When it comes to mobile social-gaming, nowhere is it as popular or as profitable as it is in Japan. The nation’s two industry giants, Gree and its rival DeNA’s Mobage, have been increasing their sales every quarter for years. Gree’s TV-commercial campaigns are second only to cosmetics company Shiseido in the amount of money spent on advertising (Kao is No. 3, P&G No. 4 and Kirin No. 5), which shows just how large the social-gaming industry has grown.
However, during the Golden Week holidays trouble began to brew for both Gree and DeNa. In a Yomiuri newspaper scoop on May 5, it was revealed that one of the biggest earners for both companies — kompu gacha, a virtual card-collection system based on a paid lucky-dip-style gaming called gacha — is now under threat due to questioning by The Consumer Affairs Agency (CAA) over its legality. On May 7, when the stock market reopened, traders reacted strongly against the social game providers — many shareholders lost 20 percent to 25 percent of their share value by the end of the day.
As a result, on May 9 and 10 most social-game providers, platform owners including Gree and DeNA (Mobage), and many other third parties rapidly announced their decisions to kill the kompu gacha feature by the end of May.
So what exactly is gacha and kompu gacha? In social card games, users collect virtual cards to advance to the next level (scene, scenario). Some cards are given away free during the play and some you purchase if you really want to win and/or take short-cuts. Drawing an unknown gacha card costs ¥300 in many games, which give you a randomly chosen one. Kompu gacha (complete gacha) is a system involving the collection of gacha cards. If you collect a certain set of cards, you will get another rare, valuable card.
On April 24, the CAA secretary stated at a regular briefing that kompu gacha could be thought of as “prizes” under the prize regulations (Law for Preventing Unjustifiable Extras or Unexpected Benefit and Misleading Representation).
Yomiuri’s May 5 scoop, which was then taken up by many other media outlets, was based on this secretary’s comment. Although the CAA never identified any companies by name, news of its intention to investigate whether kompu gacha schemes are illegal or not traveled quickly to stockholders.
If banned, some analysts suspect the loss of kompu gacha could decrease the value of some companies by more than 20 percent. But how does kompu gacha make such huge profits? There is both math and a psychological trick at play behind the scenes.
Let’s experiment with a simplified model: If there are 20 virtual cards and you get all 20, you will win a stage of the game. If you draw a random card by paying ¥300, how much can you expect to spend to complete the set?
If you are weak at math, as a child may be, you might answer ¥6,000 — which is already a lot of money. However, this figure does not factor in duplicate cards. You will (almost certainly) get cards that you already have. The more cards you get, the chance of getting a new card will decrease. To gain all 20, you actually will need to draw, on average, 70 times, which costs ¥21,000.
The real kompu gacha, however, is even more complex than this. Let’s say for example that among the 20, you only need to gain 8 target cards to win something. As you do not need to collect the other 12 cards, this sounds easy — but nothing could be further from the truth. The odds of getting most of the targeted cards will usually be, say, 5 percent. However, for two of the target cards the odds will only be half that, that is 2.5 percent.
This variation means you have to draw gacha cards a few hundred times to get the 8 cards needed, which means ¥100,000 or more. The lower-chance cards work against the user to make drawing future cards statistically very difficult. With this setting, you get half of the target cards very early, which encourages users to keep trying and spending to try to get the rest. Some players have noticed that some cards are rarer than others and share that info on BBS and Wiki. But if the provider were to change the odds slightly — for example to less than 2.5 percent — it can control how much users spend, which can be as much as ¥200,000 or ¥300,000 per set of cards.
In conclusion, kompu gacha is a system in which it is easy for players to make erroneous assumptions on probability. Players have success early on in the games, but by aiming for completion, the chances of getting the cards you need become slimmer and slimmer.
The games themselves are offered “for free,” and the majority of players do not pay for such digital items. However, within free-play you may get some rare cards — for example some games offer free gacha cards once a day, or others let you draw cards in certain game “events.” When a user notices that they already have half of the needed cards without paying, it is tempting to pay for more gacha cards to finish the game.
And once you start paying you’re hooked, as you do not want to lose what you have already paid. It is called the “sunk cost effect” in psychological terms. Once you have invested, you are often reluctant to pull out, because you feel the money you spent will be totally wasted if you stop.
The worst thing is, game providers do not make the odds public. Unless the real numbers are given, players will never know the real probability even if they understand the math in my examples above.
In a follow-up report on May 14, the Yomiuri interviewed a 30-year-old social-game engineer who revealed that he routinely adjusts odds based on the transition of real-time item sales, the number of playing users and so on. The engineer, who requested to keep both himself and the company anonymous, said to Yomiuri that the success of social games depends on how to keep the addicted paying users playing longer, without exploiting them so much that they give up.
As all the data is on the side of the provider’s’ Web server, it is impossible to know whether they actually do control the odds, but the possibility of dynamic odds adjustment has always been rumored. Although the engineer in the Yomiuri article spoke on condition of anonymity, I don’t believe that interview was a setup. Actually, some game providers have told industry conferences that they modify game events and items in real time by checking customers’ spending. If those modifications affect cards and their chances — even if they are not manipulated for single customers, but by global adjustment — it is manipulation.
So why has this issue been overlooked until now?
The particular clause in the Consumer Affairs Agency regulations that applies to this issue was added in 1977. At that time, actual physical baseball- and anime-themed trading cards were all the rage among children, and these cards used the same “set-completion” technique to encourage kids to keep buying the cards. So a regulation was introduced to ban the technique, and stop kids from spending too much money.
When all the kompu-gacha providers were rushing to stop such schemes on May 9-10, Susumu Fujita, CEO of Cyber Agent, which runs several popular card games, told the Nikkei that he never knew that section of the law — believe it or not. Though I suppose the law was made 35 years ago, and for a different technology, the system is basically identical. When sales are growing by billions of yen, you’d think a company would check things with its legal section.
Another way of rationalizing the kompu-gacha system is the logic that players do not purchase the digital data, but simply “rent” them from game companies. Therefore the 1977 law does not apply. When antecessors reported great growth with kompu gacha, almost all providers jumped on the bandwagon.
Under this ban, will businesses simply return to pre-kompu-gacha-era systems? I do not think so. The problem at this time is only with kompu gacha, which involves the collection of cards in a gambling-like environment where odds are not revealed or officially regulated. So, you probably will not see typical kompu-gacha games under self-regulation. But by their nature, game scenarios tend to have “if you have done A, B, C and D, then you can proceed to E”-type structures, and if game-play and payment use regular gacha (lucky-dip) systems then game companies will not be violating the ban. Vendors who were making easy money on kompu gacha may seek a semi-kompu gacha system that skirts the edges of regulation.
The way the CAA dragged this old regulation from its pocket could be an ingenious way to answer increasing complaints from consumers. At this point, the CAA has not really invoked the regulation, as it never named a single provider. Instead, social-game companies jumped in to self-regulate before the storm got too big. But if customers keep complaining, not only the CAA but other governmental bodies may try to intervene in this business model.