Anonymous Opinion is No Opinion

2 PM July 7, 2004

Anonymous opinion has zero information content.

Let me clarify that. The bulk of anonymous opinion has zero useful information content. Some anonymous opinion may contain useful thoughts, but it requires just too much effort to properly understand and judge.

In a perfect world, any statement by any person would be evaluated purely on the merit of the statement. In the real world, though, the identity of the speaker does matter. Here’s why.

Humans Communicate in Context

A simple statement such as “I took the bus to work” only communicates as much as we share context about what a “bus” is, what “work” is, what it means to “take” a bus, and who “I” am. For example, if I were to say “I took the bus to work”, I would have meant this:

I arrived at the bus stop early so I could wait in the cold bracing winter morn for five tons of clunking boneshaker that was probably going to be late anyway, then resented paying $1.80 for the trip to the train station, where I caught the train into the city.

The more you know me, the more of that you would have caught from the simple, “I took the bus to work” statement. If I had said it anonymously, it would have been completely without meaning to anybody.

Context is also important when we communicate across the Internet about professional matters. If Mr Anonymous were to write, “I am excited about Microsoft Longhorn,” I honestly would have no idea what he meant. If Robert Scoble wrote “I am excited about Microsoft Longhorn,” I would know right away.

Context is required in order for communication to occur. Identity helps establish context. Anonymous opinions are more difficult to understand because more work is required to establish the context.

Humans Trust Expertise

I read lonita’s links log. Every time I click on one of her links and enjoy it, I mentally deposit a few cents in her trustworthiness account. On the rare occasions that she has spun a dud, it was easy to forgive. It has now got to the stage where I just go to every page that she recommends, because it is probably worthwhile. Even though I have never met Lonita, I have come to the conclusion that she is an expert link hound, and I trust her judgement.

Similarly, when I read the opinion of somebody that I trust, who has some expertise in a field, I give their opinion more weight than I otherwise would. It’s not the only factor I judge by, but it is an important one.

An anonymous opinion precludes any possibility of being able to judge the expertise of the speaker. This makes evaluating the opinion more difficult.

Humans Protect Reputation

When asked my opinion on a matter I know little about, I am faced with a choice. On one hand, I could make something up, presenting half-remembered snippets as university researched world’s best practice. However, since I’d rather develop a reputation for honesty, my answer is, “I’m sorry, I don’t know much about that.”

Anonymous pundits don’t incur any penalty for dishonesty, so they may feel free to say things that are untrue, or not well thought out.

Because of this, it is harder for me to give anonymous opinion credence. I have to work harder to verify the statement’s truth before it crosses the threshold of believability.

Humans Converse

I find the to and fro of ideas that is engendered by a conversation stimulating. A typical blog “conversation” might stretch over ten people, three posts and twenty five comments and a few days.

A conversation is much more than the sum of the ideas expressed and it’s difficult to include someone when they are so rude that won’t even give a real name.


If you have something intelligent to say, saying it anonymously does not help your case. On top of finding your anonymity impolite, people will have trouble accepting your ideas because they don’t know your background, they cannot evaluate your expertise and they suspect your motives.

All this because you won’t leave your real name.

I recently gave up trying to evaluate anonymous opinions. It’s too hard, so I don’t bother. Because I don’t bother, they have zero information content. If you wrote an anonymous comment in order to communicate something to me, you may as well not have written it.

By alang | # | Comments (10)
(Posted to Rants and javablogs)


At 14:19, 07 Jul 2004 Anonymous Coward wrote:

bah :)

At 15:45, 07 Jul 2004 Doug wrote:

Hmm... I definitely agree with the general point of the article: that statements of opinions are worthless without knowing who made the statement.

However, I'm not completely with you when you extend that to requiring a "real name". Particularly when you point to Lonita's weblog as an example of opinions you trust, even though it's not at all obvious what Lonita's real name is (it may be somewhere on the Web site, but I didn't see it in a quick once-over).

A consistently used pseudonym is, I think, as reliable as a "real name" for the purposes of evaluating a stated opinion. Of course, it's trivially easy to forge a name on wikis and weblog comments, so reliability is a bit shaky anyway.

Personally, when I leave comments on other Web sites (as I am doing here), I sign them as just "Doug" because that's how I like to be called. It also has the salutary effect of keeping my scattered comments from cluttering up a Google search under my real name. For context (and disambiguation) I always include my Web site URL. Where, if anyone REALLY cares, my real name can also be found.

At 17:08, 07 Jul 2004 The Badger wrote:

I think you really have to evaluate the information communicated before making policy-based judgements on whether a person has anything interesting to say. Quite often, people will respond to an anonymous comment dismissively ("if you won't put your name to something, blah, blah"), yet the information could well be a ticket to board the cluetrain for that person if only they'd read the text of the comment.

These kinds of policy-based judgements are counterproductive in quite a few situations, too. For example, corporate whistleblowing gets people fired if they put their name to a complaint about unethical behaviour, whereas anonymous tip-offs convey essential information in such cases.

At 23:49, 07 Jul 2004 Alan Green wrote:


My post confounds the notions of identity and "name". I plead that the confusion is tangential to the main thrust of the post and made it shorter.

I'd like to write some more about that - if I ever get inspired. Meanwhile, Charles has a helpful policy on blog commentry that he derived from the C2 WikiWiki's policy:

At 00:00, 08 Jul 2004 Alan Green wrote:


It takes time to evaluate a statement. If I have a thousand statements to choose from - such as I get in my daily list from - I have to prioritise which I am going to bother with. Anonymous opinions start very low down that list.

Anonymously given facts can be judged more easily than anonymous opinion. But they can only be used to point oneself in the direction of a clue-train, they must be cross-checked. This is not the case when I read something written by somebody that I trust.

I have to ask: what possible reason might you have for making your comment under a pseudonym? What and why are you hiding? By replying, am I feeding some kind of psychosis or personality disorder you might have? Who are you? :)

At 08:10, 08 Jul 2004 Lonita wrote:

I saw this post in my referer logs so I came to peek. Thank you so much for the praise. It was very happy-making and very kind of you to say. :>

At 22:35, 12 Jul 2004 Mark Ziebell wrote:

These comments seem much stronger then the "real-names-please". I thoroughly agree that comments in a discussion should be referenced (possibly because I regard it as "good manners" which is part of your point).

However the comments raise the question of the identifier being part of the data, not the meta-data. The data is not be valid if not complete, but data can survive and have meaning without meta-data.

So why present a UI that allows an incomplete submission (some people write their own software, I believe, because of such annoyances).

Also the issue of when do I know the identity and how does that affect my consumption of the message is interesting.

With information overload we only have time to look at a small percentage of information presented let alone actually analyse and cogitate. Although I would confess I am not the world's greatest listener, I do think there is a danger that we choose only to receive information from those who we think do indeed think like we do.

So I hope the rejection of anonymous information is on the grounds of "bad manners", not because we wish to prejudge the status of the information on the basis of our (prejudiced) view of the source. (Although over a beer I am sure I would confide a there are more than a few commentators who I don't bother reading because I regard them as idiots -- but since when was I expected to be consistent).

At 23:57, 14 Aug 2004 Not a Child of the Web wrote:

Alan asks "what possible reason might you have for making your comment under a pseudonym? What and why are you hiding?".

It strikes me that there are several reasons, that don't require us to jump to a hypothesis of "psychosis or personality disorder".

As had been already argued, some people use pseudonyms rather than their real names because that IS their identifier. They consistently use the same name. Lonita may well be one of these people. It really isn't relevant whether Lonita is a real name.

To me, the real answer for not using my real name is "privacy + Google". Let me explain.

I use Google to check out job candidates and potential (real life and cyberspace) friends. I have to expect others will do it to me.

When I Google myself, I find the questions I asked and the opinions I expressed 10 years ago to be mildly embarassing; like having your mother show your friends pictures of you when you were 12. I can't deny that *was* me, and I was acting honestly and in good faith at the time, but it is not me *now*. I don't want others to judge me on my behaviour *then*, I want them to judge me on my behaviour now.

Furthermore, I think that every person has different facets of their personality that they show to different groups - my colleagues, my friends, my family, my lovers - they all know a different part of me. Google doesn't respect that distinction. George Costanza describes the feeling of his world's colliding.

So protect my 10-years-in-the-future me and to protect my worlds from colliding, I use pseudonyms on the Internet. In fact, my pseudonym here is based on a comment Casey made once, explaining that he was now a "child of the Internet", and felt no need for the privacy of pseudonyms. I hope he still feels that way in 10 year's time.

I can understand that humans are forced to use hueristics to decide what they will read and trust. I hope, however, that the arguments I made here are valid no matter who I am.

At 17:51, 26 Aug 2004 intentionally anon wrote:

Nonsense. Take the comments on their own merit. It doesn't matter who made them. I like anon commenters because they speak the truth much more than anyone who is protecting their name and reputation.

At 09:28, 12 Jul 2007 Burk wrote:
 I think opinions show no deference if they are anonymous or not. I think an individual who is pre occupied with this is of an antiquated and untrustworthy character.

Add Comment

(Not displayed)

(Leave blank line between paragraphs. URLs converted to links. HTML stripped. Indented source code will be formatted with <pre> tags.)

© 2003-2006 Alan Green