May 2023 General Election Count Correction

Pallaith

TNPer
-
-
-
-
I have previously brought this to the attention of the election supervisors as well as the chief commissioner, and at his urging I am posting this now, but it is clear to me from a comparison to my own personal record of the votes (consulted after I got home from work the day of the election) that the results published to the region in the most recent election were inaccurate. Several ballots were improperly recorded and registered candidates who the voters did not choose, and I understand (and @Just a Lore can clarify) that some votes had the wrong selection for RON indicated in the results. Regrettably I was unable to point this out during the challenge period, which occurred either while I was asleep or while I was at work. I had no ballots to challenge, as no ballots were wrong, but a close comparison between my spreadsheet and the one used by the commissioners would have revealed the mistakes. And I didn’t have that handy at work.

The votes are publicly visible and and anyone can count them to confirm what I have observed. Whatermelons and Sanjurika’s votes are recorded for Gorundu instead of Kasto, and Robes and Thomas Insaniac abstained but are also counted for Gorundu. And in the Speaker race, Kvylia voted for Ocean not Skaraborg.

The other reason this is being brought up here is because there is disagreement as to how such a correction would even work. To my memory, it’s never been needed before. Can our supervisors simply correct the results? Obviously these changes wouldn’t change the outcome of any of the races, but the EC considers general principles, and doesn’t make decisions that are exclusive to one election.
 
Last edited:
After the 6 hour challenge period, nothing can be done privately anyway.

Way I see it there is three options.

1. We acknowledge the issue and put a rule in place for future elections since the outcome would not be affected.

2. Lore and I use our authority and discretion under 4.3.17, correct the count and the results, and post on the results thread what we did and why. Not sure if that is the intent behind that clause, but its there.

3. We count Ghost's post as a citizen petition. Vote according to section 6 of the rules. The EC Overrules Lore and I's count using the B option, we correct the sheet and the results, without having to reopen the election.
 
To be clear. I believe 1. is our only option under the law. The results are final and no changes can be made. Should the EC choose 2 or 3 I will not participate and will vote present.
 
After now spending some time reading the actual rules instead of just relying on back and forth DMs like I shouldn't have been doing... as far as the issue at hand, the EC rules state that "[t]he Election Supervisors may modify how each ballot is counted until they certify the final results, or present the results to the commission at large to be certified." I think it's rather clear that the results as posted are final, even if they're inaccurate.

I had no ballots to challenge, as no ballots were wrong
I think this is a misinterpretation of how the challenge rules are intended to be used. You aren't challenging the ballots themselves, or how the voter cast it. You're challenging the way in which the supervisors interpreted that ballot and recorded it, including just saying "Hey, you didn't count that correctly. The vote went to the other person," which is like 99% of challenges these days.

What worries me is why these errors were not caught. Ghost was at work. I was at work too. I didn't hear of any challenges from @Fili or @Simone_Republic. So, what do we need to do here? Extend the challenges period so that hopefully there's time when working ECers can look after work?

What I also would like is an explanation from @Dreadton and @Just a Lore as to what their process was for certifying the results. Were both of you involved or did a single individual handle calling the results official and posting them? How did we go from an official result to then one of the supervisors pointing out multiple counting errors? Do we need to make it a requirement that both supervisors sign off on the result? How would we enforce this?
 
After now spending some time reading the actual rules instead of just relying on back and forth DMs like I shouldn't have been doing... as far as the issue at hand, the EC rules state that "[t]he Election Supervisors may modify how each ballot is counted until they certify the final results, or present the results to the commission at large to be certified." I think it's rather clear that the results as posted are final, even if they're inaccurate.


I think this is a misinterpretation of how the challenge rules are intended to be used. You aren't challenging the ballots themselves, or how the voter cast it. You're challenging the way in which the supervisors interpreted that ballot and recorded it, including just saying "Hey, you didn't count that correctly. The vote went to the other person," which is like 99% of challenges these days.

What worries me is why these errors were not caught. Ghost was at work. I was at work too. I didn't hear of any challenges from @Fili or @Simone_Republic. So, what do we need to do here? Extend the challenges period so that hopefully there's time when working ECers can look after work?

