Systematic Errors in NAF Complaints and Decisions
Date: Monday April 26 2010, @07:18AM
Topic: Uniform Dispute Resolution Policy (UDRP)

GeorgeK writes "In a recent article at, CitizenHawk was called out by a National Arbitration Forum (NAF) panelist for the submission of automated complaints which contained complete nonsense. Through the discussion in the comments to that article, the community discovered that the problem is far deeper.

It turns out that UDRP panelists at NAF have been churning out boilerplate cut/paste decisions of their own, with utter nonsense of their own, and that this has been going on for years.

In a recent decision involving the domain name, there was a shocking result that the complainant proved their case, but instead of transferring the domain name to the complainant, the domain name was instead cancelled. The community suspected that a cut/paste job was to blame, and indeed keen observer Andrew Allemann noticed that the decision contained text completely unrelated to the case, saying that:

Complainant held a trademark registration for “AOL” and Respondent registered the domain name “”).
This was obviously pulled out of some other unrelated case, as the complainant was WOOT, not AOL.

Looking into other cases by panelist Nelson A. Diaz (the panelist in the case), it was discovered that in the decision, there was some very nonsensical language in the decision, namely that:
Having established all three elements required under the ICANN Policy, the Panel concludes that relief shall be DENIED.
This is saying that the complainant proved their case, but is declared a loser. Through analysis of other NAF cases, it appears that originally a panelist years ago wrote the following language:
Complainant not having established all three elements required under the ICANN Policy, the Panel concludes that relief shall be DENIED.
Note the words "Complainant not" at the beginning of the sentence, which makes the sentence perfectly fine and logical. Over time, the word "Complainant" was dropped, which still kept the meaning of the sentence intact. However, at one point a panelist, probably because they were cutting and pasting from other decisions, dropped the word "Not" too! Removing that word made the sentence completely illogical.

Was this a fluke? It turns out not to be the case at all, and UDRP panelists have been systematically copying and pasting this nonsensical language into decisions. I found 41 cases of this nonsense alone, in the following cases:
1302612 1299752 1297995 1270184 1267051 1265798 1260576 1259275 1247548 1236932

1236721 1202940 1195739 1182336 1172987 1158551 1140580 1128992 1128482 1122960

1065183 1048716 873313 823027 818403 811632 804703 662187 639195 587435

555041 550345 514784 488845 469107 352423 289033 250240 237522 221239

Some of these were even 3-person panel decisions, where one would expect greater scrutiny and care.

As Zak Muscovitch has documented in a recent study, there are certain NAF panelists who handle a disproportionate number of cases. They're churning out decisions like a sausage factory. In my opinion, this use of "boilerplate text" in the decision-making process of panelists demonstrates that some of the "ingredients" that are going into the sausages are not kosher.

This use of boilerplate text goes directly to the issue of systematic bias against domain name registrants in UDRP cases by some panelists. It is obviously a lot more work for them to deviate from that template. Given the choice of doing extra work (for no extra money) or simply cutting corners and filling in the blanks, the economic incentives for panelists would lead to excessive rulings against domain registrants. I call upon ICANN to seek an explanation from NAF, who has a very interesting history. Indeed, a case might be made that they should be decertified as a provider.

For a long time I have also called for UDRP decisions to be in standardized XML formats, so that academic researchers could perform better analysis of decisions. There are over 2000 matches in Google Scholar for the term "UDRP", but only a handful that scrutinize the decisions themselves. UDRP providers successfully argued for and achieved mandatory electronic filing of complaints, in order to save themselves money, but they have failed to output their own decisions in a standard format so that their work can be scrutinized. At last week's ICANN Board meeting, the Board disussed whether UDRP providers need to be brought under contract with ICANN. As is typical, ICANN made no resolutions to fix UDRP problems or even compel providers to enter into contracts with ICANN, in order to ensure accountability. ICANN should require UDRP providers to provide their decisions in a standard format. I believe that one of the studies that can and should be conducted is the extent to which UDRP panelists are using boilerplate decisions and the contribution of this to bias against domain registrants. ICANN has subsidized the TM lobby by funding travel for IRT members (which turned out to be a big waste of money). For fairness, ICANN should fund studies like those of Professor Geist which looked into how the UDRP is being misused, and not adequately protecting domain name registrants (consumers). has made a followup article that also covers this story.

Last week another study of NAF was published by the National Consumer Law Center, in the context of consumer arbitration. With a US Supreme Court case being heard this week involving consumer arbitration, this might force ICANN to rethink the UDRP completely."

This discussion has been archived. No new comments can be posted.
Systematic Errors in NAF Complaints and Decisions | Log in/Create an Account | Top | 1 comments | Search Discussion
Click this button to post a comment to this story
The options below will change how the comments display
Check box to change your default comment view
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
A quick update on the story
by GeorgeK on Wednesday June 16 2010, @08:20AM (#17008)
User #3191 Info |
It turns out this story is still ongoing, see: past_udrp_decisions/

Now they're actually revising past UDRP decisions!
[ Reply to This | Parent ]

This article comes from ICANNWatch

The URL for this story is: