"Quality Control"

I'm pretty sure that means "editing."

UPDATE:  The unity headquarters reporting room is apparently located in a boiler room, complete with furnace for undesirable results and documentation.

They're not even pretending anymore.

18 comments:

MikeD said...

Well, they're trying to pretend. "Oh it's just a glitch with the app". But at the end of the day, anyone who is not suspicious of the results is really far too gullible.

Christopher B said...

My hot takes

As an IT professional this sounds like a serious problem with lack of testing, especially stress testing under a realistic usage scenario.

As a former Iowan, remember that the majority of the app users are going to be .. well, let's just say there's a reason for things like Jitterbug phones.

Somewhat seriously, the caucus format isn't a bad way to do things if you're willing to let the delegate chips fall where they may. I think a significant part of the problem is the Democrat viability rule intersecting with the required proportional allocation of delegates. There was also a late breaking rule change after 2016 that prohibited people from switching their vote in the second round if they had selected a viable candidate in the first round. I can see that causing a huge number of fights as people see who the voters for non-viable candidates switch to. A better methodology might be a secret ballot using a ranked order of candidate choices, where votes for non-viable candidates are then given to the second choice, and so on until all votes wind up being for viable candidates.

Texan99 said...

The meme-verse is exploding. https://twitter.com/DNC/status/1224456542478462977 I like the dumpster fire floating down the flooded street.

"Trump has to have made a pact with the Devil. You can't 4-D chess your way into the Democrats blowing up their own caucus."

And now Sanders has released his campaign's internal results, showing 40% of the results: Sanders, Warren, and Buttigieg are the only candidates to break 15%.

E Hines said...

I can see that causing a huge number of fights as people see who the voters for non-viable candidates switch to.

I've seen a report--unverified--that one batch of nonviable Biden supporters divided themselves among bottom-of-the-ballot contenders, rather than going to one of the higher candidates, even the putatively "moderate" ones. And apparently Spartacus picked up a delegate from a collection of protesting caucusers.

The IDP, too, has announced they'll release the results of "most of the caucuses" at 1600 Central Time today. Most of. Wow. Mandy McClure, Iowa Democratic Party Communications Director, had this:

The underlying data and paper trail is sound and will simply take time to further report the results.

I have to ask, then: what's different between this year's "paper trail" and past years' that prevents these guys from going manual and reporting off their current paper trail?

Eric Hines

MikeD said...

@Eric H, they're not finished forging the paper record yet?

E Hines said...

Mike,

Here's Biden's legal team's position on that:

Joe Biden's campaign management has written a letter to the IPD decrying this, and demanding "an opportunity to respond [to IDP's explanation], before any official results are released."

Eric Hines

J Melcher said...

In certain simple Business Management use-cases, there is advantage to abandoning complicated apps and programs and "Visio" diagrams and Ganntt charts etc. Set up a live webcam watching a white board, covered with sticky notes and erasable lines, and let the team monitor status and accomplishments and changes in real time.

For a caucus in a gym or church basement, same idea. Set up a camera, take pictures of how many people are in which corner, and publish. If a "re-count" is needed, you count heads in the picture, not from the room-where-it-happened.

A Kodak moment. Or faster yet, maybe a Polaroid moment.

Easy peasy.

Texan99 said...

It leaves me wondering how they really did it in prior years. The failure of the app is a trivial problem; anyone with a pencil and a piece of paper out to have been able to work it out by hand by now. It looks very much as though the truly new variable this year is not the failed app but the need to show their arithmetic. The very first time they have to do this, suddenly they're completely unable to produce a result by their own ostensible rules? Doesn't it make you think in the past they just made up the result in a black box and published the results knowing they couldn't be challenged?

Christopher B said...

E Hines - The Booker ploy might be acceptable. There's no ballot in a caucus, it's whoever you want to vote for, plus there have been instances where candidates who dropped out have been on a primary ballot. I think the Biden incident is probably a procedure violation but I'm not an expert. I don't think a non-viable candidate(s) was supposed to be part of the second round.

MikeD - I didn't remember this at first but another article reminded me that in 2016 no record was kept of individual vote counts in precincts and none were reported, just the delegate equivalents. Reporting vote counts in each round was the change this app was designed to implement.

My particular conspiracy theory continues to revolve around the 'no switching' rule since that was new. IIRC in all caucuses since it was introduced in 1996 the second vote was a free for all among the remaining 'viable' candidates. The IDP has been pretty categorical in announcing that there was no issue with precinct totals being reported but that the glitch was supposedly something in the software doing the totaling. I keep suspecting that, contrary to the rules, they started seeing viable candidates losing votes in the second round in a large number of precincts. There could also have been an issue with inexperienced caucus goers who left after the first vote under the impression they were done. The 'problem' with the app that caused leaders to phone in could have been that it rejected a second round total for a candidate that was less than the first round total with no message explaining why. This also jibes with them releasing 'most' of the vote totals, i.e. they are holding back the precincts where candidates went from viable in the first round to non-viable in the second.