What I also would like is an explanation from @Dreadton and @Just a Lore as to what their process was for certifying the results. Were both of you involved or did a single individual handle calling the results official and posting them? How did we go from an official result to then one of the supervisors pointing out multiple counting errors? Do we need to make it a requirement that both supervisors sign off on the result? How would we enforce this?
I want to be clear that I am aware challenges are not limited to bad ballots. In observing this process for years, this has been where most focus was spent because these were the reasons results were in dispute. I have kept a spreadsheet counting votes for every prior election, and the errors I caught this time have never been a factor. I have been the one making the errors, and if there was ever doubt, it was because of dubious ballots. So if there’s a silver lining here, I do believe this is a first, and this doesn’t mean past election results have had this problem, because they haven’t. The reason I make that final check of my spreadsheet is because I want my record to be accurate, and I have to measure it against the EC results to make sure that it is. It always has been.

It’s easier to look through ballots and catch questionable ones. That’s something I could do in the window I had at work to glance over the results. I did look at the official spreadsheet as well, because I knew some people changed their votes mid-election. But I couldn’t do it one at a time and couldn’t do a side-by-side. I was one of half a dozen people, I figured with no questionable ballots the spreadsheet was probably fine. That’s always been good enough before. Maybe it’s a wake up call that we should pay closer attention and one way we can do that is not to wait for the challenge period to do a check. If I had done that side by side before the voting closed, I would have caught all of those errors, because they weren’t ballots from the end of the election.

Maybe we need to formalize this - each supervisor keeps their own personal spreadsheet and makes sure they’re both in agreement with the main one? Or maybe we just get another ECer to volunteer to keep the second opinion spreadsheet? Every time I have encountered another person who keeps their own records, we always find our results diverge in some way and can usually figure out why. I think that’s how we guard against this sort of thing in the future.

As for the solution to this particular election…if the results have been certified, we have our winners, and that obviously can’t change. Perhaps a note in the OP can be added that explains what the actual numbers are. The way, the record we have in archive is in fact accurate. I’m not sure what rules we would put in place for this point of the process, final is final, so I think it will likely end up being for earlier parts of the process.
 
What worries me is why these errors were not caught. Ghost was at work. I was at work too. I didn't hear of any challenges from @Fili or @Simone_Republic. So, what do we need to do here? Extend the challenges period so that hopefully there's time when working ECers can look after work?

We could do that but even when we have 4-5 comissioner active during the period, they often do not weigh in.


What I also would like is an explanation from @Dreadton and @Just a Lore as to what their process was for certifying the results.
We updated the vote count as the come in, per standing practice. We discussed various question over certain votes, per standing practice. We found an error during this process and corrected it, per standing practice. We discussed how to phrase part of the results regarding irregular ballots.

Were both of you involved or did a single individual handle calling the results official and posting them?
Both, and the results were certified by Ghost, no other ECer weighed in.


How did we go from an official result to then one of the supervisors pointing out multiple counting errors?
Lore didnt point out mutiple counting errors, so I am assuming you mean Ghost who was not supervising the election and signed off on the election, then discovered these issues. As you are well aware, during the 6-Hour, errors like this are often encountered. They are generally caught by the non-supervising ECers. Fresh eyes generally make the difference. Here, the usual process was followed but only one non-supervisor certified the results.


Do we need to make it a requirement that both supervisors sign off on the result? How would we enforce this?
I dont see how this will solve this issue, as "both supervisors" and one other ECer signed off on these results. You should look at the 6 hour window and overall ECer activity during it.
 
We updated the vote count as the come in, per standing practice. We discussed various question over certain votes, per standing practice. We found an error during this process and corrected it, per standing practice. We discussed how to phrase part of the results regarding irregular ballots.

Both, and the results were certified by Ghost, no other ECer weighed in.

Lore didnt point out mutiple counting errors, so I am assuming you mean Ghost who was not supervising the election and signed off on the election, then discovered these issues. As you are well aware, during the 6-Hour, errors like this are often encountered. They are generally caught by the non-supervising ECers. Fresh eyes generally make the difference. Here, the usual process was followed but only one non-supervisor certified the results.

Okay, my apologies, especially to you, Dreadton. I guess I just didn't have a clear timeline of what happened and when. You and Lore finished up and certified, and Ghost said no objections. Three ECers, including myself, either didn't check or otherwise had nothing to say. Then, after the results were released, the errors were found, but at that point it's too late.

