IF THERE'S ONE thing worse than reading press coverage about scientific research on gaming, it's reading about government policy based on scientific research. And yet, the announcement last week that the government was recommending treating the much-maligned Loot Box as gambling was broadly welcomed, uniting researchers, the government and ripped-off gamers in a weirdly united coalition.
It was a good time to pick up the phone to Dr Peter Etchells from the University of Bath Spa ahead of his talk on video game research at New Scientist Live next month. Immediately, he talks about the "correlations between Loot Box purchase behaviours and problematic gambling behaviours."
"Now obviously correlations don't imply causation, but in a sense, it doesn't matter what the correlation is: either people who have pre-existing vulnerabilities when it comes to problematic gambling are more drawn to playing games that have loot boxes… or people who don't have any pre-existing vulnerabilities start playing games that have got loot box elements in them and that creates problematic gambling behaviours. So I think either way it's not good news for loot boxes."
Something comparatively black and white is highly unusual in games research, but that might be partly because of the narrowness of the question: it's no longer are video games good or bad for us, which Etchells views as somewhere between 'unhelpful' and ‘meaningless'.
"It's kind of like asking is food good or bad for us and your life depends on what food you eat and what you do exercise and all these other sorts of things."
Though even here, there are grey areas. "If you spend £50,000 - I mean no game is worth spending £50,000 on, but for the sake of argument - if you spend that £50,000 on a game and you can't afford that, then that's a problem. It's causing harm and that's something that needs to be addressed. But if you're a multi-millionaire and £50,000 is a drop in the ocean... is it the same sort of harm? We don't really have a good way of navigating around those sorts of issues at the minute."
The hunt for good data
Still, this is a relatively accepted conclusion - except possibly by game publishers set to lose revenue. But leaving them to one side for a moment, there's the far more controversial question of video game violence - now back in the news as the NRA tries to shield blame from the things that spat out the bullets at recent mass shootings. Here, the data is infuriatingly imprecise, and the researchers aren't necessarily to blame.
"You can't give a bunch of people a violent game to play, a bunch of people a non-violent game to play, and then give everybody a gun and see how many people die in those two groups because you know, ethically it's a bit dubious to do that," Etchells jokes. Instead, the most common measure of aggression in gaming papers is something called a "capacity reaction time task". In short, you tell someone they're competing with an unseen other player to press the spacebar first, and if they win, they get to blast their opponent with a loud noise. There is no opponent, of course, and it's actually set up that they'll win some and lose some to test how ‘angry' they get.
"I think there's something like 150 papers that have been published and there are 180 ways in which they actually analyse the data," Etchells says. "You can literally find that video games definitely cause aggression or that they definitely don't cause aggression and that's nothing to do with the data. It's everything to do with the decision that you make about how to analyse the data."
Even the good data isn't without its problems. One of the better data sets is something called the "Children of the 90s": a survey of 14,500 parents in the Bristol area analysing trends of their children as they grew up. "We found there is a very weak association between playing more violent games at age eight and conduct problems at age 16, but the risk is pretty tiny." Only 30 children were in this subset, he says, and even then you can't completely account for other factors.
It would be very helpful, therefore, Etchells says, if Microsoft, Sony and Nintendo made their player analytics public, although he recognises why that's something they would resist for fear of losing competitive advantage.
Still, Etchells was pleased that the DCMS acknowledged the weaknesses of the research in its report, and hopes this will lead to pressure on the publishers to play ball. "Given what's happened with the commons report, they're going to get very strongly pushed to start doing that," he predicts. This could ultimately be in-game developers interests, allowing them to build games that maximise the good points while minimising the negatives.
But for now, the data is limited and there are lots of problems in this, obviously - not least is that certain elements of the media don't, to put it mildly, like dealing in ambiguities. In fact, one dubious headline has a lot to do with why Etchells entered the field: "Computer games leave children with 'dementia' warns top neurologist." We won't link to the publication here, but it's almost certainly the one you're thinking of.
"It was frustrating for a number of reasons," Etchells recalls. "First is that the person who they claimed was a top neurologist wasn't - they were a neuroscientist and they're different things. The second is, you know, if you look at the research, there was nothing on early-onset dementia in children and video game play."
Four years later, he was in a position to publically call out bad reporting for the Guardian with a different paper. This one reportedly showed a link between Alzheimer's and Call of Duty. "Actually if you look at the research paper, it doesn't mention Call of Duty once, it doesn't mention dementia once," he explains. "It was on something that's incidentally related to those things: video games in the context of spatial navigation." Nought for two, then.
"If you're doing proper research on this, then you have to wait until people who played are at the point where you'd expect them to get dementia, and then see if the prevalence rate is higher in gaming populations than non-gaming populations."
A model for the future?
Even though Etchells is equally critical of overly positive gaming research ("I think you've got to take the line that the way we do a lot of these studies is not particularly robust") there is one project he's effusively enthusiastic about: a little app called Sea Hero Quest. Designed by Glitchers in association with Alzheimer's Research UK, UCL and UEA, the app requires players to memorise a course from a 2D map, and then translate that memory into a 3D world. All that data - when people go wrong, the player's age and gender - is amalgamated and analysed by scientists.
"I think they wanted to get about 50,000 people to do this study, to get enough people to figure out what spatial navigation abilities look like over the course of life, hoping it would give some insights into how dementia works. As of last year, they've had four million players from every country in the world."
That, if you haven't noticed, is a lot of data. "They've collected about 1,700 years' worth," Etchells explains. So much, in fact, that they're making the data set open access.
"The first paper came out last year, and showed that spatial navigation abilities start to decline from about 19 - much earlier than we thought it was, which is slightly depressing," Etchells says. The data has also revealed interesting discrepancies between genders, wealth and even country.
There's a certain irony in the fact that gaming, so often misread in gaming science coverage, has provided such a rich data set for researchers to delve into. Hopefully, it provides a template for future titles, although it's fair to say this kind of marriage between science and gaming is currently a rarity.
As for Etchells, his gaming time has recently been curtailed for the best possible reason. "I became a dad three months ago, so my gaming time is very limited at the minute," he says. "The Switch has been really good actually for lots of sleepless nights when my daughter wasn't sleeping, so I could just have her sleeping on my chest while I played Doom. Quite nice and mindless at five in the morning."
You can imagine the tabloid headlines now. µ
New Scientist Live is on at the ExCeL London from 10-13 October. Tickets available here.
Hype for HyperThreading
Hey kids, leave them iPhones alone
The Mac lady sings
Babel in yo ear