Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
November 1, 2014
arrowPress Releases
November 1, 2014
PR Newswire
View All
View All     Submit Event

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

Lucky Breaks
by Randy OConnor on 09/06/13 09:59:00 pm   Expert Blogs   Featured Blogs

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.


I'm struggling with luck these days in game design.

I have been extremely lucky in terms of my non-digital life. I have seen great concerts, backpacked a ton of miles this summer, traveled, went to PAX with Ian and showed Escape Goat 2 to great response. We sold a ton of preorders, a bunch more people are now wearing the shirt I designed. There are some great things coming down the line. This month is super busy with a series of fun activities, not least of which is hopefully releasing EG2 (if not soon after).

But are these things luck? What the hell does luck even mean? I listened to a talk at PAX Dev that mentioned lucky people are often more observant.

In games we generally consider luck to be a dice roll. We lay our troubles before the fate of the universe. As a designer, it feels wrong, though, to do something so lazy. I was told by a friend that Michael Mateas thinks dice rolls are basically us copping out of work.

So imagine my dismay at knowing that my two big personal projects right now are fundamentally ruled by randomness.

The Yeah Yeah YeahsDid I discover a new band this summer by luck? Or was it proactively seeking out concerts and podcasts? (The Yeah Yeah Yeahs)

My boardgame about pirates allows you to shoot other pirates, steal their treasure, and move around the board. But your hand randomly assigns which of these actions you'll be best at for the next round. Do you do the things you've been given? Do you want to? Did you move off to a distant part of the board and now one of those actions is inconsequential? 

Randomness in my boardgame has grown on me, despite knowing that I can get completely owned by newbies.

The randomness seems to level the playing field. You have to take the opportunities where you can, because you don't know if you'll get another shot. It creates stories. It presents massive opportunities and dearths of opportunity, and the excitement in the game is created by not knowing how the numbers are aligned in a turn. It's coming back from nothing, or losing it all, but always building up a series of moments.

I've resisted making the game shorter than its (approx) 90-120 minute playtime, despite randomness throwing long-term strategies to the wind. I think it's because I want you to have a storyline. Who cares about being a pirate for 30 minutes? You want the buildup, you want the finale. The joy of the game is living through that tempest with a group of friends, seeing who survived and who was dashed upon the rocks.

Alexander Bruce talked once (or perhaps several times) about taking chances and how that was how success came into being for him. I'm sometimes afraid to take chances, because I know that there's more work involved. Taking a chance opens the door to work. Work is tiring.

My other game is much more difficult for me to come to terms with in the role of luck. It's an iOS game where you choose the trails to make your way up a randomly generated mountain. Leave the main trail for excitement and random lucky breaks and moments, stay on the main trail for nothing but safe steady progress up the hillside. But neither really feels like the right path. So how do you choose? Is there weight to your choice?

I think this new game is supposed to be a metaphor for life, for that crazy unknowing of how a little decision now can shift the far future. I think that's what the game is. But is that fun? Does the game have to be fun? Do you want to hate me knowing that you'll play for eight minutes and a decision you made three minutes previously is why you died?

Untitled mountain ascent game

I brought up the fates of the universe earlier because I think there's a real desire that any game should slowly reveal its systems in a way that we can master or better ourselves at. We can choose our paths in life, and we have a hope or intuition or knowledge of where they will lead. We should not be so powerless as to fall down dead through a random dice roll. It certainly feels like the universe has that power against us sometimes. Forces such as class and genetics and location subject us to poverty and cancer and war, and despite our best efforts, we may not win. In games we want to not be subject to these things. 

I'm not bored of basketball. I'm not bored of chess, or Counter Strike. I know there is mastery to be found, like I want to believe there is in life itself, though life has a different set of boundaries.


Right now I'm bored of my iOS game. Luck seems to be the problem. I have thrown some code up, I have created a universe that gives you paths to choose from. I want that metaphor to have purpose.  Yet I feel no urge to take any path, less traveled or not. I think I need mastery in my game. But does that mean that the luck should go?

