While at the GameOn Finance event in Toronto, I found myself in an interesting conversation about ways to maximize the revenue generated by MMOGs. I found it difficult to fully express my thinking on the matter at the time, so during my flight home I wrote this post. Consider it a sneak previous into my upcoming IGDA Leadership Forum lecture on MBA Lessons applied to the game industry. :-)

One of the concepts I learned in business school was the “two-part tariff,” which is best explained through a simple example that we’re all familiar with: a nightclub. Most nightclubs generate the majority of their revenue from the sale of liquor. Why then do some of them also choose to charge a cover fee? Doesn’t that turn away potential customers? Well, part of the reason is simply to “keep out the riffraff,” but bouncers at the door can (and generally do) already reject anyone who looks like they won’t be a valued customer. Part of the reason is to project an aura of quality and/or exclusivity, but again, a velvet rope and an obstinate bouncer can already accomplish that as well.

Two kinds of customers

The third major reason for a cover charge at a nightclub is revenue maximization, pure and simple. Here’s the underlying rationale: nightclubs basically have two kinds of customers. One kind buys a lot of drinks (the especially valued customer buy a lot of the most expensive drinks.) The other kind buys one drink and nurses it all night, or even — heaven forbid — just a glass of water. Both kinds of customers are attracted to the nightclub because it offers music, attractive people to dance with, etc. Both kinds of customers clearly value the experience. But only one kind of customer will be profitable for the nightclub. Sound familiar?

So the nightclub does a very simple calculation. It asks, “what is the experience of being here worth to most people — or more accurately, just enough people that I can easily fill the place each night.” That estimate of worth becomes the cover charge. It extracts at least some revenue from the people who want to enjoy the nightclub but have no intention of paying if they don’t have to. The other customers — the ones who are likely to buy a ton of drinks — are not dissuaded by the cover charge because they already know going in that this is going to be an expensive experience for them. What’s five or ten more bucks at the gate?

Entry/Subscription fees and microtransactions are not mutually exclusive

It seems unfortunate to me that despite the existence of this very classic pricing example, many game developers seem to think that microtransactions and entry/subscription fees are mutually exclusive. But I’d argue that our industry’s equivalent of the “popular nightclub” — aka an MMOG with high production values and either A) strong IP and/or B) tremendous buzz — can take advantage of both. The thought process is the same for nightclubs and MMOGs: “how many people do I need to attract to make this an exciting environment for everyone to be in, and how much can I get away with charging as entry/subscription fee while still reaching that number?” Of course, that doesn’t mean you need to charge right from the get-go; there’s always the possibility of a free month’s trial (or something like that) to help build critical mass.

Maybe the subscription fee is $9.99 a month. Maybe it’s just $1.99 a month. The amount depends on the MMOG. And for many MMOGs, it’s clear that the amount is “zero.” There’s simply too much competition for customers in this space, and the competition is only going to get hotter over time. But for those MMOGs that can potentially justify a non-zero tariff, the truly important thing to remember is this: you can always drop price. Raising it is MUCH harder. If your experiment with a two-part tariff fails, eliminate the tariff. You might have lost a little momentum building towards critical mass, but odds are the delay won’t prove to be a critical error as long as the tariff itself wasn’t insultingly high to begin with.

Will it offend players? Not if handled correctly

The major objection to my argument seems to be that “people who pay a subscription fee will be offended if some players can pay to get an advantage.” It’s not clear to me that this is true, but let’s take for granted that it is. There is still a whole host of things you can sell that do not convey any sort of strategic benefit in game. Call them “status items” and “gifts.” A really cool-looking outfit, or virtual flowers. Or perhaps even an in-world home.

Now, it may be that an existing MMOG (like World of Warcraft) may not be able to institute microtransactions after the fact, because it has been around long enough that players have developed an expectation for “how the world works.” But a new MMOG faces no such preconceived notions. And if said “new MMOG” happens to be, as I mentioned, a likely equivalent of the “popular new nightclub in town,” the developers of that MMOG absolutely should consider a two-part tarriff revenue model.