We could do that but even when we have 4-5 comissioner active during the period, they often do not weigh in.

I know I do a lot of pings in the Election Commission channel, and the TNP General server in general does a lot as well. Could this be contributing to the activity problem? Are ECers muting the server, the channel, or are they suppressing all role mentions?

Or, are we running into timezone problems? As I alluded to before, is the 6 hour window big enough? Should we consider going to 12 hours, or 24? Sure, it's an inconvenience to those awaiting the results, but it would give the body generating the results time to have everyone weigh in to make sure it's right.
 
A bigger window would likely be best. Maybe a place where it can be recorded who certified and who did not particpate. The RA and the EC would have something to look at during the next confirmations, IE if an ECer hadnt certified multiple elections in their term that they were present for.
 
We should remember that challenges can be lodged at any time during the voting period. Most votes are received in the first two days of voting, so there is plenty of time to catch most mistakes. I think some Commissioners are perhaps neglecting to check the spreadsheet until the voting period has ended, which would mean very little time to look for mistakes. That said, I am supportive of extending the challenge window, as long as it’s not over 24 hours.
 
Sorry for the late reply I've been working some weird shifts irl.

I will admit my checking of the results was only cursory on my part because I was out of the house. For my part in the results I only ever gave a small lookover on my work on the election spreadsheet from my phone because I only had my notes of what I did accessible at the time.

As such I was only ever consulted on the irregularity language and that a mistake was found and fixed about an extra VD vote that hadn't been deleted properly.

Personally I find the current rules we have set up a little lacking. It gives very little flexibility for RL timing and events. 6 hours is less then the time then a full shift of work and is an amount of time that many people could easily be unreachable or uncontactable. They are honestly baiting a problem to finally occur and the fact its taken this long is honestly a little amazing.

The fact the current rules literally allow for citizens to have their votes effectively stolen by a culmination of the inactivity of non-supervisory commissioners, very bad timing in regards to IRL life, and a few mistakes.
 
I think we seem to have some hint of a consensus that the challenges period is too short.

What I propose is the following rule change:

7. Any time during the voting period and for up to 24 hours after the voting period has ended, any Election Commissioner may privately challenge a particular ballot and submit an alternate interpretation of how it should be counted. The Election Supervisors may accept or deny any challenge. The period in which challenges can be made may be terminated early if all present Election Commissioners notify the Election Supervisors that they do not have any challenges to make, but early termination of this period can occur no sooner than 6 hours after the voting period has ended.

Basically, we get 24 hours, but if everyone speaks up and certifies the election quickly, we can post results as soon as 6 hours after voting ends.
 
Out of curiosity, given the 6 hour period clearly isn't some magic bullet here, if we all don't have challenges and specifically notify the supervisors to that effect, why couldn't the results be released sooner than 6 hours?

Also, regardless of how I feel about the rule, what do we do about the obviously wrong results?
 
I'm not married to the 6-hour minimum, but I do feel like there should be a minimum amount of time to allow supervisors to not feel rushed and to have an opportunity to double check their own work. If we wanted to pick a different minimum time, I'm okay with that, but I believe there should be something, and something substantial.
 
I’m going to quote myself from Discord:

“Personally I don’t think length of time is the best guardrail. I think every member needing to take specific actions that force them to double check works better. And if they don’t all do those actions then the results don’t get published. This minimizes possible errors and makes it less likely it’s rushed. You can’t discount the possibility that someone will be lazy and just claim they looked stuff over, but we have that risk now, and it’s possible for a full slate of commissioners to do this within minutes of voting closing, and then we just have to wait 6 hours for their errors to take effect instead of it being instant.”

We need a stronger procedure for how to check the results and we need to steer supervisors into complying with it. Giving them the same arbitrary amount of time to look things over one more time that the rest of the EC has, when most of the other commissioners don’t even utilize it (either by necessity or because they don’t respond), is a system that could have led to much more of these problems and has held up surprisingly well. Obviously it only has to go wrong once to be a serious problem.
 
Unless you have another idea, I think the only solution that I can think of that fits your idea is just to make to so that the Election Commission at-large has to do a certification vote for every election, instead of just ones where challenges may impact the final result.
 
Unless you have another idea, I think the only solution that I can think of that fits your idea is just to make to so that the Election Commission at-large has to do a certification vote for every election, instead of just ones where challenges may impact the final result.
I like this idea.
 
Unless you have another idea, I think the only solution that I can think of that fits your idea is just to make to so that the Election Commission at-large has to do a certification vote for every election, instead of just ones where challenges may impact the final result.
Yeah, the idea I shared earlier. Have each commissioner keep a separate spreadsheet and compare them against each other to put together the “main” spreadsheet and then that’s the one that is evaluated by the full EC. Or…we pick a separate EC member who isn’t a supervisor to keep a second spreadsheet, and compare that one against the commissioner spreadsheet. Don’t allow results to be released until the competing spreadsheets are in agreement and the full EC signs off. And we can honestly do this whether or not we mess with the time frame of the challenge period. Perhaps throw in not a maximum wait, but instead we do not allow certification until everyone has weighed in, whether that’s a formal vote or just requiring everyone to say something.
 
I also want to again draw attention to my note about having a correction note on the results, I think regardless of the new rule we come up with, we need to put a post script on those results so the historical record still shows the correct numbers.
 
I think we seem to have some hint of a consensus that the challenges period is too short.

What I propose is the following rule change:



Basically, we get 24 hours, but if everyone speaks up and certifies the election quickly, we can post results as soon as 6 hours after voting ends.

As expressed in a few private DMs, I support everything to be done in duplicate (to double check) and I also support extending the timeframe to up to 12-24 hours (subject to other commissioners' agreement). This means once I learn how to check votes I can do it twice, once before I go to bed and the second time when I wake up and having had copious amounts of coffee.
 
Last edited:
I am concerned about the amount of conversations about this going on in DMs. I will remind my fellow commissioners of Section 7.2.

Official business of the Election Commission must be conducted in public, and votes of the commission must take place on the official forum.

This is one of the reasons I refused to change the ballots in question, after the private challenge window was completed and the results certified.
 
I am concerned about the amount of conversations about this going on in DMs. I will remind my fellow commissioners of Section 7.2.
This is one of the reasons I refused to change the ballots in question, after the private challenge window was completed and the results certified.

All Simone was asking in the DM I discovered this morning was whether we knew the time zones of all of the commissioners. I replied that I do not know, but it may be helpful to record that in the private roster.
 
All Simone was asking in the DM I discovered this morning was whether we knew the time zones of all of the commissioners. I replied that I do not know, but it may be helpful to record that in the private roster.

Yes I DMed to ask about timezones only (to try to think about whether 24 hours make sense, or less, based on the timezones of the commissioners) as I believe some commissioners may consider that information not to be "official business" and infringing on their privacy. (For example if someone answered "UTC+9.30" that likely indicated Adelaide as their physical location.)

I am sorry I did not make that clear.
 
Last edited:
I would argue that the current global distribution of the EC is actual a bad question to ask because it creates a baroque solution that will break down in the future because that exact situation is not going to exist 3, 4, 5 months in the future as commissioners leave and new ones get appointed.

Personally I'm an edge case in my irl schedule so I see the outer edges of the timeframe for challenges should be closer to 36-48 hours if for no other reason then to future proof for other possible combinations on the commission.

I work a job that fairly regularly expects me to work two standard shifts with only 8-9 hours between shifts. If the period after an election mapped to midnight to midnight people like myself would have plenty of time but the elections have a fluid end time and it's not inconceivable that a 24 hour period would perfect map over a time period I am completely unavailable.
 
I'm not in favour of making all commissioners do a certification vote for every election. Obviously commissioners shouldn't be inactive, but if there is just one inactive commissioner then that holds up the whole process. And even when they are not strictly speaking inactive, there might be other things getting in the way to stop them from checking the results fully for, say, a full day. I think picking a commissioner to keep a separate sheet and ensuring the sheets match before results are published is a good solution. It only requires the people keeping the two sheets to be in agreement, and I think it's fairly unlikely that they would make the same mistake. I'm still in favour of extending the challenge window though, so that we can get more eyes on the sheets before the results go out.

As for how we should go about correcting the results for this specific election, I would be fine with the supervisors or the Chief simply issuing a correction note. However I do think we should come up with some amendments to the EC Rules that cover this situation. And there is also the very remote possibility that such an error in certification could affect the actual outcome of the election (i.e. who gets elected), in which case perhaps it's not sufficient to just have an EC Rule for the situation, but a provision in the Legal Code as well.
 
