search

gryphonshafer / Quizzing-Rule-Book

Bible Quizzing Rule Book
Other
10 stars 5 forks source link

8.4 Annual CQLT Ratification #131

Open JoshJetto opened 2 years ago

JoshJetto commented 2 years ago

Current Rule: Annually, the CQLT will review a PR of the integration branch merged to the master branch. A simple majority of the CQLT will ratify or reject the PR. When ratified, the PR will be merged to master immediately following that season's IBQ. Whatever is in master will be considered the rulebook and supporting documentation for the upcoming IBQ.

Proposed Change: Annually, the CQLT will review a PR of the integration branch (edit: "all of the changes to the current rulebook") approved to be merged to the master branch (edit: "included in the rulebook update") by the Rules Committee. A simple majority of the CQLT will ratify or reject each rule as a whole to which changes have been made. The CQLT may not make alterations to the wording of any proposed rule. The CQLT, at its discretion, may choose to vote to ratify or reject the entire PR (edit: "block of rules changes") as a whole. Unanimous consent by the CQLT would be required to put the entire PR (edit: "block of rules changes") as a whole up for a vote. If unanimous consent is achieved, then a simple majority would be required to ratify or reject the block of rules changes as a whole. All changes to the rulebook approved by the CQLT will be considered the rulebook and supporting documentation for the upcoming C&MA International Bible Quizzing tournament.

Rationale: This current rule takes an inordinate amount of authority out of the hands of the CQLT as the elected governing body and puts it into the hands of an appointed committee. From an organizational standpoint, this is out of order. The current rule ties the hands of the CQLT from being able to provide meaningful oversight over the Rulebook governing C&MA Bible Quizzing. If the elected governing body is not allowed to approve or reject proposed changes to the rulebook rule by rule, then it is forced either to accept rules changes that would not pass the CQLT on their own merit (in order to accept those rules changes which it finds beneficial) or to reject rules that would pass the CQLT on their own merit (in order to keep rules it deems unhelpful or detrimental out of the rulebook).

The current rule also puts at risk an entire year's worth of work by the Rules Committee on a single vote on the entire PR, in which there are changes which may have passed the Rules Committee by split decisions. Forcing the CQLT to accept or reject the entire PR as a whole, which is likely to include some debatable proposed changes, does not allow the further insight of a multitude of counselors (especially those of the higher elected body) to evaluate individual proposed rules changes on their own merit. It instead forces the CQLT to weigh the harm that would be done by bad rules changes against the benefit that would be derived from good rules changes, and then choose the lesser of two evils. This kind of weighing will not produce the kind of rulebook that will best serve the Bible Quizzing community and its mission, and it is wholly unnecessary to make it a part of the change management process.

The proposed change puts the proper oversight authority back into the hands of the CQLT, while also protecting the work of the Rules Committee, and it creates a change management process that allows for good teamwork between the Rules Committee and the CQLT to produce the kind of Rulebook that will best serve the mission of Bible Quizzing.

Additional Note: I think we should remove the "GitHub-ese" from this rule. Unless a person is very familiar with the GitHub process, they will not understand what the terms PR, integration branch, and master mean.

ARMediting commented 2 years ago

I agree with the substance of this idea. It benefits both the CQLT and the RC. I think there needs to be more clarification of "each rule as a whole to which changes have been made." Is the definition of a "rule" any numbered point in the RB? How many levels deep? What if we have two separate changes to a numbered rule that have nothing to do with each other (maybe a change to the rule and a wording clarification)?

For the ease of those reading this issue (I had to go look it up as well), PR is short for "pull request," which is what an issue becomes when the Rules Committee reaches a decision about it the results in a proposed change to the rulebook. I assume "PR of the integration branch" is all of those as a collective. (Is there a way to include a glossary in github? The rulebook might need one too!)

ARMediting commented 2 years ago

It's also worth noting for phrasing that some rule changes will affect multiple rule sections.

jswingle commented 2 years ago

This is something @JoshJetto and I have discussed at length, so I present the counter-arguments below: the CQLT should be approving or rejecting each year's rulebook in full, not by each individual rule change proposal. (These mostly aren't new arguments to Josh or to the CQLT, but should be publicly available of course.)