Do I want you to never get tired of my game? I don't know. The designer says you should find endless mastery. My knowledge of the universe says that is also the case. And yet I persist in this unfair, random, lucky nonsense. The mountain awaits you, and there's a good chance you'll grow tired, you'll pick the wrong path too many times, according to how I see fit. Perhaps the bounds of the system are just not wide enough to hold all the paths up to the top that you should be able to take.

Do I want you to reach the top of the mountain? Do you think you should be so lucky?


Randy is an indie developer currently finishing up the art for Escape Goat 2.

He helped make Waking Mars with Tiger Style Games which is totally available on everything.
Randy also made Distractions and Dead End for iOS!

You can also follow his ramblings on twitter, like him on Facebook, go to his deviantArt, or I dunno. Maybe you wanna read some of his poetry?

Related Jobs

Rensselaer Polytechnic Institute
Rensselaer Polytechnic Institute — Troy, New York, United States

Assistant Professor in Music and Media
The College of New Jersey
The College of New Jersey — Ewing, New Jersey, United States

Assistant Professor - Interactive Multi Media - Tenure Track
Next Games
Next Games — Helsinki, Finland

Senior Level Designer
Magic Leap, Inc.
Magic Leap, Inc. — Wellington, New Zealand

Level Designer


Darren Tomlyn
profile image
There are two main types of randomness and luck that can be found and used within games, and recognising and understanding such differences should be part of the basic knowledge every game designer must have - since it's part and parcel of what games are in the first place.

There is randomness and luck on behalf of the player and their behaviour, and there is randomness and luck on behalf of the game itself, including its setting.

It's all about recognising the difference between things the player does, and things that happen to them - and understanding that the two can never be the same thing.

Games are defined as and by the behaviour of the player(s), not the game (and/or its creators).

Randomness on behalf of the player, (such as dice rolls), is usually about giving the player influence over such randomness in order to compete in playing the game, and therefore usually defines a game as a game of chance.

However, a lot of games, (especially card games), often involve a random starting position and capability for each player, from which they need to compete by using their own power in how they are used. It still depends on just how much power the player has over the outcome, based upon such a position, itself, however, as to whether or not such an activity can be a game. Being dealt some cards, or other such random behaviour offered TO the player, has nothing to do with any activity being a game, since it has nothing to do with the player's behaviour.

If the cards dealt, for example, are all that determines whether or not they win or lose, (are able to compete), then their influence is negligible, and it's not really a game - they have won or lost based on what has happened TO them (the deal of the cards) rather than anything they've done.

Note: Deciding whether to continue playing or not, does not count as actually playing the game itself - (see poker).

None of this means that chance and randomness, and the luck it enables, have no place in being being offered TO the player during the game itself, only that it has no place in defining it AS a game, which is currently a problem - (Wikipedia, for example, defines a lottery as a game).

An activity in which someone competes to be TOLD whether they have won or lost, (usually by a random draw or Judge's opinion), is (and/or should) be called a competition. (E.g. talent competitions etc..)

A random draw of some kind is one of the main ways in which such randomness is offered to people taking part in such activities, but randomness in the setting, such as weather, terrain etc. can also be part of such an activity.

Randomness offered in such a manner, can be part of a game, as an element the player must compete AGAINST, by doing something for themselves, as they play the game.

For this reason we therefore have three main ways in which randomness can be used within a game:

1) As part of the players behaviour, (as having some influence over), in order to compete whilst playing the game - (e.g. dice rolls). This usually defines a game itself as being chance-based. (And yes, this means that if you automate such dice-rolling then it's usually no longer a game, either.)

2a) As offering a random position to compete from, and/or with random playing pieces to use and compete with.*

2b) As offering something to the player to compete against.