The Freakanomics blog is worth subscribing to, if you haven’t already. Via it comes word of a neat experiment: in the psychology department coffee room at Newcastle University, prices for tea and coffee were posted on the wall, and an “honesty box” (i.e., in which to place your payment) was set nearby. This sort of experiment is fairly common, but the twist was that, in some weeks, a photo of flowers appeared above the price list. In other weeks, it was a pair of human eyes, staring directly at the person reading the price list. In weeks with eyes on the list, staff paid 2.76 times as much for their drinks.

This brought to mind my earlier post on the wonderful book Predictably Irrational, which noted that you could dramatically cut down on cheating in exams if you simply asked students to recall the Ten Commandments before they took a test, or — more pointedly — by reminding them of a school honor code. (You had to do this right before the test — it couldn’t happen weeks before and still retain the effect.)

A long time ago, I wrote an article for Gamasutra exploring the possible design of a feedback/rating system that would discourage antisocial behavior in MMOs. That system, which still may have merit, pales in comparison to the wonderful elegance and simplicity of these psychological tricks. And given that (1) one of the biggest challenges for online, anonymous systems like MMOs (and LIVE) is antisocial behavior, and (2) companies are spending tens of millions of dollars on these systems (if not more), why aren’t more companies hiring psychologists and behavioral economists as consultants or full-time employees? The cost seems justified, given the potential benefits.

(Actually, I’ve heard of a few MMO developers hiring economists, but I think that most are focused on the optimization of in-game economies, as opposed to tackling anti-social behavior. There is a relationship between the two, but they are not equivalent.)

I’ve been meaning to write about a web-based MMO called Travian for a long time. Travian is, to my mind, the very embodiment of the phrase “so close, and yet so far.” It has all the basic components of a perfect low-budget MMO, but a few maddening design flaws make the game basically unplayable (in the long term) for most people. The following is a very long deconstruction of the game. If you’re interested in MMOs, read on. If not, it’s safe to skip this post. ;-)

Travian in a nutshell

In a nutshell, Travian is a pseudo-real-time massive multiplayer strategy game. You build towns and armies, and use your armies to conquer and pillage other towns. I say “pseudo-real-time” because, while the game operates in real-time and you can take action whenever you wish, each action requires a variable but substantial amount of time to complete. (For example, building a granary might take 20 minutes in the real world; upgrading it might take several hours. And while you’re building your granary, you can’t build anything else. Likewise, sending your army on a raid could take as little as 30 minutes or as long as a day.) There is real genius in this — it preserves the feeling of a real-time game while effectively preventing people with tons of spare time from overwhelming competing players. The eleven-year-old who wants to can obsess over the world map and communicate with allies to his heart’s content, while the forty-year-old parent with twenty minutes to spare can quickly take his turns and tune out till the next day.

Escalating demands on your time

Or that would be the case, if Travian didn’t enable you to keep building towns, each of which require a fixed time to manage. By the time you reach four or five towns, the game no longer feels like a small commitment. Worse yet, Travian does not offer any meaningful way for you or your allies to auto-defend your towns, so players feel compelled to log in obsessively (in order to track and respond to incoming attacks.)

If Travian made player advancement more scalable (from a time-management perspective), channeled all excess time/energy of players into actions that don’t unbalance the game (like team communications, in-game personalization, optional micromanagement, etc), and made it easier for players to defend themselves when away from the computer, the beauty of the game’s basic design could have been preserved.

Alliances that live up to the term “massive”

Alliances are another (conceptually) great part of Travian. To be brief: you can’t survive Travian on your own. Players that don’t form alliances are quickly overrun by neighbors with friends. And eventually, single alliances are overrun by groups of alliances. Travian is by no means the first game to offer this sort of group functionality, but it fits very nicely within Travian, and players really don’t seem to miss the opportunity to play as a loner. (I’m one of those guys who always choose the best “loner” class in MMORPGs, so I can testify to this myself.)

In fact, I’ll go so far as to say there’s something amazing about the feeling of being part of a huge online team, and knowing that your participation really makes a significant difference. I always hated 40 man raids in World of Warcraft because I found it impossible to shake the feeling that (in general) my participation was irrelevant. 39 men were more than enough to complete 99% of the raid content. But in Travian, you can’t help but feel you count, even in a 500-man war. When your army hits an enemy town and you get the damage report back, it feels… substantial — in no small part because your army probably took one week or more (real time) to build, and losses on both sides definitely matter. And at the same time, losses are generally manageable with the help of a well-run alliance.

