Study Shows Almost Half Of Loot Boxes In Games Constitute Gambling

The Australian Senate voted to conduct an inquiry into whether purchasable random rewards in video games (known colloquially as loot boxes) constitute a form of gambling and whether they are appropriate for younger players.

Our recent paper, which was cited in the senate motion, explores exactly these questions.

We found that the loot boxes in almost half (45%) of the 22 games we analysed met the criteria to be considered psychologically similar to gambling, even though they are rated as appropriate for adolescent players under the age of consent for gambling.

What is a loot box?

Loot boxes are digital containers of randomised rewards, and are available in a number of video games. The box may contain rewards ranging from cosmetic items which alter the appearance of in-game characters to functional items that increase the player’s power in some way (for example a gun that fires faster or does more damage).

In our research, we sought to answer two questions: are loot boxes like gambling and, if so, what should be done about it?

First up, we want to clarify that video games are not evil. Games companies are not evil. Making money from video games is not evil. And playing video games with loot boxes is unlikely to result in young people flocking in great numbers to casinos.

However, simultaneously, it may also be true that loot boxes represent a troubling and potentially inappropriate monetisation strategy, with the potential to cause short and long-term harm to some players. Our intent is to educate readers about loot box mechanisms, and promote a reasoned, evidence-based discussion about ethical practice in video games.

Loot box rewards may be highly desirable or valuable (for example, a particularly valuable cosmetic item or very powerful weapon), or virtually useless and undesirable (items referred to as “vender trash”). Most importantly, the contents of the box are determined by chance. Some (but not all) loot boxes are purchasable for real money. In some cases, items earned from a loot box can also be “cashed out” for real world money.

The gambling problem

The problem is that spending real money on a chance outcome that results in some people “winning” and others “losing” is fundamental to gambling activities. Thus, we analysed the loot box features in 22 console and PC games released in 2016 and 2017, with a view to understanding how psychologically similar they were to gambling.

We used five criteria to distinguish gambling from other risk-taking activities. These have been developed by University of Trent psychologist Mark Griffiths in his work on behavioural addictions and gambling disorders. To be considered psychologically similar to gambling, loot boxes must involve:

  • an exchange of money or valuable goods takes place
  • an unknown future event determines the exchange
  • chance at least partly determining the outcome
  • non-participation avoiding incurring losses
  • winners gaining at the sole expense of losers.

We took a reasonably strict interpretation of the final criterion; assuming that people only “won” if they gained some form of in-game competitive advantage (for example more powerful weapons). Arguably, this approach ignores the subjective value that might be created by the scarcity of, or player preference for, certain cosmetic items. However, it appeared to us to most closely resemble Griffiths’ intent.

Loot boxes in just under half of the games (45%) met all five of Griffiths’ criteria and, thus, could be considered psychologically akin to gambling.

All of the loot boxes operated on a variable ratio reinforcement schedule – a technical term for a reward given to a person on average every so many times they engage in a particular behaviour. This type of reward schedule results in people quickly learning new behaviours (for example buying loot boxes) and repeating them often in the hope of receiving a reward. The strategy is effective because the next time a box is opened it might be the “big win”.

Adolescent gamers face a higher risk of developing problematic gaming behaviours if they are exposed to gambling at a young age. Lauren C. Hall, CC BY-SA

Perhaps most concerning was the fact that at least five of the games had mechanisms available to on-sell virtual items, allowing players to cash out their winnings (though four of these five had terms and conditions explicitly prohibiting this).

The ability to cash out winnings is something that many consider a legal requirement for an activity to be considered gambling. Although the legality of loot boxes is a question for individual regulators and governments, exposure to mechanisms which closely mimic gambling in a psychological sense is concerning to us, especially since all of the games we examined were rated as appropriate for those under the age of consent for gambling.

The short and long-term consequences of engaging with these mechanisms are unknown. Plausibly, short-term consequences may include overspending on loot boxes. The potential for long-term consequences also concerns us because males (a particularly large group within gamers) exposed to gambling when young are particularly at risk of developing problematic gaming behaviours.