*This is the hardest to get right whilst consistently being used to enable a game - if the player has no true influence over the randomness, and has no true power over how the results of such randomness can be used to compete, and therefore win, then it's the randomness itself that has already determined the outcome, which is not part of the players behaviour, and therefore it cannot be a game. It's possible, since it's a random position, for this to change on each and every play. If this is the case, then it's the DEFAULT behaviour that must be used to define the activity itself. (This also applies to judge's opinions too, which is why boxing should be defined as a competition, even though it's within the player's/competitor's behaviour to determine the outcome, and therefore make it a game.)

Example: In hold'em-based poker, the player has absolutely no influence over the cards/playing pieces themselves. If the player cannot compete by their own behaviour, (betting money etc.), then the cards themselves determine the winner. Such an outcome has nothing to do with the player's behaviour and therefore cannot define it as a game - it is, instead, defined as a competition. However, in draw-based poker, the player does have influence over what cards they have, as part of their behaviour, and since it's still a random draw, it must be defined as a game of chance.

It's therefore possible for an individual activity to be DEFINED and LABELLED as and by the result of such randomness, before and during it's existence - such as a draw of cards/numbers - because the behaviour enabled can be affected and therefore also be random.

(Note: Solitaire is one of the other main activities for which this is also true, but that's a subject for a blog post in itself.)

Alex Boccia
profile image
both I guess, it depends on how you do it. I had a friend who was 100% against chance. He just got angry/incredulous when confronted with chance in games and simply couldn't handle it. I think they are fun, especially when applied to stats/diceroll combat such as in popular MMOs. I think on the other side of the coin it becomes a crutch though because look how homogenous the gameplay in the current MMO market is.

I think chance is also fun because it's not a guarantee. When you get a buff or ammunition for a weapon, say it gives a CHANCE on hit/swing to do something extra. I think that's exciting, because you are going into combat betting/relying on that buff to work, and when it does, it feels like magic.

Plus, chance occurs in real life. People have been demanding more realism in games for years, no?

Maria Jayne
profile image
I tend to think of dice rolls as representing the organic element in games. Factors that cannot be defined with numbers that can be reliably reproduced.

For example you cannot swing a sword exactly the same way every time, because you are human. The positioning of your feet upon the terrain, the way your muscles and tendons perform, the environment you are within and the object of your aggression add up to make the outcome uncertain.

Now you can train all your life to swing that sword accurately, but one day you could pull a muscle, slip on some moss on the ground, be blinded by sunlight in your eyes or have your opponent do something unexpected at the last second.

A dice roll prevents the outcome being a sure thing, it allows for an element of luck to be a factor. Either lucky if you are unlikely to succeed or unlucky if you are an expert and have decades of experience. The mechanics and engineering of a machine or a computer facilitates the ability to remove luck from the majority of its usage.

If a machine is functioning correctly, it is predictable for a significant period of time. Natural wear and tear or environmental degradation may well loosen the tolerances of the machine over time but for the most part, it does what is expected if it has been designed and manufactured correctly.

But something organic, is prone to external and internal influences, that cannot be predicted or compensated for under every circumstance. A dice roll with or without modifiers is a representation of this element of uncertainty. It makes the outcome potentially exciting or catastrophic.

I think if you can predict an outcome 100% every time, perhaps some of the potential is lost. Consider Chess, a game largely created with set tolerances that never change. The only variable are the players, if the players are both computers, the outcome becomes the same, would you continue to be engaged watching two computers of equal power duke it out over and over resulting in the same outcome?

It is the element of luck or uncertainty that makes all sport engaging, because if the outcome was guaranteed, we just wouldn't spectate.

Tristan Angeles
profile image
Some games, especially multi-player games use luck to give new players a chance against more experienced players.

Carsten Germer
profile image
Does "randomness" have to be really random?
Just a little example how I, as a designer, have previously dealt with rolling dices for calculating loot in a chest-like-thingie.
I used this technique for something comparable to "critical hits" and on multiple other occasions.
It can easily be abstracted in code and only requires holding a few more variables for every not-really-random-randomness.

