Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
October 21, 2014
arrowPress Releases
October 21, 2014
PR Newswire
View All

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

Opinion: The Problem Of Choice
Opinion: The Problem Of Choice Exclusive
April 10, 2009 | By James Portnow

April 10, 2009 | By James Portnow
More: Console/PC, Exclusive

[In this Gamasutra opinion piece with bonus video, designer and Divide By Zero founder James Portnow highlights why most game choices are really problems, and why knowing the difference is key.]

All the time in our media we hear people railing about what it will take for our medium to become an art form. I should know: I’m particularly guilty of shouting from the rooftops.

The problem is that we want to race ahead to becoming an art before we’ve done the groundwork that will allow us to develop a unified art form.

This leads to pieces of art emerging out of the milieu of ‘games’ but keeps us from becoming a coherent art.

Today I’d like to offer up one of those groundwork definitions that may help us move a little further down the road towards becoming an art. I’m open to criticism and correction. At the bottom of this piece you’ll find contact information – please send me your thoughts.

[EDITOR'S NOTE: Animator Daniel Floyd has also provided a v.entertaining video version of this opinion piece, available on YouTube - go check it out!]

The Problem of Choice

Game design is about designing decisions. It’s the first thing we learn when we embark down the path of becoming a game designer. The problem is that we are rarely taught that decisions come in two forms: problems and choices.

We often muddle these two ideas. We use them loosely. Even the most experienced game designers will haphazardly apply these terms, often referring to everything as a "choice."

As pretentious as it sounds, and as a much as I hate to say it, this linguistic confusion is holding back our medium. Why? Because at the core of our craft as game designers there are two very distinct and very different ideas that we’ve conflated and now try and address as one. Until we unravel that distinction we’ll still be lumbering in the dark.

So, without further ado, let’s define the types of decisions found in games.


Any decision that has a definitive best answer is a problem and not a choice. Anything with a ‘solution’ is a problem and not a choice. Anything that is a calculation is a problem and not a choice.

The easiest place to identify problems is in the math problems that many games are riddled with. For example: in World of Warcraft deciding what gear to wear is a problem and not a choice. Why? Because for any given objective in World of Warcraft there is a set of gear that will best facilitate accomplishing that objective.

Or, more simply put: if you want to kill a monster in WoW there is an optimal set of gear for doing so.

Gear in World of Warcraft is a puzzle. It is an interesting and complicated puzzle, one whose solution may not be immediately obvious, but it is a puzzle.
The dexterity challenges in first person shooters are also problems.

There is very little “free will” involved in these types of problems, for the most part you know what you are trying to execute (moving your reticle over your opponent without letting them move theirs over you) and the challenge is in the execution.

Anytime where you present the player with a clearly defined goal and the “fun” or challenge is in the execution of that goal you have presented them with a problem.


Choice appears when you are asked to decide between two things of equivalent or incomparable value.

Choosing between an apple and an orange is a choice. Choosing between friends or lovers, choosing between roses and lilies, choosing between anything that’s six to one half dozen -- these are choices.

Choice appears in games less frequently then problems do, but that does not make it less important to the game designer. The easiest way to identify choice in games is to find decision points that aren’t problems (yes, it’s as simple as that). These usually come when the player is offered multiple, exclusive options that can’t be weighed against each other.

Choices tend to be much harder to design than problems because they don’t have a clear right answer. They also tend to be much harder to fit into games because their divergent nature tends to broaden the scope of games.

Problems direct the player towards a goal; choices let the player choose their goal.

Are problems a problem?

Are problems bad? No, absolutely not. In fact, most games are, and should be, built around problem-solving. To date few games have been made without problems (Facade and a handful of indie games being the only things that come to mind), but many great games have been made without choice.

Problems are at the heart of what we consider gaming today, and there’s nothing wrong with them. It’s just important that we distinguish them from choices.

So why is this such a big deal?

Because without making this distinction we tend to reduce problems to choices.

How many times in a game have you been presented with something that should be a choice but instead was simply a problem with a clear right answer. Bioshock and Mass Effect are two recent examples of this. In both of these games what should have been ambiguous moral choices were reduced to elements of larger game problems.

How? By making them commensurate with other gameplay mechanics.

What does that mean?

Let’s return to our apples and oranges example. Deciding between an apple and an orange is a choice, but if you become informed that an apple is worth five dollars and an orange is worth ten dollars, and you know the goal is to get the most dollars, deciding between an apple and an orange suddenly becomes a problem rather than a choice.

This is the issue we often run into with faux choices in games. Many games attach rewards that effect gameplay problems to choices, thus reducing the choice to a simple equation.

