Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

[STDS-802-11] 802.11 LB voting - some suggestions for improvements



--- This message came from the IEEE 802.11 Working Group Reflector --- Hi Adrian (and WG members),

When I asked in the last plenary about the (to me) missing yes/no/abstain voting aspects of the report on attendance vs voting status, I was told that this was not part of the subject of a motion that caused the Vice-Chair's study and subsequent report.

Be that as it may, per your (Adrian's) advise during the last plenary session, I agreed to write this topic up.

So here it is...

When we vote on 802.11 letter ballots, and one chooses to vote "abstain", there are multiple abstain reasons offered by the ballot tool.
Yet picking any reason other than "for lack of technical expertise" results in an invalid 802.11 LB vote.
I presume that the other options are valid within other groups in IEEE that use the tool.

Problem part 1: The only abstain reason that 802.11 accepts is "Lack of expertise".

Some people are not comfortable saying that they have a lack of technical expertise for any ballot. I know this from personal knowledge (I have had people tell me this first hand).

Problem part 2: Skewed Vote results.
However, the indirect effects of the situation do cause me to have some concerns.

Instead of saying "lack of expertise", a significant number of people pick another path.
Because they have to return a valid vote to maintain voting status, their remaining choices are to vote yes or no.
They tend to vote "yes" as the easiest way to return a valid LB vote (since a no vote requires one to say why the voted no).
Essentially "yes" is being used as a way to say "I don't care but I am forced to respond to this LB, so here is a 'yes' vote, now leave me be".

This behavior inaccurately skews LB technical results toward passing.

We know there has been evidence of this pattern for many, many years - anyone that has run a LB has seen Yes votes show up very shortly after the LB announcement goes out - well before any non-super human could have even read the posted document.

Suggestion: Allow abstain reasons other than "lack of technical expertise" to be valid 802.11 LB votes.

(Frankly I don't see much value from multiple abstain reasons.  Any abstain vote is a vote that says "I wish not to influence the outcome of the vote".   Since the tool already has other reasons, letting any of them be valid in 802.11 seems to me to be the easiest thing to do).

Problem part 3: Yes/No/Abstain ratios
I also know that 802.11 chairs would prefer to ignore this vote skewing pattern - because current rules say we need a minimum % of votes to be yes/no to make a valid ballot. 

Alas, I have essentially no sympathy for that as a motivation to encourage gaming the vote technical results. 

Consider(these (slightly contrived for sake of making the point) examples:
1) Under current rules a ballot might result in
    75 yes, 25 no, 10 abstains (for lack of technical expertise)
We would pass that ballot.

2) The same ballot under revised rules might result in
    50 yes, 25 no, and 35 abstain (for any abstain reason),
We would not consider that ballot to be passing.

I think that we should not be passing ballots via abstain->yes vote skewing.
Instead we should honestly be asking (based on abstain ratios) if the project that generated the LB results should be continued.

Suggestion:
Change the rules so that any ballot with Yes/(yes+no) > 75% be a (technically) passing ballot,
AND
if the (yes+no)/(total votes) ratio is too low, then ALSO automatically require that the WG vote on whether the project continues or not (effectively a WG vote of whether or not to stop expending WG manpower on a low support level project).


Summary:
An abstain vote should be a vote that does not influence the technical result of the ballot.
An Abstain vote should count for determining if a project still has sufficient support to be worth continuing.

David Bagby

_______________________________________________________________________________

If you wish to be removed from this reflector, do not send your request to this reflector - it will have no effect.

Instead, go to http://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-11 and then press the LEAVE button.

If there is no LEAVE button here, try http://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-11-RO.

Further information can be found at: http://www.ieee802.org/11/Email_Subscribe.html _______________________________________________________________________________