American Squirrels
Play with Fire: Post Mortem

Designing Luck

Dice All card games rest upon chance as a primary source of play, as do the vast majority of commercially successful board-games – so why are videogames less commonly found relying on chance and luck as a central mechanic? Let us explore briefly the question of designing games of luck.



This post is part of an ongoing sequence developing a new (and disposable) taxonomy of videogames, largely as a spur for discussion. The following posts are worth a look if you haven’t seen them already:

  • The Rituals of Alea: the original post about Roger Caillois’ play pattern “alea” (games of chance) – essential groundwork for this discussion.
  • Emotions of Play Revisited: a new look at the connections between emotions (and Nicole Lazzaro’s work) and Caillois’ patterns of play.
  • Wargames: the first of the new taxonomy pieces.
  • Rushgames: the second of the new taxonomy pieces


The Appeal of Luck

Believe it or not, games of luck have wider appeal than games of skill – evidence for this can be found in the vast revenues made by the gambling industry. The global gambling industry had a turnover of $950 billion in 2001 (this figure has almost certainly risen in the last six years). By comparison, the global videogame industry is worth about $30 billion, and the global movie industry is worth $27 billion (just based upon box office receipts; DVD sales add another $5 billion or more). In other words: gambling is worth fifteen times more than videogames and movies combined!

Furthermore, TV game shows based on luck are phenomenally popular. For almost as long as TV has been around, Mediterranean nations have had game shows which under the hood are essentially games of chance. Even in the United States, where the dogma of competition and meritocracy dominates all aspects of the culture, “Wheel of Fortune” remains the most watched syndicated show on television, and the successful export of Endemol’s franchise “Deal or No Deal” to the US shows beyond a doubt that luck sells.

The secret: universality. Only a competent player can be victorious in a game of skill, but anyone can win a game of luck. But there is more to chance in games than just this aspect of accessibility.


Sources of Play

There are essentially two sources of play that any game can draw upon: the first is chance, which can be seen most clearly with a game like Beggar-my-Neighbour (AKA Strip Jack Naked) or Snakes & Ladders which use only chance to drive play. The second is choice (or design), which we see in any FPS game, for instance, which eliminate all random elements (excluding, perhaps, as a driver in AI) and place the player in a tightly controlled environment.

There are few if any card games that do not use chance as the source of play; this is inherent in the nature of card games. The player learns a small set of rules, and then operates within these mechanics to process randomly generated hands in a certain way, wherein the fun of play can emerge – either by offering interesting choices for play, or by generating interesting situations within the play. (Even though the player does not choose to play an Ace in a hand of Beggar-my-Neighbour – it is turned over as the next card – it is still fun to do so; there is the schadenfreude of making the other player suffer, for a start!)

If we extend our attention to include board games, and in particular the hobby games which are the more interesting part of that commercial field (since selling “reskinned” Monopoly sets is about marketing, not about game design), we can see that successful games in this market almost always draw upon chance. There are classic strategy games like Chess, Draughts and Go that have no random elements, but these are invariably a smaller market than those games that do draw upon chance to drive play – a new non-random board game can't offer much more than Chess or Go already offers, so it would be competing with a dominating established brand. Most, such as the short lived 1979 board game Kensington, vanish without a trace. I should perhaps acknowledge that chess sets probably do make up a big market share in the board game market, but the point remains: if you want to make a new commercially viable boardgame, it will include chance either via random tiles, a deck of cards, or by dice.

Does the same thing happen in videogames?


Chance in Videogames

The second most successful videogame of all time uses chance as its source of play – Tetris (33 million units on GameBoy, admittedly on the back of bundling with the handheld unit, but even without this, Tetris remains one of the most popular videogame of all time). Indeed, almost all puzzle games share this facet: Bejewelled and Bust a Move also draw upon chance as the primary driver of play (although there is a strong design factor in the latter), and it is difficult to find exceptions. Lemmings is a notably unusual case – although the appeal here can be traced in part to other factors.

What’s interesting about this link to puzzle games is that the connection – both in terms of play and in terms of popularity – between card games and puzzle games becomes apparent. Both are formed around similar principles: a small set of rules that mediate the way the player deals with a randomly generated sequence of cards, tiles, bubbles or what have you. I might go so far as to say that if you want to make a successful “Casual game”, you would do better to study the way people play solitaire card games than to study anything that happens on a Sony, Nintendo or Microsoft platform.

So why do we not see more use of chance in other videogame genres? The answer appears to be twofold: firstly, a widespread (but not universal) prejudice among game designers and programmers against chance that is probably connected to a micromanagement mentality, and secondly, a pragmatic quality assurance (QA) problem.

Games which use chance such as card games and puzzle games essentially represent atomic systems, which need to be tested and balanced in a variety of different ways. This is a viable QA task, because the scope is constrained. But consider what happens when subsystems in a larger game switch to random sources – consider (for instance) what would be implied in a subsystem that gave randomly generated monster distributions in the rooms of an FPS: the QA workload would be exponentially higher as a result of this. A single configuration could be checked in one pass, while a random configuration might not be fully checked in a thousand passes.

