11 ms·

Usernames reveal the age and psychology of game players

63 points by limbicsystem 11y ago

Natsu 11y ago
From reading the results, they found out that you could guess someone's name when they put a year in the username and that antisocial words meant the person was more likely to be a jerk. Am I missing something here? The idea that SatansDick2007 is probably an 8 year old troll won't come as a shock to much of anyone.
- feral 11y ago
  That's an appealing and intuitive idea to me too, yeah; but they have validated that it's actually the case on a large population of users. Their paper seems very methodologically careful/thorough. It's also interesting to me that they correlated this with the post-game player-feedback. Its easy to say 'we knew that' post hoc, but they've done the science. (Also, hindsight bias is a real thing.) This seems like the opening step in a line of research. Its like in Feynman's 'cargo cult science' address about the (apocryphal?) experimenter 'Young': this is learning how to carefully do science in this domain. http://calteches.library.caltech.edu/51/2/CargoCult.htm http://calteches.library.caltech.edu/51/2/CargoCult.htm
  - Natsu 11y ago
    While it's good to validate the things we all know, any forum mod could've told you that it's not usually hard to predict the troublemakers or spammers by name.
    - feral 11y ago
      There's the things we all know, and there's things we all know because they have been scientifically studied and validated, and these two sets of things are not the same. I'm sure the research in this area will uncover more unintuitive things later!
      - mreiland 11y ago
        I once saw a study that concluded heterosexual couples were more likely to have children than homosexual couples. I mean, there's the things we know, and then there's the things that have been scientifically studied and validated. And then there's the things that really don't need to be validated. I once had someone claim that if you drove a vehicle through mud, it would get muddy. But since it hasn't been scientifically proven and validated, do I really "know" that to be true? Is this really going to be an existential crisis where we don't really know anything until science tells us it's true?
  - downer70 11y ago
    Bah! Cargo cult is a much more nuanced concept. That primitive peoples would blindly imitate in pantomime, the behaviors of an industrialized society, disembodied from the original roots of said behaviors is a deeply shocking outcome. There are so many perverse implications to that, that one almost hates to admit it happened, and still continues to this day. This on the other hand is like profiling someone's personality based on their choice of ringtone, t-shirts or bumper stickers. It's not very interesting to learn that such choices map well with demographics.
    - feral 11y ago
      I was not direct referring to the sociological/anthropological concept of a 'cargo cult'. I am instead making reference to Richard Feynman's commencement speech, titled 'Cargo Cult Science' (see link I provided) in which he talks about what is necessary for good science.
AnkhMorporkian 11y ago
As someone who runs a fairly large multiplayer game with a good mix of children, teenagers, and adults, I've always been convinced of this[1] but haven't had any data to back it up, and I certainly didn't have the time or drive to conclusively demonstrate it like these fine folks did. I would love to see someone tackle some solutions to this problem with some A/B testing. Sadly I don't think my userbase is large enough or antisocial enough for me to effectively test it. [1] https://www.reddit.com/r/leagueoflegends/comments/3op7e2/whats_up_with_these_riot_surveys/cvze54n?context=1 https://www.reddit.com/r/leagueoflegends/comments/3op7e2/wha...
cjslep 11y ago
I played LoL briefly during the beta and before the Tribunal was implemented, and even won their first contest (logo lookout) when Nikita was still an intern for Riot. I was very interested in the community and the lore since Riot was proactive in words and, for a brief time, actions. I even have an email dated Oct 12 2010: > 2. Has the idea of a "Summoner Spotlight" ever been discussed? I think a small article written by a Summoner about their own character, and approved by Riot Staff could do wonders to enrich Roleplay almost to the point of The Hollow.[0] Suddenly, on Oct 22 2010 Summoner Spotlights showed up! Probably a coincidence since the content was totally different, but for a fan like me at the time it was exciting. [1] To say I was drooling the Kool-Aid at the idea of having a positive community would be an understatement. Then something must have happened internally because Babaganush was soon fired and nothing was really done to stop the community from suddenly becoming overwhelmed with adolescents describing their urges surrounding "Nikasaur" once the Spotlights did show up. That, coupled with the in-game behavior, made the whole community smell of rot. I left after it became clear Riot was not actually focused on really cleaning up the community. Riot took the game towards the e-sports competitive scene which, sadly, helped shape the community into what it was and is. I've heard the Tribunal has helped, but it's not one of those features that is going to make me come back. [0] http://imgur.com/VfUvarl http://imgur.com/VfUvarl [1] http://leagueoflegends.wikia.com/wiki/Summoner_Showcase http://leagueoflegends.wikia.com/wiki/Summoner_Showcase EDIT: To the downvoters, if you disagree with how I am recounting history, please do elaborate. Riot got rid of a significant amount of their moderate playerbase by their early business decisions, which was a contributing factor to having a negative-attitude playerbase in the 2010-2012 years. That negative attitude at this point has become a stereotype, and perhaps it could have been avoided.