For example: how many times have you been offered a choice to be nice to an old man or to ignore him and had the reward for being nice to him be X experience (or ammo or money) and the reward for ignoring him be Y experience (where Y is less then X, and often zero)?

This decision is a problem and it’s pretty automatic, you simply ask yourself “Do I want that experience, or would I rather spend my time doing other things?" Helping the old man never even enters into the equation.

For many of us in our mental state, we place the tangible benefit of solving problems over the more nebulous (and work-intensive) satisfaction of making a choice. This does not mean we get more pleasure out of solving problems or that problem solving is a more “fun” activity, it simply means that reducing choices to problems is one of the easiest ways to approach them, and thus what a player will do if given the option.

So what was this hoopla about art?

In order to become an art we have to be able to address “the human experience”. There are a whole range of human experiences that are better expressed through choices then through problems. If we want to move this medium forward we have to be able to distinguish between the two and choose the one that is appropriate for the experience we are trying to craft.


We face today a problem – a problem of choice.
As an industry we are trying to learn how to make our games more compelling. As an art we are trying to learn how to make our games more expressive. These challenges will not be overcome without utilizing all the tools we have before us.

The first step to doing so is to understand exactly what our tools are. Choices and problems are both vital to crafting compelling gameplay. Both are vital to expressing the human experience. Not every game needs to include both, but better understanding the distinction is essential to creating the best products we can.

[James Portnow can be reached at]

Related Jobs

Zynga — Chicago, Illinois, United States

Senior Software Engineer (Front End)
Cloud Imperium Games
Cloud Imperium Games — Santa Monica, California, United States

Marketing Director
Trion Worlds
Trion Worlds — Redwood City, California, United States

Senior Gameplay Engineer
Trion Worlds
Trion Worlds — Redwood City, California, United States

Senior Graphics Engineer


Robert Farr
profile image
This is definately something thats been in need of being said for some time now. Another way in which I see this problem rearing its ugly head is in quests in level based MMOs, its not a question of choosing which quest sounds like it might be fun or interesting, is usually a problem of choosing which quest is most appropriate to your characters skills/level or what quest reward is best for the character. All of a sudden, your not doing the quest because it is inherently interesting in of its self, you're doing it because you feel you need to and it stops being a choice.

Glenn Storm
profile image
I can't tell you how much I love seeing an important but complex and amorphous concept boiled down so gracefully and presented so well. Excellent start to the discussion, James. Well said. I might add that some other popular terms that have confused the issue include, "agency" and "customization". Re-framing those descriptions in light of the, "Does it have a solution?", benchmark should help clarify what we're doing. Thanks for pointing out that tool we can all use.

It strikes me that recently the designer's decision to include choice has been made to deliver a venue for player expression. Whether you're talking about a social venue like Sony Home or an RPG like Fallout3 or even character selection in Street Fighter 4; the character avatar choice, item collection, combat style, custom advancement of skills; these are all examples of player expression that arose from choice and not problems, by the definition you laid out here. (And please understand, I realize some will see player choice in SF4, for example, as a problem. Let's just skip that semantic argument for the moment.)

I think what I'm getting at is the idea that choice between apples and oranges may not have a solution, it may be, in strict terms, an arbitrary decision, but to the player who decides, "This time: orange", it is not completely arbitrary. Its value is weighed by the player on some level and the "solution" reflects the player's expression. While problems may have a solution that's meaningful to the mechanics of the game, choices have solutions that are meaningful to the player's expression.

Another point that strikes me is that you've mentioned just a few examples, yet they are all pretty cut-and-dry, either problems with solutions or choices with no solution. And this leads me to wonder about systems that might offer more fluidity between the two decision types. Can a decision be both? Can it change from a problem to a choice and back to a problem? If a casual choice later affects another casual choice and changes it into a problem, does the entire system become one big problem? These are interesting new ideas to explore for me. Thank again, James.

One final thought about this opinion: If one subscribes to the Butterfly Effect, the theory that all systems are interconnected, that the flap of a butterfly's wings in south Asia can eventually lead to a tornado in North America; this could lead one to conclude that (in RealLife™) there are only consequential problems and no casual choices. So should all casual choice in games find some way to effect the system and be, in fact, a problem? Fun stuff to think about.

Glenn Storm
profile image
Sorry, one more thought. And this might be analogous to, "If a tree falls in the woods...", with no definitive answer, but if an apparent casual choice is actually a problem, where the game consequences of the decision or the connections to the game mechanics are actually hidden from the player at the time of the decision, is it still a problem?

