Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Environment and Communications References Committee
17/09/2018
Gaming microtransactions for chance based items

CAIRNS, Dr Paul, Reader in Human-Computer Interaction, University of York

ZENDLE, Dr David, Lecturer in Computer Science, York St John University

Committee met at 17 :07

Evidence was taken via videoconference

CHAIR ( Senator Steele-John ): I declare open the hearing of the Senate Environment and Communications References Committee in relation to its inquiry into gaming microtransactions for chance based items. This is a public hearing and a Hansard transcript is being made. Before the committee starts, I remind all witnesses that in giving evidence to the committee they are protected by parliamentary privilege. It is unlawful for anyone to threaten or disadvantage a witness on account of evidence they give to a committee and such action may be treated as a contempt by the Senate. It is also a contempt to give false or misleading evidence to a committee. You should also note that, while evidence to Senate committees is protected by parliamentary privilege, this protection may only apply in Australia. You should therefore be aware of the limitation of this protection.

The committee prefers all evidence to be given in public but, under the Senate's resolutions, witnesses have the right to be heard in a private session. It is important that witnesses give notice if they intend to ask to give evidence in camera. In addition, if the committee has reason to believe that evidence about to be given may reflect adversely on a person the committee may also direct that evidence be heard in private session. If a witness objects to answering a question the witness should state the grounds upon which the objection is taken and the committee will determine whether it will insist on an answer, having regard to the grounds on which it is claimed. If the committee determines to insist on an answer, a witness may request that the answer be given in camera. Such a request may, of course, also be made at any other time.

On behalf of the committee, I would like to thank all those who have made submissions, and sent representatives here, for their cooperation in this inquiry. Having said all that and taken a breath, I would now like to welcome Dr David Zendle and Dr Paul Cairns via videoconference. Information on parliamentary privilege and protection of witnesses has been provided to you. The committee has your submission, gentlemen, and I invite you to give a brief opening statement, at which point we will then ask some questions.

Dr Zendle : I will read our joint opening statement. As experts in the effects of playing video games, we welcome the opportunity to provide evidence to the committee regarding the potential for harm present in loot boxes. In our work we have looked at relationships between people spending on loot boxes and whether they have problems with gambling. We've found that loot boxes are linked to problem gambling. The worse that people's problem gambling is, the more they spend on loot boxes. We have demonstrated and replicated this relationship in studies with over 8½ thousand participants. The link between problem gambling and loot box spending is neither small nor trivial. Our research has shown that this relationship is comparable in size to links between problem gambling and important factors like alcohol dependence, drug abuse and depression.

Our professional opinion is that this relationship indicates one of two things. Loot boxes may well be acting as a gateway to problem gambling amongst gamers; hence the more gamers spend on loot boxes, the more severe their problem gambling becomes. Alternatively, it may be the case that individuals who are already problem gamblers instead tend to spend more on loot boxes. There are good reasons why that might be the case. Loot boxes share key similarities with other kinds of gambling. Since problem gambling is characterised by uncontrollable and disordered spending on gambling activities, this lack of control and excess in spending may apply to loot boxes too. Hence the more severe a gamer's problem gambling, the more they spend on loot boxes. It may even be the case that both things are true. Engaging in loot boxes leads to problem gambling amongst gamers and this problem gambling, in turn, leads to more and more spending on loot boxes, trapping gamers in a vicious cycle of gambling-related harm and disordered spending.

It's therefore our opinion that loot boxes should be regulated. Our research suggests that loot boxes either literally cause problem gambling or, alternatively, allow games companies to exploit serious gambling problems amongst their customers for massive monetary gain. It is important to remember that loot boxes are projected to generate as much as $US30 billion in revenue this year alone. The nature of regulation and how it should be applied in different regions and even for different games obviously varies in respect to existing laws and regulations that surround gambling. We feel that it's not necessary to consider loot boxes to be gambling to view them as potentially harmful. That is our opening statement.