Let's say the chance of finding a golden ticket, besides the usual sell-loot, in a chest should be something like 1:10000.
If you use true randomness the player may eventually find a GT in three consecutive chests and after that she will find no GT at all in three weeks of constant play.
As a game designer, just thinking of this should make your stomach cramp.

What we do first is to reduce the number 10000 by 100 on every try, store this variable in howLikelyIsPlayerToFindGT and use this variable to calculate our random chance. 1:9900, 1:9800 and so on.
The chances get better with every chest until, after 100 tries, a GT is granted.
Remember, it is still possible the player finds a GT with a chance 1:8400 or all other combinations, the odds simply get better with more tries.

When the player has found a GT we want her to not find another GT too shortly after.
We set another variable absolutelyNoGTFindable to, say, 10.
With every chest looted we make sure that no GT can be in there and decrease the variable by 1.
Once absolutelyNoGTFindable reaches 0 we go back to our initial 1:10000 and reducing by 100 with every try.

A really cool thing is, that you can influence the variables withs buffs, in-game events or whathaveyou.
* For a level that is playing in a "poor" area I can set it that the player never finds a GT and finds fewer of the loot-tabe "rare" etc. without having to define a completely different loot-table for every occasion. I simply tag the area or the specific chest as "poor".
* If it fits better, you can set absolutelyNoGTFindable to an amount of time that counts down.
* You can influence variables for e.g. "hard hit" and "critical hit" depending on the level-difference between opponents to even out the playing field in a MOGA.
* You can generate filler enemies with not-so-random-randomness strengths and weaknesses based on the players performance in the game thus far.

Voila, randomness harnessed.

Based on what you as a GD want to do, you can introduce more or use other variables.
The important thing is that the mechanics in question are still "random", can be heavily influenced by game design and, in my experience, are far more easy to balance then having, for instance, 100 different loot tables and switch them around.

Cheers /Carsten

nicholas ralabate
profile image
Thanks for sharing that trick, seems like it would work wonders with procedural animation as well (when to show the "special" idle, etc.).


Alfa Etizado
profile image
Chance is good, it makes the future harder to plot. For me that's the whole point of it. It can be fun, just like creating elaborate plans and acting them can be fun. It makes planning more interesting when you can't see everything clearly.

I think luck is better used when it casts a dark cloud, not when it is there just to balance an action that's too good or to simulate how things should be in real life. Not that these two uses are bad, it's just that maybe luck isn't the best tool.

Adam Bishop
profile image
It's interesting that you mention basketball as being a non-random game; there's actually a ton of writing on the role of luck (or probabiities, if you prefer) in professional sports. All kinds of results in pro sports are subject to some degree of randomness and many statistics in sports show a tendency to regress toward the mean. This is a big part of what Moneyball and statistical analysis in sports in general are about.

Katy Smith
profile image
I think it all depends on the type of game you want to make. As you said, chess is a game where the design has no random luck element. However, if you are making a chess AI, do you really want it to be perfect all the time? Randomness in AI behavior allows the AI to be more "human", and give the player a chance to occasionally win.

There are many games that focus heavily on luck and are still fun. Randomness in games is on a spectrum. Would "Flux" be as fun the deck was in a set order every time? Some games do luck elements very well. For example, Settlers of Catan uses luck to distribute resources, but everything else is up to player strategy. Monopoly is an example of a game that I think uses luck and randomness poorly. In Monopoly, it all comes down to who can get the most properties in the first one or two times around the board. There is very little player strategy.

One thing that was mentioned in the article is whether it is a good or bad thing that a player can lose an 8 minute game for a decision that happened 3 minutes ago. My answer would be that it could go either way. Can the player trace the failure back to that condition, or is the failure presented as "what the heck just happened?" . If the player can recognize his/her mistake, randomness is fine. Games like Temple Run or Super Hexagon use randomness for level structure. More importantly, they have excellent user feedback. You know why you died / are going to die as soon as the bad decision is made.

Jess Groennebech
profile image
Pretty much any competitive game today relies on calculated risks from the beginning, often the first 2-3 minutes determines the winners with ~75%. It's the casual part that destroys monopoly, the idea that you get enough moneys to buy anything you want the first couple of rounds in order to make it family friendly.