Christopher B said...

Tex - I largely agree. In prior years a lot of poor procedure to outright cheating at the precinct level was probably swept under the rug of just reporting a final vote total. The big change this year was having to report the delta between the first and second rounds.

Tales like the one Eric told have always swirled around how people react to having to change their initial vote.

I will give them the benefit of the doubt that the overall results are not made up out of whole cloth. A certain degree of incompetence is a better explanation.

Texan99 said...

I agree that incompetence is usually the most likely explanation for most of the confusion in a debacle like this. But I continue to wonder whether the results in past years reflected anything like the rules that supposedly were in place. Whether by finagling or by incompetence, it seems likely that the leaders simply announced results and made it impossible for anyone to audit them. It's too much of a coincidence that the first time they agreed to shine a spotlight on the mechanism, they found it impossible to explain their results even after more than 18 hours. The app isn't a good enough explanation for that failure.

MikeD said...

Tex is right. Incompetence is absolutely the most likely reason, and thus should reasonably explain everything. But the question then becomes how is this the first modern Iowa caucus unable to have any results turned in nearly 24 hours after it started? What does it say about the IDP that they're incompetent to this level?

I guess what I'm saying is that if "the IDP just blisteringly incompetent" is to be considered the better of the two possibilities, wouldn't you rather be caught cheating were you in their shoes? I'm pretty sure I would. "Better to be a fool than a cheat", perhaps, but I think they're WAY past "fool" at this point.

E Hines said...

...there is advantage to abandoning complicated apps and programs and "Visio" diagrams and Ganntt charts etc....

Per the updated link, the IT guy is being a stereotypical IT guy and blaming the user.

It looks very much as though the truly new variable this year is not the failed app but the need to show their arithmetic.

That's not necessarily nefarious. We can grunt through a calculus problem, for instance, but it's hard to program a computer to do that--and if your programmers are used to operating in a database environment rather than an arithmetic environment, it's harder.

The Booker ploy might be acceptable. ... I don't think a non-viable candidate(s) was supposed to be part of the second round.

Certainly, and protest votes on actual ballots are common, too--a significant fraction of 3rd party votes, if not most of them, are protest votes.

Viable candidate delegates are supposed to be locked in to their guy, but non-viable candidate delegates are free to change to another candidate--hence the Spencer campaign's ire over the non-viable Biden delegates' behavior--which was legit, if perhaps outside the spirit of the "second choice" opportunity.

I share, though, Tex' general suspicion, but I'm less inclined to write any of this off to incompetence. Sometimes malevolence really is the reason.

Eric Hines

Grim said...

As for prior years, Hillary supporters made a documentary about the cheating in caucus states by Obama’s team in 2008. Acorn? I think that was it.

Texan99 said...

"That's not necessarily nefarious. We can grunt through a calculus problem, for instance, but it's hard to program a computer to do that--and if your programmers are used to operating in a database environment rather than an arithmetic environment, it's harder."

I don't see that here. This ain't calculus. It's simple arithmetic, made difficult only by the need to compile almost 1700 calculations. So I can understand that they expected the app to deliver results by, say, 9pm Central and were disappointed, and they might need another hour or two to crunch it all by hand and give several layers of people a chance to double-check it before they went on air. By midnight, that excuse was starting to stink up the place, and by late this afternoon it's risible. Now they're promising maybe half of the results by 4pm Central. That suggests a problem with failure to preserve the underlying raw data, or a meltdown over how the rules of calculation are supposed to work when people are actually able to double-check the work--or both.

Maybe it only ever worked in the past because we had to take their word for how they got to the delegate totals. Meanwhile, the Republicans already were in the habit of producing the underlying vote totals, having somehow broken the mystifying code of arithmetic that apparently is beyond the Democratic leadership in Iowa. Frankly it's just not that hard. There's a first vote, then a chance to change votes, and a second vote. You count them all up, and you report them. What in the heck is the big deal? The only hard part is the horse-trading involved in persuading people to change their votes, not in tabulating them.

Aggie said...

It's pretty obvious that the problem has to be related to competence, or rather incompetence. The question is: What are they incompetent at? Running a caucus, or high-jacking one? Either answer provides the same end result.

Grim said...

It's pretty obvious to me that they were gaming this as hard as they could. But as you say, if you should choose to turn your hand to thieving, you should at least strive to be good at thievery.

E Hines said...

By midnight, that excuse was starting to stink up the place, and by late this afternoon it's risible.

Yeah, that was my question about the paper trail that McClure had touted. Lots of high school interns in Iowa that could have done the grunt work there. And quickly.

Eric Hines