This resistance from QA is actually a powerful and necessary force inside the games industry, and arguably the reason that games drawing up on chance are found principally outside of the console market. Game designers love to posit design ideas that will generate “a different game every time” – but such games rarely make it through to the upper market. Games of this kind are much more commonly found as independent games, which by nature of their lack of budget have reduced scope, and thus inevitably fall into the comfortable arena of just one major randomly-driven system to balance and tweak. (Strange Adventures in Infinite Space springs to mind as an example).

The exception? Treasure tables in cRPGs and MMORPGs (think: Diablo, the game that spring-boarded Blizzard into the big time). These do employ chance as the major driver, and the QA implications are bridged by the simplicity of the mechanics involved (which are rarely more than the equivalent of a dice roll compared to a table). Animal Crossing manages to use this same random "treasure" mechanic without any violent play elements, and has outsold all but the most successful cRPG titles (8 million on DS, beaten only by Final Fantasy VII's 10 million and the Pokémon franchise's sales which go up to 14 million for Gold and Silver). Note how these exceptions reinforce the idea of the inherent appeal of chance as a mechanic within games.


You, Chance and Games

And now I must turn the discussion over to you. I have noted several instances of the use of chance to drive play – outside of videogames we have card games, board games, luck-based game shows and gambling games; inside videogames, puzzle games and treasure in cRPGs and toyplay sims. But there are doubtless situations that are using chance I have overlooked.

What other situations have you seen where a game has drawn upon chance to drive play? How successful was the attempt? Did it create play you enjoyed, or play that you hated? What is the most fun you have had with luck? What is the least? How much do you think your experiences – positive or negative – reflect your taste in games, rather than a deeper fundamental aspect of play?

At some point, the decision has to be made as to whether this new (disposable) taxonomy requires genres equivalent to Caillois’ “alea” – what I would end up terming luckgames. But while Caillois’ alea is still quite clearly a powerful force in play in general, it is not so clear that within the context of videogames a genre based on luck will hold water – especially when the main contender (puzzle games) can also be seen as rushgames. But then again, Bejewelled is not obviously a rushgame – but it is clearly a luckgame.

What do you think? Discussion in the comments is both welcomed and encouraged!


Feed You can follow this conversation by subscribing to the comment feed for this post.

Luck in games seems to work best when applied to the world generation process, and to specific, easily-comprehensible mechanics like combat. Chance elements that sit in between the extremes - like movement behaviors - tend to frustrate the player and reduce the strategy of the game to the "average case". For example, a shoot-em-up where all enemies appear and move in random patterns borders on unplayable; there's nothing to memorize or to react to, and there's nothing which one can plan ahead against.

On the other hand, Counter-Strike uses semi-random fire patterns as a key mechanic: the really great players can control these patterns, but most will rely on them like a chance element - which really helps the addictiveness of that game.

Chance seems to work best, as others have described, used in small blocks that appear regularly in the game. Loot tables, next block, fire patterns... these can be QAed because over a QA cycle they will be viewed tens of thousands of times. Though I *still* don't understand how a gnat drops a polearm as loot :-). To roughly quote Pratchett, "He'd seen Twoflower put a seven foot ceremonial pig-tickling stick into the Luggage, and bits didn't stick out anywhere."

I have a pet dislike of games that feature large amounts of chance that I can't ring-fence by my actions. A canonical example: Clock patience. It's random. I cannot affect the outcome - why would I play the game? Card games such as Bridge, however, are far more interesting - or Black Maria ("Hearts"), where there's the possibility of winning with a sufficiently *poor* hand. These are akin to the landscape-generating comments above: the deal generates the random landscape for play, but there is a high degree of skill in playing.

My own bias: I dislike luck games. I'm known in my RPG group for having an amazing habit of failing die rolls at critical moments while making the non-critical ones - this is, of course, both subjective and anecdotal, but "Pete's bad dice mojo" has reached legendary status within the group. Where I create games, I tend to create them so that any luck is tempered by skill before being applied, rather than being directly applied. For example, I have a space combat game with no luck at all in the game mechanics. I like it; most playtesters have declared it cold and unplayable. So I've introduced a deck of "random effect" cards where the players hold three cards and can play them at any time. This seems to introduce an acceptable amount of randomness for the playtesters, and it requires sufficient skill to play the cards that I don't feel too worried about it. By contrast, I'd be quite upset at the idea of introducing to-hit rolls or similar.

I think that random elements certainly have a place in games, however, elements of chance must be carefully watched and managed against other factors to be effective.

First I'll start with a short aside about the prevalence of luckgames in both the media and monetary arenas. I agree with your point that games of chance are great levelers, anyone can play and has the same odds of doing well for themselves. However, I don't thing that is as much of a driving factor as the money that is at stake in these games. Casinos, luck-based gameshows, and anyone else tied to luck games for the purposes of gambling are so because the odds that they will win are better than that they will lose. The best odds you have at casino are 50/50 (playing Blackjack by the book). This is by design. Gameshows are geared to give away some money, but on average less than they make in advertising. Neither would be profitable or sustainable otherwise. Games of skill, on the other hand, can be mastered, and then bilked by the players and will, over time, result in a loss for the stakeholder. Even gameshows of skill are geared to have enough chance that the players are more likey to lose over time than win.

But I've strayed on that point enough. Back to game design using random elements. I personally dislike any game in which the main element deciding the outcome is a random one. The more the random elements of the game affect how well I do, the less I can play. However, using random elements can be essential to providing lasting play.

As an example, match-3 games. Recently I've been replaying Puzzle Quest on my DS. And I've been noticing things the second time through about how the straegies in the game interact with the random elements of the board. The better I get at playing the better I do, overall. By playing smart and using the skills, modifiers, and intelligent selection of moves to make I can win more often than I did the first time through. However, the random gem drops are still the most annoying thing in the game. No matter how well I play, there is always that chance that the oponent may make a spectacular move that was unknowable beforehand. Each time this happens (especially in battles where the enemy is powerful enough to make it challenging in normal play), there is a break in flow when either side makes a spectacular and unforseeable move. When the luck overshadows the skill, and reduces its effect to nothing, then the luck removes the draw for me to play. Puzzle Quest seems to ride this line very close for me.

I think that this is where the line gets drawn on luckgames. All effective luck games either have some element (or appear to have some element) of skill to them. If they do not, then there is no draw to play them other than the elation and addiction of winning. The more that the game relies on luck, then the more that the game relies on fueling your winning urge to keep you playing. If the frustration beats out the elation, then the game stops being fun except if you are winning. This differs from games of skill, where losing can inform you on how to play better. In luck games, there is nothing to learn, and no way to improve.

Interestingly enough, I wrote a blog post about random numbers in game design, and I used Tetris as an example of a game with very *little* randomness. The only thing random in Tetris is which shape of piece will be next (after the one already waiting). And I thought that in general, we were using too many random numbers in MMO games. Everything else in Tetris is strictly deterministic.

My takeaway was that the operative element was not randomness per se, but *uncertainty*. In a single-player game, the only possible source of uncertainty is random chance.

Gambling, on the other hand, seems to involve something else, a process I've started calling "Intuitive Magic". Like intuitive physics, it's tied to our hardwired presumptions about how the world is supposed to work.

Interesting comments! Let's have a pick through some snippets...

James H.:

"Luck in games seems to work best when applied to the world generation process, and to specific, easily-comprehensible mechanics like combat."

Yes, there are two separate ideas here: luck as the generator of the play field (e.g. Bejewelled, Strange Adventures) - the "landscape" function - which I feel can never be applied to the bigger games, and luck as noise in an otherwise deterministic system - such as the fire patterns in Counter-Strike.

But these are two very different functions: one creates unique landscapes of play, one simulates reality by incorporating noise. Are these the only things chance can be used for?

Peter: An emerging pattern in the descriptions of your play is that you need to be in control. Rushgames rub you the wrong way, because they thrive on placing players on the edge of control - and that's not what you're about.

I particularly like your anecdote that you designed out chance in a game and your players disliked it - I can well believe this! With all chance removed, a game becomes a purely strategic affair - great for you (and other Rational-dominated players), not so great for the player-on-the -street. :)

I have faced a similar problem in games where I have designed out direct-conflict (for the benefit of players who dislike this); the players who like direct-conflict aren't happy about being forced into diplomacy. ;)

I like using card systems as the chance element in a game design, for exactly the reasons you say here: the play is controlled by skill, but a source of variation and surprise gets included.

I'm not sure if in the general case all players necessarily need to exercise skill in a game they are playing. Monotony, for instance, has very little skill (the dominant strategy is buy every property you land on; the closest to skill in the game is trading) but it's "popular".

I think this is part of the boundary condition between mass market board games and hobby games, actually: hobby games always centre upon skill, but will include random elements to ensure each game is different (the landscape function) or to provide excitement via die rolls et al (the noise function).

Conversely, mass market board games are only skill focussed when the skill is highly tactile e.g. Operation, Jenga. In all other cases, an illusion of control is more important than actual control, and it's worth remembering that most players do believe that they control the action if they roll the die. This is an aspect we've been unable to get into videogames at all!

As for your unlucky die rolls - once a story like this starts, it's hard to escape it. That doesn't mean there's not an objective case under the hood, of course, but stories are much more powerful than truth! :)

Duncan: Is money the only thing driving the appeal of luck? Well, clearly the appeal of gambling *is* that you can win money, but the appeal of luck goes more deeply than this, I feel. Remember that people *watching* a game show have nothing to win - yet there are still engaged. Plus, I'll mention Monotony again. People do play this game, despite Yehuda's groans (and mine for that matter!)

I don't get the impression that all players are looking for skill. In fact, I'll go further: I'm almost certain from the audience studies we've done that many (although a lot less than half, I think) players are not interested in proving their skill, and would prefer not to be placed in a situation that tests their skill. For to be tested in this way is to risk failing, and such people don't want to face this.