It's the lack of challenge in getting moneys to pay for your properties that makes the difference, not the randomness.

Katy Smith
profile image
I don't see calculated risk as the same thing as randomness. I keep going back to chess, but you can take a calculated risk based on the opponent's board, and it doesn't have anything to do with a luck factor. It has to do with how you as the player are interpreting what your opponent will do.

In Monopoly, I do agree that it comes down to getting money to pay your rent and avoid landing on owned property in the first place, but that is determined by dice rolls, which are inherently random. Adding to the dice rolls, you have the chance and community chest cards which add extra randomness on top of that!

I would also argue that Monopoly is not a casual game, but that's another post :)

Altug Isigan
profile image
It may sound like the wrong thing to do, but I advise my students to avoid the use of chance-elements until into the later stages of design. The reason I do this is that I want them to focus on the core mechanics of the game before they start to introduce randomizers that have an impact on that core mechanic. It turns out that many games do not even need such randomization when the core mechanic is neat and fun enough.

Also I believe that randomization of movement mechanics is different from the randomization of resources. They change different things and need to be treated differently (in relation to the design as a whole, of course).

A way to give flexibility to the player without using randomizers and to create variety thereby is to allow to create certain combinations with already established mechanics or tokens. It gives players a feeling of control, yet due to the randomness element within the combination aspect, the game always develops differently, and this without the need to introduce artifical randomization.

The advise that I give to my students is that they use randomization methods only when the core design always results in a stalemate. This means that there needs to be a force that helps the system to behave so as to decide the game in favor of one of the players. If your core design doesn't achieve this by itself (overbalanced), then using randomization can be brought into the game. Also, when a game seems to be underbalanced, that is, if an early leader in the game always tends to win, then again, randomizers can be used to slow him down and maintain the illusion of winnability for the other players, because randomizers then serve as sort of a "catch-up" mechanism. Noone can really escape and the heat is on until the end.

This "no randomizers until later" rule may not be the best suggestion for those who are already advanced professionals, but I think it helps learners to become professional, because they learn to differentiate between various system types and their functions and they develop an understanding of what randomization is and is not good for.

For some genres it is of course not possible to simply say "no chance factor allowed" etc, especially for the casino type of games, but I think even then it is important not to confuse external randomization factors with those that are already inherent to the core mechanism. They're of a different natüre and serve different needs. You can tell an inherent chance factor from an external one in the following way: If you remove the chance factor and the core mechanism itself collapses, then it is inherent to the way it functions. If the core mechanics still is intact after you removed the chance factor, then it is external. Try it with Black Jack for example, and you see what I mean. Chance is also a means of distribution and circulation in these games, so it plays a much more central role than just adding a luck factor, although noone can deny that it *is* a luck factor too.

Please see Ian Schreiber and Brenda Brathwaite's Challenges for Game Designers. It has some neat chapters on Pure skill and püre chance games. Always helpful!

Good luck! :)

Anthony Botrel
profile image

Sorry for my bad english :).

I think that luck is bad only when you want the player to have control over the game and it becomes so dominant that it overrides that control. In a general context luck(randomness maybe would be a better name) is good when it forces you to adapt to fast changing situations.

In a competitive game, a bit of luck is good to create uncertainty, and most importantly, to create surprise. It makes you search for the best fit at a certain time, not the perfect solution. It must create uncertainty but not destroy a good strategy, only generates ripples that you need to adapt to.

If you want the player to experiment "jumping in the unknown" with your mountain climbing game, then luck is fine, because this is what the player will get. If it's "what you advertise on the box" it's ok, because the player is not looking for a win, but just following a feeling at a crossroad.
If you present your game as a climbing contest, where the player must find the shortest path or that there is a prize when it reaches the top then it becomes a problem because the player will want control.

It's all about the experience you want to provide.
I think that the question to ask here is : Does luck match your vision of the game ?