- Natsu 11y ago
  How long have you been gone? I think you might be out of date. Most games aren't negative (compare to, say, voice chat in CS:GO...) and they've rolled out plenty of features to report people who are being jerks. That said, honestly, you can't prevent people from saying things you don't like, especially a bunch of anonymous kids. They can always sign up for a new account or whatever. Better just to punish and move on, especially as they have plenty of things to lose for being a jerk (IP, rank, champions, money they've spent, etc.). The Tribunal is gone, but you can file reports, some punishments are automatic (did they spend the entire game swearing at their team and get reported?), you can mute anyone you like, etc.
kincardine 11y ago
> Data from League of Legends reveals links between real-world and online personality. Surprising?
- Laaw 11y ago
  That's something I think people forget; the vast majority of folks don't have some kind of split personality disorder, and their behavior online is the same behavior they exhibit in real life, except perhaps toned up a bit.
  - lsc 11y ago
    We all sometimes have the urge to do or say inappropriate things, no? A big part of what makes you, you know, you is the degree to which you have a "filter" that screens out those inappropriate urges before they turn into inappropriate actions. This filter is rather different in different situations, because what is appropriate (and what has consequences) is different in different situations. In a very real way, I would argue that changing that filter enough (and pesudoanonymity is usually 'enough') makes you, essentially, a different person. Most people, in a pseudonymous situation, have much less of a filter, in part because bad behavior is more acceptable, and in part because bad behavior has fewer negative consequences, both to themselves and others, than it does in a real life situation. (I'm not saying that there are no consequences; people are hurt. But usually not as much as people would be hurt if the same people removed the same filters in real-life interactions.)
whoopdedo 11y ago
What about cultural references in user names? The use of non-English words can reflect cultures other than American. Someone with the name "Browncoat" (a Firefly reference) may be younger than the name "Krell" (a Klingon in the original Star Trek). I imagine there are other types of information that can be inferred from an alias. It supports the argument that metadata contains private information. As well whether use of pseudonyms provides anonymity.
- ChrisGranger 11y ago
  Perhaps we'll see Diceware-or-similar usernames in addition to passwords if enough people care about keeping their pseudonyms as anonymous as possible.
  - qbrass 11y ago
    Most people will re-roll if they get a name they don't like.
jakejake 11y ago
The most interesting part of this to me was that fact that players tend to use their birth year in their username. And that the number 88 was a statistical outlier because for some reason people must really like that number.
- coke12 11y ago
  One possible reason that isn't discussed: https://en.wikipedia.org/wiki/88_(number)#In_white_nationalism https://en.wikipedia.org/wiki/88_(number)#In_white_nationali... On Reddit a username with 88 in it can be interpreted as a dog whistle for racists. Although considering how popular League of Legends is with Chinese people, the other hypothesis seems much more reasonable.
  - feral 11y ago
    Actually, they do mention that, just before the start of Section 4.1 (I hadn't heard of this before, maybe why I noticed it.)
  - whoopdedo 11y ago
    > On Reddit a username with 88 in it can be interpreted as a dog whistle for racists. Or NASCAR fans. Which contrary to popular belief is not the same thing.
HappyTypist 11y ago
This analysis is certainly interesting, but I can't help but think supervised machine learning (e.g. neural networks) would have been a far better way to analyze this data. We're using neural networks to predict future actions of users from weakly-correlating metrics and achieved surprising accuracy from one day of tinkering.
Animats 11y ago
It's a weak result. When I saw the title, I thought it might be useful for detecting jerks on blogs and Wikipedia. It's not. The mean age of their sample is 15.9 years. This result may not apply to higher ages, even college age. Running a similar analysis on Wikipedia might be interesting, minus the age analysis.
CM30 11y ago
Yeah, it's pretty obvious that online usernames say a lot about someone's age and personality. Heck, they often say a lot about a variety of other things too, like their gender and political affiliation. Either way, while it was fairly obvious to anyone who runs forums or other communities, it's nice to see some actual research that backs up our common sense intuitions here. On another note though, do you what would also be interesting to study here? The correlation between avatars and age, psychology, etc. Because let's face it, the picture someone choses to go with their profile (or the in game model they design or choose for an online game) says at least as much about them as their name does. A lot of people do tend to choose pictures or characters that look a bit like what they look like in real life (or in a lot of kids case, what they wished they looked like). P.S. Am I the only one somewhat surprised about the 'ages figured from usernames tend to match what was entered on registration' aspect? Given that a lot of games and communities are 13+ only (or perhaps older), I'd expect a lot of kids to lie about their age when signing up.
NicoJuicy 11y ago
I really wonder what they would think of my counterstrike nickname: Right Ball of Legolas, which is just meant as something funny

