In this final post on emergence in games, I want to talk about how emergence impacts the business model.

Discovery

Emergent games are inherently social.  The more emergent a title is, the higher the volume of novel content generated by the players:  there are more surprises, more varied progress, and more opportunities for self expression.  As a result, there’s not only more to share, but more that players want to share.

Just look at The Sims:  that product has more shared content than almost any other on the planet despite being single player and lacking built in viral hooks.  Or Minecraft, a game with no marketing budget and a ridiculous amount of commerce and distribution friction.  Giving players interesting things to share — and by interesting I mean things that appeal to non-players — dramatically increases the native viral coefficient for the product and reduces the cost of discovery.

Player Life Expectancy

In previous posts I’ve noted that emergence increases the possibility space for players, which in turn increases their engagement (by reducing the pattern matching problem and increasing the likelihood of finding something interesting to play).  Greater engagement increases player life expectancy, and players who play longer are more likely to monetize.  The data Kongregate has data been sharing the last few years is particularly compelling on this point.

Development Expense

Launching a live service-based product puts you on the content creation treadmill:  you’re in a race with your players to author content faster than they can consume it.  That’s a very expensive proposition for most games if they want to maintain player engagement (it’s not enough to re-skin content or re-purpose mechanics since players will have already pattern matched the play dynamics).

In an emergent game every tiny bit of new content combines with all the old content, refreshing it and exploding the possibility space all over again.  Balance issues aside, both initial and ongoing content creation costs are small and manageable.

The Business of Play

So emergence lets us use the underlying mechanics of the game itself to drive discovery, increase player life expectancy, and reduce development expense.  That’s a much better approach to long term sustainability than optimizing LTV > CPA and trying to re-capture players in another game when they churn out.  In fact, emergence allows us to re-think core assumptions about how we make games.

Our industry is in the business of selling experiences through play.  Traditional game companies make products with a finite life, and they offer players more play by creating new products.  This approach comes with some inherent problems:

• A short product life cycle means the threshold for success is high.  You need a large audience to return enough revenue to cover expenses, which in turn drives up customer acquisition costs.
• It’s very difficult to find this level of repeat success, so companies use portfolio strategies to offset risk.  This raises the success mark even higher, since now a few hits have to cover a bunch of duds.
• Milking the hits you do find can help, but sequels only take you so far (see Eidos and Tomb Raider as one of many examples).  These product lines get tired over time, due in part to the small possibility space that offers little in the way of new play.

This is what we call a hit-driven business, because if you fail to keep making hits, you die.

With emergence, you offer players more play in the same product.  There is more play in one emergent game than in 100 traditional games.  I’m being arbitrary with that ratio, but my point is that the possibility space is so large players never run out of interesting things to do.   As a result:

• These games last forever.  A long product life cycle requires a smaller audience to generate the same return, thereby reducing customer acquisition costs and making less competitive niche genres viable.
• With a long product life and a large possibility space, you have a lot of flexibility to adapt your product.  You take multiple shots on goal with one game, not many.  That’s less expensive than building new products until one hits.
• Games of this type stay fresh for a very long time — there’s less risk of player attrition due to product exhaustion, and there’s no need for sequels.

You have to re-prioritize many of the metrics we use for evaluating success if you’re going to build products this way, but if you’re willing to take the long view emergence can get you out of the high risk hits business and into a safer, more sustainable model for making games.

The entire series on emergence can be found here:

Games Have An Attention Problem
Attention and Emergence
Emergence and the Build-Try-Fail Loop
Key Challenges Creating Emergent Games
How Emergence Changes The Business Model

As much as I’ve been touting greater use of emergence in games, it’s a non-trivial problem in terms of execution.  Developers face a few challenges if they go down this path.

Balance

The massive possibility space that’s so beneficial to player engagement also happens to be impossible to balance.  It’s simply too large to test all the permutations.

Talent, experience and iteration go a long way toward dealing with this issue, but a simpler approach is to re-think what it means to be balanced.  If unusual, lop-sided situations are re-cast as part of the experience — and the build-try-fail loop is short — players will find them entertaining instead of frustrating.

Being a live product also helps– anything too extreme can be quickly adjusted if it’s having a large impact on play (and by large impact, I mean it’s driving all players to pursue a single strategy, thereby undermining the benefits of an emergent system).