1. The original Change Management Process was ratified by a vote of the districts, which should be the highest priority vote possible.

Technically speaking, the RC has the right to amend the Change Management Process (CMP), but other than grammar or clearing up the language (and I agree with clarifying the GitHub-ese), we shouldn't edit it.

I think this is an unfortunate loophole in the construction of the RC and we really shouldn't have the right to amend it in the first place. The CQLT kicked this vote to the districts because it was so important that we wanted a more representative vote than the CQLT can supply with only 6 members. It's rare for the CQLT to ever conduct a vote this way, so if the CQLT saw it as important enough to go through the trouble of a district vote, the CQLT (and its appointees, such as the RC) should respect what was voted for. This is still true even if there are technically no institutional safeguards that prohibit us from overriding the vote.

This first point concerns why we shouldn't change the current CMP, even if we think it's lacking. But the rest of the points below will argue why the CMP is not lacking, and give the reasons for why the present relationship of the CQLT and the RC was designed the way it was.

2. The CQLT already has more than sufficient oversight of the RC.

First of all, anywhere from 1-3 members of the RC must also be members of the CQLT. That means the RC never lacks a CQLT presence, and the CQLT can have as many as half the votes, if it wants that to be the case. Note that those 3 votes are, by themselves, enough to block any proposed rule from entering the rulebook.

Second and much more importantly, the CQLT appoints the RC, and has unlimited latitude in how to do it. If the CQLT gets 15 applicants to the RC, 3 of whom are from the CQLT itself, they can (and likely will) prioritize themselves and appoint 3 of themselves to the RC. No harm, no foul -- I'm fine with the current RC appointment procedure, but it has to be acknowledged that this is already a large amount of oversight. With only 1-year term limits for the RC, the CQLT can engineer the people in the RC in favor of a desired outcome, if it really wants to. If they don't like a certain vote, they can just choose for that person not to return and replace them with someone who they know will vote a different way. (Whether or not the CQLT would actually scheme like this is not the point; I'm just saying that it theoretically could, therefore it has a lot of oversight and power already.)

Lastly, the CQLT still has to approve the rulebook in full. Nothing ever enters the rulebook without the CQLT stamping it. The point of this CQLT vote is so that the CQLT can make a general assessment as to whether or not the RC is doing its job properly and on the right track. If the CQLT thinks the RC is completely off, it should simply reject the rulebook revision. Under the present structure, the RC can never just do what it wants; the CQLT is the gatekeeper, which is a substantial amount of oversight.

3. As a corollary to #2, this rule change would give the CQLT too much power over the rulemaking process.

Delegation of authority is organizationally healthy, and the purpose of the CMP rests on the understanding that the CQLT does not need to make every single decision themselves. CQLT members are voted into their positions by the districts, and they are the authoritative governing body for C&MA quizzing... but the districts ALSO voted that it was appropriate for the CQLT to delegate the rulemaking process to a separate governing body.

By giving the CQLT the power to override any vote by the RC, we are de facto rendering the RC's votes useless. The RC effectively becomes an advisory committee which proposes rule changes to the CQLT for them to vote on instead. The RC vote becomes purely ceremonial and non-actionable. Besides the RC votes being a waste of time, this also disrupts the intended structure: the CQLT is not supposed to be intervening with the RC's work unless they have an especially strong disagreement with it.

For example, there is in discussion right now the issue of whether we should remove the requirement to say all unique words in order to be counted correct. Let's say the RC votes for this 6-0, and the CQLT rejects this change with a 4-2 vote. Is this really the sort of nitpicky detail we want the CQLT to be overriding? Under the current CMP, the CQLT wouldn't bother rejecting the full rulebook revision over this even if they disagree 4-2, because it's just not that important of a rule. But if they were allowed a vote on every issue, they would in fact override the RC frequently on minor details, and the rulebook effectively becomes a product of the CQLT's vote, not the RC's.

Furthermore, it needs to be said on a practical level that the RC is quite a lot of responsibility and time commitment, and the CQLT members who aren't on the RC are not going to spend 10 hours before a meeting catching up on the full arguments of every single issue that the RC votes on. (And it would be ridiculous to expect them to; the CQLT is already plenty of work.) There can be no uncertainty that the RC members will end up as the experts on the various pro/con arguments for the issues being voted on. This doesn't make the RC automatically correct, but the more informed group is likely to make the better decision. It's a large and dedicated task, and the RC was appointed by the CQLT to take it on.