There is a way to make a at-large vote that wouldnt be ground to a halt by inactive commissioners and that would be to stipulate a minimum of 5 votes to certify. If 5 commissioners can't be expected to vote on certificate then I would argue the commission is acting willfully against the intentions of the law.

The current law expects 5 commissioners to be present for an election. Hell our current rules expect that if you are not "absent" you need to still be present if there are challenges to votes for an at-large certification. If we can't expect that for every election that is a terrifying prospect cuz we are asleep at the wheel.
 
I'm more behind the idea of the seperate spreadsheet, although I'd have 2 commissioners maintain it, mostly for any unexpected events that could hold any of them inactive in any amount.

I see also propositions of votes, which would greatly slow down the election process, and (unless I misunderstood the concept), wouldn't strictly require the commissioners to precisely look over all votes.

The second spreadsheet idea has the advantage of having to actively write down all votes, in contrast to just looking over something someone else wrote, which simply forces higher concentration.

There is an option for someone outside duos maintaining the sheets to confirm together with maintainers of both spreasheets, but is sounds a bit redundant.

And since the votes should be counted not only during the challenge period, but actively during voting period, I'm not in favour of extending the time, but I also don't see a major reason to be against it.
 
Just wanted to revive this again since it’s still unsettled. We clearly want to make some changes but what shape should that take? And is there some action we can take to make sure the correct results are not lost in the shuffle of all of this and that our archived information is accurate?
 
Here's a new section for the Election Commission Rules that I've written up to create a formal procedure to amend election results after they have been certified:
Section Seven: Amendments of Certified Results
1. Citizens may petition the Election Commission at large to amend the certified results of a past election, provided the proposed amendments would not alter the outcome of the election.
2. When such a petition is submitted, the Election Commission at large will discuss the petition. Any Election Commissioner may propose a different amendment to the results for the election in question.
3. Election Commissioners will then vote, with all proposed amendments available as separate options, in addition to an option to uphold the current results.
4. Any Election Commissioner who was a candidate in the election in question will not be eligible to vote.
5. The results will only be amended if all Election Commissioners present vote in favour of one proposed amendment.
I've proposed a unanimity threshold here because I thought we should have a high bar for going back and changing results that have already been declared, so that we don't go around changing results all the time, but I'm possibly amenable to alternative arguments.
 
I don’t like that it can only be done if it doesn’t change the outcome of the races. If the results are wrong they’re wrong and should be corrected. I think the process should allow for correction while assuring that certified results stand regardless of if the actual results wind up changing. And I don’t think that should require a citizen petition to happen - I believe citizen petitions can request this even if we have a process for it.
 
I don’t like that it can only be done if it doesn’t change the outcome of the races. If the results are wrong they’re wrong and should be corrected. I think the process should allow for correction while assuring that certified results stand regardless of if the actual results wind up changing. And I don’t think that should require a citizen petition to happen - I believe citizen petitions can request this even if we have a process for it.
I could see that inviting R4Rs and general chaos if errors flipped the election and there was no way to correct it to make sure the correct people were in the seats. I think we should also be looking into thresholds for requiring EC At-Large certification when an election is close, say with the candidates within 10% of each other, 55%-45% or closer in a heads-up match-up, or whatever is within 10% in each round for instant runoff, so that we catch errors that could flip the election before the results are certified.

Instant runoffs that are close, even between backmarkers that have the potential for a comeback could really screw things up if an inaccurate mouse click puts the wrong name in a box.
 
Last edited:
But the thing is Sil, at a certain point an election is over. Once winners are declared and oaths are taken and they start doing the job, it’s much too late for taking it back. Correcting the record is not inherently a controversial thing. Controversy comes from that correction revealing things to be not how they should have been. If we have a clear unambiguous process for certification, and a clear process for correcting results that leaves intact the certification, I’m not sure what an r4r would accomplish.

Ideally mistakes would be caught before certification but that may not always be the case. I’m not sure about the threshold idea, I think making this process more complicated will just lend itself to more potential errors. If an election is that close I would hope we’re already taking extra care with the counting, and our problem is more that we’re not getting the most mileage out of all the sets of eyes we have, and not having a consistent or predictable second opinion when releasing the final results.
 