The helping an old man scenario you mention is a common example of this situation, where the old man later turns out to be a critical mentor to your development in the game or your long lost father or something. At first glance (or during the first play experience), this was a side-tracking choice, but if it shifts the game system in some way, it's technically a problem with no obvious solution. The player who expresses themselves as a helpful character and helps the old man isn't deciding on a solution or solving a problem because there was no context that there was a definitive solution to come to. This was purely a choice by the player's perspective at that time. (standard video game narrative tropes aside)

I know this highlights the importance of communicating clearly the mechanics, available actions and their consequences of the game, but it also seems to suggest that as we discuss, design and implement both choice and problem in our games, we are forced to face another subtle and complex aspect of our medium: Management of audience expectation in an interactive system. But I think that's another topic altogether, sorry.

Seth Rediker
profile image
Excellent write-up. This is a matter that often comes up when I am discussing game design with people.

I would go further to say that another problem with "choices" in games are they are just too black and white and you often can guess the outcome before you even make the decision. For instance, while there are great games, look at the Knights of the Old Republic series. Most decisions in those games are "You see a baby that someone lost, do you (A) return the baby to it's mother or (B) kick the baby." With that it is never really a choice either since the decision is so transparent.

The only gave that has really hit me with hard choices was The Witcher. All decisions were pure shades of gray. Often times it would take me a great deal of time to pick which option to go with because the decisions weren't transparent and all and each option had merits to it.

I think the best choices are the ones that you can not decern the solution to from the outside. Choices should be more about "What would I/my character do?" not "What would make me the most powerful."

Dave Mark
profile image
Oddly, this mentality needs to actually apply to AI NPCs as well as the player. See my writeup in my Gamasutra Expert blog.

profile image
The WoW gear example is true to a certain extent. Each specific character is going to be seeking one or two main stats to raise primarily in any given gearing up problem. Some classes have different specs that they need totally different stats to play as so now the gear problem also becomes a choice because to the player both class specs are equally important at what they do, but each class also has specific max stats that they're looking to increase.

Ben Snyder
profile image
This is a great article. I think a lot of players get into a mindset where choices do not matter only problems. If apples and oranges are both worth 5 gp then the decision is unimportant. It seems like the 'art' lies in being able to make choices important to the game experience but not turn them into problems. Choices can add to the richness of a game experience, particularly MMO's were you are able to make an expression to other players through your choices.

On the other hand, I do like the butterfly effect where every choice has some sort of consequence eventually even if it is very subtle. I think the way to achieve this is to have a rich world and a balance of out comes.

Ryan Hibbeler
profile image
@ Glenn,

I think James' World of Warcraft gear example actually can fit your request for an example of a system which includes both problems and choices. It wasn't explored in the article, perhaps because most of the choices take place outside of the game mechanics, but this is an excellent instance of choices leading to problems. The choices here aren't which specific gear is most optimized for a certain purpose, but what purpose the player actually pursues.

The first and most basic choice the player makes in this regard is which class to play, which will lead directly to a problem of which gear is best for that class, but there are many other choices the player can make about his play style: Which role do I want to play? Do I want to excel at that role, or do I want to be more versatile? Do I want to play primarily solo or in groups? PvE or PvP?

Of course, the player also chooses their own goals in the game. Am I trying to be the best tank? Beat Naxx? Top the PvP leaderboards? Complete all the quests? Get a certain achievement or title? Or just make it to 80? Once the player makes these choices that have no programmatic solution, then problems arise concerning how best to meet that goal. These problems often bring up more choices, like "How much time do I want to spend filling this armor slot, or raising this stat?"

Since most of these choices happen outside of the scope of the game mechanics, one could argue that many of the same choices could be made in any game. However, World of Warcraft is an exceptional case simply due to the breadth of content that allows these choices to carry more weight. Contrast the differences in player experience from the choice of class in WoW with that of a game like Too Human, where the game plays quite similarly no matter which class you pick.

Enrique Dryere
profile image
First let me say that this is an intriguing subject and well-crafted article. I just have one concern:

I know it may be arguing semantics, but isn't the difference here between problems and choices more like the difference between important and unimportant decisions? Or perhaps even "interested" versus "uninterested" decision making.

Using your example of apples and oranges, all other factors remaining equal, my preference should determine which fruit I choose. If I want an apple, I'll have an apple. If I want an orange, then that becomes my goal. The choice between the two only becomes "fifty-fifty" when I didn't really want either, or wanted both exactly equally.

My point here is that the twitch-based aiming of FPSs or the stat-based equipment choices of RPGs are only really "problems" as you describe them, because players are motivated to "win." If they didn't care to win, they would merely be choices, as you describe by this article.