4. More specifically than #3, this proposal gives those individuals who are members of both the RC and the CQLT far more power over the rulemaking process than anyone else on either team.

The current rules allow for anywhere from 1-3 of the RC members to be CQLT members. Practically speaking, this means that whoever is on both committees gets two chances to persuade a vote in favor of whichever rule changes they are passionate about, but everybody else on the RC only gets one. We emphatically do not want this.

Let's say there are three CQLT members on the RC. A rule change passes the RC 5-1, with one of the CQLT members as the lone dissenting vote. Then, at the next CQLT meeting, this one dissenting voice is able to persuade the three non-RC members of the CQLT to their position. The rule change gets vetoed with a 4-2 vote, despite the fact that both the majority of the RC and a majority of the CQLT members on the RC voted in favor.

Or another situation: this time there is only one CQLT member on the RC. A rule change passes the RC 5-1, with the one CQLT member as the lone dissenting vote. Then, at the next CQLT meeting, the one RC member gets to persuade the entire CQLT of their position. The RC member means well, but accidentally misrepresents the arguments for the other side which persuaded the rest of the RC, and since this one RC member is the only one who's actually read the entire discussion, the rest of the CQLT goes along with the misrepresentation and vetoes a rule that they may have otherwise accepted.

These are all really bad outcomes which could completely demoralize the RC, or could even pit members of the CQLT against each other. Again: we emphatically do not want this. If we were going to make this change to the CMP, it would be paramount that we also remove all CQLT members from the RC as well.

5. It is not a big deal if the CQLT chooses to reject the entire rulebook revision in a given year.

This is a less significant point than 1-4, but still worth pointing out for those who are rightly worried about the big load of RC work getting swept away by a "no" vote

A "no" vote from the CQLT would not mean all the work of the RC over the past year goes away. The CQLT would naturally indicate to the RC which rule(s) they think are so egregious that it made them unable to vote in favor of the whole thing. All the GitHub discussions and edits are still present, and the next year's RC can reconsider and communicate with the CQLT about the disagreement. So a "no" vote doesn't invalidate the RC's work in general unless they did terrible work, in which case such an invalidation would be justified.

The only negative to a "no" CQLT vote is that there is simply no rulebook revision that year. Before the CMP was instituted, rulebook revisions only happened every few years. There's no necessity for yearly revision. A hesitant CQLT indicates that it would be entirely appropriate to go back to the drawing board a bit before rolling out any new revisions, but it doesn't take the drawing board and throw it in the dumpster.

6. Lastly, the CQLT can already publicly comment on RC votes for 3 months and attempt to change the RC's mind before they conduct their final vote to move the pull request into the integration branch for the year.

The RC has to take a 2nd vote on any given issue 3 months after initially voting on it. Any and all members of the CQLT can publicly comment on the RC's decision during that entire 3-month period. So if the CQLT notices a potential unintended consequence of the RC's choices, they have a good amount of time to ask the RC to reconsider. The RC, for their part, may also want to reconsider if it senses the CQLT has wide disapproval of a certain rule and may strike down the annual rulebook revision if it's not changed. Such feedback and collaboration is a baked-in part of the process.

Furthermore, as already highlighted, there is always at least 1 member of the RC who is also on the CQLT. Naturally, whoever this is will want to alert the CQLT's attention to potentially controversial rulings for their public comment, or any rule change which may have logistical implications for the CQLT's work (say, for example, the RC wants to re-institute the marriage clause, or change the geographical district eligibility requirements which have been a headache for the CQLT in the past.)

Thanks for bearing with the long comment.

jswingle commented 2 years ago

Another thing we may want to amend is specifying certain circumstances under which the RC can do things without a vote.

For example, we've been closing out (without a vote) issues that we deem identical to other open issues, or inextricably related. I think doing so is true to the spirit of what the Change Management Process is about, and helps avoid the Rulebook Committee getting bogged down. But it would be best to spell this out explicitly in the CMP.

jswingle commented 2 years ago

To-do list for an eventual revision: