[gnso-rpm-wg] FW: Inferences (was Re: Mp3, Attendance, AC recording & AC Chat Review of all Rights Protection Mechanisms (RPMs) PDP Working Group)

[George Kirikos]

Hello,

On Mon, Jul 17, 2017 at 2:00 PM, Greg Shatan <gregshatanipc at gmail.com> wrote:
> Of course, this proves that we actually "know" nothing of the sort.
>
> First, as far as I know, we have no idea whether the HOTEL or HOTELS TMCH
> registrations represent "gaming/abuse" (see my earlier email in this thread
> on "false positives").  These could be bona fide trademark registrations and
> bona fide TMCH registrations for all we know.  (As also noted in my earlier
> email, HOSTING and THE do appear to represent gaming/abuse.)

This re-raises the issue of who is in denial here, or whether this is
"Groundhog Day" or "50 First Dates". "HOTEL" was analyzed back in
March 2017, in the post at:

http://mm.icann.org/pipermail/gnso-rpm-wg/2017-March/001119.html

(section 6) and the same analysis applied to "HOTELS" (i.e. the
screenshots showed HOTELS.TLD strings also gamed and for sale), and
brought up other times too.

> Second, trying to manufacture a percentage of sunrise domain registrations
> based on "gamed/abused" TMCH records in the "average Sunrise" is artificial
> and useless.  We have no idea how many "gamed/abused" TMCH records have
> actually been registered in any given Sunrise (or the "average Sunrise"),
> and no information on which to base any assumptions.  So that "0.13" is a
> junk number, and shows nothing about "higher rates for gaming of the
> sunrises."  The only percentage that we could even hazard a guess at is the
> total number of "gamed/abused" records in the TMCH as a percentage of all
> TMCH registrations.  Taking the two known likely gamed/abused marks (THE and
> HOSTING) and dividing by the total number of TMCH records (approximately
> 40,000) yields a percentage of 0.00005.  Of course, that doesn't prove your
> point, but it's a more valid number than 0.13.  Even you assume 200
> gamed/abused records instead of 2, the percentage is still only 0.005
> (one-five hundredth of one percent).


You're trying to choose a larger base/denominator (percentage in the
TMCH, vs. percentage from the sunrise), when we've never been given
the opportunity to examine the entire TMCH database (which would
change the numerator). The minimal data we've had access to is from
the actual sunrise registrations, so that's why it belongs in the
denominator, as opposed to the total TMCH registrations.

Furthermore, your math is off by a factor of 100. When taking a
percentage, one divides the two numbers and then multiplies by 100.
e.g. 2 (your number, not mine) divided by 40,000 = 0.00005, = 0.005%
Of course, the correct numerator is not just "2" (it should be
higher), and the correct denominator is not 40,000 (it's more like
130).

> infringing.  Assuming for the sake of argument that this is also 20%, each
> UDRP represents roughly 25 infringing domains.  So, taking your assumption
> of a 0.1% rate for bringing UDRPs, that means that roughly 2.5% of all
> domains could be considered infringing.

I wasn't assuming 0.1% is the exact rate --- that was an upper bound,
for simplicity. UDRP rate is at best 10,000 domains per year in a UDRP
divided by 150 million domains, which is 0.007% (i.e. 0.00007 x 100%).
I used highly conservative upper/lower bands previously, just to get a
rough picture.

Also, one needs to understand that not all crime reaches a level
beyond "de minimis" damages (i.e. annoyance vs. damage). In retail,
there's the concept of "shrinkage", due to employee theft,
shoplifting, vendor fraud, etc. A certain percentage is inevitable,
and won't be prosecuted because it's so minor. The focus is on the
more serious cases, e.g. someone steals 100 iPhones, instead of the
granny eating a grape at a grocery store and not paying for it.

Many of the RPMs appear to be a disproportionate response to the
"granny eating a grape", rather than the the theft of 100 iPhones.
When the theft of 100 iPhones does take place, there are already laws
on the books to handle that situation (e.g. ACPA, or other
litigation).

Sincerely,

George Kirikos
416-588-0269
http://www.leap.com/