Consider MMO players who ignore the quests and socialise - they're in it for the mimicry. A lot of players are the same.

"I personally dislike any game in which the main element deciding the outcome is a random one."

In videogames, everyone would tend to agree with you. In board games, not so. This comes back to the point I made before that people believe in their own capacity to affect the outcome when they roll dice. We haven't managed to make that work in videogames, I doubt we ever will.

As for Bejewelled clones ("Match-3" games as you and others have dubbed them), Puzzle Quest is trying to sell a higher level of skill over the landscape function. But Bejewelled - which is far more popular - has a fairly low level of skill, certainly in untimed. (On timed, you have the skill of finding the patterns; but in untimed even this skill element is removed). Yet it is incredibly compelling, for the same reason that patience games are compelling despite having little or no skill: the player still takes the actions, and the player still has choices; that the dominant factor in outcome is the landscape function is somewhat hidden from the player, and since the player can reach a flow state (focus-immersion) just from running the process, it's "fun" for those players.

"All effective luck games either have some element (or appear to have some element) of skill to them. If they do not, then there is no draw to play them other than the elation and addiction of winning."

Your own bias came into this sentence, I think. You are presuming the superiority of skill to the draw of winning. The draw of winning is far greater (in general terms), especially when there is money at stake, as you hinted at before. A pure-chance gambling game *is* effective. But it need not have skill.

Now I accept your point that the money is what elevates these games to a level of interest, but people stake money on games of skill as well as games of luck - but the games of luck are the ones that are vastly more popular. So there must be some aspect of games of fate (alea) that is more appealing that games of skill, at least in the context of monetising the outcome. I feel the lack of control *is* the appeal - because you get all the excitement and fiero of a game of agon (competition) but if you lose, well, that's just bad luck and not your fault. That's an appealing combination of emotional factors.

Perhaps, therefore, it would be fairer to say "when no tangible prize is offered, games of skill are more compelling; when prizes are available, games of pure chance have wider appeal". What do you think?

Dave: Interesting post - thanks for the link! But when you say "Random numbers are used to bring uncertainty into gameplay" I feel this is a half-truth.

When a player plays a game for the first time, there is uncertainty even in a fully designed game (with no chance). So while what you say is correct (it describes the noise function) I think it's also fair to say that random numbers can be used to create a different experience every time you play (the landscape function) - which I don't think is fair to characterise as uncertainty (although I can see how you could make this claim if you wanted to).

Do you agree, or do you think the landscape function is usefully characterised as uncertainty?

Tetris has little randomness? You say:

"The PRNG in most FPS games is used so little it might as well not be there, the only thing random in Tetris is what shape will be the next to appear."

Yes, that is the only thing - but is also EVERYTHING. That the next piece you get is unknown is the essence of the game of Tetris; one develops methods to stack the pieces as they come, methods that are robust to whatever may come. The random number that determines the next piece is the essential driver of play in Tetris.

You could make a game of Tetris without random elements - all the tetronomos arrive in a cycling pattern, say - but it would become boring very rapidly. Tetris is never boring because it draws its play from chance, from uncertainty (to coin your term).

"In a single-player game, the only possible source of uncertainty is random chance."

I agree with what you're gesturing at here, but as I mentioned before: the first time you play (say) an FPS, there is uncertainty - you don't know what's around the next corner! There's greater uncertainty in this situation than in the next piece to fall in Tetris, actually, as in the former case *anything* could be lurking around the next corner, but in the latter, you at least know what the options are.

So I would say chance is not the only possible source of uncertainty in a single-player game, but it is the only method of reliably and repeatedly producing uncertainty. Which I rather suspect is what you were gesturing at anyway. :)


Thanks for the comments everyone! Very interesting reading.

"My takeaway was that the operative element was not randomness per se, but *uncertainty*. In a single-player game, the only possible source of uncertainty is random chance."

He's hit the nail almost square on here (sentence 1), although he seems not to realise it himself (sentence 2).

The thread of experience that ties all the games mentioned, gambling, luck, board and simulated-world video games is uncertainty. The source of the uncertainty is variable. The effect is to titillate the brain with nuggets of novel but interpretable information ('oh no, he's onto information again!').
By information, I just mean input, data, stuff, whatever you want to call what it is that comes into our heads and makes us think. No maths here.

Sid Meier thought he had it as well, but like you've said, sometimes there are no choices to be made. The point was never the choices, it was the uncertainty inherent in the choice. Thats how I got to Information Theory, by the way - it describes the act of choice formally (I like formality :D).

So, when I see mechanics for generating uncertainty, they look like a continuum. On the one end you have massive* luck-based events, like roulette wheels. The whole game, ALL the uncertainty, rides on one random event. Then a ways down from that, there are board games, with fixed worlds and dice rolls for progression, maybe some card-based skill/luck crossover mechanics. There are table-top wargames, which have so much dice/table mechanics that it becomes a skill game to get each set of rolls to go in your favour (charge the knights! protect the archers! etc). Video games are a leap, because they offer the provision of uncertainty through unknown worlds. You dont know whats around the corner in the simulated world, or where the opponent is in the FPS arena. Random events can be used, but aren't needed (and are pseudo-random anyway).

