Minutes | Meeting of the New gTLD Program Committee 29 May 2012

Note: On 10 April 2012, the Board established the New gTLD Program Committee, comprised of all voting members of the Board that are not conflicted with respect to the New gTLD Program. The Committee was granted all of the powers of the Board (subject to the limitations set forth by law, the Articles of incorporation, Bylaws or ICANN's Conflicts of Interest Policy) to exercise Board-level authority for any and all issues that may arise relating to the New gTLD Program. The full scope of the Committee's authority is set forth in its charter at http://www.icann.org/en/groups/board/new-gTLD.

A telephonic Meeting of the New gTLD Program Committee of the ICANN Board of Directors was held on 29 May 2012 at 18:00 UTC.

Committee Chairman Cherine Chalaby promptly called the meeting to order.

In addition to the Chair the following Directors participated in all or part of the meeting: Chris Disspain, Bill Graham, R. Ramaraj, George Sadowsky, and Mike Silber.

Thomas Narten, IETF Liaison and non-voting liaison to the committee, was also in attendance.

Rod Beckstrom, Erika Mann, Gonzalo Navarro, Ray Plzak, Kuo-Wei Wu, and Thomas Roessler, TLG Liaison and non-voting liaison to the committee, sent apologies.

Heather Dryden, GAC Liaison, attended as an observer.

  1. Batching Process


  1. Batching Process

    The Chair introduced the agenda, with a focus on the batching process.

    Mike Silber noted that the Risk Committee is discussing risk identification and mitigation work in relation to the batching process.

    Kurt Pritz presented background on the consideration of geographic diversity in formation of the batches, including the round robin process. When the round robin model was discussed with the GNSO in Costa Rica, there was concern raised about its fairness, as regions with smaller proportions of numbers in the it would assure that that geographic regions with a smaller percentage of applications overall would have a greater number of applications in the first batch than a strictly proportional system would assure. There was also a suggestion that the process could be susceptible to gaming. For example, if there was a favoring of geographical location or a community TLD, the applicants may go back and adjust applications to take advantage of that. In addition, the issue of fairness in demoting applications that would have otherwise made the first round for those who "won" at digital archery, in favor of those who were promoted in the first round by virtue of round robin, was a community concern.

    Kurt continued that the issue of geographic diversity seems to be geared to ensure that, at minimum, if x percent of applications from a particular geographic region, at least x percent of the applications in the first batch should be from that particular region.

    George Sadowsky asked Kurt if any statements were made in Costa Rica that would encourage applicants to go back and restructure their applications, such as a change to round robin or moves to proportionality.

    Kurt noted that the proportional model had been posted as the method in the "Batching Basics" paper that was posted after Costa Rica, and no statements had been made about changes to that model. Though the proportionality model was posted, if the Board wishes to change the model, it could. There would be risks to adopting a different model, however, particularly with the community comment on fairness and potential demotion if round robin was used.

    George commented that all members of the Committee believe that we should be as fair as possibility to applicants. But that depends on whether there is a level playing field to start with. Currently, there are no IDN gTLDs, no gTLDs from certain regions, so even trying to move forward with a proportional representation model perpetuates a nonlevel starting point. George noted that he voted for the digital archery process because he believed that it was a round robin method that would ameliorate the disadvantage of some of the regions. He is against proportionality.

    Bill Graham commented that he does not see why going to the round robin method would be potentially discriminatory, and does not see the basis for claiming unfairness unless all digital archery scores are released. It would actually serve to remove questions about fairness, as it would remove concerns about whether a response from one geographic region could get in as fast as a response from another geographic region. There seems to be a game in defining unfairness within the community.

    Chris Disspain commented that the more human interference that is introduced into the process, the more likely it is that a problem will arise. If there's a simple statement at the outset about the rules of the competition, that's fine. But if we start changing the outcome based on the results, there is a risk to that. Chris also inquired about how ICANN determines which geographic region an application is from.

    Mike noted that digital archery had been explained as a full concept, not the digital archery and a methodology for manual batching to be applied afterwards. At this time, we cannot just change one portion of the process. The proportional representation is a manual intervention, and the Board would need to vote to approve that change. The round robin method should remain.

    Ramaraj agreed with Mike.

    Chris asked about the possibility of gaming and whether it is possible for applicants to change their applications now to try to take advantage of any change.

    Kurt responded that applicants already submitted their response to a question seeking the place of organization, where there has to be some sort of license to do business there. The ability to change that location would close with the close of the application window.