Dysfunctionally massive

Unfortunately, Travian doesn’t actually offer alliances any meaningful way via which to coordinate attacks or defense. Players are left to rely on external message boards and/or chat clients, which means obsessively logging into at least TWO systems (Travian and the comm system) if you want to play the game effectively.

You’d have to play Travian to truly understand what I’m about to write, but: imagine having to coordinate an attack with 50 (or even 250) people, all of whom need to take a precise set of actions at a precise set of times, without having any built-in communication or planning mechanism for this — not even something that helps manage the different time zones that players (from all over the world) are operating in. Suffice to say, in several months, I never managed to participate in a single, well-executed attack. Not one. And while my alliance may not have been the very best, it was by absolutely no means the worst.

With a simple set of tools, Travian could enable alliances to coordinate attacks and auto-execute defensive actions (so players wouldn’t feel compelled to login constantly, as mentioned previously.) Without those tools, participation in an alliance becomes (at best) a very necessary evil.

Conversion incentives for non-paying customers

I was also intrigued by the ways that Travian gets customers to upgrade to “Travian Plus” (the paid service). Travian is completely playable as a free game. Upgrading simply makes playing Travian more convenient; you can execute resource trades more quickly, queue build orders, etc. In other words, upgrading simply saves you time on a daily basis. Guys like me (with more money than free time) can pay for convenience. Personally, I much prefer that to casual MMOs that sell performance-enhancing items to my opponents, which feels inherently unfair and manipulative. But perhaps that’s just me.

Unfortunately, Travian ultimately turns into such a gigantic time-sink that the convenience of Travian Plus barely helps in the long-term. There’s a balance to be struck here, and I’m convinced that a “better” Travian will strike it.

Stickiness, or lack thereof

On one hand, Travian is relatively sticky (by virtue of the bonds you form with your alliance members.) And on the other hand, quitting Travian is relatively easy (as compared to World of Warcraft), because there isn’t much opportunity to personalize your in-game presence. You’re just a name on a set of villages. At the end of the day, if you’re on the losing side of a conflict (or just getting bored), quitting means saying goodbye to your teammates, but not to a carefully-constructed, lovingly-maintained image. Travian players who convince their teammates to try another game with them, in other words, have basically nothing to lose.

And despite all these flaws…

I witnessed some amazing things while playing Travian. Alliances planting spies in other alliances. Breaking into each other’s private forums. Political intrigue at the micro and macro level. Incredibly passionate groups of people, desperate to defend one another and climb the ladder to #1 alliance. Travian has many thousands of dedicated players, most of whom would probably agree that the game is deeply flawed.. but they’re still playing. It’s a testament to the power of a relatively simple game that enables many people to meaningfully join forces.

There’s much more I could write about the design of Travian, but this post is already too long. Suffice to say, I’m glad I played the game, and even more glad that I quit. I think I learned a few valuable lessons about MMOs in the process.

Last word: someday, a better Travian will come along… and when it does, it’s going to make a giant pile of money. You can bet I’ll be one of the customers tossing bills onto that pile.

Real-time communication in modern MMORPGs is a funny thing. With rare exception, it tends to resemble anything but “role-playing”. MMO user text generally consists of acronyms (LOL, ROFL, etc), poor grammar, and a million little references to the outside world (“hang on, my dog is barking.”) Speech is, in some ways, even worse — nothing like the screech of a petulant 10-year-old (or the sound of a toilet flushing in the background) to disturb the illusion of fantasy.

Outside the context of self-policed, dedicated role-playing servers, this may be impossible to “fix”. I put “fix” in quotes because it’s unclear that this is a problem of any real significance — it’s quite possible that the majority of potential players really don’t miss the opportunity to role-play more deeply, even in the “perfect” environment for it. But my gut tells me that, at a bare minimum, there’s room for something more than what’s available today.