While I think I do understand the point of the article as a whole, and this may just be nitpicking, I think it's important to ask: Can a game really offer a player choices whose outcome doesn't matter to the game, but matters to the player?

My conclusion to this is that ultimately a game of "choices" provides multiple final results and goals (or none at all), where a game of "problems" really only contains a single end result.

As you say, there has been a lot of "hoopla" about games as art over the past few years, but isn't the purpose of art to elicit emotion? I can't think of any "art" form that has drawn more intense feelings from me than games. Whether it be frustration, pride, satisfaction, happiness, disappointment, or any other color of the rainbow, seeing out the results of your decisions and actions within a game will always produce intense emotions, but only so long as you care about the outcome.

Tom Newman
profile image
Great topic for sure.

From a players perspective this can be very confusing, as often times you do not fully know weather a decision is a choice or a problem until the game progresses past the point of the decision. The difference as laid out in this article is spot-on, and something every designer should be concious of.

Jeff Beaudoin
profile image
I think the most important point is the fact that choices with discernible, weighted, gameplay rewards are problems. As in the Bioshock example.

If you really want a choice in a game to result in some reward for the player, then the two options need have the same reward, have rewards that are of equivalent power, or have rewards that are undetectable to the player.

Bart Stewart
profile image
Excellent article -- I love seeing a distinction given a good airing-out to check whether it's meaningful. In this case, I'm convinced.

Not that I needed much convincing on this one, but still. :)

One question I have is how players may respond to game designers deliberately adding more choices to their problem-rich games.

Ask a random selection of gamers (or even folks off the street) to define "game," and I'd happily bet that more than nine times out of ten you'll hear something that translates to "rules-based competitive action."

And rules for achieving competitive advantage imply gameplay that's designed as problems, not choices.

So how do you persuade existing gamers that choices, as that term is distinguished in this article, have anything to do with playing a "game?" If play isn't entirely about taking actions to follow a set of rules that confer some competitive advantage, how can it be presented to gamers that will lead them to conclude that having choices made the game more fun rather than just interfering with problem-solving?

Is such an effort to find a satisfing way of presenting "choice" content to today's gamers a necessary effort? If it is, does it have any reasonable chance of success, and what can designers do to improve the odds of such success?

Reid Kimball
profile image
Bart, like you, I agree with James' distinction between problem and choice.

Related to the issue you bring up, how players will react to more choice in problem focused games. I think designers will need to decide, "Is my game a problem focused game or a choice focused game?"

I think adding more choices to a problem focused game will cloud the essence of the game and the choices may end up not being meaningful or lacking depth, because the goal is to solve the problem, everything else is "fat" that needs to be trimmed.

At the end of the day, the game won't give a damn if you acted like a gentlemen or a jerk, as long as you defeat the antagonist that oppresses them. I don't know if that makes sense, but I think the two distinctions lend themselves to two very different kinds of games.

Steven Conway
profile image
Superb article James.

Kyle Brink
profile image
As others here have suggested, one must take into account player goals when determining whether a given set of options are truly, well, optional.

The essence of the player experience in a game is fantasy. Each player decides for himself or herself what sort of fantasy it is. Do I want to be powerful? Loved? Have the best stats? Find all there is to know? Feel unique?

What seems a "problem," in the article's parlance, may in fact be a true "choice" -- and vice versa. I chose my pet in WoW because I thought that a mangy, nasty vulture was appropriate for an orc. It was not numerically optimal. But I achieved the play experience I desired by choosing it. By being visually interesting and consistent with "orcishness," I made friends with like-minded players, and they in turn helped me and were good company. Similarly, in Fallout 3, I often chose sub-optimal gear because of how it looked or how satisfyingly it played out. The sound of a gun and how it visually impacted the world was as important to me as how large its damage value was or how easy it was to find its type of ammunition.

When desiging any interactive entertainment system, it is essential to consider the diversity of possible player enjoyment vectors when weighing the equity of outcomes (or, indeed, what outcomes constitute "rewards" at all).

Entertainment is a subjective measure, varying from person to person. Know the mindset of the people for whom we are crafting a game experience, and then measure how truly optional any given decision point is against the intrinsic value of the possible outcomes in the mind of those intended players.

I find the distinction between "choice" and "problem" in the article to be largely artificial because it presupposes a universal set of values by which a set of outcomes can be measured. In reality, there are as many values to a given set of outcomes as there are players who consider it.

That being said, there is definitely value in considering the worth of different outcomes in the minds of all the possible players when designing these decision points, to ensure that we provide problems and choices where we intend to rather than by accident.

Zaid Crouch
profile image
Great article, James, and it's been great to read through the discussion it's provoked too.

