Oct 032008

Economics of Virtual Worlds: The Rarest of Lawyers.

Great graphs there explaining why it is that a microtransaction-based model ends up hitting a far larger market than a sub-based one.

  24 Responses to “Economics of Virtual Worlds: The Rarest of Lawyers”

  1. […] that one can make more money with free-to-play/virtual asset sales than with subscriptions. Raph referenced that post as well, stating that the graphs Dan includes in his post demonstrate why a microtransaction model ends up […]

  2. Take a piece of string 10 inches long. Hold one end on a table and lift the other end 6 inches off the table, so the ends plus the table form a right-angled triangle. Now, lower the end from 6 inches to 3 inches. The string will then trace the approximate outline of the actual graph you get when you offer free-to-play plus microtransactions. From a business perspective, you want to charge just enough to cut out all the freeloaders, to cover your costs; however, those freeloaders are making the game fun and vibrant, and give the people who pay for in-world advantages someone to have those advantages over. In other words, the players themselves are part of the calculation of “willingness to pay”.

    Likewise, in the flat rate model we have people who are willing to pay more, but who wouldn’t be willing to pay more if other people could pay even more and have them at a disadvantage. In this view, virtual worlds are games and should therefore be played on a level playing field, and just because they’re so much fun that you’d be happy to pay more to play them, that doesn’t mean you would be willing to pay more if people can pay extra for cheats. Again, therefore, the “willingness to pay” is dependent on the players and what it is they’re paying for, it’s not a fixed curve.

    The tension between micropayments and flat fees is one of fairness, and only really applies to game worlds, not social worlds. If you think it’s fair that people can buy success in a game situation, play one with micropayments. If you don’t, play one with a subscription (or with micropayments only for things that have purely cosmetic value).


  3. So — I am not saying anything about WHAT is sold. It could be access, it could be game perks, it could also be cosmetic things or even side services. So it might not affect fairness at all.

    Matt Mihaly’s post (linked in the trackbacks) is worth reading too… he brings up good points about the nuances. I am just saying that the nature of a pay-what-you-want model can hit larger markets because it removes the first price sensitivity threshold. Not that it necessarily makes more money (right now, there’s no proof that it does).

    Most mature industries seem to have settled on a compromise with tiered pricing plans, which brings in some of each model. More than one sub level, that sort of thing.

  4. One of the biggest flawed, persistent, incorrect assumptions about Free-To-Play games is that paying gives an advantage. This is not true in one that is well-designed. Most of the items that are sold in Korean F2P games are purely cosmetic or alternately designed items that are available for free.

    There is an interesting question that games like Runescape raise – which is that low-cost subscriptions seem to have a very high conversion rate (I think they quoted 12 percent!) which is higher than the number for F2P games.

    Low-cost subscriptions may do better because the perceived barrier to entry may actually be lower ($5 per month vs. a payment card) AND you get the “Health Club” benefit – where people keep paying whether they show up or not… and the perfect customer is one who keeps paying without actively playing.

    There has been very little thought about the whole “impulse purchase” model for games.

    What is not discussed is the cost to service the free customers. We spend blessedly little time discussing how to lower the operational cost for games. Reducing downloads and bandwidth, minimizing server utilization, etc. Free customers do “cost” something. I think the CEO of NHN stated at a conference that they had a game they brought to the US that lost money every time someone played because of the infrastructure costs (… the differences between the US and Korean bandwidth and data center costs).

  5. >So — I am not saying anything about WHAT is sold.

    But the graphs DEPEND on what is sold. They’re only valid if what’s sold doesn’t affect the willingness of people to pay.

    >it might not affect fairness at all.

    It might not. However, in practice they seem to do so with remarkable frequency.

    >Matt Mihaly’s post (linked in the trackbacks) is worth reading too…

    Yes, I already read it, and it is indeed worth reading.

    >I am just saying that the nature of a pay-what-you-want model can hit larger markets because it removes the first price sensitivity threshold.

