Hi Tapani, As per the EC meeting yesterday does this constitute the instructions to be sent to the list and if so has this been approved by the EC as the formal communication? -James On 25/08/2016, 07:42, "NCSG-Discuss on behalf of Tapani Tarvainen" <[log in to unmask] on behalf of [log in to unmask]> wrote: >Dear all, > >While the voting rules confirmed by NCSG EC yesterday may be familiar >and clear to oldtimers and mathematically-minded people like myself, >they may not be obvious to all. I try to clarify them a little. > >In particular the actual effect of None-of-the-Above vote to >candidates' chances of getting elected (that is, beating NotA) in the >councillor election is not quite intuitive. Chair election is clear >enough so I won't discuss it now, and I'm ignoring any symbolic >meanings as well. > >Key point: in the present situation, the only thing that matters in >deciding if a candidate gets elected is whether or not their vote >count is less than NotA's. > >With that in mind: > >In the councillor section of the ballot there are four boxes >one can tick: one for each candidate, let's call them X, Y and Z, >and one for None of the Above. > >This gives in effect eight different ways of filling the ballot: > >(1) Leave it empty - tick no boxes. > >This has no effect on the outcome, but will be counted as a valid >vote. > >(2) Select only NotA, none of the candidates. > >This will decrease all candidates' chances of getting elected equally. > >(3) Select one candidate, say X (but not NotA). > >This will improve candidate X's chances of getting elected >and has no impact on the chances of candidates Y and Z. > >(4) Select two candidates, say Y and Z (but not NotA). > >This will improve both Y's and Z's chances of getting elected >and has no impact on the chances of X. > >(5) Select one candidate, X, and NotA. > >This will have no effect at all on X's chances but will >reduce Y's and Z's chances of being elected. > >In effect the NotA vote will cancel the positive vote to X, >leaving only negative vote against Y and Z. > >(6) Select two candidates, Y and Z, and NotA. > >This will have no effect on Y's and Z's chances but >will reduce X's chances. > >Again, the effect of NotA is canceling out the positive vote >to Y and Z, leaving only the negative effect on X. > >(7) Select all three candidates (but not NotA). > >This will improve all candidates' chances of getting elected. > >(8) Select all three candidates and NotA. > >This will invalidate the ballot and it will have no effect on the >outcome. Invalid ballots will be separately counted, however. > > >The most counterintuitive cases are (5) and (6): if voting >for one or two candidates, adding NotA will actually >reduce the chances of your chosen candidate(s) being elected >(just as much as those of the other candidates'). > >If everybody votes that way, that is, selects one or two candidates >plus NotA, no candidate can ever be elected. > >A simplified hypothetical example to illustrate this: > >Assume we have 400 voters with one vote each. > >The assume 100 people vote for X+NotA and 300 vote for Y+Z+NotA. > >Result: X gets 100 votes, Y and Z 300 each, NotA gets 400, >and nobody gets elected. > > >I find this system so bizarre that it actually never occurred to >me that it really was the intent in previous elections. >Indeed I failed to believe it even when people kept yelling at me that >yes, that was the intent. Well, I was wrong. Not for the first time. > >Good reminder that what someone thinks is obvious may not be so at all >to another. And in things like elections that can be dangerous, so all >such assumptions should be made explicit and written down. > > >Anyway, this is the system we have, longstanding practice, >and we are not going to change it for this election. > > >So, vote - but take care that you understand the real effect of your >vote, especially when thinking about voting for None of the Above. > > >****** > >For the future, however, I recommend reconsidering this and looking >for better methods, even changing the charter if need be. Perhaps some >type of approval voting, separate NotA for each candidate, or a vote >threshold would work. All such systems have their own pitfalls though, >it takes care to do them well. I will not go deeper into that now, but >I suggest it would be best to define the rules at a time sufficiently >far from any actual elections that thoughts of current candidates and >strategies will not cloud people's thinking. Of course election-time >discussions such as the present brouhaha should be very useful >material to review then, so by all means let the debate continue if >people aren't tired of it yet. It may prove worthwhile in the end. > >-- >Tapani Tarvainen