    Chris clarified that this means that an organization would already have to have set up within a specific region in order to be associated with it as part of their application process.

    The Chair queried as to whether the country identification selected would apply in either the proportional model or the round robin model, and Kurt confirmed this was the case.

    Kurt further responded that there may be a difference in how effective a geographic designation is in getting into the first batch of applications based upon the different method used. For example, if there are 200 applications from one region, and a total of 2000 applications, the proportional model would assure that an applicant for that region would have a 10% chance to get into the first batch. Under round robin, an applicant from that region would have a 50% chance.

    Chris asked if allowing proportions to be modified later was more acceptable than a straightforward rule that says that we're taking these applications in regional order.

    Kurt repeated that there was a risk associated with demoting applications who achieve a digital archery score among the top 500, who merited position in the first round under a pure digital archery system.

    Chris countered that demotion does not occur if the rule is based upon having slots associated to geographic regions. There's little difference between setting that rule, or allowing applicants who achieved a lower digital archery score to move into the first batch by virtue of string contention. There's little difference, whether you are adding to the batch or not including people in the batch. Chris requested that the Committee also discuss the possibility of batches of larger than 500.

    Thomas Narten agreed that we have to be clear about defining "fair" and "fairness." One definition of fairness could be very different from another. Second, the place to have the discussion about the issues that may arise with the round robin method is to have the debate within the community. This seems to be change to the batching process as it understood by the community, and making a change of this type without making it clearly understood and assuring it is supported is not good. If a change is made to the process, that is OK, but it has to be done in a way that is clearly communicated.

    Kurt noted that the round robin was never published to the community as a method of formulation of batches; there has not been a change.

    Thomas Narten countered that it is well-known that the Board was discussing a round robin.

    Kurt responded that we've been very careful on what was communicated on the process, as the process had not been approved by the Board. The round robin method is not well-known, and was not really discussed until the days leading up to the Costa Rica meeting.

    Thomas Narten noted that it was discussed in Costa Rica and it is the operating assumption of the community. If we change from that model, we are making a change, so even if it is not technically policy in the GNSO sense, this is the kind of thing where you do not make a change unless you have a good reason.

    Kurt commented that the Board did not talk about geographical diversity in Costa Rica; the meetings were concentrated on the robustness of the digital archery model and the comparison of digital archery to other models. To reiterate, this is not a change to what has been discussed with the community.

    Thomas Narten countered that because the round robin was discussed publicly in Costa Rica, it seems that that is what the community thinks is being implemented.

    Kurt noted that there has never been a written document stating that ICANN is implementing the round robin model.

    Chris clarified his understanding that the topic was raised in a discussion was raised in a session with the GNSO in Costa Rica and Kurt received significant pushback.

    Heather Dryden offered comments to build on what Thomas Narten was saying, regarding the assumptions within some parts of the community. There is not a GAC view that Heather is able to communicate on the round robin versus proportionate model, as the GAC is in the process of compiling questions. There needs to be attention on what the information flows and communciations are on this topic and to be sure the community is well informed.  If there a decision was made to support digital archery on the basis of certain assumptions, then that is what the community is going to process, including inputs from those at the GNSO session referenced. The main concen is that we seem to building decision upon decision, while not being clear about the assumptions into the first decision. In relation to the GAC, they have a lot of questions, and sensitivities around making sure that batching works and its timed in a way that makes sense. Hether noted that today's discussion is giving her pause.

    The Chair noted that many are being given pause with this discussion, as there is disagreement about the assumptions in the first decision. Thomas Narten stated tha the community has an assumption that batching will be done on the round robin model. Kurt noted that that round robin was only discussed in a GNSO meeting as a potential method, and a lot of people at that meeting raised concerns about it being open to gaming. The Committee needs to ascertain if this is going to be a run on a round robin basis of not. The Chair asked Thomas Roessler and George if the subgroup that looked at batching after Costa Rica discussed the methodology.

    George recalled that round robin was presented as the means for assembling the queue for digital archery, however the real focus of discussion was between auctions and digital archery.

    Mike and Chris agreed with George's recollection.

    Mike continued that the round robin solution was explained to the subgroup and that is what the subgroup agreed to.

    Chris continued that in the choice between digital archery and auction, there was a discussion about regionalizing the digital archery and a region-by-region basis. There was not an explicit statement that a regional round robin was going to be used, but there was not discussion of the proportional model at that time.