And given that, I just don’t understand why better voice-masking technology hasn’t found its way into MMORPGs yet. I know it’s complicated, but I refuse to believe it’s impossible to create efficient software that makes a 10-year old girl sound like an orc (or a 30-year old man sound like a female elf, for that matter.) And I have to believe that such software would not only be extremely popular in MMORPGs, but in many online virtual environments. The benefits:

• Greatly enhances role-playing.
• Eliminates “I hate the sound of my own voice on the answering machine” syndrome.
• Greatly reduces shyness. (Voice chat is far more intimate than text chat. There’s a reason most non-hardcore gamers still don’t indulge in much voice chat with strangers online, and that voice-chat is less popular in more reserved cultures.)

Many people are wondering what will define “the next World of Warcraft“. It would be egotistical at best to imply that I hold the answers to that question — I certainly don’t. But if I had to take a wild guess, I’d say there’s a decent chance that the next WoW will offer:

1. Significantly better integration of user-generated content. (A virtual world is really not much of a world at all if its inhabitants can’t change it. And virtual worlds are growing too large for even the best-funded dev team to populate with content.) And/or,
2. Better and more numerous opportunities for users to express themselves. That brings me back to user-generated content, but it also brings me back to my original point — the most basic form of self-expression — direct communication. Virtual voice.

And yeah, I really want to sound like an ancient dragon. Or an English-speaking cave serpent. So sue me.

PS. My friend Tom Cadwell raises an excellent point: it’s unclear what the ethical, legal, and PR implications of virtual-voice technology would be if it were used, for example, by sexual predators to more effectively prey on children. I haven’t immediately thought of a great response to this; I can only ask “where do you draw the line?” Should we shut down MySpace because it makes life easier for some deviants? Eliminate webcams? This is a far larger debate.

Lots of interesting news this past week about real money transactions (RMT) in MMORPGs. I’ll get into specifics shortly, but first, I encourage you to view the following information through this lens: are traditional MMORPGs first and foremost a game, or first and foremost a social networking service? (And assuming you think the distinction is even meaningful, what bearing does your answer have on RMT, user-generated content, cross-cultural communication, and “virtual property rights” in these games?)

For-profit power-leveling

First, from Raph Koster’s blog, a look at the power-leveling industry. As Raph points out, the average market value of a WoW level is $8, and an hour of WoW play is worth under 75 cents. Clearly not enough to turn most US citizens into WoW entrepreneurs, but certainly enough to keep inspiring “level farms” in China. What I found more interesting was a comment by Raph outside the article:

Disneyland doesn’t tell you “Sorry, you can’t get on Space Mountain until you have ridden A Small World five times. Your friends have, so they can go on ahead. you have to sit and listen to inane music a while longer.” If I met that situation, damn right I’d pay someone else to listen to that annoying song endlessly.

Is it paternalistic to assume that the shared experience of leveling up is necessary to preserve the overall enjoyability of an MMORPG for all its participants? If so, is it then a mistake to design MMORPGs around the concept of leveling up? If not, what other design mechanisms can be used to create shared experiences? And lastly, does any of this even “matter” as long as users are given the opportunity to segregate themselves into communities that respect a given set of social and economic rules?

EQ2 Station Exchange: “legalized” RMT

Speaking of: Sony recently released a whitepaper (.doc link) exposing some interesting stats from the Everquest 2 “Station Exchange” — the proprietary RMT market run by Sony itself, which is only accessible from two special EQ2 game servers. (Thanks, once again, to Raph for highlighting this.)

Some key information from the whitepaper:

• Net cash includes any money that passes through the Station Exchange auction service. Total cash collected between June 2005 and June 2006 was $1.87M. Recognized revenue for the first year was $274,083. (DJE: theoretically it could have been more, since Station Exchange is limited to just two EQ2 servers. It depends on user demand and the extent to which some long-time users may have felt rooted to their existing, pre-Station servers.)
• The average active Station Exchange buyer is 32 years old. This is much older than the average age of EQII players overall, which is 25 years old. (DJE: probably supports the assertion that most buyers are simply busier individuals — i.e. with advanced careers, families, etc.)
• The top 15 sellers all earned upwards of $10,000 from trading.
• 40 percent of customer service time was spent on disputes over virtual item sales. Since the debut of the Exchange, the overall customer service time spent has dropped 30 percent. (DJE: in other words, the best way to manage a potentially disruptive, deeply-rooted social practice is to carefully regulate it, not outlaw it.)

