RE: [gnso-vi-feb10] Group on documenting "harms"

[Milton L Mueller <mueller@xxxxxxx>]

OK, need to partially retract my last message. (flip) If all Avri is advocating 
is a rough assessment of likelihood, and an analytical separation of harms and 
likelihoods, that's fine. Of course, we will all disagree on the likelihood, 
and there will be no objective way to settle those disputes - I think that's 
what Tim was saying. (flop) But at least we will know where we disagree. And 
those who moan about harms will have to be a lot more specific and make their 
case. (flip)

--MM

> -----Original Message-----
> From: owner-gnso-vi-feb10@xxxxxxxxx [mailto:owner-gnso-vi-
> feb10@xxxxxxxxx] On Behalf Of Avri Doria
> Sent: Wednesday, July 28, 2010 10:50 AM
> To: Gnso-vi-feb10@xxxxxxxxx
> Subject: Re: [gnso-vi-feb10] Group on documenting "harms"
> 
> 
> Hi,
> 
> In one sense you are right.  No one ever knows what the future holds. So
> maybe we should just quit where we are ahead and say, gee no idea.
> 
> But in another sense this is a normal risk analysis process in the face
> of unknowns.
> 
> And indeed what we are talking about in terms of harms is risk analysis.
> And no analysis of the future of a system I know of, avoids risk
> analysis because there are unknowns.
> 
> The reason I advocate that everyone do this exercise and that there be a
> function of two values (harm, likelihood) is because in averaging the
> results of many views and in determining the range (or stddev) we get an
> idea of what the risk might be and to what extent there is agreement of
> that determination.  This is what people (especially those in business -
> if i remember my 25 years in industry correctly) do in the face of
> uncertainty.
> 
> We can also do sub group analysis if it is of the value to see if there
> is a different average from one type of respondent than another and to
> determine whether their range/stddev is any different.
> 
> a.
> 
> On 28 Jul 2010, at 16:26, Tim Ruiz wrote:
> 
> > For me the point is, that regardless of experience, background, etc.
> of
> > any of us, there is no objective way to determine which harms are more
> > likely to occur than others. So much of it relies on business models,
> > interests of the players, etc., and new ones crop up all the time. I
> > think trying to determine likelihood of a harm will get us skewed off
> > into never never land again.
> >
> >
> > Tim
> >
> > -------- Original Message --------
> > Subject: Re: [gnso-vi-feb10] Group on documenting "harms"
> > From: Eric Brunner-Williams <ebw@xxxxxxxxxxxxxxxxxxxx>
> > Date: Wed, July 28, 2010 8:58 am
> > To: "Neuman, Jeff" <Jeff.Neuman@xxxxxxxxxx>
> > Cc: "'avri@xxxxxxx'" <avri@xxxxxxx>, "'Gnso-vi-feb10@xxxxxxxxx'"
> > <Gnso-vi-feb10@xxxxxxxxx>
> >
> >
> > On 7/28/10 9:31 AM, Neuman, Jeff wrote:
> >
> >> ... those that never operated a registry and a registrar together ...
> >
> >
> > I agree that assertions offered without operational experience have
> > limited value. However, the scope of relevant operational experience
> > goes to more than registry and registrar operations. The secondary
> > market, in which some (few) registrars function as registries, and
> > some (many) domainers, including some (few) registrars, function as
> > registrars, offers insight into the design of systems in which the
> > operational experience has value.
> >
> > Eric
> >
> >
> >> ----- Original Message -----
> >> From: owner-gnso-vi-feb10@xxxxxxxxx<owner-gnso-vi-feb10@xxxxxxxxx>
> >> To: Gnso-vi-feb10@xxxxxxxxx<Gnso-vi-feb10@xxxxxxxxx>
> >> Sent: Wed Jul 28 01:48:29 2010
> >> Subject: Re: [gnso-vi-feb10] Group on documenting "harms"
> >>
> >>
> >>
> >> On 28 Jul 2010, at 03:06, Antony Van Couvering wrote:
> >>
> >>> Let's prioritize for harms that are dangerous AND most likely to
> occur.
> >>
> >> I think after the Harms Sub Team lists all of the possible harms,
> setting these two values may be a good use for another of Mikey's polls
> were we each rate the degree of harm (H) and the likelihood of the harm
> occurring (L) on a 5 point scale.
> >>
> >> then to arrive at the ranking factor = H * L
> >>
> >> and then averaging and showing range for each defined harm.
> >>
> >> cheers,
> >>
> >> a.
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >
>