    The Chair commented that it is hard to launch the system and allow submission of digital archery, and then announce later how the batching will work. The announcement of the method has to be done at the same time. We have to provide the rules.

    George noted that the sooner we get the rules out, the better everyone will feel, but we could launch the batching system without having provided the notice in advance.

    The Chair commented that if he were to call for a vote today among the members of the committee present, the vote would be for round robin, as some believe that is what was already agreed to, or what the community is expecting.

    Chris commented that it may be of help to have staff try to address any issues that may arise if the Committee were to direct staff to proceed with a round robin. The Committee heard that there was pushback that this system could be gamed. Are there are any other consequences to returning to the round-robin model?

    The General Counsel and Secretary noted that the "Batching Basics" paper describes the proportionality method, so there are some who believe that the process will be round robin, but we've published that we'll be following proportionality. Of course, we've also stated that we are not providing operational details until later, but we've already published some operational details. The recommendation at this time is go back, determine what has been said where and when on this, and see if that provides further clarity.

    Mike stated that it would take a lot of nuance in reading the "Batching Basics" paper to try to argue that round robin is not a proportional system as stated in the paper. Mike noted his opinion that "Batching Basics" could describe either proportional or round robin distribution.

    Bill agreed with Mike, that "Batching Basics" does not unequivocally state that only a proportional representation model will be used. Based on the language in the paper, you'd have a hard time insisting that it refers only to one or the other.

    Chris and George agreed with Bill. Chris noted that there had been articles written describing the five region approach.

    The General Counsel and Secretary confirmed that a number of articles were released that summarized Kurt's discussion to the GNSO; some people in the community took the position that the process was defined, though ICANN was saying operational details were still to come. Kurt clearly said that it was a proposal, though that was not how the media reported it.

    The Chair asked again what does the Committee believe is the right system to introduce, including considerations of gaming that have been raised? Does proportional give a fairer result? What about George's comments regarding a level playing field?

    The General Counsel and Secretary confirmed that applicants are not able to change their address in the system at this time; the last day to change an address was 29 March 2012, so it is not now possible to change an address as an aspect to gaming.

    Mike noted that round robin will produce greater geographic diversity than the proportional method.

    George reiterated his point that we do not have a fair situation now, so basing a decision on an assumption of fairness will not hold.

    Kurt re-stated Thomas Narten's point that everyone has a different sense of fair. The question now is what should the result be? A lot of the pushback to date come from a perception of relative demotion, from those who scored within the top 500 but did not get into the first round.

    Chris queried whether, if you're clear going in about how the batch will be comprised, then there is no real "demotion" – that is only if we are going around and moving slots around.

    The General Counsel and Secretary returned to Chris' earlier example of the contention sets being moved into the same batch; there, only on of the applications in contention can be delegated, which allows us to expand the size the batch. 

    Chris asked if there was any concern in proportionalizing the batch manually, as opposed to using round robin? How will the numbers of strings in contention sets be counted for the purpose of identifying proportion percentages? These are the types of questions that have to be answered for clarity in a proportional system. With a round robin, you just set out the rules.

    Kurt agreed that using the proportional model makes it more complicated, as so many of the items within the New gTLD Program have been complicated. Many details have been worked out and are presented in the companion paper. It is complicated, and refinements have to be made. 

    Chris agreed that round robin may be the easiest and most clear method, though it is not clear that it is the fairest.

    The Chair asked if the simplicity of the round robin method was preferred? A decision has to be made soon.

    George suggested that the Committee pass a resolution affirming confidence in the first resolution and confirming that round robin would be the method to arrange the final queue.

    The Chair asked if round robin was mentioned in the original resolution, and George confirmed that it was not.

    Chris asked if a resolution was needed at this time.

    The General Counsel and Secretary responded that if the Committee wished to simply advise and guide staff that could suffice, but if they would like to a resolution that is fine too.

    George noted that he believes a resolution is necessary given the state of discussion in the community, and providing more information to counter any ambiguous communication that may have taken place.

    The Chair polled all persons present on the call to see what each person believes the best system to be. George, Bill, Chris, Mike, Ramaraj, Akram Atallah, Jamie Hedlund and the Chair all supported round robin. Heather stated there was no GAC view at this time. 

    Kurt noted his preference was for a way to do the round robin while not "demoting" anyone from the batch. The General Counsel and Secretary agreed with Kurt.

