TR Member Perks!

The choice for mankind lies between freedom and happiness and for the great bulk of mankind, happiness is better. -George Orwell, 1984.

Okay, that’s a massively melodramatic opener for a commentary about a small story from SocJus propaganda e-rag Recode. If it comes to pass 10 years from now you get a timeout at 0930 for 2 minutes to shout hate at “thought criminals” and prisoners of war from East Asia (we’ve always been at war with them; Eurasia is our friend), you can look back on July 7, 2015 as the day Thought Crime was implemented for real.

Riot has a long history, by internet standards, with Thought Crime.  In September 2014, Riot rolled out live testing of what they called Ranked Restrictions.  Here’s an example of how they work:

Let’s say a person solo queues for Ranked play and is placed on a team with a terrible player.  The team with the terrible player gets rolled in 20 minutes.  It’s obvious, based on statistics collected as a part of play (Kill/Assist/Death ratios, gold earned, items in inventory, etc.), the terrible player is the reason the team lost, and the person who solo queued decides to be honest with the terrible player about the terrible player’s quality of play.  The terrible player can then report our hapless solo queue player for “negative behavior.”  If this particular false flag report sends the solo queuing player over an arbitrary threshold, the reported player is Ranked Restricted.  The Ranked Restricted player must now win a certain number of games in Normal Draft, where he/she is chat restricted, and must demonstrate they’ve improved their in-game behavior (in this case, suffering players that treat bad play and getting carried to wins in Ranked as entitlements) before being allowed back into the competitive environment.

With one overly broad stroke of the brush, Riot made it worse to be honest with a terrible player than it is to be a terrible player and cause your team to lose in the competitive play mode.

The Recode article is from Jeffery Lin, Lead Social Systems Designer for Riot. It discusses tackling the “Impossible Problem” of abuse in online games.

This Space Reserved for Non-Controversial Disclaimers

This next paragraph is for all the try hard, LoL fanbois out there, as well as pro-thought-crime members of SocJus who earnestly think criticism is harassment. Yes, I am aware there’s a problem, in general, with people being harassed online. Yes, I am aware this problem can be exacerbated in online games, especially MOBAs. Yes, I am aware depending on which research is cited, traditionally marginalized groups may suffer more harassment. And yes, I acknowledge any group trying to develop a means for diminishing the worst online harassment is, in general, a good thing.

The Massive “but,” Part 1

Trash talk is neither harassment, nor is trash talk a negative behavior. In fact, trash talk is integral to what League of Legends is desperate to be: sports. Don’t tell this to Lin, who included trash talk as a negative behavior:

Our team found that if you classified online citizens from negative to positive, the vast majority of negative behavior (which ranges from trash talk to non-extreme but still generally offensive language) did not originate from the persistently negative online citizens; in fact, 87 percent of online toxicity came from the neutral and positive citizens just having a bad day here or there.

Why do we use trash talk? Trash talk is used in the mental part of any competitive endeavor. I shouldn’t have to convince you that players and developers of LoL think it is competitive, given all the pomp, circumstance, and full-of-sound-and-fury-signifying-nothing bravado that goes on at every major LoL tournament—as if saying there are LoL tournaments isn’t, in and of itself, sufficient evidence for LoL being a competitive environment.

Trash talk is used to gain an advantage over your opponent. It is, in fact, a useful tool mentally and emotionally strong people use against mentally and emotionally weak people to induce the weak people to not perform optimally. Indeed, entire books have been written toward parents to guide them through training their children in mental toughness to maximize the positive aspects of youth athletics.

And, of course, Sports Psychology is an actual industry dedicated to maximizing performance.

If LoL is a sport, then why is the psychological aspect of sport been turned into Thought Crime?

My first issue with Riot’s Thought Crime Algorithm is the definition of negative behavior is either naïve about competitive environments in the real world or far, far too broad.

The Massive “but,” Part 2

Returning to this quote:

Our team found that if you classified online citizens from negative to positive, the vast majority of negative behavior (which ranges from trash talk to non-extreme but still generally offensive language) did not originate from the persistently negative online citizens; in fact, 87 percent of online toxicity came from the neutral and positive citizens just having a bad day here or there.

Doesn’t that seem like an enormous amount of false positives? If 87% of incidents are one off incidents of trash talk, then what are we talking about? Further, what’s the 87% part of? Is it 87% of 67 million? Is it 87% of the total number of claims, without regard to whether a claim of harassment or “negative behavior” is legitimate, even by Riot’s impossibly broad definition?

Clarity, of a sort, comes a couple of paragraphs later. Lin says thus:

To deliver meaningful consequences, we had to focus on the speed and clarity of feedback. At Riot, we built a system called the “Tribunal,” which automatically created “case files” of behaviors that players reported as unacceptable in the community.

This is not good news. League of Legends was designed for the lowest of the lowest common denominator of people who play games. There’s no structure in place to induce the most important behavior from players: good play. As I showed above, Riot made it worse to tell people they are terrible at LoL (the only feedback they’ll get from LoL is that they’re bad at it) than it is to be terrible at LoL. Of course, advocates for terrible LoL players celebrated.

My second issue with Riot’s Thought Crime Algorithm is the “criticism is harassment” crowd has had a hand in teaching the algorithm what is and isn’t bad behavior.

Can’t Argue With Success, Right?

Of course, the commentary concludes with a set of metrics people can’t help but get excited about.

…incidents of homophobia, sexism, and racism have fallen to a combined 2 percent of all LoL games. Verbal abuse has dropped by more than 40 percent, and 91.6 percent of negative players change their act and never commit another offense after just one penalty.

That is complete success, right? Maybe. What’s missing is a measure of whether people are actually communicating in these games anymore, or if players are so scared of the Telescreens and Big Brother that they’ve stopped talking in LoL at all. Dr. Lyte, it would be, perhaps, an interesting metric to see if the amount of communication present in aggregate in LoL games is more or less post-Thought Crime Algorithm than pre-Thought Crime Algorithm.

Surprising no one, the worst people on the Internet have already advocated a similar system be used everywhere. This advocacy demonstrates galactic naivety that all online communities are created equal, when it is obvious a video game community designed specifically for players who refuse to get better is completely different from social media communities which is different from anonymous message boards, and so on.

Further, it’s no surprise the article above spends zero words on the impossibly broad definition of negative behavior employed by Riot, or the fact that the worst of the worst actors in the community (bad players in competitive play modes) get a voice to determine what is, and what is not negative behavior.

That said, I’m all for the LoL community determining it’s own standard for what is and isn’t negative behavior. Not being able to be honest with terrible players wasting the rest of their team’s time is just one more reason not to play LoL. However, before Riot’s Thought Crime Algorithm is employed everywhere, let’s get some industry people who understand the value of communication, including sarcasm, wit, aggressive communication and trash talk, to contribute to what’s actually negative  behavior and what’s people with a victim complex crying wolf.

Ending racism, sexism, and homophobia in online communities sounds like a noble goal. Ending sharped wittedness, aggressive communications, and trash talk as a part of that goal is unacceptable.

Todd Wohling

A long time ago on an Intellivision far, far away my gaming journey started with Lock n' Chase, Advanced Dungeons & Dragons The Cloudy Mountain, and Night Stalker. I earned both a BS-Physics and a BS-Mathematics from the University of Wisconsin-Eau Claire. Today I spend most of my time on PC. I left a career of 14 years in aerospace in Colorado, so I could immigrate to Norway.