A breakdown of panelists’ decisions and what went wrong.
I’ve called String Confusion Objections a clusterf@ck based on seemingly conflicting opinions by different panelists.
It’s clear that which panelist is assigned to a case has a lot to do with the outcome, which is a sign of a poorly designed system.
Whether it’s ICANN’s fault or International Centre for Dispute Resolution’s fault is up for debate, although I’ll point out ICANN gave very little instruction for determining if two strings are too similar to both be delegated.
Let’s look at the numbers (see embedded table for details).
So far 31 cases have been decided by 15 panelists. Only 3 panelists have found in favor of objectors, approving 6 objections in total. Of the 3 panelists, two have found both in favor of objectors and applicants.
Here are the three panelists that have found in favor of objectors:
M. Scott Donahey
Scott Donahey is a familiar name in the domain industry as he’s an active UDRP panelist for World Intellectual Property Organization.
He has found two cases in favor of applicants: .immobilien/.immo and .epost/.post).
But he also found that .sports will cause confusion with .sport and should be placed in a contention set.
This isn’t as simple as a typical plural vs. singular, so we can’t draw any conclusions on how Donahey would decide other cases involving plurals. In this case, he pointed out that ‘sport’ and ‘sports’ are pronounced the same in French and that .sport is a singular and collective noun, essentially making them synonymous.
Nau is the panelist that I think would find in favor of just about any objector.
He has found in favor of the objectors in both .tours/.tour and 通販/.shop.
It’s the latter decision over the Japanese IDN 通販 that’s the most baffling. Amazon claims that 通販 translates to “online shopping”, although some Domain Name Wire readers who speak Japanese say it actually means a bit more than that.
Nevertheless, I suspect most panelists wouldn’t find “onlineshopping” similar to “shop”, and when you add in a non-Latin characterset it is an even zanier decision.
If you’re an applicant and get Nau as a panelist, watch out.
Laeuchli has decided four cases, three of them in favor of objectors.
He has found .pets/.pet, .cam/.com, and .ecom/.com too similar.
He also found that .company was not too similar to .com, finding in favor of the applicant.
In the .ecom decision, he said it wasn’t fair that internet users would have to carefully scrutinize each .com domain they saw in order to verify it wasn’t .ecom. It seems that domains one letter off from each other are too confusing, and he gives special weight to .com.
In the .company decision, however, he said that internet users “are not foolish” and that “On average, they will be able to distinguish “com” from “company” – even under adverse conditions such as time pressure, distractions, and using a foreign language.”
I’m sure some applicants would suggest that if internet users “are not foolish”, then they can handle .ecom and .com and plurals.
Differences of opinion
So far both panelists who have found for both an objector and applicant give some justification for this in their decisions. It’s not like they’ve found .pet and .pets similar but .car and .cars dissimilar.
Where the problem comes in is the differences in opinion between panelists.
Some people will certainly argue that it is the panelists that are finding in favor of objectors who are the right ones and the other panelists are wrong. But it’s undeniable that different panelists would have decided the same case differently. With this in mind, perhaps the biggest mistake in the string confusion objection process was assigning only one panelist instead of three.