Interesting Games

Sid Meier is often quoted as saying that “a good game is a series of interesting decisions”. When building the economy for a social game, it often pays to bare this in mind.

An interesting decision is a choice between options where no single option is clearly better than the others. As it is more difficult to trade off factors without a direct exchange rate, interesting decisions involve choosing between loosely or unrelated factors. This is perhaps even more true in real life that in games – think of the job offer with poor pay but interesting work vs. the one with good pay and less interesting work – it’s a difficult choice because it’s hard to work out an exchange rate for pay and interesting work.

In social games this equates to having multiple factors influencing each decision, so that when players weigh up their options, the optimal choice is not obvious, or is not the same for everyone depending on their style of playing.

The economy in Atlantis Fantasy is an interesting one (we’re pretty proud of it at Kobojo J) – it successively challenges players to balance the three main resources: coins, population and workshop materials (there are over 20 different workshop materials, but for the purposes of designing the economy, they can generally be considered a single resource). When building a house a player is left trying to balance the following options:

–          The cost in coins and materials

–          The population it will provide

–          The rent (in coins) that it will generate

But really this is just a short list. If you take into account that houses take up space on the map, and players are often space constrained, as well as the fact that players are constrained by the amount of energy they have to collect rent then you realise that savvy players will be trading off more complex metrics:

–          Population density

–          Cost per population

–          Rent per energy (and therefore amount of energy to pay back)

–          Rent per hour (and therefore number of hours to pay back)

Not to mention how different houses fit into their artistic vision of their city…

Whilst the basic metrics need to improve with each house that the player unlocks to give them a sense of progression, balancing the secondary metrics to give each house its own strengths and weaknesses means that each house is interesting to players for longer.

Take the following two houses, unlocked by players at level 10 and level 18 (i.e. at significantly different points in the player lifecycle):

Broadly, the Mansion is better for rent, and the House on the Seabed (HotS) is better for population. But that’s only true if players are both energy and space constrained. Relax the energy constraint and HotS is best for both, relax the space constraint instead and the Mansion clearly superior.

This means that even though HotS is unlocked later and at first glance appears to offer more population and rent, the Mansion is still an interesting choice for more advanced players in certain conditions.

Finally, I’d life to mention my favourite piece of wisdom from Tom Perkins (the founder of VC firm Kleiner, Perkins, Caulfield & Byers): that a difficult decision means the options before you are equally attractive, and therefore the opportunity cost of a “wrong” choice is actually small. In this context, a game that is finely balanced for hard core players is actually still very playable for casual gamers – because whichever house a casual gamer builds, they won’t make a wrong choice and break their experience.

Wrap Up

To create an interesting game economy, consider the following points:

  1. Make your basic resources as independent from each other as possible, to prevent players calculating a direct exchange rate
  2. Make sure the headline figures of buildings improve with players’ progression to give them a reason to level up
  3. Balance options on secondary metrics as well as headline figures to make options more interesting
  4. Check the experience for casual gamers – will any choice break the experience?

Solving distribution of wealth in digital communities

This post follows a recent conversation with 4chan’s creator Chris Poole.

Many communities and especially games use points systems to increase engagement – Foursquare shows you a rolling one week total of your points earned, compared to your friends. People have a natural inclination to accumulate as many points as possible and have more than those around them. Some communities reinforce this urge by allowing points to be traded for virtual items, used to unlock new features or otherwise giving them some value. For example,’s DJ Points unlock different avatars skins.

However, as soon as points are converted in other items, distribution of wealth starts becoming a problem. Whilst new users in the community enjoy seeing other wealthier users around them to begin with, it is because they aspire to that wealth themselves. If that wealth seems unattainable, then they may well lose interest in the whole community.

Furthermore, the community dynamics can become disrupted by trading, even where this is not explicitly allowed. Users tend to find a way to reciprocate actions to effect trading very quickly. I wonder how many users on are voting up songs, and thereby awarded DJ points to others, not because they like the songs they are playing, but because they want others to reciprocate this. The ecosystem can even spill out into real world currency – you can buy virtual items of other players in World of Warcraft for USD. When many people use online communities for their egalitarian nature, allowing users who are richer in real life an advantage is hugely destructive.

So how to avoid the problems of distribution of wealth, without removing the highly positive influence that a points system can have?

Make wealth less conspicuous

If it is easy to see how much wealth other users have, and how far you are behind them, then the problems associated with wealth distribution are exacerbated. Conversely by reducing how conspicuous people’s wealth is, these effects can be minimised. In many communities (as in real life), a user’s wealth of points cannot be directly observed, but the items they can purchase with these points can be.

Making items that users can buy with points exponentially more expensive for marginal gains in functionality or form can allow “richer” users to have some value to their wealth, without putting them too far ahead of less well off users. Most level systems adhere to this principle, with the first few levels easily achieved, and higher levels requiring a far great number of points to reach.

Expire points after set time

One of the simplest ways to prevent some users from getting too much richer than others is to have points expire after a certain amount of time. This way experienced players need to keep engaging with the community to maintain their wealth, and the maximum wealth that anyone can achieve is limited to the amount that they can earn in the lifetime of the points. Foursquare keeps a record of only the points you have earned in the past week, and British Airways reset their frequent flyer programme members tier points (but not the actual miles) to zero at the end of the year in this way.

Adopt a welfare system

Many community based points systems have some form of welfare system, where users are given a certain amount of points each day or week automatically, or simply for logging in. This allows users to get started immediately and taste the benefits of accumulating more points, but also means that users without the skills to accumulate huge numbers of points just need a little more patience. A welfare system alone is unlikely to be a solution, but it can be helpful in many communities to get users started.

Create seasonal virtual goods

Instead of expiring points after a certain time period, virtual goods can be given a temporary nature instead (or indeed, as well). This could either work in the sense that users pay a certain number of points to lease virtual goods for a set time period, that virtual goods only exist for a finite time period and need to be renewed, or that new virtual items are constantly being released in seasonal cycles like fashions. In this last case, the items might have different forms but identical functionalities. In any of these cases, users have to maintain their points earnings to have the latest virtual items, and new users can get up to speed quicker, by only buying the most recently released items.

Migrate more experienced users to a new area

Some games migrate more experienced, richer users away to a completely separate environment from new users. This way the environment that users interact in always has a limited variation in wealth. Call of Duty allows users access to a new set of hard core multiplayer servers in return for handing back all the weapons and virtual items they have accumulated and starting again. Many MMORPGs have a newbie town where new players are sheltered from the wealth (and dangers) of the wider, virtual world.

There is no reason that a large number of different areas cannot exist, each for a particular wealth of user. Furthermore, users don’t need know that it is possible to enter other, more advanced areas before they have accumulated enough points to enter them. Indeed, having “secret” areas that reveal themselves just as users think they are exhausting the possibilities of the original one is undoubtedly a hugely compelling feature for many.

Wrap up

Giving users in a community points has a tremendously positive effect on engagement, at least initially, and this can be amplified by allowing users to trade points for virtual items or features. However, communities with a wide distribution of wealth can quickly run into problems with new users feeling points “wealth” is unattainable, or puts them at unfair advantage. There are several ways in which the negative effects of distribution of wealth can be moderated, and these can be used in conjunction with one another if needs be. The right solution for any given community is likely to reflect that its overall value system and culture, and should augment rather than supersede this.

Image from zzzack’s Flickr photostream