5 ms·

When it's OK to manipulate users w psych A/B testing

12 points by steven 12y ago

falcolas 12y ago
It's a vain wish, but I wish there were more oversight when companies do experiments with our emotions. Who is to say that Facebook was wrong while Riot was right? Who is to say that long term impact Riot's "tests" that produced negative consequences had on people? Who is there for those people who were affected negatively? Nobody.
pmalynin 12y ago
In my opinion trash talk is part of the game, part of the fun I'd even say. Part of the reason I still play Heroes of Newerth sometimes is to have the undiluted experience of arguing, or fighting trolls, and other various degenerate behaviours. Hell, even Michael Jordan commented on the importance of such behaviour in basketball. More generally, however, this makes me a little scared. Feels like eventually it will get to the point where the machine learning engine will be able to recognize if I'm having bad thoughts and serve "quick punishment".
typicalrunt 12y ago
To answer the title: it's never OK to manipulate users. Let's look no further than the dictionary to see why. Here is the OED definition (3rd entry): manipulate: To manage, control, or influence in a subtle, devious, or underhand manner. It's sad people are seriously talking about human experimentation like it's a noble thing. Regardless of the fact that it's using bits and bytes, these are peole's minds being experimented on. Since nobody signed up for taking part in the experiment, this PhD should lose his credentials for performing immoral experiments. Instead he is praised because we euphemize it as "A/B teting" and most people don't know that these experiments are immoral. Even the article accidentally lets it slip that these experiments are immoral. The article mentions that when Mr. Lin was receiving his PhD he had a team of students as test subjects. My guess would be that these students signed waivers and acknowledged that they would be part of an experiment (hence "test subject"). What Mr. Lin is doing now is using gamers as test subjects without their consent. If this were any other industry, or if the experiments targeted a physical (as opposed to mental) part of the body, I think people would be more angry at the callousness of these so-called scientists. Heck, what if GE or Monsanto came out with a glowing article stating the benefits of A/B testing on farmers' mental states and their crops? edit: added definition edit 2: To explain why I get so outraged when I see articles like this... receiving my Bachelor of Science meant that I had to understand the right and wrong way to conduct experiments. And experimenting on humans is the most cumbersome because, to be honest, you can really screw someone up if you aren't careful. Even the most innocuous experiments can lead to bad outcomes for the test subjects, so one doesn't simply jump to human trials until you can conclusively argue that it won't damage another person.
- drabiega 12y ago
  Well, I'm not entirely sure that subtly influencing people to be nicer to each other is a bad thing and I don't see how anything they are doing is devious or underhanded.
  - typicalrunt 12y ago
    > I'm not entirely sure that subtly influencing people to be nicer to each other is a bad thing This would be a subjective opinion. The ends do not justify the means. I agree with you that getting people to act nicer to each other is a nice goal, but you can't do it by enlisting them all into a hidden experiment. > I don't see how anything they are doing is devious or underhanded. Well that's the point of A/B testing isn't it? The user doesn't realize that they are part of the experiment. That's the devious part. If they had signed up to be part of an experiment, then at least it wouldn't be an underhanded maneuver to start changing game UI or semantics on them without their knowledge.
- jbob2000 12y ago
  The user agrees to a terms of service when they sign up, which I'm sure includes a section detailing that riot can do whatever they want with user data and the in game environment. Don't like that Riot is experimenting? Don't play the game.
drabiega 12y ago
Manipulation is, ironically, a loaded word. The primary differences between 'manipulation' and the subtle behavioral clues we give to others we communicate with is that the latter is generally done unconsciously and thus not based on testing and intent. This applies to games as well, as in this case. The design of your UI is going to influence your user base, so why is it a bad thing to try to ensure that it influences them in a positive way?
PeterWhittaker 12y ago
The headline is incorrect, it should read Inside the Largest Virtual Psychology Lab in the World.

12 points by steven 12y ago

It's a vain wish, but I wish there were more oversight when companies do experiments with our emotions. Who is to say that Facebook was wrong while Riot was right? Who is to say that long term impact Riot's "tests" that produced negative consequences had on people? Who is there for those people who were affected negatively? Nobody.

pmalynin 12y ago

In my opinion trash talk is part of the game, part of the fun I'd even say. Part of the reason I still play Heroes of Newerth sometimes is to have the undiluted experience of arguing, or fighting trolls, and other various degenerate behaviours. Hell, even Michael Jordan commented on the importance of such behaviour in basketball. More generally, however, this makes me a little scared. Feels like eventually it will get to the point where the machine learning engine will be able to recognize if I'm having bad thoughts and serve "quick punishment".

typicalrunt 12y ago

To answer the title: it's never OK to manipulate users. Let's look no further than the dictionary to see why. Here is the OED definition (3rd entry): manipulate: To manage, control, or influence in a subtle, devious, or underhand manner. It's sad people are seriously talking about human experimentation like it's a noble thing. Regardless of the fact that it's using bits and bytes, these are peole's minds being experimented on. Since nobody signed up for taking part in the experiment, this PhD should lose his credentials for performing immoral experiments. Instead he is praised because we euphemize it as "A/B teting" and most people don't know that these experiments are immoral. Even the article accidentally lets it slip that these experiments are immoral. The article mentions that when Mr. Lin was receiving his PhD he had a team of students as test subjects. My guess would be that these students signed waivers and acknowledged that they would be part of an experiment (hence "test subject"). What Mr. Lin is doing now is using gamers as test subjects without their consent. If this were any other industry, or if the experiments targeted a physical (as opposed to mental) part of the body, I think people would be more angry at the callousness of these so-called scientists. Heck, what if GE or Monsanto came out with a glowing article stating the benefits of A/B testing on farmers' mental states and their crops? edit: added definition edit 2: To explain why I get so outraged when I see articles like this... receiving my Bachelor of Science meant that I had to understand the right and wrong way to conduct experiments. And experimenting on humans is the most cumbersome because, to be honest, you can really screw someone up if you aren't careful. Even the most innocuous experiments can lead to bad outcomes for the test subjects, so one doesn't simply jump to human trials until you can conclusively argue that it won't damage another person.

drabiega 12y ago
Well, I'm not entirely sure that subtly influencing people to be nicer to each other is a bad thing and I don't see how anything they are doing is devious or underhanded.
- typicalrunt 12y ago
  > I'm not entirely sure that subtly influencing people to be nicer to each other is a bad thing This would be a subjective opinion. The ends do not justify the means. I agree with you that getting people to act nicer to each other is a nice goal, but you can't do it by enlisting them all into a hidden experiment. > I don't see how anything they are doing is devious or underhanded. Well that's the point of A/B testing isn't it? The user doesn't realize that they are part of the experiment. That's the devious part. If they had signed up to be part of an experiment, then at least it wouldn't be an underhanded maneuver to start changing game UI or semantics on them without their knowledge.
jbob2000 12y ago
The user agrees to a terms of service when they sign up, which I'm sure includes a section detailing that riot can do whatever they want with user data and the in game environment. Don't like that Riot is experimenting? Don't play the game.

drabiega 12y ago

Manipulation is, ironically, a loaded word. The primary differences between 'manipulation' and the subtle behavioral clues we give to others we communicate with is that the latter is generally done unconsciously and thus not based on testing and intent. This applies to games as well, as in this case. The design of your UI is going to influence your user base, so why is it a bad thing to try to ensure that it influences them in a positive way?

PeterWhittaker 12y ago

The headline is incorrect, it should read Inside the Largest Virtual Psychology Lab in the World.