We can start by reviewing the ideas we have in mind...

Pallaith's idea:
Have each commissioner keep a separate spreadsheet and compare them against each other to put together the “main” spreadsheet and then that’s the one that is evaluated by the full EC. Or…we pick a separate EC member who isn’t a supervisor to keep a second spreadsheet, and compare that one against the commissioner spreadsheet. Don’t allow results to be released until the competing spreadsheets are in agreement and the full EC signs off.

My idea, amended:
Any time during the voting period and for up to 24 hours after the voting period has ended, any Election Commissioner may privately challenge a particular ballot and submit an alternate interpretation of how it should be counted. The Election Supervisors may accept or deny any challenge. The period in which challenges can be made may be terminated early if all present Election Commissioners notify the Election Supervisors that they do not have any challenges to make.

Lore's idea:
Five election commissioners, including the two election supervisors, must approve the final results for the election to be certified.

Gorundu's idea:
See the Amendments of Certified Results rule proposal above.

I want to be clear; given the whole controversy over the idea of just going into an archived thread and editing it to put the correct results in, even as a footnote, I am not willing to do that without establishing a rule as to how it is to be done, and that we follow that rule if we're to go back and edit the 5/23 general results.

I think one of the three ideas regarding original certification needs to happen, and we need to address what needs to happen if we discover that the amended results under Gorundu's proposal means a different candidate should have been seated. I am not going to accept the idea that the wrongly seated candidate can't be unseated if the corrections happen in a reasonable amount of time.
 
I don’t think we should adopt Gorundu’s proposal at all. The others all have merit, why choose between them?

Two commissioners each track their own results, compare and agree on a single set of data, and that’s what the final spreadsheet has. You require the full commission to certify the results after the challenge period ends. Extend the challenge period, let’s say make it 12 hours instead of 6. More eyes being forced to compare, more time to check, and requiring a formal consensus I think will eliminate these issues in the future, or greatly refuse the chance accidents happen. Finally, we have a provision that if errors are discovered in the future, we allow an addendum to the official count.

I also agree that if the errors are discovered quickly enough, we should have a provision for making sure the actual winner takes office. But it’s hard once that oath is taken. We have officials who often are quick on the draw and take that oath right away. I’m not sure how we settle this particular aspect, but I think the rest of it is easy. That’s what I suggest we do, and the rest of this discussion should be fine tuning the scenario of incorrect results being corrected changing the result. It’s admittedly unlikely but we might as well have a plan for it.
 
I am in favour of at least extending the challenge period to 24 hours, possibly 48 hours if there is consensus for it. Waiting for 24 hours when there's like four months of toil and sweat ahead doesn't seem too much. This would cover weekends plus those of us on the other side of the planet and not US based.

The rest of the proposals I feel are for the views of the more senior ECs.
 
To be honest, I am more inclined to support Lore or Sil's idea. The idea is basically to ensure that a fair number of Election Commissioners have read through and verified the tabulated data. When or how the data is generated shouldn't necessarily be defined by the rules we are going to change.
 
With another election approaching, I want to bring this up again. We should try to figure out a resolution here. I reiterate my suggestion from my previous post as the best way to modify procedure, and that we should focus on how to respond to known instances of the results being incorrect.
 
Does anyone have any thoughts on these potential changes? They encompass some of the reforms suggested earlier in this thread.
Section Four: Election Procedures
[...]
9. If the results of challenged ballots could change the outcome of the election, t The results must be certified by the Election Commission at large. Otherwise, the Election Supervisors will promptly certify the results after the voting period ends.
10. To certify the results at large, the Election Supervisors will promptly present the results of the election to the Election Commission for certification by majority vote. Simultaneously, the Election Commission will also vote on the outcome of each challenged ballot. If the vote to certify the results fails, then each challenged ballot will be counted according to the outcome of their respective votes. The results will then be considered certified. The vote to certify the results and each vote on every challenged ballot will end as soon as a majority of non-absent Commissioners have voted either in the affirmative or the negative or forty-eight hours have elapsed since the start of the vote, whichever is sooner.
 
Last edited:
Change it to "absolute majority" instead of majority. That way the vote can end quicker and therefore be more efficient if a few ECers are absent.
 
Back
Top