What to do about it

Middle-earth: Shadow of War announced it would be patching its loot boxes and microtransactions out next month, following backlash against its implementation.

There is cause for hope. Electronic Arts (one of the largest game studios in the world) has recently announced the removal of loot boxes from upcoming titles. This suggests the games industry is taking consumer and expert feedback seriously, and may take steps to self-regulate.

In our view, this is the optimal solution, given the diverse policy landscapes across the countries in which video games are sold.

Where industry is not willing to self-regulate, and loot boxes are most similar to gambling, regulators may need to consider additional steps, although this should be undertaken selectively. Belgium and the Netherlands have declared at least some loot boxes to be illegal, while the US and UK have decided that they are not a form of gambling.

As noted above, the Australian Senate unanimously supported a vote on the 28th of June to refer an inquiry into the legality of loot boxes in video games to the Environment and Communications References Committee.

Most importantly, we recommend that loot box mechanics should be added to content warnings to give users and parents the information they need to properly assess whether particular games are appropriate for themselves or their children. Ensuring that users can make well informed decisions about the appropriateness of content remains one of the strongest consumer defences.

We hope that this work will form the basis for a well-reasoned, evidence-based policy discussion about ethical and sustainable practices in video games. Our intent is not to stigmatise games or gamers, but to spark a discussion about what mechanisms are and are not appropriate for particular audiences, games and the industry more broadly.

This article was written by Aaron Drummond, Lecturer in Psychology, Massey University and James Sauer, Senior Lecturer in Psychology, University of Tasmania. It first appeared on The Conversation. Read the original article.


    Without loot boxes, bye bye Injustice 2 mobile! lol

      Strangely enough, that wasn't one of the 45% they concluded was gambling. Neither was Battlefront 2, the game that started all this.

      CoD:IW was gambling, while CoD:WW2 wasnt. Interesting list, and importantly most of the games looked at were rated 17+, which I think is appropriate.

      Its worth clicking through to their paper, its only 3 pages long and theres a good table on page 2 laying all the games out.

        Battlefront 2 was included originally - as was Shadow of War - but they pulled it after changes to the microtransactions in both those games.

          That makes sense. Seeing things like this though only reinforces my worry that the end result isn't going to be what people expect.

          10 of those 22 games tick all 5 boxes they looked at (that's the 45%), but 11 others ticked 4 of the 5. That's 21 out of 22, or 95.5% and they aren't small name games either. Its not a stretch to see any rule change basing on hitting 4 out of 5.

          I worry that we'll end up with a process that only reinforces our nanny state attitude with gaming, and doesn't really do anything to solve the problem itself. It hits sports games like FIFA and Madden. Is that where it should be heading?

            Hopefully the government actually reads the report and understands why the researchers drew the line they did. All five criteria are needed for it to be gambling by the most accepted standard, the ones with three or four are missing essential parts of the definition.

            Of course, I still advocate for a separate regulatory approach to this stuff, not pigeonholing it in existing gambling definitions and regulations.

              As a career public servant, hope is something I have little hope with this. Good intentions have a habit of ending up with bad processes, just look at We Happy Few failing its drug test, South Park having a koala tell us a story, or Fallout 3 needing a ridiculous bandaid solution that really did nothing different.

              Because there are politics involved here, there are compromises you should expect are going to be made. Like getting the Nick Xenophon Team vote in the senate...Right there, its easy to see this ending up being based on the wider 4 out of 5 net.

              This isn't an easy thing to sort out.

                I agree, it's not easy. I share your concern that the government will try to go the easy (and wrong) route, which is to just slap a 'gambling' sticker on any and all loot boxes and call it a day. It needs a more nuanced touch than that.