    Thomas Narten stated that he was undecided and wanted more information about the risks to the round robin system and how they could be mitigated.

    Akram noted that simplicity is the most important thing, and if there was an ideal formula as suggested by Kurt, that would be great; however, someone will always be unhappy, so its best to go with the simplest form. Jamie agreed with Akram.

    The Chair asked if Kurt had any proposals for the round robin "adjusted" he mentioned previously.

    Kurt noted that he has a couple of ideas for models in his head, but they are complex, and could require a great increase to the size of the first batch. Kurt stated that he may not push too hard on further modeling. But the model would essentially be to do the round robin and, on top of that, include those in the first batch that merited a first batch by scoring in the top 500 across all applications.

    George noted that this could make the queues larger for every region, and make the first batch absolutely major.

    The General Counsel and Secretary noted that one of the things that we will all know soon is the proportional division of applications, and that may provide an opportunity for change. However, simplicity is probably the best for now, and making a rule that will be re-evaluated later will lead to havoc. The round robin makes sense. If we find a need to change it, or are able to incorporate Kurt's suggested model at a late time, that can be considered.

    George asked if the General Counsel and Secretary's comment meant that changes would be introduced later to achieve other objectives?

    The General Counsel and Secretary clarified that he is advocating for a simple model, though there always remains a possibility of an anomaly that would allow for a larger batch or to allow a means to address what Kurt has identified as a risk. In that event, that would always be taken back to the Committee for consideration. The rule would be firm, though ICANN should remain open to mitigating risk if there's a means to do so. The intent is to set a clear path that round robin will be used. If something comes up later that changes how the risks have been identified, that would be brought to the Committee.

    The Chair summarized that all of the Committee present on the call, other than Thomas Narten and Heather, would prefer to proceed with a round robin as it's simple. The rules can be set before batching launches on 8 June 2012. The decision today is round robin.

    The General Counsel and Secretary noted that he and Kurt removed their reservations about finding an adjustment to the system and support the round robin.

    The Chair then called for discussion on whether a resolution was needed.

    The General Counsel and Secretary advised against a resolution at this time, as it may confuse communications as opposed to improving communications on this topic.

    The Chair inquired about what would be communicated when.

    Akram noted that a draft communication had been provided to the Committee, including some visibility on the reveal date and the batching schedule. There are not dates in there for the GAC Early Warning process, as more discussions are needed with the GAC on that item. In the announcement, information could be added that a geographical round robin would be used to form the batches.

    Chris asked for clarification, as there were previous statements that there would not be operational details released on digital archery until after the application window closes. In that event, should details be released at this time?

    Bill suggested that further communication may not be necessary, as this is a decision that the Committee began with and decided not to change.

    The Chair noted that clarification seems to be needed, and the communications team will have to decide what to say. The decision is that round robin will be used, and the mechanism for communication will be determined by the staff.

    The Chair then turned to the issue of string contention, and asked if the Committee members agreed with the proposal that contention sets go into the highest numbered batch, while no applicants get demoted. No objections were raised.

    The Chair turned back to communication of all the rules, and noted that they should be communicated before the launch of the batching system.

    Akram noted that the plan is to close the application window on 30 May and shortly after, announce the rules of the digital archery, possibly including a Q&A. Round robin would be included in this information.

    The Chair then asked about the work in the Risk Committee on gaming of digital archery, and if there are any communications about that.

    Mike noted that the sooner we communicate, the better it will be.

    Chris noted that there's a lot of information already available on technical steps that an applicant could take to raise their digital archery score, such as the ability to test. We have to be clear about the protections put into place by ICANN, such as placing the servers behind a VPN, and the use of a secure platform.

    Kurt agreed that this could be included in communications, and he will work with Akram on that point.

    Thomas Narten spoke against addressing those types of security measures in a communication. Instead, providing the details of the digital archery system would be better, so that the community can study and understand it.

    The Chair noted that there is no negative side including statements about the security placed around the system, as well as an explanation of how digital archery works.

    The General Counsel and Secretary noted that, as digital archery is intended to be a game of skill, the baseline is a level platform. Different skill sets will result in different results.

    The Chair then asked if the Committee wished to comment on the communications draft before it was released.

    George noted that it is the prerogative of staff to implement this, though if there were any doubts about the correctness of the communication as it relates to the Committee discussion, the Committee should review.

    The Chair requested that the Committee be given a 24 hour advance preview within which to read and respond.

    The Chair then closed the meeting.