even with a narrow definition of “game”, the role of randomness should always take into account the number of human players, in my opinion.

  • There’s a big difference between “input randomness” (“procedural generation”, stuff like maps, stuff that informs your decision) and “output randomness” (stuff that decides the outcome of a decision, like dice-roll combat for instance). This article is really only talking about the latter.

    My position on # of players is that for single player, INPUT randomness is a requirement, and output randomness… can potentially be acceptable. In multiplayer, input randomness is often good, but not necessary, and output randomness is almost always game-destroying.

  • dpeg

    Ah, that sums it up very well! Thanks.