    One of the more ludicrous arguments about why some people consider some loot boxes ok is that they contain ‘cosmetics only’, when they would tend to drive a lot of the sales, and many mobage games work purely on cosmetics because a large amount of players of all ages would keep throwing money at the games just to try and get that one character or one skin.

    Also, even if the items themselves aren’t sellable, people do sell accounts based off the loot box rewards they’ve obtained.

    Last edited 29/06/18 12:08 pm

      It's not a ludicrous argument though, it's one of Griffiths' five criteria for defining gambling. Cosmetic rewards give no competitive advantage, so winning is not at the sole expense of losers.

      Selling accounts is against TOS in all the inspected games. The onus is on the person in that case, not the vendor, the same way a bottle shop can't be held accountable if you buy booze from them and then give it to someone underage.

        It was meant to be a comment on the defence at large I’ve seen across the internet, but it really highlights that the standard definitions of gambling really don’t fully capture the nature of mobage and loot boxes in general, and that it should be broadened.

          No, it definitely should not be broadened. The definitions people come up for in their own heads while discussing this end up inadvertently including things like TCG card packs, Kinder Surprise style toys, lucky dips and toy capsule machines. This needs separate regulation tailored for the subject matter, not being shoehorned into gambling laws.

            I mentioned in another thread, that at least some TCG card packs could be considered gambling and easily cost people far too much money hunting for a single rare card. One of the big differences though, is that they (or at least some of them) display the actual chance you have of winning. Like "1 rare card in 10 packs" or similar.

              I refuse to accept any definition of gambling that includes TCG card packs, that's plain and simple overreach. Whether odds are displayed isn't in the list of criteria for or against the definition of gambling used in the study.

                Just because it's not listed in the study doesn't mean it's invalid, or irrelevant to research (or potentially legislation). Potentially their definition should include "purpose" or "intent".

                The reason I believe some TCGs are gambling is that some of them had such valuable cards people were buying the packs not because they play MtG or WoW or whatever, but because they were trying to get the rare cards to resell them. When you have super rare cards that are (were) selling for thousands, even tens of thousands of dollars.


                I think if you have any form of chance (be in loot boxes, lucky dip, TCG packs or whatever) where they have a rare item that is *so desirable* that people who don't even play the game are actually treating it as a kind of investment then it's tipped over into gambling.

                  I am saying that it's irrelevant to the definition of gambling. It's not relevant to Griffiths' definition of gambling, it's not relevant to the dictionary or legal definitions of gambling.

                  Published odds are good information for consumers, but that's all it is - information. It doesn't decide if something is gambling or not, including in China where published odds are part of the regulations for online video gaming. Plenty of games where the odds are known are still gambling, like roulette and craps. Knowing the list of odds has nothing to do with whether something is gambling.

                  Rare MTG cards are worth what they're worth because they're old, in mint condition, and sourced from a small print run. The odds of getting a Black Lotus in an alpha or beta booster pack were no different from getting any other rare card in those same packs, the only reason their value has changed since is because those packs are no longer in print. That is not, nor should it ever be, a factor in determining whether something is gambling at the time it's sold.

                  I stand by my earlier statement. TCG card packs are not gambling and classifying them as such, including your own attempt to do so, is overreach.

                  I notice you didn't comment on the WOW TCG cards. Their prices were highest when they were new (spectral tigers have dropped in price now). It's still a valid point.

                  Swift Spectral Tiger's bid value is higher now than it has ever been, and has consistently gone up for the last 5 years. Its buyout value is about three times higher now than it was when it launched.

                  The attraction (and price) for Swift Spectral Tiger is the same as for Black Lotus, or pick any classic car you like that's not being manufactured any more - collectors are willing to pay a premium for the sake of completion. The price is driven pretty much entirely by availability, the randomised nature of cards in the booster packs is orthogonal - it'd be as high today even if Blizzard had released it as a guaranteed but limited run item.

                  Case in point, sealed copies of the vanilla collectors edition are $4500-5000, despite having nothing to do with gambling at all.