The difference between choices and problems is one that has gone undistinguished for some time, and this is a point that I think needs to be considered when implementing choices into a game. I would assert that games have trained gamers for a long time that choices are problems, and to relate to them as such.

Certain genres do this more clearly than others, but as someone who grew up with Japanese RPGs, I came to relate to any dialog choice, no matter how innocuous, as something that, if answered incorrectly, could prevent me from obtaining some bonus 20+ hours later into the game. Clearly this is problematic. So long as the player views any given choice as a problem, they will treat it as such and try and solve it.

I think another aspect of this is also in how choices are and have been offered to the player. One thing I would love to see more of, is instances where a game recognises and reacts to how a player plays through the game and the subconscious choices they make as opposed to the ones where they're hit over the head with the "it's time to make a choice" stick. For example, some players will choose to explore every nook and cranny of an area, whilst others will try to choose the shortest path to their objective. In certain story situations, this could be construed as a meaningful choice, with a commensurate set of consequences (again, the nature of these needs to be carefully considered.). This is the kind of gentleman vs. jerk choice that Reid mentioned above.

profile image
How's this for a horrible problem in a game: Zodiac spear in FFXII. You had to get to a certain point in the game without opening a specific set of chests leading up to that point. If you opened any of the chests that you weren't supposed to then the item wouldn't be where it was supposed to be, instead the item would be in a regenerative chest with like a less than 1% chance of appearing (even with an item to increase that percentage) AND later in the game than you would have attained it had you not opened any chests. The game penalizes people that do search every nook and cranny of the game by drastically reducing their chances of ever getting the item because of their dedication to meticulous searching. Part of the RPG genre is exploring and finding treasure so to be penalized for doing just that was just horrible.

@ Kyle

I did the same thing in Fallout 3. I played the entire game melee based even though I knew I could be using guns and explosives to make some things easier.

James Portnow
profile image
Fantastic! This discussion has helped me refine my own thinking.

Jason VandenBerghe
profile image
I do love you, James, but I disconcur.

The choice of weapons in Gears of War is largely a matter of personal preference as related to emotional satisfaction and personal play style. It is only a "puzzle" as it relates to your ability to predict which weapons will work best for you before you have another chance to re-select your gear... and you don't have much chance to do that.

So, that's a choice. That said, it should be acknowledged that it has maybe 20-30% of "puzzle" in it.

Your argument would actually work much better if you sliced it along genre lines. Say, between adventure and action mechanics. "Choices" with adventure mechanics tend to be puzzles, while "choices" in action mechanics tend to be actual choices.

p.s. Further, it is a fallacy to say that the confusion around semantics in design is "holding the industry back". That's grandstanding, plain and simple. Confusion around these concepts in an _individual game designer_ can hold _that designer_ back, but the industry itself has produced a metric ton of work that demonstrates a clear comprehension of the distinction between these terms on the part of their authors.

Still, an excellent discussion-starter, as Glenn said. :)

Nestor Forjan
profile image
One problem I have with the issue as presented is that a classic solution to the moral choice turned problem of rewarding choice with in-game rewards has been to reward the "evil" option more, since players do often show a natural tendency to the "good" option. An alternative also usually found is to reward the "evil" option with hard currency and the "good" option with indirect benefits like NPC recognition of their moral standing.

In fact, pure choice as described here, apples vs. oranges, is often disruptive on narrative-driven or otherwise goal-oriented games. It's great for The Sims, where you set your own goals, but in a linear environment, the open choice takes away from players by preventing them from experience the content that remains locked by a choice that had no good answer. I bet I'm not the only one that's heard Fallout 3 players whining about not being able to obtain all achievements because some are locked to different Karma standings in the game.

In that sense, the difference between problem solving and choice making is in fact very useful... because choice should not be obvious in an otherwise problem-based game. Whenever this point about non-linearity is brought up I always point people to the amazing branching path story in Westwood's Blade Runner, in which players made choices without being aware of it, which completely removed the feeling of loss for not having seen the flipside of their elections.

William Holt
profile image

Truly, the point that any given game is not designed for a global audience is a valid one. However, I find it a tad analguous to your well thought out post that you find the distinction between "choice" and "problem" as it is laid out in the article to be largely artificial. The distinction between choice and problem is not one that exists universally, in a game or in real life, and I don't think that making a statement of such a grandoise nature was ever James' intention. The article simply points out that, within the value set for a games' target audience, the choices have to be significant to the audience.