CHAIR: Thank you, gentlemen, for that. I will start with an observation that has been made by Dr Daniel King and Professor Delfabbro who have described loot boxes as predatory monetisation schemes. Would you agree with this characterisation?

Dr Cairns : The problem is that loot boxes are many different things. In some games they are just simply a way to buy something fancy in the game that makes you look nice and in other cases they are heavily monetised with a huge range of prices. It all can be very different from just buying something nice inside a game and that's the problem. To say there's one type of loot box, I think, is difficult. To say that there are loot boxes like thatI think I would feel comfortable with that, actually.

Dr Zendle : I would concur with Paul. I'd say that not all loot boxes may be equal but certainly some of them could be characterised as predatory in nature.

CHAIR: Yes. This is an observation that we've heard previously: the importance of distinguishing between the mechanism itself and the drivers on either end of that mechanism. For instance, a loot box mechanic in the context of a group dungeon is relatively harmless but, when you have real-world currency on one end and the potential for something which has real-world value or significant social value on another, then that very much changes the dynamic.

Dr Cairns : That's right.

Dr Zendle : Yes. Our research has linked loot box spending in general, as a whole, to problem gambling. We don't have any data on how different types of loot boxes might affect people differently at this point in time; that research is under way. So, even though it seems like that, we don't have any evidence from our studies as to what types of loot boxes are worse or whether they're all potentially harmful.

CHAIR: Did your study differentiate between these mechanics functioning on, say, mobile devices as opposed to consoles or PCs and that kind of thing?

Dr Zendle : No. We looked at the overall relationship between just how much people spent on loot boxes and how severe their problem gambling is and we found an overall relationship. I think the literature is just that it's beginning with loot boxes, which is one of the things that make your decision-making very hard. There is very little evidence for you to go on. Certainly when it comes to empirical studies there is very little. We found an overall relationship. It will take months, if not years, for the literature to gain the nuances that you're talking about and be able to inform you in any empirical way.

Dr Cairns : Can I just add that we did look across games. We deliberately targeted players from a range of games where loot boxes were known to be present. It is a case of, whilst we don't know which particular platforms are worse or better in terms of loot box spending, it was across a very wide variety of games.

CHAIR: A lot?

Dr Zendle : Yes, it was a lot.

Dr Cairns : It was sort of hundreds of types of games.

Dr Zendle : Yes.

CHAIR: Are you aware, yourselves, of the steps being taken in Belgium and also the Netherlands in response to these issues? Do you have a view on the regulatory approach that has been taken there or suggested there?

Dr Zendle : Paul and I are experts in the harm that loot boxes may or may not cause. We could tell you all about how they may or may not lead to changes in society. We are aware of the legislation that is going on in Belgium and the Netherlands, and we will have opinions on it that you can have from us, but we don't have anything based on the data to recommend whether the approach taken in Belgium or the Netherlands is the correct one or not. It's our opinion that regulation is appropriate.

CHAIR: Can you expand for us upon the harm side of things then? Did you look at things in your studies such as average amount spent by problem gamblers on loot boxes?

Dr Zendle : Yes. One of the interesting things about the data—we have run two studies. The first study was run on about 7½ thousand gamers, and the second study was run on about a thousand gamers and replicated the results. We got the same thing both times, which is always nice to see in science, because it suggests that the effect you are seeing in the world is real and it is robust. The first time we measured categories of spending. We asked people: 'Do you spend less than a dollar? Do you spend between $1 and $5? Do you spend between $5 and $10?' That makes it harder for me to make absolute statements from that data about how much they're spending because it is divided up into categories.

In the second study we asked directly: 'How much are you spending in dollars? Give us the absolute amount.' I can make statements about differences in spending. If memory serves me—the data is publicly available—it is about a $10 to $15 difference per month in spending on average between problem gamblers and non-problem gamblers in that study.