        I'm not entirely sure that the last criteria is entirely appropriate though. Does someone else need to lose or just the gambler? Take poker machines as an example, they are specifically programmed that the club/pub/casino won't lose. Sure they might lose on a single push but overall they're designed that the only person who can lose is the person playing it.

          Someone else needs to lose. In the case of poker machines, other gamblers are the ones losing.

            That's not actually the case. Or rather that's not necessarily the case. I could literally be the only person who ever plays a poker machine. And still lose money on it. The machines are designed to only average a return that's a loss (about 80% from memory). It's irrelevant whether it's just me playing it or twenty different people.

            Another example is scratchie tickets, me buying one has no effect on anyone else's ticket (other than the fact my ticket is taken out of the pool).

              It is the case, I think you're misreading the criteria. It doesn't say you can't lose or that net losses can't be higher than winnings, it says winners gain at the sole expense of losers. The entirety of winnings are funded by losers over the life of a poker machine. The same is true for scratch-and-win tickets, winnings are funded by the people who bought tickets and won nothing.

    Im surprised that a lot of the mobile based gatcha games are not really mentioned in here. Things like
    - Love Live School Idol Festival
    - Honkai Impact
    - Dragon Ball Z
    - Final Fantasy
    - Fate/Stay mobile

    Is it just because the user base is so much lower that these are not getting a mention?

      No, its because the authors limited their study to a limited number of high profile AAA games to highlight, not every game out there. Its about establishing what the problem is, not what games are problems.

      If you want a list of what games have lootboxes, its going to be a looooooong list.

      They are noe open to public submissions now if you'd like to pit those games forward.

    I've always understood gambling to involve a risk of loss.

    Since there's no such risk with loot boxes, as you're guaranteed to always get something, then they're not akin to gambling.

    Sure, there's a compulsion to keep buying them until you get a specific thing you want, but that's still not gambling.

      The study seems to consider the purchase cost as the 'loss', which I think is a long bow personally, but that's their interpretation. They do point out that not all risk-taking behaviour is gambling, a distinction I think more people discussing this topic over the last few months need to understand.

        I think that's exactly right. If you weren't paying for the loot box you couldn't incur a loss and there'd be no issue. That's why stuff like random loot from a boss, or the Christmas presents in WOW aren't an issue.

        As for the idea that @comixfan puts forward that you always get something, that's true but you have to look more closely at it. Look at a scenario where you have 5 different items in a loot box. When you buy the box there might be a 24% chance of it being one of the first four (common items) and a 4% of it being the last (rare item). It could (and is likely to) take 20+ loot boxes to get the rare item.

        So while the other items are "stuff you get" after the first time they're probably useless and potentially valueless. Especially if they're non-tradeable. In that case you've made 20 purchases and got 5 useful items the other 15 purchases were wasted. And that's assuming the four common items have any value to you whatsoever.

        Obviously, if the rarity increases the more loot boxes you're likely to have to buy (statistically speaking) and the more money you spend and effectively waste.

        Bearing that in mind there are solutions other than simply banning lootboxes or limiting them to 18+ games. For example, you could make the items self-limiting. ie: once you've got an item you can't get it again until you've gotten all the other possible items. So in the same "five items" example you are guaranteed to get one of every item after you've bought five boxes. While that's still technically gambling it limits the potential loss to reasonable levels.

        Another solution is to make all the items similar "worth" and allow them to be traded. eg: I really wanted the blue hat but I got the red one. But with a large playerbase someone will trade me a red one for a blue one.

        And I'm sure there are other possible solutions that could be uncovered with a little creative thinking.

          All digital items are valueless in the absence of a means to trade or cash out. It's impossible to apply an objective standard of value based on the subjective wishes of each individual purchaser taken in isolation, and there's no way to evaluate value across multiple purchasers without a common point of reference (ie. external currency).

            I don't think that's a valid argument since the same can be said for physical goods. A loaf of bread is valueless if you can't sell it to someone else.