But, you're right - even within a target audiences value set there is some variation. To use a simple example, when trying to get from point A to point B within a time limit, with a scaled reward for a faster or slower time, the player might take the fast path and receive a nifty looking hat as a reward for time taken, but nothing from the environment he passed through to get to point A. Alternatively, the player may explore path AB a little bit, and find an item worth 60 gold, finish the race, and receive a slightly less nifty looking hat for time taken. Lastly, the player may explore the environment and completely ignore the hat race, and find a hidden NPC with a quest whose reward includes an item worth a bit more time investment, but at the cost of any hat whatsoever. Admittedly, this may seem like a choice to someone who values both the hats appearance and the power the quest reward brings, but to one who values one over the other, it is a simple problem of choosing which they value more.

This brings me to the thought that, when designing a reward system, it's important to include rewards whose sole intrinsic value is not their power. In other words, stroking the players' ego as they play is just as important as equipping them for the fight ahead. Mechanics allowing the player to get married (Fable), or allowing the player to choose their own arbitrary look and customize the power of the look they've chosen (Morrowind or Oblivion), or having the players' moral-based choices affect how the game world perceives them (KotOR, Jade Empire, etc), or even doing something as simple as providing a way to acquire 'flavoured' items at a cost of additional time investment (Fallout 3's dart gun or railway spike gun) give the player the ability to not only play the game as the designer intended, but also to make a portion of the game their own. The players' ability to identify with the game via branching decisions (This is MY save game because I made all the decisions (problems OR choices) important to ME) or via the story line (This game is so awesome because, as I played through the game through the eyes of the lead character, the things that happened to him/her inspired me, in some way, to better myself) is an indisposable aspect of any games' design, and is, I feel, what differentiates the instant classic from the one-play wonder.


While you are right in saying that kicking the baby is often rewarded more in the scope of the single quest, it's important to note that the rewards (or lack thereof) intrinsic in any given gameplay path may be compensated for in other aspects of the game where the player is prompted to make a decision whose choices are similar to the ones they've been making all along. To simplify that thought, let's look at KotOR. Taking the light path in dialogue choices may gain you the admiration of the people and a lessened monetary reward, but the rewards intrinsic in taking the light path involve gaining powers that you would not be able to achieve (as easily) if you were to take the dark path.

Also, I find it strange that you consider Fallout 3 to be linear. The player does have the ability to play through the game again, and make different choices, and see how those choices impact the environment differently that their first playthrough. I've gone through the game at least three times, now - once as my traditional goody two-shoes, then as an evil brawler, and now as a morally ambiguous heavy weapons expert. I've yet to get to playing through as a pacifist. Suffice to say, however, that if players are complaining that they need to play the game more than once to get all the achievements, it's likely that those specific players might not be precisely who the developers had in mind when they were designing the game. Also, the players' decision to go after one achievement over another, and refusing to attain the other one because "I already kicked the baby, why should I save it, now, to get all the achievements" is indicative that the player might enjoy other aspects of the game more than the morals-based ones, such as the customizability. "Replay Value" means more than just "You can choose this or this, but not both in the same playthrough". It can also mean "If you like kicking babies, you can make a mod that adds 50 hidden babies into the game, and if you kick them all you get a reward." I suppose, though, that it all basically boils down to creating a game for your main audience, and including aspects that derivatives of your main audience might enjoy. To put it into terms of Magic: The Gathering, make a game for the greens, but also include aspects for the whites and the blues.

In any case, thank you, James, for a very thought-provoking article.

Ruthaniel van-den-Naar
profile image
James is right. If you want choices, need nonlinear game (5 independent choices in Mass Effect is little) =>lots of work for gamedesigners, too lots..

Resolution is imho in most robust AI system (include knowledge system). Something as GM in PnP. Manualy write thousands variants of diaglogs - its imposible, we need some automatical solution.

AI is "Lapis philosophorum" (Phislophers Stone) for games and not only for they.

Nestor Forjan
profile image

To clarify what I mean about the problem of losing an option when a choice is presented, I only chose the achievement example in Fallout because it is a very simple structure: you're good, you get one; you're bad, you get another one; neutral, you get the third. Regardless of the non-linearity and replayability of Fallout, my point here is that your choice creates a "shadow" on top of some rewards that you know exist but can't access as a result of your choice. This creates uncertainty in players and some frustration. Which is not always bad, but it's certainly there. I've heard players warning each other to save before that award is given to artificially change their karma meter and get all achievements in a single playthrough. That's an example of a choice that is making them think about the mechanics and handle stats and the metagame, which is pretty much the opposit of what a system to gauge morality should be doing.