Again: is the Station Exchange an “unhappy solution” to a core design problem — the need to keep pace with your friends — or is it a good solution for an affluent segment of the market? Does user self-selection into different social/economic environments (i.e. different servers) address most major concerns with RMT?

Are there “alternatives” to RMT, from a user perspective?

Lots of interesting questions. I’ll leave off with one final thought: in sports (and indeed, in some video games), we have the concept of a “handicap” — an advantage given to players with less skill and/or experience. Handicaps enable players of different levels to face off and make an otherwise boring game much more interesting.

MMORPGs typically try to segregate players of different levels, not indulge them with handicaps; real-money purchase of powerful virtual items is a (generally unsanctioned) workaround. And frankly, not a very effective workaround for the user, since items in MMORPGs are generally a poor substitute for skill born of experience or innate talent. Same with power-leveling; inexperienced users frequently fail to fully take advantage of their unnaturally-advanced characters.

Is the general absence of handicaping in traditional MMORPGs a good, bad, or neutral thing? Could handicaping reduce the demand for RMT, and is that even a desireable outcome for traditional MMO developers that may profit from RMT? And again, can MMORPGs sidestep this entire debate by reducing or eliminating level-centric player segregation altogether (which may or may not mean eliminating levels altogether?)

Your answer may depend, in part, on your own personal skill level and/or available free time. ;-)

I was reading Raph Koster’s Theory of Fun and something caught my attention. A quote: The only real difference between games and reality is that the stakes are lower with games.

Interestingly, this statement is becoming progressively less true over time. For example, the extremely successful Diablo franchise offered, among other things, a (“hardcore”) mode of play in which character death was permanent. People who opted for hardcore mode played in constant fear of the single mistake that would end their character (which may have survived through hundreds of hours of gameplay.) For them, every moment in-game was deadly serious.

MMORPGs have taken in-game consequence to the next level; not by introducing permanent death (in general, recent generations of MMOs have made death less punishing, not more), but by creating an environment in which the social consequences of your actions may be severe. As in real life, poor teamwork, cheating, etc in an MMORPG can cause people to avoid and/or badmouth you in the future. As social systems in some MMORPGs continue to evolve, this will become an increasingly powerful phenomenon with very real consequences for players.

The emotional connection that people have to their virtual characters and pets is quite strong. It’s remarkable to see descriptions or video clips of the elaborate rites that some people go through (in-game and out-of-game) when finally quitting an MMORPG. They lay their virtual pets to rest in the same way they might a real dog or cat. They quit their character with as much regret as they might quit a failed romantic relationship.

Of course, Raph knows all this. He was almost certainly talking about stakes in the broader sense; i.e. when you go mountain climbing, there’s a very real possibility that some accident could kill you. There’s a real chance you’ll wreck your car every time you test yourself by driving aggressively (or drive at all, for that matter!) So what he wrote is true, for the most part. I just felt a counterpoint was worthwhile.

A couple of days ago, Sony announced a new MMOG bundle. Starting March 14th, you’ll be able to buy EverQuest Evolution, EverQuest II, PlanetSide Aftershock, The Matrix Online, and Star Wars Galaxies for just $39.99. The move (among others: 1, 2) has its fair share of detractors; this Penny Arcade comic sums it up.

While I do believe that the game industry generally over-emphasizes the importance of pricing signals (we desperately need more $30-$40 high-quality games), I have to agree with Sony’s critics on this one. The MMOGs that Sony is pitching are (mostly) “high playtime” games which, when they’re successful, tend to encourage long and frequent gameplay sessions. And they all appeal to a hardcore audience. The problem is, a hardcore gamer who’s going to play an MMOG for ten, twenty, or thirty hours a week isn’t going to care overly much about the initial price of the game. He/she is going to care about who’s playing it, and how it compares to other MMOGs on the market. When you’re going to pay $10 or $15 a month for something, and play it constantly, the first $50 doesn’t mean a whole lot.