    And I’m just saying that when part of what you’re selling is based on the notion of competition, pay-what-you-want could have pay-what-other-people-don’t-want-you-to-pay effects that mean you can’t rely on those graphs.

    >Most mature industries seem to have settled on a compromise with tiered pricing plans, which brings in some of each model. More than one sub level, that sort of thing.

    Is this true of industries in which tests of skill are an important part of product use?


  6. From a Tufte point of view I would have to call them horrible graphs. Humor me and explain what the axis labels and units are…

  7. Richard Bartle:

    But the graphs DEPEND on what is sold. They’re only valid if what’s sold doesn’t affect the willingness of people to pay.

    The graphs model consumer behavior as related to subscription and microtransaction pricing strategies. What’s sold is not relevant. We could be talking about memberships or magazines, and the model would remain applicable.

    There is a direct correlation between price and propensity to buy. A high price generally correlates to a low propensity to buy while a low price generally correlates to a high propensity to buy. How “low” and “high” are defined depends on the market segment, but a middle price (e.g., flat monthly fee) ignores preferences in favor of an easily managed standard and a stable income from a smaller piece of the pie.

    In other words, the players themselves are part of the calculation of “willingness to pay”.

    Right, the players are part of the value that consumers can derive from use of the product. This only reinforces the idea that demand is variable. Relying solely on a subscription pricing strategy, where the market is heterogenous and demand is variable, is suboptimal. Microtransactions fill the gaps in the curve by enabling consumers to customize the value that they can derive from product use.

    There are other solutions and quick fixes, of course.

    From a business perspective, you want to charge just enough to cut out all the freeloaders, to cover your costs

    For the sake of clarity, are you saying that firms should only charge enough to break even? Sounds like bad business to me. Even charities aim higher. Cost-plus pricing has many problems and is primarily used when market research is lean.


    Humor me and explain what the axis labels and units are…

    The author explained the graphs in the text.

  8. Morgan Ramsay>For the sake of clarity, are you saying that firms should only charge enough to break even?

    No, this was looking at it from the micropayments perspective – I wasn’t saying that a subscription model should aim for break-even only. I was saying that if you just looked at the graphs as if they were independent of what was being sold, then you would see a vast length of people who were costing you more in overheads than you were getting from them – the “freeloaders”. If the graphs were static, then you would want to cut those out because they were costing you money – even from a microtransactions point of view. Therefore, you would have some initial micropayment that covered the cost of processing them. You could then charge micropayments as usual. Actually, I seem to recall that Puzzle Pirates does make a stab at this, but only as an option – they tempt you to hand over $3 or something for some doubloons, but won’t get cross if you don’t.

    The reason I said this was to show that although such an action would make perfect sense if the graphs were static, they’re not static. There are three main reasons for this:

    1) Some of those people who don’t pay now will convert to payers in future.
    2) The payers need some non-payers around, in order to make their payments more meaningful.
    3) As Matt says, with a dual-currency system you can have economic interaction between the payers and non-payers, which leads to a more equitable symbiotic relationship between the two (ie. the system seems fairer, and therefore less unattractive as a game).

    The Puzzle Pirates solution seems a reasonable compromise between the economics (freeloaders directly cost you money) and the practice (freeloaders indirectly gain you money).


  9. I don’t have any data to back it up, but I think that the things that are sold can be purely cosmetic and it wouldn’t affect people’s willingness to pay.

    My wife pays Gaia Online, and she and other people like her will bay small amounts frequently for things that are purely cosmetic (new shirts and the like). Heck, people would probably pay a good amount for something that adds particle effects to your armor.

    That’s just one source of cosmetic revenue. If you get into new character slots, or extra bank or housing space, it’s even more microtransactions.

    Where do you draw the line? I’m not sure – it lies in the definition of what is cosmetic and what is functional.