Initially that might not seem much but there are a couple of things to remember. First, the way we take averages in that study discounts the effects of very, very extreme problem gamblers. We saw five or six people within that sample who were claiming to spend $2,000 or so a month. In general, we see about one per cent of the people in each of our studies is spending $300 or so, or upwards, per month on loot boxes. So you have got this long tail in the data where at the end you have a group of people who are spending really, really large amounts.

The average between the groups isn't that much. It was just that if you put something into place to protect the people who are spending extreme amounts, then for the other people it might not be too bad. However, that is only a problem if the thing you care about is the spending on the games. If you're saying, 'Oh, the effect that we care about protecting people from is overspending on loot boxes,' that's what we're talking about. However, if you are talking about loot boxes acting as a gateway to problem gambling it doesn't matter how much they are spending on the loot boxes.

CHAIR: Yes, of course.

Dr Zendle : The fact that they are linked is the important thing, not the amount they're spending on it. The fact that only a $10 spend might act as the difference between these groups is potentially more worrying.

CHAIR: Yes, of course. Can I ask: in your interactions with the people that you worked with on this study, what was their self-perception of the expenditure that they were giving in these processes? Did they perceive that they had spent, say, $15, or however much money, and got value for money? Or did they feel that they were participating in something that was negative for them?

Dr Zendle : That's a very interesting question. We have data on why people bought loot boxes and how they felt about it—whether they felt positive or negative about it. We haven't done that analysis yet. From eyeballing the data—this is in a published form or even in a pre-print over—the one thing I could say is that there appeared to be a variety of different reasons why people bought loot boxes, ranging from being better at the game, equalising themselves with other players, to looking good and social reasons, and that a surprising amount of gamers' opinions about loot boxes was negative; they talked about them quite negatively. That is about as much as I can say on that subject.

CHAIR: And given your background—I imagine you would be very familiar with the counterargument from the industry that these things are no different than Kinder eggs or trading cards, that kind of line—do you have a view on that? Has your data given you anything to suggest empirically how rubbish that is?

Dr Cairns : Our data doesn't have that. However, from what we are in the world, there is a very big thing in science at the moment called big data. And the thing that computers allow us to do is to do things differently because of the speed and quality with which we can process data. It is characterised by various things but usually lots of Vs. Two of the Vs in big data are velocity and volume. What computers can do is do things quickly and in large quantities. So if I wanted to go and buy a Kinder egg I have to pop down to the shop and buy a Kinder egg, or I can buy a box full of Kinder eggs. But once I've spent that and opened them, I'd have to go back to a shop again, and that slows things down, because it is a physical action. When I am on a computer, I can keep pressing 'buy' at a rate as fast as my finger can click. So there is velocity is there. And of course the volume is that I can spend as much as I think is reasonable as well. The physical world puts natural barriers in the way of people's behaviour, which makes life more complicated. As a point to intervene in, say, gambling there are regulations around what betting shops look like, in order to regulate that. There is no such thing in the virtual world. If I was an addicted book reader I could buy a book a minute on Amazon without any problem whatsoever. Nothing would stop me. And it is the same with loot boxes and games. There's nothing stopping people spending at that volume and that velocity. It is a difference in nature, not a difference in quality.

CHAIR: I don't want to hog all the time. I might pass to my colleagues.

Senator DUNIAM: Thank you gentlemen for joining us. I want to go back to the first question the Chair was asking and, in light of the recommendations you have in the submission you provided, just to go through them, which talk about placing a parental adviser on a game that features a loot box—a descriptor outlining that the game itself features gambling content and then further recommending that serious consideration is given to restricting games with loot boxes to players of legal gambling age. I may have misunderstood or not properly followed. I think the term is I am a noob in this area. But I just wanted to understand, given this distinction between types of loot boxes, some might be less insidious—let's use that term—than others. In terms of the recommendations you provide—and I note what you said before about those nuances will be borne out by extensive research over perhaps many years—but in light of that and your recommendations, do you have anything to add?

