question about reputation

Timothy C. May tcmay at got.net
Mon Sep 11 09:44:42 PDT 1995


How reputation systems work is an important issue. I hope we can discuss it
further.

At 10:32 PM 9/10/95, Wei Dai wrote:

>In an economy based on positive reputations, how does one acquire a
>reputation capital?  One way may be to initially perform services at a
>price below cost, but this has some problems.
>
>For example, Alice starts a anonymous consulting service, and announces
>that she will answer the first ten queries for free.  Upon hearing this,
>Mallet immediately starts another consulting service, and announces the
>same offer.  At this point Mallet can simply forward his customers'
>queries to Alice and Alice's answers back to his customers.  Thus, he gains
>reputation at no cost.
>
>On the other hand, this "man-in-the-middle" attack can also work against
>conventional True Name based services, but perhaps with less effect.  Has
>anyone ever heard of this being done?

This line of reasoning is a variant of the "Chess Grandmaster" scheme,
wherein one gains the reputation of a chess grandmaster by echoing the
moves of a chess grandmaster playing in another game. I believe this is
described in various crypto books, but I haven't looked it up here.

In practice, nothing so simple as an "Ask any 10 questions" would be a
practical way to gain reputation. If Alice is trying to build up a rep,
she'll choose her customers with some care, or make sure that alternate
channels also exist.

>Is there a better way to acquire a good reputation?

Directly contacting a larger group, such as this list, is usually a faster
and better way to build a reputation than in, say, answering only questions
directly made. Admittedly, many consulting cases involve direct contacts.

However, the reputation of a Pr0duct Cypher, or a Black Unicorn, or a Tim
May, or a Wei Dai, is usually made in a public forum, not primarily in
one-on-one contacts. While not all of us sign our posts, the principal is
roughly the same: we are communicating directly, so man-in-the-middle
attacks, or "Chess Grandmaster" attacks, are fairly ineffective.

And practically speaking, if someone hires _me_ as a consultant, it is
probably based on past achievements, through multiple channels. This would
apply to digital pseudonyms as well, though not as directly.

(Humorous Sidenote: an informal variant of the Chess Grandmaster approach
is to use the best arguments found on one list on _another_ list. Another
variant, widely used, is to adopt the best arguments of others and use them
one's self (oneself?). This is how memes spread, and is central to the
advancement of knowledge.)

--Tim May

---------:---------:---------:---------:---------:---------:---------:----
Timothy C. May              | Crypto Anarchy: encryption, digital money,
tcmay at got.net  408-728-0152 | anonymous networks, digital pseudonyms, zero
Corralitos, CA              | knowledge, reputations, information markets,
Higher Power: 2^756839      | black markets, collapse of governments.
"National borders are just speed bumps on the information superhighway."








More information about the cypherpunks-legacy mailing list