Seeing this continuum, I find it difficult to contribute to a taxonomy for splitting off luck-based games. All I know is the luck is almost secondary to the games' raison d'etre, from the POV of the players brain.

*(on the scale of the game's lifespan)

"people stake money on games of skill as well as games of luck - but the games of luck are the ones that are vastly more popular."

Are you sure? Have you forgotten spectator sports?

Thought about previous comment further - didn't cover everything. I believe social play can also fall under the cognitive addiction to uncertainty, since what is more generative of interpretable uncertainty than other people? And therefore even Mimicry can be thought to be satisfying this, although I would call this a type of mannered Mimicry. The mimicry that is about explorative release from the self, carnivale-style, tends more toward illinx, which is where play breaks away from the uncertainty function, to my mind (unless the uncertainty is now located within the players themselves, ie they dont know what they'll do next, self-control is abandoned...thats a hard claim to make).

Peter: An emerging pattern in the descriptions of your play is that you need to be in control.

Damn, rumbled ;-).

zenBen: some thoughtful additions to the discussion!

First, let me deal with professional sports - to which I must utter a hearty "whoops!" :) Just a month ago I was fully congnisant of the popularity of sports (which are agon to gambling's alea), yet when I wrote this piece it slipped my mind completely.

But, let's not be so hasty here... the US makes the most revenue from professional sports, and that only appears to be $16 billion. If I can make the assumption that European sports are worth the same, the total revenue from professional sports can't be more than $50 billion or so - a long way shy of the $950 billion for gambling. (I may have underestimated - if anyone has figures, please share!)

It's worth asking: do more people play the lottery than watch professional sports? I don't have the answer to that.

I accept that you can link Dave's "uncertainty" to your information theory, but I caution you not to let your love of reductionism carry you away. You, like so many Rationally-minded games people, want to find a single underlying mechanic - but I don't think you will succeed. Or rather, any single underlying mechanism you find will apply outside of games just as well, and therefore not tell the whole story about games. :)

I like your description of the relationship of luck and uncertainty, and the continuum that results... but I question a few things.

Mimicry. I don't think mimicry can be fully understood in terms of uncertainty. Now it's probably true that play involving mimicry involved uncertainty, but here we have to be careful because there are at least *two* forces in human enjoyment: novelty and familiarity. And the appeal of the familiar - which is intimately connected to mimicry - is not rooted in uncertainty.

Social play. I don't think this can be fully explicated in terms of uncertainty, either. The compelling aspect of social play appears to be less about not knowing what will happen next, and more about aligning oneself with the group. I'm struggling to find the substructure of this (in terms of brain areas, proteins etc.).

To be clear: I think you are on to something, but I caution you not to try and shove everything into the information theory box - the desire for a neat theory might blind you to the observational facts. :)

Novelty - uncertainty - is something humans do respond to, but routine - familiarity - is also deeply embedded in our behaviour. And in terms of research we've previously conducted, familiarity beats novelty 3:1 in games and elsewhere (witness the power of the sporting licenses!) But of course, this doesn't mean that uncertainty isn't a factor inside familiarity - if every sporting game played the same way, it would be very boring.

So to reiterate: I do think you and Dave are onto something, but I caution prudence in assessing uncertainty as sufficient for a complete description of play. It is something we should dig into more, though, don't you think?

Finally, the question of whether we need a luckgames branch of our (disposable) taxonomy... Well the fact of the matter is, people are playing poker online, and that alone could be used to mandate its inclusion - unless one makes the arbitrary decision to exclude gambling. (But some of the people playing poker aren't playing for money, so...)

The bigger question for me is: can Bejewelled be understood as a rushgame? And if it cannot, which I suspect, should it be understood as a luckgame?

If anyone other than zenBen is still reading, it would be great to hear a few extra voices on the question of whether luckgames will form a tangible genre group inside videogames. :)

Best wishes!

Sports vs gambling - I guess it depends how far you wanna stretch the definition of game. I train 2 or 3 times a week at the dojo...but I don't gamble or watch any sports. Do I even count in comparing their relative popularity?

"I accept that you can link Dave's "uncertainty" to your information theory, but I caution you not to let your love of reductionism carry you away."

Now hang on, I did say explicitly: "By information, I just mean input, data, stuff, whatever you want to call what it is that comes into our heads and makes us think. No maths here."

Maybe I wasn't clear enough, since I went on to mention Info T. anyway. No matter. can I go on record to say: I DO NOT think a reductionist approach will provide the 'final answer'; I just hope one reduction at a time might provide insight into one module of brain operation at a time (even a tiny amount of insight is worth the funds - people are spending money on evolutionary psychology to explain why girls prefer pink, for gods' sake).

As for uncertainty - weeeeell, it has to be interpretable novelty, otherwise it can't be decoded and enjoyed. So familiarity is definitely in the mix. And there is the temporary disalignment of illinx - what can be predicted or reasoned about that?
Trying to untangle the mix would be fascinating!