Dr Cairns : I guess the thing is that it's like we are in a brave new world of digital games and things going on there; so you've got to draw from what we already understand. For instance, with films, you can get a Disney film which can have violence in it but it is toned down. The Lion King has a fight between lions. Nothing terrible seems to happen. It's okay. Then you have got—I don't know; I don't even begin to watch those sorts of films—the slasher-gore horror films, and of course you regulate those. They are very different in nature. But in some sense they depict violence.

What you have got to do is think: 'Well, what's the range of activities here? And what is it that is considered acceptable?' That is what parental guidance or a rating system does, it puts in that nuance. It says, 'Okay, you can buy things in this,' but it doesn't look like that. It is not random stuff. You buy things, you'll get things back; whereas in other cases you might say, 'This is unrestricted loot box spending, you can spend more than $500 a month on this and rewards have this probabilistic distribution.' Then you are in a position of saying, 'Well, that looks really like gambling.' You might want to consider very carefully about putting a game like that in front of your children or what have you.

Senator DUNIAM: So you are talking about scaling it? Is that the point you are making? In this game it is at this standard, but in this game over here it is more costly and perhaps higher risk, more like gambling?

Dr Cairns : I would say yes. The problem is: what we don't know is what the kudos of scale is. Of course the film industry has had 80-odd years to fine-tune it, and it is always fiddling with it. No-one would say it has got it completely right. We are at that really early stage of saying, 'Well, perhaps we could do something.' But what that might look like, I don't know. David.

Dr Zendle : I will follow on from what Paul was saying. There are different kinds of loot boxes. One of the big distinctions you can make is pay-to-win loot boxes. You can say, 'Does this thing I am potentially getting from the loot box have some sort of capacitor value in the game I am playing?' Or you could make the distinction of loot boxes where you can cash out. 'Is there some sort of grey market where I can sell the things that I am getting in the loot boxes for real-world money?' And you can say, and people have said—I imagine you probably had submissions from people—'Some of these things are worse than others.' Our data doesn't tell you about that. Our data tells you that there's an overall link between loot boxes in general and problem gambling in general. Any recommendations we make will be our opinions and not based on that data. In a few months time we will have some data for you and we will be able to come back to you and say, 'Look, these ones are very strongly linked, these ones are less strongly linked.' But right now all we know is that there is an overall relationship.

Senator CHISHOLM: Did you have any findings in terms of the data around demographics? Was it particular age groups that were more susceptible to the loot boxes than others?

Dr Zendle : We haven't done that analysis but I would say that the data that we have taken is typical of data of games that you tend to get, which is skewed young and skewed male. It seems like it should generalise to gamers in general, but that's the sample we've got. We didn't do any subgroup analyses to look at, say, if particularly young gamers are particularly affected by it. Another thing I would like to note is that all the research that we've done has been on people aged 18 or over. Mechanisms may operate differently for younger people.

Dr Cairns : It is also worth mentioning that with the problem gamblers of course—we have a nice, large sample of 7,000 people in one of our studies but even then the incidence of problem gambling is low, typically in the population it is sort of one to two per cent—we are finding exactly that in our study: a detailed analysis, as it turns out, of about 130 people who have a severe problem with gambling. We are starting to get the level where you might just get chance variations due to how we have recruited participants. It is quite a small sample as opposed to, say, a population sample of the problem gamblers. But that's just the nature of finding people like that.

Senator CHISHOLM: I suppose it is expanding the problem gambling market, for lack of a better phrase, though. You are giving more people the opportunity to get hooked, therefore problem gambling becomes a bigger problem.

Dr Zendle : You are either giving more people an opportunity to be hooked or you are giving people who are already hooked, who are already problem gamblers, more opportunity to spend. We don't know which one it is. It could be both.