    How does free to play dampen the percieved value of the game, though? When people commit money per month to play, there’s an expectation that it might be worth the money. When it’s free to play, you have your marketing, your word of mouth buzz about the game (which will probably ultimately be more important) and it’s easy to toss the game in a large pile marked ‘there are too many of these to try’. How much does the fact that someone’s paying up front for your game get them past the threshold that they’re playing it rather than sampling it?

    Reducing the previous paragraph down, what are people here expecting from a game that you pay monthly vs a microtransactions game? Even if they’re the same exact game with two different models, the perception is vastly different.

  10. I kinda take issue with the fairness argument honestly.

    Your traditional advancement via time spent MMOG is not inherently any more fair than one that lets you buy success. A time based model is only fair to players that have large amounts of time to spend. There’s only equity if the time commitment is small; otherwise it vastly favors people with large amounts of free time. Time is just as limited a resource as money is for many people. Just because the traditional market for MMOGs is made up of people with more time than money doesn’t mean that buying success as a mechanic is any less valid, it simply means it’s targetting a market with more money than time. In order to reach an actually equitable situation you need to effectively remove character advancement from the equation entirely and rely exclusively on player skill. While there are things to be said for doing this, there are a lot of downsides too, and for our more traditional gameplay models this is kind of like throwing out the baby with the bathwater.

    But a game world run entirely off of microtransactions that has no valid way to advance via time is just as “fair” as a game that’s run entirely off of advancement over time with no ability to purchase that advancement. Or rather it’s that they’re both equally unfair.

  11. The author explained the graphs in the text.

    Tufte groans…

  12. @Tim,

    Both your comments made me chuckle, even though they (the graphs) looked like a willing-to-pay thing to me at first glance (yay for having taken Econ 1A and 1B way back when).


    I have to strongly disagree with you on fairness. I have little personal/moral objection to the replacement of time with cost (when considered separately from the other effects RMT-type things have on games). However, and it’s a BIG however, I do object to a buyable sword being flatly better than what’s available without buying. Once a cash shop makes the leap to offering game-affecting enhancements, they are under no obligation to keep the time-channel open, and in fact would have a stronger incentive to make sure the best stuff is ONLY available for cash. I do have a very large problem with this, which I consider part of the fairness question.

  13. Oh, and as a followup thought, I’d probably be most comfortable with game-affecting cash transactions if two things were done:

    1. Purchases were solely for in-game currency, and

    2. Some upper limit was placed on how much could be bought in a given unit of time.

    This could be a smooth curve, tiered subscription model, what have you. But, this way, cash-transactions do not bar access to the same resources from players earning game currency in other, more traditional ways, and it guards against abuse of those without cash by those with cash (people attempting in-game monopoly of resources, for example). “For an extra $5 a month, your characters can have some extra cash in their pockets” is a much less obnoxious setting than what’s usually envisioned, in my opinion.

    Also, so that I’m not misunderstood, this only applies to things which have an in-game effect. Extra character slots, cosmetic stuff, that’s all fine alongside these proto-rules.

  14. @Peter, why? If the game is structured that way from the beginning, if it’s extremely *clear* about the fact that this is not a game in which you primarily advance by time, but rather one that you advance in via expenditures of real money, why would that be any less fair? I’d go as far as to say that it’d be as equally unfair as the version in which the only legal advancement is via time.

    It may not be something you’d *enjoy* but that’s fine isn’t it? There is plenty of room for games for all sorts of different types of people. They don’t all have to cater to one taste do they? You just don’t play the ones that aren’t compatible with your personal playstyles.

  15. Oh, and to clarify, I don’t think you should take a game that’s not built to account for RMT and attempt to shoe horn it in on the issue of “fairness”. I think that’d be a horribly bad idea.

    It’s simply that from a global context, looking objectively at game worlds, there’s no substantial difference between one limited resource or another in terms of how your advancement scheme works. Time isn’t inherently more fair of a metric than money is. It’s simply *a* metric, and while consistency is important for maintaining fairness, that’s only useful from the specific context of a single game world. Until you’ve got a stable playerbase, the concept of one being more fair than the other is pretty much irrelevant; there’s no difference except in how the players view them.

