Profile cover photo
Profile photo
Audrey Penven
736 followers
736 followers
About
Audrey's posts

Post has attachment

Post has shared content
#nymwars Professionally, I work in email antiabuse and deliverability for a marketing automation company. I attend antispam conferences, and spend much of my day explaining to marketers why and how not to be a spammer.

That's a big part of why the real name policy bothers me.

In the spam world, it has become considered a truism that content filtering alone does not work. Filtering out every instance of the word "anal" just causes you to block every time someone sends the word "analysis" - and the spammers will spell it 4nal or ana! or ana| anyways. Sure, it will block the dumb ones, but those weren't the ones you really had to worry about anyways.

Most antispam systems these days use multiple tactics to block email abusers. These systems use honeypot trap accounts, traffic analysis, reputation systems built around domain/IP ownership and sending patterns, user response, and some (light, and evolving) content filtering. Using content filtering alone is easy to evade, and is widely regarded to be an antiquated technique for dealing with email abuse.

When I worked at my last employer, an antispam service provider, every false positive was considered a black mark on our reputation. Pure content filtering has an extremely high rate of false positives. We used all these various systems to block abuse and - equally important - to keep our false positive rate as low as possible. Every email that someone wanted that was blocked made us look like jerks, so we tried to keep those as few as possible.

If pure content filtering is not very good at blocking abuse and has a high false positive rate when dealing with email abuse, why would Google think it is a good strategy to deal with social networking abuse?

A real name policy amounts to a pure content filtering strategy in that it a) makes a determination on what terminology is acceptable or not, and b) defines abuse through use of unacceptable terminology rather than unacceptable behavior.

In the case of a name, it's very difficult to even make a determination about what is "real" or not. For more on that, see this article:

http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-believe-about-names/

In the course of my work, I have frequently had to deal with customers who have issues delivering email that users have signed up for because their company name contains the word Analytics ("anal"), or they talk about money in the course of discussing the financial services their customers signed up for. Content filtering alone is dumb. It makes no exceptions, takes no clues from context, does not respond to what is or is not wanted.

Filtering users by their names is an equally poor strategy. We have recently seen a huge number of cases of false positives from this strategy - see the legally mononymed +Sai . or the unfortunate Blake Ross for examples. It's also a poor way of stopping abuse. An old friend of mine has a highly motivated insane internet stalker who has already made his way here with more than one account that are not under his real name. This policy punishes innocent users and misses abusers.

Abuse in a social network should be identified with multiple strategies. Names may be one vector of abuse, but a real name policy has too many false positives to be used alone. Additionally, it implies that once the pseudonyms are gone, we're all safe and cozy - a clear falsehood, given the numerous examples of abusive behavior from individuals who use their real names.

In the interests of ending this rant, I'll stop myself from going into the other, more social reasons I dislike this policy. I could write another rant about the right to self-define or the plight of the oppressed, but I'll spare you. I wanted to post this here because I feel like this point isn't hammered home hard enough:

The "real names" policy on Google+ has an extremely high false positive rate, and is a poor means of stopping abuse. This is too high a cost for too little benefit.

Post has shared content

Post has shared content
One common rebuttal I have heard to activism around Google+'s real names policy is that if we don't like the service we shouldn't use it. I have also heard that the real names policy couldn't possibly be such a big deal, because it has not caused a massive exodus from Google+.

So, why are pro-pseudonym activists making such a big deal about the policies of a private service that we don't have to use? Because Google+ is in a limited Beta. For a relatively brief time, Google has made this service available to some users, asked us to kick the tires, and actively solicited our opinions about how the service can be improved. Unlike Facebook, there is a possibility that Google is not wed to the real names policy. Activism on this topic right now may actually change their minds, resulting in a social networking site that is safer and more comfortable for the many classes of people who are silenced by a real names policy.

No one is seriously disputing that Google has the right to set their own policies on Google+, but they have explicitly chosen to involve Beta users in the process of deciding how this service is going to work. We would be missing an enormous opportunity right now if we simply voted with our feet.

Post has attachment

Post has attachment

Post has attachment
Audrey Penven has a new profile photo.
Photo
Wait while more posts are being loaded