But the uncertainty and frustration come from the simple, short tem way in which the choice is presented: do this, get that. Be good, get this achievement. What I liked about Blade Runner's idea of hiding the choice, making it transparent, is that maybe going to your apartment instead of rushing somewhere else triggered a different set of events. There was a layer of "reading" the player through his actions that was deeper than just asking him if he wants to be good or bad, choose apples or oranges. The choice developed organically and players weren't tempted to exploit the system to get different results... because they didn't know how.

The Bioware crowd are great at what they do, but there's promise in the basic frame of their RPGs that is unrealized. There are so many metrics that one can set up in a narrative game to try to analyze the player through his behaviour in-game rather than by asking questions. Does he run away from battle often? Does he prioritize some quests over the others? Does he talk to some characters more often? I'm not saying all choices should be transparent to players, but doing things this way is a resource that isn't used often enough and it creates a real sense of wonder and consistency in players when it's done well without requiring more writing or planning than simply presenting the same choice openly.

William Holt
profile image

From the way you describe it, Blade Runner's hiding of the choice would really make for an interesting play through - it seems it would give the player an illusion of a living world, and make them approach the game completely differently. In the case of Fallout 3, the linear progression of the achievements almost invites the player to play in a 'rules lawyer' manner, using the karma system in a manner contrary to its (arguably) intended use.

I wonder what applications Blade Runner's opaque rewards system could have in a session-based, online game such as CSS, TF2 or L4D? Plenty to think about, in any case. Thanks for clarifying.

Michael Rivera
profile image
While I agree with the basic premise of the article, I think it's dangerous to say that a choice can't affect gameplay or have different rewards. True, a choice between a $5 apple and a $10 orange has only one right answer in the player's mind, but that doesn't mean the developer should make the apple $10 too. Gameplay-wise, a $10 apple $10 orange choice basically tells the player "your decision doesn't matter". A much more emotionally engaging choice would be between say, $10 or 50 experience points. If both rewards are different, but roughly equal in value it tells the player that their choice does matter even if there is no right answer.

I guess this is also why I have a problem with how the article labels the Bioshock example as a problem rather than a choice. Yes, the player's decision affects gameplay, but the rewards are roughly the same. Either they get extra Adam or they will get an exclusive plasmid, both of which are roughly the same value.

@Will: Your "hat" example and discussion about tailoring rewards to your audience is pretty interesting. Certainly, what might seem like a choice to one person may be viewed at a problem to another. It all really resonated with me because that idea is really important to a game I'm working on right now =)

Scott Gardner
profile image
Good write-up, but I'm not sure I understand the distinction you're drawing between a problem and a choice. Is a problem a decision with a specific "optimal" solution? Or is any goal-directed decision a problem? For example, is a player's choice of move in a game of chess a "problem" or a "choice"? It is certainly a goal-directed decision, but often doesn't have a specific optimal solution.

I also don't understand how this distinction fits into the larger issues you raise. For example, you argue that the decision whether to harvest or rescue the little girls of Bioshock is a "problem," rather than a "choice." What makes it into a problem? As I recall, harvesting the little sisters gives you (1) a lot of Adam and (2) a "bad" ending, whereas rescuing them gives you (1) less Adam, (2) a powerful plasmid, and (3) a "good" ending.

So what's the problem? Is it that one choice gives obviously better rewards from a purely gameplay standpoint? If so, that's just a gameplay balance issue and could be corrected by evening out the rewards.

Or is the problem that the player is required to weigh goal-directed rewards (Adam and plasmids) against narrative outcomes? If so, I don't understand why that isn't a "choice" under your definition, given that extra Adam is incommensurate with a "good" ending. I.e., extra Adam is an apple, and a desirable story ending is an orange.

Or is the problem that, because the choice involves both gameplay and moral considerations, that the player will always make a choice based on gameplay rather than morality? This seems like it's what you're really getting at, but not everyone prioritizes gameplay goals that much. I knew many people who rescued the little sisters even though they believed they'd get fewer gameplay rewards for that. I also know many Bioware RPG players who make quest decisions based on roleplay rather than rewards.

Ultimately, it asks like you're asking developers to divorce gameplay rewards entirely from narrative decisions, and I'm not sure why that's desirable.

Anyway, these are really interesting and thought-provoking issues. Thanks for the article.

Sean Parton
profile image
Fantastic article. Props to you James; this actually has appeared just as I was thinking about the designs of a personal game I've been kicking around ideas for lately, and has been delicious food for thought.

On the matter of "Undetectable Rewards":

Having rewards undetectable to the player will only be concidered "choices" up until the player realises he could have gamed the system for a different result. At that point, it's not a choice, but a masked problem. The main issue with then doing this is that players who find this out will attempt to metagame _instead_ of trying to make meaningful decisions with choices, which ruins the point of having choices to begin with (having emotional connection that isn't based on solving problems).