    That being said, I don’t think a game should necessarily try to mix the two without being very careful about it. The culture of the gamer is one that values money a lot more than time, at least in general. Any sort of shortcuts in a game that allows advancement over time will tend to be looked down on, regardless of how they’re structured.

  16. I believe there are several subtle but significant differences between advancement by money and advancement by time. And certainly, if someone were to make a game where advancement were based on money rather than time (The Merch Online, anyone?), I wouldn’t hold it against them, I’d just avoid it (went through that once already with Magic: The Gathering).

    Still, to give you an answer (and I do have one 🙂 ) is going to take a bit of formulation, so it’ll have to wait until I’m home from work.

  17. I believe there are several subtle but significant differences between advancement by money and advancement by time.

    What? “Time is money”, no longer true? 🙂

  18. Well, I await your response then, because as I see it, the only real difference between time and money is whether you have more of one or the other. Gamers tend to have more time than money, but that doesn’t mean that there aren’t potential markets that go the other way round, and for members of those markets, the necessary time commitment makes current time advancement games very unfair.

  19. The discussion brings to mind Project Entropia. I don’t know how many of you have played that game, but I can tell you this: if you don’t spend real money, it’s no fun. The entire economy is designed to shove you with unsubtle force into buying in. The constant broadcasts of players getting huge loot drops or resource extractions have the same function and effect as the flashing lights and bells of a Las Vegas casino… and they are just as reliable as a source of income (actually, if you have some understanding of probability theory, Vegas is a safer bet).

    As Entropia is still afloat, I’d have to concede that it’s a viable model. But from a player perspective? Unless you’ve got more disposable income than common sense, fuggitaboutit. It’ll never be more than a niche product, subsisting only because it caters to a well-heeled niche.

    The difference between money-based and time-based advancement is that in a time-based system, everybody can eventually max out a character. It may take me a couple years to achieve what a twinked-out powergamer can accomplish in three weeks, but with patience, it is attainable. In a money-based game, on the other hand, I know I can never compete at the same level as the guy who can buy his own island and personalized nano-armor. What incentive could exist for me to play in the same game (unless I’m being paid good money to be second-banana)?

    Peter’s comparison to Magic: The Gathering is spot on, and is the reason that the only competitive format I play in M:TG is sealed-deck… where Mr. Suitcase is forced to face me on roughly equal terms. And I confess to a certain dark relish when confronted with a Mr. Suitcase who is less than familiar with the format…

  20. CRUD! Non-wall-of-text version below. I blame Word. 😛

    Game players by definition have some surplus of time, which they spend playing games. They may not have a surplus of money (that is, time surplus is always > 0, money surplus can be = 0).

    A game is designed for time to be spent playing. Spending time is required irrespective of spending money. Put another way, if you spend only money and not time, you are not playing the game. (time cost is always > 0, money cost can be = 0)

    In the types of MMOs we are discussing, players are rewarded for time spent playing. An hour of time spent “purchases” both the immediate reward of an hour of entertainment and the cumulative reward of character advancement.

    In games that allow (or illicitly have) cash transactions, for which those transactions have a material effect on the game, a unit of money spent purchases some unit of character advancement. This advancement is usually, but is not required to be, fungible with the character advancement purchased with time. On its face, though, one is tempted to say that cash and time are equivalent, since the same rewards are available to both.

    Purchasing character advancement with cash creates an odd synergy, though. Like cheat codes in single player games (and I beg you, I am NOT making a comparison of cash = cheating), it allows the value of the game to be extracted more quickly. Put another way, for someone able to spend cash, each hour then spent buys more, both in terms of greater (more “condensed”) immediate enjoyment and in terms of more rapid time-purchased character advancement.