"the desire for a neat theory might blind you to the observational facts. :)"

Definitely, I'm very wary of this. Unfortunately I have far more access to theory than fact :(

Never played for luckgames in a taxonomy, I think you have to. But could the taxonomic categories have fuzzy boundaries somehow?

As well as luck playing a part in 'generating a landscape' (I understand Tetris would be placed in this category by you?) or providing 'noise', perhaps random numbers governing non-influential factors could be a separate category?

Though you say it's not true alea, the 'wrapper' of the game (non-game-affecting graphics, names, messages etc.) can certainly affect the experience of playing and seeing a weapon that behaves identically but has a unique name and graphic could still elicit pleasure.

Another suggestion for a 'category' of luck - what about 'luck' as a way of 'locking' content - e.g. Chris Houlihan's room in Zelda 3?

I don't see why luck-games WOULDN'T be a category.

But could the taxonomic categories have fuzzy boundaries somehow?
Seems to almost always be the case.

Random events can be used, but aren't needed (and are pseudo-random anyway).
I never understood why a computer's 'pseudo random' pattern would be seen as being intrinsically different to rolling a die. Unless you're using an emulator and TASing, I see no real difference.

zenBen: hearty apologies for overestimating your commitment to reductionism here; I read between the lines, but as is so often the case I found my own prejudice there. :)

You are of course correct that looking at revenues doesn't tell the whole story. After all, how many millionaires contributed to the cash turnover of the gambling industry? :) I think as a general indicator, though, it is illuminating.

Thanks for the link - I am conducting something of a literature review for biological mechanisms that link to play experiences, and this will be interesting. I am at the point of suggesting that the trigger protein behind the fiero experience is dopamine, possibly modulated with norepinephrine. I have no real ability to investigate this sort of thing further, so I need papers to help. :)

And I agree with Bezman, both that genre categories are always fuzzy, and that luckgames may be an inevitable part of this taxonomy. But I should like to identify a few non-gambling luckgames - and Bejewelled is my front runner.

Bezman: "non influential factors" - aesthetics driven by random elements? A tangential topic to be sure, but worth noting!

Luck as a means of hiding content? This is an interesting one, as recently a few games have come to rely on the fact that an audience of hundreds of thousands or even millions of players guarantees the discovery of something "locked" by chance. Another tangential topic, but equally worth noting!

Got any more? ;)

"I never understood why a computer's 'pseudo random' pattern would be seen as being intrinsically different to rolling a die."

That's probably because you're assessing this through a fairly detached paradigm. But people are anything but detached about dice rolling (I mentioned this before in The Rituals of Alea)! They believe in their capacity to influence the outcome of a die roll, but rather less so in their capacity to influence a random number generator (even though, ironically, there's more scientific evidence of the latter than the former, but that's another story!)

There is a sense of ownership in the die roll for most people - it's a fascinating part of the psychology of play.

Best wishes!

I probably phrased that wrongly - I fully understand the psychological difference between rolling a dice or pressing a button - there's the tactile element for a start... I dislike being given random cards or being moved random spaces on a computerised version of CCGs/roll-and-move board games, whereas I'm far more willing to live with - or even enjoy - such elements in real life.

For a start, there's always the suspicion that the RNG really IS cheating - short of breaking/entering and stealing the source code (or decompiling), can we really know otherwise?

I just wanted to know why zenBen doesn't consider 'pseudorandom' luck to be equivalent to 'true randomness'.

Back to the 'uses of luck' - would you call Roulette using luck to 'generate the landscape'? It seems to fit, but is a stretched case...

Maybe it's worth making a distinction between cases where luck generates a playfield and choices are made based on that (e.g. Hearts,Spades, M:tg, other card games, Tetris, later parts of Minesweeper) and those where the playing is about predicting how the 'playfield will be generated' in the future - e.g. Deal or no Deal, Roulette, Blackjack.

I didn't realise equivalents to Chris Houlihan's room existed in other games and look forward to reading about them. (Hopefully...)

"I just wanted to know why zenBen doesn't consider 'pseudorandom' luck to be equivalent to 'true randomness'."

Naturally, I get caught on the loosest assertion!
I don't see much of practical difference, past the psychological one you've both mentioned. Just threw in the parenthesised comment to feel smart :D

On the other hand, I did once program a random-topology maze game, where the pseudo-random maze layout was built in real-time in reaction to the player's efforts to find the exit. Over the testing, I noticed high-level predictability coming out. But no sane person plays a game for pleasure the way I did to test that thing.

Personally, I think that the best approach to taxonomising games that include random factors is to look at how players have to deal the randomness in the mechanics/cognition* of play.

*(not sure of the right word there)

Say you have very random events which form a very large part of the mechanic, like roulette. Its totally central but players have zero influence. They think of it as totally external to themselves.
Then something like craps, thats maybe the same but the player thinks they have more influence. On to card games, then maybe tabletop dice-based war games, then skill games with uncertainty, then games of complete information.

That would be my approach, and draw the lines around the whole of luck games...well, where do you think?