63 points by limbicsystem 11y ago

Natsu 11y ago

From reading the results, they found out that you could guess someone's name when they put a year in the username and that antisocial words meant the person was more likely to be a jerk. Am I missing something here? The idea that SatansDick2007 is probably an 8 year old troll won't come as a shock to much of anyone.

feral 11y ago
That's an appealing and intuitive idea to me too, yeah; but they have validated that it's actually the case on a large population of users. Their paper seems very methodologically careful/thorough. It's also interesting to me that they correlated this with the post-game player-feedback. Its easy to say 'we knew that' post hoc, but they've done the science. (Also, hindsight bias is a real thing.) This seems like the opening step in a line of research. Its like in Feynman's 'cargo cult science' address about the (apocryphal?) experimenter 'Young': this is learning how to carefully do science in this domain. http://calteches.library.caltech.edu/51/2/CargoCult.htm http://calteches.library.caltech.edu/51/2/CargoCult.htm
- Natsu 11y ago
  While it's good to validate the things we all know, any forum mod could've told you that it's not usually hard to predict the troublemakers or spammers by name.
  - feral 11y ago
    There's the things we all know, and there's things we all know because they have been scientifically studied and validated, and these two sets of things are not the same. I'm sure the research in this area will uncover more unintuitive things later!
    - mreiland 11y ago
      I once saw a study that concluded heterosexual couples were more likely to have children than homosexual couples. I mean, there's the things we know, and then there's the things that have been scientifically studied and validated. And then there's the things that really don't need to be validated. I once had someone claim that if you drove a vehicle through mud, it would get muddy. But since it hasn't been scientifically proven and validated, do I really "know" that to be true? Is this really going to be an existential crisis where we don't really know anything until science tells us it's true?
- downer70 11y ago
  Bah! Cargo cult is a much more nuanced concept. That primitive peoples would blindly imitate in pantomime, the behaviors of an industrialized society, disembodied from the original roots of said behaviors is a deeply shocking outcome. There are so many perverse implications to that, that one almost hates to admit it happened, and still continues to this day. This on the other hand is like profiling someone's personality based on their choice of ringtone, t-shirts or bumper stickers. It's not very interesting to learn that such choices map well with demographics.
  - feral 11y ago
    I was not direct referring to the sociological/anthropological concept of a 'cargo cult'. I am instead making reference to Richard Feynman's commencement speech, titled 'Cargo Cult Science' (see link I provided) in which he talks about what is necessary for good science.

AnkhMorporkian 11y ago

As someone who runs a fairly large multiplayer game with a good mix of children, teenagers, and adults, I've always been convinced of this[1] but haven't had any data to back it up, and I certainly didn't have the time or drive to conclusively demonstrate it like these fine folks did. I would love to see someone tackle some solutions to this problem with some A/B testing. Sadly I don't think my userbase is large enough or antisocial enough for me to effectively test it. [1] https://www.reddit.com/r/leagueoflegends/comments/3op7e2/whats_up_with_these_riot_surveys/cvze54n?context=1 https://www.reddit.com/r/leagueoflegends/comments/3op7e2/wha...

cjslep 11y ago

I played LoL briefly during the beta and before the Tribunal was implemented, and even won their first contest (logo lookout) when Nikita was still an intern for Riot. I was very interested in the community and the lore since Riot was proactive in words and, for a brief time, actions. I even have an email dated Oct 12 2010: > 2. Has the idea of a "Summoner Spotlight" ever been discussed? I think a small article written by a Summoner about their own character, and approved by Riot Staff could do wonders to enrich Roleplay almost to the point of The Hollow.[0] Suddenly, on Oct 22 2010 Summoner Spotlights showed up! Probably a coincidence since the content was totally different, but for a fan like me at the time it was exciting. [1] To say I was drooling the Kool-Aid at the idea of having a positive community would be an understatement. Then something must have happened internally because Babaganush was soon fired and nothing was really done to stop the community from suddenly becoming overwhelmed with adolescents describing their urges surrounding "Nikasaur" once the Spotlights did show up. That, coupled with the in-game behavior, made the whole community smell of rot. I left after it became clear Riot was not actually focused on really cleaning up the community. Riot took the game towards the e-sports competitive scene which, sadly, helped shape the community into what it was and is. I've heard the Tribunal has helped, but it's not one of those features that is going to make me come back. [0] http://imgur.com/VfUvarl http://imgur.com/VfUvarl [1] http://leagueoflegends.wikia.com/wiki/Summoner_Showcase http://leagueoflegends.wikia.com/wiki/Summoner_Showcase EDIT: To the downvoters, if you disagree with how I am recounting history, please do elaborate. Riot got rid of a significant amount of their moderate playerbase by their early business decisions, which was a contributing factor to having a negative-attitude playerbase in the 2010-2012 years. That negative attitude at this point has become a stereotype, and perhaps it could have been avoided.

Natsu 11y ago
How long have you been gone? I think you might be out of date. Most games aren't negative (compare to, say, voice chat in CS:GO...) and they've rolled out plenty of features to report people who are being jerks. That said, honestly, you can't prevent people from saying things you don't like, especially a bunch of anonymous kids. They can always sign up for a new account or whatever. Better just to punish and move on, especially as they have plenty of things to lose for being a jerk (IP, rank, champions, money they've spent, etc.). The Tribunal is gone, but you can file reports, some punishments are automatic (did they spend the entire game swearing at their team and get reported?), you can mute anyone you like, etc.

kincardine 11y ago

> Data from League of Legends reveals links between real-world and online personality. Surprising?

Laaw 11y ago
That's something I think people forget; the vast majority of folks don't have some kind of split personality disorder, and their behavior online is the same behavior they exhibit in real life, except perhaps toned up a bit.
- lsc 11y ago
  We all sometimes have the urge to do or say inappropriate things, no? A big part of what makes you, you know, you is the degree to which you have a "filter" that screens out those inappropriate urges before they turn into inappropriate actions. This filter is rather different in different situations, because what is appropriate (and what has consequences) is different in different situations. In a very real way, I would argue that changing that filter enough (and pesudoanonymity is usually 'enough') makes you, essentially, a different person. Most people, in a pseudonymous situation, have much less of a filter, in part because bad behavior is more acceptable, and in part because bad behavior has fewer negative consequences, both to themselves and others, than it does in a real life situation. (I'm not saying that there are no consequences; people are hurt. But usually not as much as people would be hurt if the same people removed the same filters in real-life interactions.)

whoopdedo 11y ago

What about cultural references in user names? The use of non-English words can reflect cultures other than American. Someone with the name "Browncoat" (a Firefly reference) may be younger than the name "Krell" (a Klingon in the original Star Trek). I imagine there are other types of information that can be inferred from an alias. It supports the argument that metadata contains private information. As well whether use of pseudonyms provides anonymity.

ChrisGranger 11y ago
Perhaps we'll see Diceware-or-similar usernames in addition to passwords if enough people care about keeping their pseudonyms as anonymous as possible.
- qbrass 11y ago
  Most people will re-roll if they get a name they don't like.

jakejake 11y ago

The most interesting part of this to me was that fact that players tend to use their birth year in their username. And that the number 88 was a statistical outlier because for some reason people must really like that number.

coke12 11y ago
One possible reason that isn't discussed: https://en.wikipedia.org/wiki/88_(number)#In_white_nationalism https://en.wikipedia.org/wiki/88_(number)#In_white_nationali... On Reddit a username with 88 in it can be interpreted as a dog whistle for racists. Although considering how popular League of Legends is with Chinese people, the other hypothesis seems much more reasonable.
- feral 11y ago
  Actually, they do mention that, just before the start of Section 4.1 (I hadn't heard of this before, maybe why I noticed it.)
- whoopdedo 11y ago
  > On Reddit a username with 88 in it can be interpreted as a dog whistle for racists. Or NASCAR fans. Which contrary to popular belief is not the same thing.

HappyTypist 11y ago

This analysis is certainly interesting, but I can't help but think supervised machine learning (e.g. neural networks) would have been a far better way to analyze this data. We're using neural networks to predict future actions of users from weakly-correlating metrics and achieved surprising accuracy from one day of tinkering.

Animats 11y ago

It's a weak result. When I saw the title, I thought it might be useful for detecting jerks on blogs and Wikipedia. It's not. The mean age of their sample is 15.9 years. This result may not apply to higher ages, even college age. Running a similar analysis on Wikipedia might be interesting, minus the age analysis.

CM30 11y ago

Yeah, it's pretty obvious that online usernames say a lot about someone's age and personality. Heck, they often say a lot about a variety of other things too, like their gender and political affiliation. Either way, while it was fairly obvious to anyone who runs forums or other communities, it's nice to see some actual research that backs up our common sense intuitions here. On another note though, do you what would also be interesting to study here? The correlation between avatars and age, psychology, etc. Because let's face it, the picture someone choses to go with their profile (or the in game model they design or choose for an online game) says at least as much about them as their name does. A lot of people do tend to choose pictures or characters that look a bit like what they look like in real life (or in a lot of kids case, what they wished they looked like). P.S. Am I the only one somewhat surprised about the 'ages figured from usernames tend to match what was entered on registration' aspect? Given that a lot of games and communities are 13+ only (or perhaps older), I'd expect a lot of kids to lie about their age when signing up.

NicoJuicy 11y ago

I really wonder what they would think of my counterstrike nickname: Right Ball of Legolas, which is just meant as something funny