* *** POSTMORTEM_2019-11-05 TXT - 11 Nov 2019 01:26:13 - JKNAUTH Postmortem Report for Precinct 19-19 for 11/5/19 Election ---------- ------ --- -------- ----- --- ------- -------- > SUMMARY < ======= The turnout was fairly heavy for a municipal election, although smaller than the last such election at this polling place. We were kept busy most of the day. The weather was good; we had no humidity problems with the ballots. The new Tabulator worked well except for a significant closing problem described below. This was the fourth election at the Sanford Creek Elementary School (SCES) polling place. The SCES Media Center is a very good voting enclosure. The support by SCES personnel was again excellent. The 19-19 precinct officials gave me many good suggestions. I have tried to include them in the sections below. > STATISTICS < ========== The calculations below use the number of voters registered as of 11/4/19 in Rolesville and Wake Forest; registered unincorporated voters are not included. The turnout for this municipal election was larger than we had expected -- about 13.3% at the polling place and 0.5% via absentee voting (mail-in or one-stop) for a total of 13.7% of registered voters. Tabulator vs. ATV counts: 764 vs. 764 Provisional ballots: 3 Spoiled ballots: 11 Emergency bin ballots: 0 Challenged ballots: 0 Curbside voters: 8 AutoMARK voters: 0 Observers: 0 Registered voters: 5763 (only in corporate town limits) Monday-list absentees: 25 Last-minute absentees: 2 ---- Total absentees: 27 Polling booths: 23 (includes 1 provisional and 1 ADA) Total chairs at tables: 0 --- Total "booths" 23 Ballot Count Reconciliation web page: http://jgkhome.name/WakeBOE/Ballot_Count_Reconciliation_2019-11-05.htm > TABULATOR < ========= The new DS200 Tabulator worked very well, at least until pack-up time (see below). The setup and takedown instructions were good. The DS200 never kicked out a ballot that was entered too gently. That had been a big problem with the old M100 Tabulator. Many voters had problems locating the right slot to slide in the ballot. The DS200 Tabulator has two different ballot slots. A small black one above a large gray one. Both have nearly invisible arrows depressed in the plastic. Although clearly our ballots could go only in the large slot, many voters stopped and tried to figure out what to do. Eventually we taped a paper slip over the small slot with a note and arrow pointing to the large slot. Our Tabulator Monitors also started pointing to the gray slot for each voter. The small slot is obviously a source of confusion. I wonder why the WCBOE didn't try to put something over it. Next time we'll probably start right away with the covering note if the WCBOE hasn't addressed this. Also, we recommend that the WCBOE apply some color (maybe yellow or white) to the arrow in gray plastic to make it more visible. One time a ballot was rejected with an ambiguous "Too Long" message. We then found that the voter had spilled something on the ballot causing a small wet spot. We spoiled the ballot; the replacement worked fine. There were 764 ballots this time. We did not have to stop the voting process during the day to empty the bin. With the M100 we would have had to empty it once during the day in this election. When we opened the bin at poll closing, we saw almost all the ballots had stacked neatly where they should have. All this bodes well for 2020 when we will probably have to open the bin only once during the day for the large elections. With the M100 we would have to stop the voting process many times during the day to empty the bin. Unfortunately the alarm on the DS200 Tabulator is almost non-existent, worse than on the M100 Tabulator. Fortunately the DS200 screen is a lot larger and clearly displays to the voter if the ballot was accepted or rejected. However that only works if the voter waits to see the result, as the Tabulator display says to do. But we found that many voters do not wait; they can be out the door and hard to catch if an error occurs. (We had at least one such runner this time; fortunately she was caught.) Our Tabulator Monitor should tell each voter to wait, but that is not always possible if a number of voters appear at once and the Monitor is also trying to hand out "I Voted" stickers. We need a more "In Your Face" way to tell the voters to wait to see the result. If the WCBOE doesn't come up with anything better, I am going to make a large sign to hang from the Tabulator lid: "WAIT HERE! until the Tabulator says you are finished." We didn't have any trouble putting in the Tabulator's power connector, although I did not hear or feel a click when I pushed it in firmly. The Tabulator powered up when it should have. The plug came out easily at the end of the day vs. one of my previous experiences. Months ago I had emailed a recommendation to put some white vinyl tape around the plug and write an "UP" on the flat side. The plug orientation is not easy to see in the cramped area behind the Tabulator. I don't recall seeing the Tabulator User Guide on the back of our Tabulator Base. Should it have been there? I did like the pictures and descriptions in that document -- very helpful. NOW FOR THE BIGGEST PROBLEM. When packing up the Tabulator at poll closing, the round key would not lock the closed screen, no matter how hard the key was turned and the screen was pressed down or was not pressed down. Finally we opened the screen; the key now turned easily, showing the latch movement. But opening the screen rebooted the Tabulator, leading to more problems. After a shutdown we closed the screen again and tried locking it again. After some effort it worked this time. Because of previous Help Line difficulties and the time crunch the 9:00 