Noise

Having a large possibility space isn’t much good if it’s hard to find the interesting bits.  Similarly, if players can’t differentiate one approach from another, then being emergent isn’t going to increase engagement.

In both cases it helps to make the core elements of the game — the ones that can be combined during play to produce different results — as orthogonal as possible.  Having a +1 sword, +2 sword, and +3 sword as options is not being orthogonal.  A better approach would be attributes that are more binary in nature:  sword vs. not-sword, AOE vs. not-AOE, DOT vs. not-DOT, ranged vs. not-ranged, etc.   The ability to combine any of those together creates a much more discrete, understandable and useful set of permutations for the player.

Representation and Visual Fidelity

Emergence works by combining core elements in many different ways, but in a large possibility space it’s not possible to create unique assets for all permutations.  Emergent elements are often represented by simply visually stitching together smaller bits.  The challenge is making the combined element quickly understandable, even if the player has never seen that particular combination before.

Having only a handful of core, orthogonal elements helps, since the player only has to learn their function and not all the permutations, as does limiting the number of elements in any combination.  Really strong art direction makes a big difference too, particularly on smaller gaming devices where core elements might be very small.

Take poker as an example.  The individual cards are core elements, and a hand is an emergent element.  The uniqueness of each hand comes from stitching together a few cards, not a new asset.  The number of cards in a hand is constrained, each card is visually distinct, and the number of possible hands is very large.  But the player can immediately recognize what they have, even if that particular set of cards is new to them.

When I talk about the build-try-fail loop, I’m talking about the time it takes a player to set up a strategy, try it, and determine it’s success.  The build-try-fail loop applies to any game, emergent or not.  For example, Clash of Clans has a very long build phase (it can take several hours to collect the resources and create the units for a single raid), a short try phase (at most three minutes), and failure is moderately difficult to assess (you can’t build units specific to the defensive target’s setup).  By contrast, Candy Crush Saga has a modest build phase (the time it takes for energy to accrue), a lengthy try phase (several minutes), and failure — which isn’t known until the end — is hard to evaluate due to the random initial state of the puzzles.

The examples above have artificially long portions of this loop for monetization purposes, and it’s hard to argue with their success at doing so.  Nevertheless, the more emergent a game gets, the more it benefits from shortening this cycle.  Specifically:

• Experimentation is encouraged. The longer it takes to set up and make an attempt — a new level, puzzle, raid, whatever — the less risk a player is willing to take.   Reducing this time investment gives players the freedom to try new strategies and explore the entire possibility space the game has to offer.
• Learning is accelerated. The more frequent the attempts, the faster the player will come to understand the game and it’s potential.  And if you agree with the thesis that fun is primarily about learning (solving problems, figuring out new strategies, etc.), then more learning means more enjoyment for the player.  An emergent game with a large possibility space has more available to learn and therefore more potential fun to be had.
• Failure itself becomes fun instead of frustrating. The less invested a player is each time they try something new, the more likely the results will be treated as interesting or entertaining instead of unpleasant.

This is the second part of a two part series on games and attention.  The first part can be found here.

Last week I talked about how games compete for a consumer’s attention, and how existing casual and hardcore products represent flawed solutions to this problem.  We ideally want to enable casual and hardcore play in the same product, but doing so means taking a different approach to game mechanics.

How?  Make better use of emergence in our designs.  To enable emergent play, a small set of core components are recombined to produce an unlimited number of novel play dynamics.  Here’s what we get with this approach:

• Simple and easy to understand game mechanics (ideal for casual play)
• A massive possibility space, with complex dynamics (ideal for hardcore play)

Games of this type avoid rule complexity (and burdensome learning curves/commitment) by working with only a few mechanics.  The rich output gives the game legs and enables deep, engaging play.

The classic example is Go, a game with only two rules but an incredibly varied output in terms of games and play styles.  A modern example would be Magic:  The Gathering and other CCGs.  More recent:  Little Big Planet, Minecraft, The Sims.

Examples of games that are minimally emergent, if at all:  Pac-man, Heavy Rain, the board game Life.

Most game designers are already familiar with emergent concepts, and indeed, all games have some degree of emergent play.  But we tend to shy away from designing heavily emergent systems because the large possibility space is difficult to balance (many designs introduce top-down constraints to control this, like classes in RPGs, but that simply reduces the possibility space and undermines the benefits of an emergent system).

