Video games have had their fair share of controversy over the years.
One of the current moral panics surrounding them has to do with loot boxes and whether or not they constitute a form of gambling.
A recent study, by David Zendle from the University of York, has added a little fuel to the fire, finding that the presence or absence of loot boxes affects in-game spending by problem gamblers.
A loot box is a form of randomized reward in video games. Contents are typically just cosmetic, but in multiplayer games — particularly those considered esports — rarer loot box items can be an important status symbol.
Depending on the game, loot boxes can be earned through in-game achievements, purchased for real money, or both. Naturally, it’s their availability for real-money purchase that’s the primary issue for those concerned.
Blizzard, a major game developer, removed paid loot boxes from its game Heroes of the Storm in April this year. The removal was announced in advance and gave Zendle the inspiration for his study.
The researcher created an online survey, which received hundreds of responses, asking players about their general gambling habits and their spending on the game before and after the removal of loot boxes.
Loot boxes clearly offer players the thrill of uncertainty, as certain items are more rare than others. Upon opening one, a player will typically hope for a particular outcome. When they have paid real money for the chance to open a loot box, it’s not hard to see why many see a resemblance to gambling.
It also seems intuitively obvious that players prone to problematic gambling habits would be more drawn to spending money on loot boxes. But when it comes to science, especially psychology, what seems obvious isn’t necessarily correct.
Studies like Zendle’s are, therefore important to confirm these intuitive suspicions.
Numerous studies have shown a correlation between problem gambling and loot box purchases. None of these have proven a causal link, however.
In introducing his work, Zendle points out that there are three possible explanations for the correlation:
Zendle’s study was aimed at ruling out the last of these by comparing the spending of the same players, in the same game, with and without loot boxes available. Crucially, the same items randomly awarded in the loot boxes were available for direct, non-randomized purchase through the in-game store — both before and after their removal.
Zendle’s carried out his survey in two parts. The first came just before the removal of loot boxes from the game, and the second two months after their removal. Each asked participants to estimate their spending on in-game purchases over the preceding month and their maximum spending in a single session during that time.
The first survey also included a nine-part questionnaire called the Problem Gambling Severity Index (PGSI). Zendle used this to divide respondents into four categories:
The study showed that problem gamblers all reported lower monthly spending on the game after loot boxes were removed. This was not simply due to decreased spending over time, as none of the other three categories reported a statistically significant change in their spending.
These results, however, should be taken with a grain of salt. Although 371 people responded to the initial study, not all came back to take the second part. Others were ruled out due to inconsistencies in their answers.
There’s also the question of how accurate players’ estimates of their own spending are. In fact, 111 of the 371 had to be ruled out for the reason that their reported maximum spending in a single session in the past month was higher than their estimate for the entire month. That’s clearly impossible.
Tellingly, the PSGI classified more than half of those players as problem gamblers.
Another thing we can take from the study, then, is that problem gamblers’ estimates of their in-game spending are particularly unreliable. That makes sense since they’re also known to be bad at estimating their total gambling losses.
It should, however, diminish our confidence in the study’s primary conclusion.
If we take that conclusion as valid, however, we can see that there is some relationship between problem gambling and high spending on loot boxes specifically — rather than in-game purchases, more generally.
This isn’t merely a subject of academic interest. Lawmakers around the world have been asking questions about loot boxes and whether they’re sufficiently gambling-like that they need to take action.
Belgium and the Netherlands, for example, have both placed full bans on loot boxes. That has led to some companies simply refusing to sell their games in those countries rather than remove the feature. Gamers are understandably unhappy about this outcome.
Conversely, the United Kingdom Gaming Commission has rebuffed calls from that country’s politicians to take action on loot boxes. The UK regulator says that they do not meet the legal definition of gambling.
There hasn’t yet been any formal action against loot boxes in the US. There have, however, been calls for further studies and age restrictions on games that make them available for purchase.
Most recently, Sen. Josh Hawley (R-MO) proposed a bill that would make for-purchase loot boxes illegal in games aimed at children.
The superficial resemblance of loot boxes to gambling is undeniable. Some games even use a spinning reel similar to a slot machine to select the item awarded when the player opens a loot box.
Defenders of loot boxes, however, argue that they don’t constitute gambling because the prizes awarded don’t have any cash value.
That argument breaks down when the game allows players to trade items between themselves, which is not uncommon. Although the players can’t exchange money in-game, it inevitably happens that unofficial trading networks pop up on forums and chat sites.
When a game facilitates such “skin trading” directly, the most desirable items often end up being sold for real money.
Blurring the line further, third-party sites are offering a service known as skin gambling.
These services allow players to send their loot box items (skins) to a bot account in return for virtual tokens on the skin betting site. Players can gamble with those tokens and eventually redeem them for other skins. They can then potentially sell those skins to other players.
At that point, the difference between loot boxes, skin betting and real-money gambling is mostly a matter of technicality.
Valve, a major game distribution service, has attempted to crack down on skin gambling sites. Its cease-and-desist notices did take some such sites offline, but new offshore sites simply popped up in their place.
This year, Valve was hit by a lawsuit from the Quinault Tribal Nation, alleging that it deliberately allows skin gambling to continue in order to profit from loot box sales.
Game companies will claim that skin gambling has nothing to do with them and that they can’t prevent it. All it would take for them to do so, however, is to remove the ability to transfer items between players.
The reason they don’t, presumably, is that it would lead to fewer loot box purchases.
Zendle’s paper concludes with some musings about the actual harm of loot boxes. Their removal did decrease the in-game spending of problem gamblers, but only from an average of $85 per month down to $50.
Compared to the amounts problem gamblers lose at actual casinos, online or land-based, $35 a month isn’t all that much.
The critical question — and one that hasn’t been answered by Zendle or anyone else — is whether frequent loot crate purchases contribute to the development of problematic gambling habits that didn’t exist before.
That question is of particular concern when it comes to underage players. Teenagers have poorer impulse control than adults. If loot boxes contribute to the development of gambling habits in later life, then the actual amount of money becomes even less relevant.
That’s certainly a concerning possibility and one that merits further study.