"For a start, there's always the suspicion that the RNG really IS cheating - short of breaking/entering and stealing the source code (or decompiling), can we really know otherwise?"

Now it's funny, because almost all RNG's in games are on the level. I say it's funny, because it strikes me we could design a "fake RNG" that would give the player a more entertaining ride. But this is a space that no-one particularly wants to get into, it seems. And, to be fair, when random chance works pretty well, and is what people are expecting, you'd have to be pretty confident to push a solution that massaged the values in an intentional manner. :)

I wouldn't consider Roulette to have any kind of landscape - as suggested here, it's more a case of single unitary random events. (On the other hand, the player might take a landscape out of the play of a game of Roulette; still, I feel - as do you - there's a distinction here).

Oh, and as for Chris Houlihan's room, I'm not sure I have anything tangible to offer in this respect. But it struck me that the Pokemon games have been leveraging the huge size of the player base to hide content within the game space, counting on chance to uncover it. I remember thinking this about Silver and Gold in particular; I've not stayed abreast with the franchise since, though.

As for how to draw the luckgames part of the taxonomy, I was thinking along similar lines to what zenBen suggests here: a continuum from those in which the chance is the core of the game (such as Roulette) though the mediated cases (Poker) to the games where player action is the dominant factor, and luck just has a landscape or a noise function (or some similar peripheral role). These latter games may not be best understood as being part of luckgames.

But I find myself in an awkward situation, as I need to write a piece on luckgames to move forward but I don't feel that I've put quite enough pieces into place: I'd mostly be recapping the content of this piece and Emotions of Play Revisited.

I could come back to it, I suppose, but it would be nice to put this one in place as this should complete all of the "stress games", I think.

Oh, and lastly: many thanks for that paper, zenBen. It's just what I was looking for: neurobiology underlying interest/curiosity. I had most of the neurobiological underpinnings for play sorted out (or at least, sketched) but I was missing a way to tie interest/curiosity into it. This paper not only give this to me, it makes the biology of play quite neat as it doesn't use any biological mechanism I'm not already using somewhere else (i.e. limbic system, dopamine, cortex pathways).

If you find anything more of this kind, do let me know!

Best wishes!

"Oh, and lastly: many thanks for that paper, zenBen. If you find anything more of this kind, do let me know!"

I like Rauterburg...especially this.

My own theoretical opus is a converted literature and draws heavily on him and Csikszentmihalyi. Thing is, although its over a year old, I haven't found much that causes me to disagree with my main points.

zenBen: for some reason, although the mouseover to your paper looks like the right link, when I click on it, it goes to this link:

*shrugs* Not sure what the problem is.

Fascinating comments! I've got a lot to read and learn about here ;) But turning back to the original article; this is going to be a bit of a random brain-splurge of comments.

As I've commented before, I don't generally enjoy a large aspect of luck in games. I'm not nearly as against it as Peter is, I am definitely aware that I DO enjoy the uncertainty and randomness to a point, but it does seem to depend on the level of control I feel I have over it though. Essentially: Do I feel there has been enough skill in manipulating the odds?

"card games, board games, luck-based game shows and gambling games" is a nice list of games that I do not generally enjoy playing at all... such as Battleships or Monotony ;)

I do however have a lot of experience with luck based elements in games such as CCGs, tabletop wargames and rpgs, cRPGs & MMOs. How much I enjoy these experiences does all depend on the level of control I had. If I felt I had sufficient influence over it, then I tend to find it is ok, if not, it frustrates or even bores me. For example in a tabletop wargame there will be a huge amount of dice rolls that will usually even out pretty fairly over the course of a game, and I will have a reasonable (to me!) amount of control over when and where and how these rolls are being made. However what frustrates me (even to watch as much as experience myself) is any huge 'swing point' when a game turns on a single dice roll. For those familiar with games such as Warhammer, a close game often seems to boil down to a single morale check or saving throw deciding the game; which irritates me no end. I find it very unsatisfying, win or lose.

Pokemon (original Red&Blue! I've not played any since): Capturing a rare / hard to get Pokemon. I did find this a lot of fun actually, despite its randomness, as it always felt like there were enough things I could do to increase my odds. ie: collecting more and more potent poke-balls, putting the enemy to sleep, getting them to very low HP without killing them... In fact I took it so far as to designing a specific "capturing" Pokemon party that was designed with all of this in mind & was extremely efficient at it, hence I found it a lot of fun. If it had been purely tied to the amount of pokeballs I threw.. I am not so sure I would have enjoyed it.

MMO's and cRPG "rare drop hunting". I think this is right on the fun/frustration luck borderline for me. I've certainly done a fair bit of it (as I am a bit of a collector/hoarder in games as well!) and sometimes the amount of which I can manipulate my chances to get a rare item or drop of some kind satisfies me. For I have definitely enjoyed finding the most efficient way to maximise my chances to collect a specific item in Phantasy Star Online or in various Castlevania games. But equally I've become quite frustrated at times when the chances seemed too low and too much repetition was needed. Sometimes by the time I even got the 'drop' I had been working for, I had become so dissillusioned by the process I did not even feel that happy about it. However one of the strengths of these games for me is that this pure "luck game" portion was never the be all and end all of the game - even while intentionally playing it, you are generally also collecting more money/experience/points etc which helps you in the 'real' game also, or even still socialising (in an MMO). You can also choose to completely ignore this aspect of the game if you wish to do so.