On the other hand, I think there is room for experimentation on the recurring fee side of the equation. Significantly reduce the recurring fees, and/or increase the “free” initial play period, and people may choose to try your game (and/or maintain a mostly-inactive account) just because it’s “cheap”. This is different from waiving the fee but capping the player to a newbie area or restricting their level growth; that strategy might work for casual gamers, but not the hardcore gamers who want to level up fast, and who are the most likely audience for a game like EverQuest II. Of course, if you’re battling for the hardcore audience on price, you’ve probably already lost the war…

Long story short, I think this is one case where a negative price/quality signal will be sent. There might very well be an influx of new players when the bundle hits the shelves, but I bet they don’t stick around for long — unless subscription fees go down. And even then, I wouldn’t hold my breath.

I just sat in on the “Continued Growth of Gaming” panel at the MBA Media and Entertainment Conference in New York. Moderator: Cyrus Beagley (Engagement Manager, McKinsey Entertainment Practice). Speakers on the panel: Greg Costikyan (Founder, Manifesto Games), Chris Di Cesare (Director of Marketing, Xbox), Nique Fajors (VP of Brand Management, Atari), Frederic Markus (President, eRelevant Games), Joseph Varet (Sr. Director of Biz Dev & Strategy, MTV Networks). I managed to catch most of what was said, except in the case of Greg Costikyan, who speaks two to three times faster than most normal human beings.

Topics discussed: What makes a franchise successful, MTV’s role in the video game industry, innovation, the attractiveness of various game markets, and some questions for Microsoft about portables and shortages.

Read the full transcript here.

A particularly interesting post on Terra Nova reveals the results of an eight month, detailed study of World of Warcraft. The dataset includes information from over 150,000 characters, so it’s certainly thorough enough. The post claims that the results of the study contradict the commonly-held assumption that people play MMOGs primarily for the social interaction they offer. However, it isn’t clear to me that the data really supports such an argument More on that.

Among the study’s most interesting findings: early-stage players (level 40 and below) spend only 30% of their time in groups, and less than half of WoW players belong to a multi-person guild. Furthermore, the average guild member collaborates (in quests, etc.) with only 11% of his/her guildmates for more than 10 minutes over the same month.

First, this data could indicate that many players rush through the early levels in order to enjoy end-game content with their friends. (Indeed, the study also found that end-game characters spend far more time in groups.) Second, guild members may form strong relationships with a small percentage of their guildmates and choose to group with them whenever possible. (They may not even have a choice, if those guildmates are the only guildmates who regularly play WoW at the same time of day.) Of course, all this could be incorrect as well. I’m just saying it isn’t clear.

The post also states that players favor “soloable” classes like hunters and warriors. (Data here.) That certainly has merit as an argument against social inclination. However, Warlocks are soloable, and they’re the least-played Alliance class. Players may prefer warriors and hunters for any number of unrelated reasons (for example, some may choose warriors because that class enjoys access to the broadest variety of weapons and armor.)

Lastly, from the post: “despite features like WoW’s ‘group xp bonus’, grouping is an inefficient way to level, which naturally steers the more ‘hardcore’ players away from groups (at least, in the early stages of the game).” But if the system is inherently biased against group play at early levels, I don’t see how you can make any major assumptions about social inclination from the data.

Long story short, I’d like to know more before accepting any pronouncements. That said, I’d bet a few developers would be pleased to learn that social interaction is not so important, since accommodating social interaction tends to engender the thorniest design problems.

As promised, I wrote an article about designing an MMORPG feedback rating system, which Gamasutra was kind enough to publish. You can read the full article here.

I’ve already received a few thoughtful emails about the article, some of which I’d like to address here. One person asked how my proposed design might relate to previous designs that took PVP activity into account. The answer is: it does not, and should not. My design is intended to help ensure that people enjoy group play (which, in many MMORPGs, is 90% of the fun at later levels). PVP is extraneous and should have its own control mechanism(s). After all, someone who mercilessly hunts you down in a PVP setting may still be an excellent group member.

Another person asked how my system might compare to the feedback system in EVE-Online. Unfortunately, I haven’t yet had the opportunity to try EVE; I’ve been meaning to for weeks. That said, the EVE system sounds interesting, if somewhat more complex than what I’ve proposed. It also appears to dictate some NPC behavior, which I’ve left open in my proposed system. It isn’t clear to me that a user feedback system necessarily has to influence NPCs, since that could intensify user desire to try to game the system, and it could (in general) amplify any unanticipated negative side-effects of the system.

Thanks to everyone for all your feedback thus far.  :)