Senator CHISHOLM: I suppose the issue with gaming is that if they are socially isolated and don't want to go to a pub and play pokies, for example, which is in an Australian context, but stay home and gamble and do it in their own privacy without anyone noticing, it could lead to problems.

Dr Cairns : There is a lovely work by a lady named Natasha Dow Schull, where she looked at—what do you call one-armed bandits?—slot machines. She said across the population the problem gambling is about one or two per cent, which is what you see unless you have regular access to slot machines. When you have regular access to slot machines, the problem gambling around slot machines gets to around 25 to 30 per cent. There is an example of a gambling situation where you make the mechanisms available for people and then the problem gambling goes up. There's the chicken-and-egg situation: are the machines put there because they will get the gamblers, or the gamblers become gamblers because of the machines? You can see where there is the opportunity problem gambling goes up beyond the normal parameters of the population. The question then—which we don't have the answer to but you need to think about it—is: are loot boxes something like slot machines, which if you get the features right will drain you dry in, I think, an evil way?

CHAIR: Do you subscribe to the view that fundamentally many, many types of them function using the same variable repeating ratio that sits behind a lot of things like poker machines?

Dr Zendle : It is my opinion that that is definitely true.

Dr Cairns : I would assume so. It is evil in my view. But the research in slot machines is very clear. It's highly effective if you get those ratios right in what's called offering a smooth ride to extinction; in other words, literally taking all the money off the gambler. They worked over decades to get these proportions right and to get the balance right in order to monetise slot machines. My guess is that it may be early days of loot boxes but there are people looking at these analytics, and if their job is to increase monetisation they will be doing exactly the same thing in the loot box context.

CHAIR: Which, I guess, is not inherently evil, given that the job of any and every game designer is to make people want to play their game. The danger is that if you put money on one end or both ends of this, then the harm is created through the monetary loss, but also you are observing that, regardless of the monetary loss, you are actually building in a behavioural pattern. Is that also what you were observing earlier?

Dr Cairns : Obviously we don't know, from our data. But yes, if we are again drawing from a domain which is well understood, like slot machines, that would be entirely reasonable to move that way in loot boxes, even though it may be very early days and it is not at the same industrial level at which slot machine developers are working.

Dr Zendle : It's worth noting that one of the main theoretical pathways to problem gambling is via that kind of conditioning. It's getting used to and learning to expect that kind of feedback. So there are good reasons why loot boxes might cause problem gambling. From our data we don't know if they cause problem gambling or if they attract problem gamblers, but there are good reasons to think that they might.

CHAIR: Have you got a time line on your additional research?

Dr Zendle : Yes. I would expect a pre-print of a paper showing different effects of different kinds of games within the next month or so.

CHAIR: Oh, that is tantalising!

Dr Zendle : Then I imagine another thing we would like to do is a longitudinal study where we have a look at how problem gambling develops over time amongst people buying loot boxes, because that could help us work out the causal argument—whether it's causing it or whether it's just an effect of problem gamblers being drawn to loot boxes. That's more speculative. I would imagine that's within six months, but I would say a hard copy within two months or a pre-print of how different games affect things differently, because we have the data. We just have to analyse it.

CHAIR: Would you be willing to provide it to the committee in that preprint form?

Dr Zendle : Of course we would. We would be happy to share it.

CHAIR: That may be a question of our time lines more than yours. But it seems that the critical piece of clarifying information is just about to arrive. Thank you for that, gentlemen.

Senator DUNIAM: Are you asking for the report?

CHAIR: We could do it, couldn't we?

Senator DUNIAM: Yes.

CHAIR: Our report date is 17 October.

Dr Zendle : We will make it by the 17th, and that is a solid promise.

CHAIR: Could I ask you officially to provide that information to the committee on notice?

Dr Zendle : Yes, you can.

CHAIR: Gentlemen, is there anything further you would like to illuminate us on?

Dr Cairns : No, I don't think so, unless you have further questions. No.

CHAIR: Thank you so much for your time.

Committee adjourned at 17 : 36