MagicTG - the luck level in a single game certainly feels like it ought to annoy me, but this is generally mitigated by simply playing enough games. I don't like decks/formats that tend to enhance the luck / randomness factors though. eg. Land Destruction decks, or "bombs" (overly powerful single cards) especially in limited formats (smaller random sets of cards are used, rather than fully predesigned decks). Oddly though, I'm so cognisant of the luck factor when playing Magic it was very rare I would get frustrated with a single game's result. In fact I could often be rather pleased with it, even when on the losing end. eg. when it allowed a much younger/weaker player to beat me - and seeing how much they enjoyed it! I also could not fail to note that at most levels of play, the same players would have a tendancy to win, even in the most supposedly "luck based" situtations, so I couldn't reasonably argue their wasn't a huge skill element to it.

A thought re: the pRNGs etc. I wonder if people tend to view the luck element very differently in MagicTG Online as opposed to with physical cards.

I wish I'd played Bejewelled now to join in the commentary on that.

Overall though in video games, I do think whilst luckgames can hold up as a genre, once you define it as a videogame where "chance is the key driver for play" I
don't think it will be a very wide genre right now, as most of them are simply going to be video game adaptions of other kinds of games. However I think it needs to be fairly strictly defined, as if you widen the definition too much, I think you can find elements of alea in almost any video game if you look hard enough. Even Street Fighter 2 actually has some random damage factors hidden behind the scenes! (something I was quite surprised to learn recently).

Thanks for a detailed commentary, Rik. If you write this large a comment this fluidly, you really might want to consider having your own blog. ;)

"Do I feel there has been enough skill in manipulating the odds?"

I think this is a common experience among gamer hobbyists - those that truly love games of luck are generally not gamer hobbyists, and the perspective of the hobbyist is generally to enjoy luck provided the balance between luck and control sways sufficiently towards the latter.

"In fact I took it so far as to designing a specific "capturing" Pokemon party that was designed with all of this in mind & was extremely efficient at it, hence I found it a lot of fun."

Yes, I did the same. :) There is a tremendous amount of luck under the hood in the Pokemon games, but they give you just enough capacity to influence (but never quite to control) that chance element that it somehow works.

"Sometimes by the time I even got the 'drop' I had been working for, I had become so dissillusioned by the process I did not even feel that happy about it."

I recognise this experience in my own play, too - especially with Castlevania games. There's clearly some kind of balance issue here that can step too far into boredom and thus undercut the reward to be gained. Perhaps games with random factors in their treasure should ensure there are abilities to be gained (perhaps in the later stages of play) that can push the random elements further into one's favour, to sidestep such a problem?

But then, perhaps there are players for whom the longer path to the "drop" is enjoyable? I'm doubtful of this - boredom does not enhance fiero like anger does - but I'm open to the possibility.

"Oddly though, I'm so cognisant of the luck factor when playing Magic it was very rare I would get frustrated with a single game's result. In fact I could often be rather pleased with it, even when on the losing end."

And isn't this part of the appeal of the luckgame? That the outcome - win or lose - can be *interesting*? That's certainly the case (say) in Texas Hold 'Em and other poker games. The random element of the deck in M:TG is the essence of what is absorbing about these games, and I tended (back in my tournament days) to favour decks which could throw up strange and confusing combinations.

"I do think whilst luckgames can hold up as a genre, once you define it as a videogame where "chance is the key driver for play" I
don't think it will be a very wide genre right now, as most of them are simply going to be video game adaptions of other kinds of games."

I agree - this seems to be what will happen. I will post on luckgames at some point, but it isn't going to contain much inside modern videogames, it seems.

Those games that use luck as part of their mechanics should probably not be considered luckgames if the prevailing tone of the play does not match the tone of Caillois' alea, but I can talk about luck based mechanics as well in the relevant piece, as its as good a place as any. ;)

Best wishes!

I just noticed this - my host doesn't allow links straight to files, only html. Fie on it.

So here are all my publications, but because the host is so terrible I have started an update but not finished it.

If it still doesn't work and you still want the paper (beware, its long), I can PM it.

Thanks! Since your new link takes me to a list of papers, can you give me a clue as to which one you were originally linking to? ;)


I'm trying to build a board game where the players have the choice of randomness level.
So we can choose between 100% and 0% randomness.

That's an interesting challenge, Minopoli! I think this should be manageable... certainly, the mechanics are workable.

Does each player get to choose their degree of randomness, or do the players collectively choose it? If each player chooses their degree, then the problem will be game balance. Interesting challenge though!

Let me know how it turns out.

[Martyn Stanley's comment has been moved to Rushgames, and my response can be found there.]

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Your Information

(Name is required. Email address will not be displayed with the comment.)