    This effect is what underlies the fairness issue. The hour of someone able to spend cash is worth more (it buys more) than the hour of someone unable to spend cash. When people talk about fairness, it’s often in the context that it’s “fair” for someone who spent the hours to have cooler stuff; essentially, as long as their hour is worth the same as everyone else’s, they don’t mind having fewer hours to spend.

    Ergo, the common complaint, “all that time I spent is made worthless by RMT,” is actually somewhat true. The disparity between the relative values of the hours is at the root of this, exacerbated by the feelings of someone who would have spent some amount of cash to increase the value of their own hours, but understood it to be against the rules (and thus feels punished for good behavior). (Also note that this applies to most circumstances that mess with the relative value of the hour, but cash transactions have by far the largest potential for disparity.)

    I would first conclude, then, that spending cash and time has a synergy that spending either alone does not. Since spending cash alone is not logically possible, it can safely be discarded from consideration. Since some time must be spent, the question becomes one of how big of a “multiplier” units of money can become on those hours, and how may “standard hours” a person actually has when considering both their cash-surplus and time-surplus. I’d say that there is a very real gap between a cash-surplus and an equivalent time-surplus due to synergy, and that spending money simply isn’t one-to-one equivalent with spending time.

    Put another way, you’re not just buying the hours you would have spent gaining your character traits, or waiting on random chance for powerful loot. To use WoW for this example, you’re also able to then turn around and run that high-level raid dungeon for the first time, but with all of the advantages that normally come from having seen it before, making it quicker, easier and more enjoyable (above and beyond the prior hours bought/skipped by the purchase). “Quicker, easier, and more enjoyable” are extremely subjective, but it is a way of stating that the demand for cash transactions is not solely a desire to “skip the boring parts” as some state. It also makes the fun parts more fun, in a general sense.

    I would secondarily raise the point I alluded to about fungibility. If cash can acquire some form of character advancement that is both unavailable to and better than what can be acquired with time, suddenly that “multiplier” becomes locked in. Either you’ve bought it or you haven’t. This is not functionally different than having a tiered subscription model where an extra amount per month enables 2X Gold, XP and Loot, Plus Higher Drop Chances (or such), except that it obfuscates the mechanic. Arguably, this obfuscation, particularly in regards to minors, is potentially a deceptive business practice (it would have to be a severe case, of course, but they are out there).

    Third, the above point of fungibility as it relates to competitive player interactions, where a separate, better “tier” ends up existing for those able to spend the cash (and man, I hadn’t heard the term Mr. Suitcase in years 🙂 ), raises its own set of issues. But, those are pretty clear and this is long enough as it is. 😛

    But, in parting, I’d also agree with the point that discussing “shoehorning” cash transactions into a game not designed for them is very different from a game designed to include them, and I think it’s the former that has bigger issues (both objective and emotional) than the latter.

    Oh, but as a side note, requiring a purchase to be competitive or to get comparable enjoyment or advancement out of an hour of time is not, in the abstract, different than requiring some other behavior that has an opportunity cost. There are comparisons to be made to gameplay issues such as the power Haste in old City of Heroes or to social issues such as those that have existed around the Ninja class in Final Fantasy Online since its introduction. So yeah, there’s a lot to discuss here.

  21. @Peter, if you take current markets sure. But I wasn’t discussing curret MMOGs in terms of how they work gameplay wise right now. Again, I’m not for backporting RMT into something that wasn’t built for it.

    Because spending cash alone *is* perfectly feasible in terms of advancement. You don’t get to dismiss that 🙂 Understand that advancement in every single game we play currently is 9 parts filler and 1 part actual content. Replacing the filler of need to spend more time with the filler of need to spend more money isn’t in anyway impossible. It may cater to a smaller audience, but it’s by no means difficult or wrong to implement. You simply cannot advance without spending money, and the only time you spend is on consuming real content instead of filler.

