Re: Peer Performance Ratings LO10400

From: John Conover <>
Subject: Re: Peer Performance Ratings LO10400
Date: Wed, 9 Oct 1996 21:32:32 -0700

Alex J. Muro writes:
> I would like feedback from any companys that have implemented peer-rating
> systems for front-line employees, particularly those in craft areas.  What
> are the results?  What can a company expect when it starts?  How long
> before the employees start feeling comfortable with giving and receiving
> performance-improvement comments?

Peer review is a zero-sum game.

The "points" one player gives away, another player wins. There is a
finite number of "points" that can be won, and more than one player,
which makes it a zero-sum, (and, presumably, iterated,) game. The
optimal strategy in such affairs is to give away all your "points" as
long as all other players do the same and cooperate in the same
fashion. As soon as any player defects and does not give the same
"points" to you that you gave to the player, you defect, and give away
no "points," in retaliation, a la the Prisoner's Dilemma.

This probably accounts for why peer review does not work very well
over an extended period of time-ie., it is not a durable process,
(except where all the players are thinking-challenged, and can not
figure out the strategy,) and there is only one stable, final,
solution to the process-everyone playing a defection
strategy. Unfortunately, that solution is not amenable to a team

Any time an iterated process has a single, negative, stable solution
it is called a "system problem."

Other than that, I don't see anything wrong with peer review.


BTW, I worked for a company that implemented peer review at the
executive level. The above scenario played itself out, as predicted.
To begin with, everyone gave everyone else "points," and nothing was
accomplished, (in the way of problem solutions,) since everyone wanted
to be cooperative. (It was a pleasant environment, though.) Then, the
inevitable happened, and someone made an insignificant defection,
(which, IMHO, was justified, in the specific case,) and in a
landslide, everyone defected.  (And, it became a hyper-political
environment-just as predicted.)  That's what made me start to consider
peer review as a zero-sum game. Similar comments could be made about
the "budget process," for that matter. If you want to make a political
agenda that turns in on itself, and becomes the corporate agenda,
(assuming a life of its own,) set up zero-sum processes.

As a passing theoretical note, it is possible to set up a process for
such things that is not zero-sum, but it requires enormous intuitive
skill and capability, and the process can not be axiomatized. It can,
also, be shown that it is impossible to make a set of rules that can
be used to implement such a process, so it can not be generalized, or
taught. (This is a consequence of Godel and Turing-if we could do
that, we could write a computer program that would invent the next
theorem of mathematics-and they formally showed that this was
impossible.) Also, there may be formal issues regarding whether such a
process would ever finish, or whether there is any formal process by
which priorities could be defined, (which is doubtful, because of
Arrow's so called Impossibility Theorem.) At least if you expect the
process to be self-consistent, (ie., not contradictory within itself,)
and/or complete, (ie., handle what it was designed to handle.)


John Conover,,

Copyright © 1996 John Conover, All Rights Reserved.
Last modified: Fri Mar 26 18:56:08 PST 1999 $Id: 961009213254.1075.html,v 1.0 2001/11/17 23:05:50 conover Exp $
Valid HTML 4.0!