That is a very interesting observation German.
Does anyone have any suggestions of an alternative approach?
On Wed, 25 May 2022 at 21:17, German Valdez Aviles german@apnic.net wrote:
Hi All
I’d like to raise for your consideration the use of the Schulze methodology in the selection process.
While the Schulze method could provide good results when multiples candidates are running, it should have a bigger universe of voters to have the desired effects. With only 15 or less voters I question the value of having such method for election. Furthermore, I have the feeling that it’s not well understood by many ASO AC members and this has led to question the validity of the results in the last Seat 9 election (at least by one of the ASO AC Members) even though the secretariat drafted a step by step and explanative report of the calculations . One final point, it’s that the current procedure based on the Schulze methodology doesn’t offer a fair process to break a tie. The current process request a random process to select the winner (it doesn’t specify what random process to use). But that’s it’s my humble opinion 😊
However if the committee would like to keep it I’d like to have the opportunity to know early in the process. So far the secretariat calculate the results manually (with deep review of the calculations) and rely in third party online calculators to validate the results. As you could see is not an ideal situation so I’d like to raise to the NRO EC to budget for next year a RIR in-house or external certifiable consultant solution that could elaborate an online program that could tabulate the ranking preference based on the Schulze method and include some random process to break a possible tie (which could happen given the limited number of voters and candidates). This could bring more trust and reliable results in the calculation of a winner in future elections.
This is just a mere observation from the Secretariat perspective for your consideration.
Regards
German
-- Procedures mailing list Procedures@aso.icann.org https://aso.icann.org/mailman/listinfo/procedures