    And buying in in a competitive environment seems to work out okay (for the most part) for TCGs, so I don’t think you can even posit that. You just have to make sure the gameplay system is built around consistency and is structured to work with your metric. Because having your competition be based on who can sink the most hours into the game isn’t any better than who can sink the most money into the game. If you need to spend 20 hours a week to remain competitive in a game, there’s no substantial difference between spending half a year’s salary. Having *any* advancement metric beyond the player’s skill inherently breaks a competitive environment though, so it’s not even really worth talking about. Character advancement is anti-competition, it’s an inherent unleveling of the playing field.

    @Yukon Sam,

    The difference between money-based and time-based advancement is that in a time-based system, everybody can eventually max out a character.

    This is not true. In a money-based system everyone can also max out a character even just spending a few dollars a month, provided that the ceiling doesn’t keep moving. It’s rarely in the developer’s best interest to have a static ceiling in a game world though, so the ceiling is always moving, and even in time based situations, you cannot max out a character unless you put a certain, typically very high, amount of time into the game. Thinking that you are even remotely competitive in time based games simply because they’re time based is just not true. You’re not.

    To use WoW as an example: Less than 10% of WoW’s playerbase will *ever* finish the high end raid content in the game, and by the time they’re getting close to being able to even participate, the game will move on and adjust the ceiling upward while in the process devaluing all of the previous raid instances, meaning that they never even get to experience that content due to lack of player interest. The number of people that ever have maxed raid equipment is a tiny fraction of the playerbase, and there’s no real growth in that size as a percentage except when content gets delayed. As long as the ceiling keeps moving, the players outside of the hardcore raid elite remain safely out of reach of the top.

    The only way you can catch up is to put more time into the game than most people are capable or willing to spend… just like the money situation.

  22. @Eolirin,

    I think you think I disagree with you more than I do. 🙂 I’m positing that, in general, 1) money and time are not identical in effect when spent within an MMO setting, and that 2) the differences betwen them can (not necessarily will) raise questions about fairness. In any specific case, well, it depends on how it’s handled.

    It’s not impossible, difficult or in and of itself wrong to build for the wallet not the watch. Where I disagree is with the statement that no new or different issues are raised. Novel and distinct issues are raised, but they can certainly be addressed. As you said, you just have to make sure the gameplay system is built a certain way.

    (For the record, with the M:TG reference I was restating that I am not going to buy into that sort of game as a purely personal choice, making precisely the point that I do not object to such games and that I would simply abstain myself without thinking of them as somehow morally bad or offensive.)

    I guess my overall point is something we do both agree on: games that accept cash have to be built slightly differently than games that don’t. I say simply that it is because there are different (similar perhaps but not identical) issues raised in that context. I did discuss at length how things might be done wrong (or might simply go wrong), but that doesn’t mean I don’t think they could be done right.

    Oh, and on philosophical principle of not leaving a raised point unresolved (can you tell I enjoy debate? 😛 ), I stand by my statement that you cannot solely spend cash. Someone can certainly buy or collect Magic cards but never open or use them – this person is not playing Magic: The Gathering (and yeah, I knew a couple folks that were investing in the cards just to resell them). That would be the distinction.

  23. On your last point, because I think we’re in complete agreement on everything else, sure you can, because the money vs time thing has nothing to do with play; this is about advancement.

    And you most definitely can only allow advancement via expenditure of cash. That’s effectively what M:tG does. You can only marginally improve your skills via the expenditure of time, and if it were a less complicated game, even that’d disappear. You can run a game where you have certain activities that you can perform, and you have to buy the ability to do anything else, and I’d argue that with the right activities, you can actually make that compelling gameplay that has no advancement component at all. It’d be harder maybe, because there’s very little example of it being done, but I think it’s a very surmountable problem. I’m not entirely sure it’s *worth* doing though, but it’s very possible *to* do.

  24. […] Lum says it, does it have more weight than my articles?  Maybe Raph Koster?  I hope so, but whatever the case, the corporate beancounters will have to pull their head […]

Sorry, the comment form is closed at this time.