            A better way to look at it is they're not valueless if they have a real world cost to obtain them (not talking time/labour - just dollars). Take a mount of pet in WOW that you purchase through their store, they have a literal, real world dollar value because you pay real money to get them. As do the loot boxes in games.

            You'd also need to define value, after all that means different things to different people. Prestige, status and popularity are all important and effectively a form of value to certain people. So wealth isn't the only measure of value. In fact for some people it's far more important to be popular than it is to be wealthy. Obviously this is a problem when dealing with younger kids because most of them understand popularity better then wealth (or rather they want to be popular and don't understand money).

            Probably a better way to discuss this is to use two separate terms, VALUE - which is the perceived value/desirability of an item and COST - which is what it costs in real dollars to obtain said item. I think it's possible to monitor VALUE and then map that figure to COST.

              Replace all instances of 'value' with 'monetary worth' if you want to be pedantic, a loaf of bread would have no monetary worth if it couldn't be sold but it can be sold so it's a moot comparison. The loot box itself has a monetary worth, its contents do not. I stand by my point, subjective value is impossible to assess without a common point of reference. I do not agree with your assertion that such a reference can be mapped to monetary worth.

      If you pay $1 to play the slots, but it always pays out $0.05 even you lose, is it gambling? You're always guaranteed to get something.

        Lootboxes and slots are not the same thing, though.

        With lootboxes you might not get what you want but you still get x amount of items for the money you've paid. You haven't lost any money like you do with slots.

    I think we sort of missed the point with this loot box deal though. As it stands all we seem to gain is a gambling label on games that are already currently rated at mature audiences which is essentially a mute point. Just slap it under violence on our shooters and no one could bat an eye.
    There were several seperate but directly related things that we wanted to change, which i believe individually hold no strength but the combination of them which is truly disasterous:

    Rng chances at items we pay money for. (Both cosmetic and ptw especially when its "fluffed" with useless gains).

    Low conversion for game time to ingame currency (unrealistic time invested incentivised real world spending).

    The way lootbox "wins" relate to gambling in the display which triggers the reward/good feeling part of our brains (learnable and then addictive behaviors).

    Real life cost to in game worth is almost always too expensive. Typically most you can spend in a single transaction is double if not triple the retail value of a game, not including dlc and you can still not get what you were after.

    Influential behaviour being broadcast directly from said games via popular youtubers with sometimes sponsored "100 pack openings" with skewed drop rates.

    As developers/publishers/designers the curve to earn with time youve invested needs to be increased over time not decreased as it becomes akin to skinner box mentality.
    Actual rewards need to be reasonably earnable in loot boxes or at least 1 to 1 equiv exchanging/ no duplicates. Not as a loss to exchange.

    As government/regulatory services we need to closely observe and be involved with more than just what things appear to be on the surface. Context is important as well as combinations of design that cause misdirection, deception and anything that can cause loss to a consumer considering all games come with a EULA that we pretty much have to agree to even after shoddy design choices are added or changed after initial purchase.
    As consumers we will do as we do and get angry about poor design choice when necessary and force change where appropriate but we also must make sure that the change is going to be different rather than the same thing with a different name.

    Slapping a sticker on a box changes nothing.
    Fix the reward/time invested ratio.
    Stop giving us shit for our time and money.
    Dont give rich people better gear.

      If it's categorised as gambling it would mean the game cannot be sold to minors. Sure there are plenty of R rated games around but having a high rating isn't always a good selling point.
      There would be the potential to loose sales, which could make publishers think twice about including them. Especially of it made a game go from PG to R.

        I can understand this but considering our ratings for games dont carry the same understanding/penalty when it comes to parents (that dont know games) purchasing for thier kids as it is, i dont see this making a difference. Experience being an older lady asking to buy until dawn for her 9 yr old because he wanted it, "you know this is rated r18 right?" "Isnt it just a scary game?"... And picture magazine has writing in it but its also r18

Join the discussion!

Trending Stories Right Now