It’s also important to recognize that emergence by itself will not help a game attract players interested in both casual and hardcore play.  To be effective:

• You really have to constrain the number of core elements and their respective functions.  Many games are deeply emergent but they get there with a crazy amount of rules and systems that undermine any potential for casual play.
• More output doesn’t mean more choice.  Farmville has great emergence in terms of user expression, but it’s functionally meaningless.  Backyard Monsters’ non-orthogonal design elements lead to a relatively small set of strategic choices (a few dominant play strategies).
• Emergence doesn’t automatically mean accessible and deeply engaging.  You still have to avoid other annoying frictions (like bad UI design), and a large possibility space isn’t much good if it’s boring.

Ultimately, making stronger use of emergence won’t expand the potential audience for your game:  a player still has to like abstract strategy games if they’re going to play Go.  But it will increase how many of those potential players choose your game over all the other media choices they have at their disposal.

Last week I talked about how mid-core games aren’t the answer to consumer engagement issues.  This week I want to take a closer look at some of the underlying design problems with casual and hardcore thinking, and how we might better address them.

It’s been several years since I first talked about games and attention.  At the time, I noted that games compete for our attention with other games, other media, and interruptions from friends, family and work.  Despite all the advances in friction reduction over the past decade, games are hard pressed to thrive — nevermind get noticed — in this noisy media environment.

Looking ahead, the trendline is not positive.  Content creation costs keep dropping, so the amount of content available keeps rising.  Our choices are unlimited but our time is not.

Traditional solutions to the attention problem fall into two buckets:

• Reduce friction (a.k.a. the casual solution)
• Add depth (a.k.a. the hardcore solution)

In general, good casual design affects a player’s willingness to play, while good hardcore design affects their desire to play.

The Casual Solution

The friction reduction method requires games to:

• Have few barriers to entry
• Be commitment friendly
• Be attention span agnostic
• Have a short build-try-fail loop

That’s all good.  However, while these characteristics reduce friction and increase a player’s willingness to play, they don’t actually generate a desire to play (or to keep playing).  In an effort to meet the requirements above and make the game more accessible, most designers simplify the game’s mechanics and the way those mechanics combine, effectively reducing the play dynamics as well (in the MDA sense).  The result is a shallow product with a small possibility space and little long term retention.

Casual games succeed by demanding less from the consumer — they don’t ask you to sacrifice attention you might prefer to devote to other things, complementing rather than competing with other media.  The best of these products tend to have large audiences with low unit economics and short user life cycles.

The Hardcore Solution

The depth approach, on the other hand, means the game:

• Has extensive and varied play dynamics
• Has a lot of content to consume
• Gives the player a reason to make a deep personal investment (often through persistence, identity and relationships).

Again, a good list.  Games of this type tend to create a strong desire to play.  Unfortunately, they also erect a lot of barriers to someone’s willingness to play.  To create a wide range of play dynamics, many designers simply pile on the game mechanics (i.e. they keep layering on the rules and systems).  That’s a lot for someone to learn and then remember from session to session, and with a lot of mechanics to comprehend, 100% focus is required.  The only players willing to do that are the few that will make a large personal investment in the product.   The result is a deep product with a large possibility space and great long term retention, but it bounces most consumers at the start.

Hardcore products succeed by being more compelling than other media — they ask for your undivided attention and tell you it’s better spent on them than competing options.  High quality hardcore products tend to have small audiences with high unit economics and long user life cycles.

Hardcore + Casual

There’s nothing inherent to these two approaches that makes them incompatible, but you can see how solving for one can easily lead to problems with the other (it doesn’t help that, as an industry, we’ve set up a false dichotomy that places casual and hardcore products at opposite ends of the same spectrum;  mid-core is the latest iteration along these lines).

Much of the problem starts with the game’s mechanics, where most casual games have simple mechanics and simple dynamics, and most hardcore games have complex mechanics and complex dynamics.  There’s a reason for that:  they’re much easier to design and balance (simple dynamics imply a small possibility space, which presents fewer outcomes to balance;  complex mechanics make it easier to isolate systems for independent tuning, particularly when dealing with the large possibility spaces associated with complex dynamics).

What we really want are games with simple mechanics and complex dynamics, because simple mechanics are easy to learn and remember, while complex dynamics are deep and engaging.  How we do that is the subject of next week’s post.