An example of this is Effort Value training (aka "EV training") in Pokemon. Each time you defeat a monster, your Pokemon gets EV's based on what kind of monster you defeated. Any player who knows this will no longer be looking at a _choice_ about where to train even for XP (because you always get EVs when you defeat a Pokemon, and there is a limit to how many a Pokemon can have). Instead, players will look at EV training as a _problem_ to overcome (where should I go to maximise EV's for the stat I want in this Pokemon?).

However, one intersting point to make about the above scenario I outlined is that if the user is in the dark, they see it only as a choice. Similarly, if the user doesn't care about this particular part of the game (if they didn't care about this specific stat-gaining technique), then it is also a choice, because the mechanical benefits are not really cared about.

This implies that a choice is not actually about something that has equal benefit to other options (which remember, can include no benefit), but that the user finds the benefit of all related choices to be equal (again, can include no benefit).

William Holt
profile image

Too often, as the nature of the beast involves quite a bit of mathematics, it's understandable that we might fall into a pattern of only considering the numbers of any given game play problem, and ignore the human element. I think that the problem you cited, the killing or saving of the little sisters, is a fantastic example. The issue with that example is not 'what is the problem', but instead 'what does the player value'. In this case, the answer, evidently, was 'Power' and 'Knowledge'. The power aspect is the additional Adam, where the Knowledge aspect is 1 - being able to see the special cut scenes and 2 - getting access to that special plasmid. Both of these two attractive aspects revolve around making the player feel special in some way.

Once it is decided which valued aspects to incorporate into the reward, it's a simple matter of making both rewards worthwhile to their respective audiences, the question 'what is the problem' solves itself. The problem becomes that of the player. Namely, "What do I value more?" As the choice (from the devs view) is presented to the player, the player is confronted with the problem (from the players' view) of solving it. Since the choice was designed with two different value systems in mind, the answer is immediately evident to the player. That's what makes this scenario a 'problem' rather than a 'choice'. It's the player who decides whether or not to buy the game, and it's the player who'd better be kept in mind when you're making the game. ;)

The article's intention was not to persuade developers to divorce game play and storyline, but the exact opposite - to consider what the people playing the game might value, and tailor the choices-come-problems specifically to their audiences. Some people might value the hat, some people might value the quest, and some people might value the hidden cut scene. It's not a matter of mathematics at all. Simply a matter of undeclared, unsigned, horrendously flawed at times human values.

William Holt
profile image
Good lord. Apologies for the awful grammar.


Once it is decided which valued aspects to incorporate into the reward, it's a simple matter of making both rewards worthwhile to their respective audiences. Your question, 'what is the problem,' solves itself, relying on the player to turn the 'choice' into a 'problem, from their point of view.


gren ideer
profile image
I disagree with the statement that pure choices are more difficult to create than problems. In fact, in many cases the opposite is true. FunCom's new MMO made the decision that all the armor and clothes in the game would offer no stat bonuses. This changed all clothing decisions from a problem into a choice. It is much easier to implement because there are no stats to balance when making clothes - just new textures to create, and the desired effect is that players will customize their avatar's look based on preference instead of stats.

In games with many weapons, the weapons should usually be a choice, ie. up to personal preference. This requires heavy weapon balancing but the point is to let the player pick the method and weapon that he likes best. Of course, having enemies that are susceptible to certain weapons poses and problem and can make things more interesting.

The fundamental issue with choices in games is that they can make the decision irrelevant if there is no gameplay significance, but as far as story concerns maybe it can be an interesting path to take.

Jason Bakker
profile image
I'm getting in on this pretty late, but I think a lot of the later comments miss the point of the making the distinction between problems and choices.


'Gameplay-wise, a $10 apple $10 orange choice basically tells the player "your decision doesn't matter".'

Yes, but story-wise, that decision - that choice - could matter.

For example:

Your character in the game has a friend, and throughout the course of the game you (the player) begin to like this friend as well. That friend, later in the game, commits murder, and someone else is wrongly accused.

You now have a choice - you know there are no gameplay changes either way (you've already looked it up on the internets just in case), but the choice still matters; are you going to do the morally right thing and turn in your friend, or protect him even if he may not deserve it?

In this case, is it really necessary to dangle a carrot in front of any one of these options in order to make the player care? If it is necessary - if you need to tempt them with rewards into caring about that decision - then you haven't created a truly engaging emotional experience in the first place.

Aaron Matthew
profile image
I appreciate this thought provoking post, but I completely disagree with the separation of the two concepts, and it sparked me to write further on the matter than I could fit in the comments here: