On Wed, Apr 15, 2009 at 2:10 AM, Tony Li <[email protected]> wrote: > William Herrin wrote: >> On Tue, Apr 14, 2009 at 11:33 PM, Tony Li <[email protected]> wrote: >>>> Again, the poll can be found here: http://doodle.com/9sybb8dmk5phvp99 >>> >>> The poll has now closed. The ayes have it, 10 to 4 (71%), which is >>> reasonably clear consensus. The definitions listed are hereby adopted. >> >> 10 people would have to hum awfully loud to form a consensus among the >> crowd RRG draws at the meetings. > > Not if the rest of them were silent, which is effectively what has happened.
Hi Tony, Had only two people responded and both said "yes", you'd know better than to call it a consensus. I'm not sure why 14 respondents including dissent should suggest a different conclusion. A consensus check is not a vote. Consensus is an affirmative finding, not a balance between affirmative and negative. Unlike a vote, there are three possible outcomes: consensus for, consensus against and no consensus. For a weak consensus (or "rough consensus" if you prefer the IETF parlance) you're looking for one of two situations: Situation A: 1. The stakeholders are present 2. A simple majority join the consensus. 3. No more than a single-digit-percentage join the dissent Situation B: 1. The stakeholders are present 2. A significant minority join the consensus. 3. No dissent. If the meeting attendance is any guide, 14 respondents is not a majority of the RRG. Further, there was nearly 30% dissent. Your methodology was in error. By failing to consider the stakeholders present but non-responsive, you pre-eliminated the possibility of "no consensus," requiring respondents to either vote for or against the particular definition. That's not a consensus; that's a vote. Corrected for non-response, the proper conclusion from your data is: No consensus. Regards, Bill Herrin -- William D. Herrin ................ [email protected] [email protected] 3005 Crane Dr. ...................... Web: <http://bill.herrin.us/> Falls Church, VA 22042-3004 _______________________________________________ rrg mailing list [email protected] http://www.irtf.org/mailman/listinfo/rrg