deadline causes (see later for both), I did not try to report all this Tuesday night. I emailed a detailed report to the WCBOE on Wednesday. > AUTOMARK < ======== There was a major problem with the AutoMARK. Two of our officials set it up as usual and I turned it on with the red key. I then went away to do other things while it booted. When I came back later to check on it, I saw a ballot was sticking out of it. It was a normal N009 ballot and had been marked, presumably by this AutoMARK. The ballot feed tray was closed, so the ballot had difficulty getting out and was curled. I pulled it out the rest of the way. There was an "Insert Your Ballot" message on the screen. It appeared the AutoMARK booted normally and then tried to eject a marked ballot that had still been inside. I attempted to call the Help Line (unsuccessfully -- see below) to report all this. I then called our Coordinator. She then reached the WCBOE via a different number. They later called me and said someone would replace the AutoMARK on Tuesday morning; also that person would take custody of the ballot. The replacement was done early on Tuesday and I turned over the ballot at that time. > HELP LINE < ==== ==== We tested our mobile phone early during setup and it worked normally. When we tried to call the Help Line via that and several other phones to report the serious AutoMARK problem, we could not get help. The call went thru and we heard the recorded message to press 0, 1, or 2 for the stated reasons. However no matter what button was pressed, we were sent into a black hole. All you heard was some static, no matter how long you waited. Finally we called the Coordinator to report the AutoMARK problem. (We had not called her earlier because of the seriousness of the AutoMARK problem, which we wanted to report directly to the WCBOE.) She discovered that the Help Line people went home at 7:00 and got us to WCBOE people by using a different number. 1) Why was the Help Line unmanned at 7:00 when there must have been many sites still possibly needing help? 2) If they did shut down Help at 7:00, why didn't the phone message say that instead of telling us to push buttons that led nowhere? All that wasted a lot of our time, not knowing whether it was some phone problem or whether it was a button 0, 1, and 2 problem at the WCBOE. We tried all of the buttons several times. I had not seen such Help Line problems in all my prior Monday setups. > DELIVERY DEADLINE < ======== ======== As I have pointed out before, it is very unrealistic to expect me to get this large and complex polling site closed: the furniture arrangement restored, all the "bookkeeping" done, everything packed in bins or my car, etc. and then drive all the way to Knightdale by the deadline. From now on assume I will be driving to New Hope in Raleigh. It is not a much longer drive, but at least that facility should not be closing before I get there. It's usually about a 30 minute trip either to New Hope or to Knightdale. > STAFFING < ======== We had 11 officials. This is about the right number for a large election at this polling place. The WCBOE wants to train as many people as possible to prepare for the large elections in 2020. We moved some people around to give them a background in various areas. For poll closing I assigned one person to coordinate the packing up. I assigned another person to prepare the input for the Reconciliation Form. The intent for both of these was to have organized work being done while the three judges were tied up doing all their poll-closing items and were less available to direct the other officials. > POLLING PLACE SETUP AND PRE-ELECTION PREPARATIONS < ======= ===== ===== === ============ ============ Although we expected relatively few voters compared to what is expected in 2020, we set up a part of the voting enclosure as if this were a large election. This let us experiment with some solutions to problems we encountered in the 2018 large election. These problems/solutions are described in the two following sections. In 2020 we will set up a number of tables and chairs as voting "booths", as we did in 2018. This had worked well in 2018 and should also in 2020. We didn't bother doing all that furniture rearrangement this time since we already knew it would work and the large number of booths from the WCBOE was more than adequate for this municipal election. For some pre-election preparation, in August I had posted an article on Nextdoor about the upcoming election; it had a link to a web page which described some special voter considerations for the SCES polling place. See http://jgkhome.name/WakeBOE/Nextdoor_Post_2019-08-18.txt and http://jgkhome.name/WakeBOE/19-19_2019-11-05.htm. Then there was a short reminder article the week before the election. In late October I contacted the SCES Media Center director to verify a diagram showing how furniture should be restored after the election. The Monday before the election I had set out direction signs along the long cross street leading to the school. This time I taped paper mileage data on some of the signs; it is a long distance from the main intersection used by most of the voters to reach our polling place and it is easy to wonder if you missed it. See http://jgkhome.name/WakeBOE/Sanford_Creek_Elementary_Signs.pdf. All these pre-election preparations worked well and will also be done in 2020. > POLLBOOKS AND REGISTRATION TABLE LINES < ========= === ============ ===== ===== Although the turnout was light compared to what is expected next year, we still had four officials work at the RT. This helped with the training and also let us configure things as they will have to be for the 2020 elections. The pollbook letter-splitting problem still exists, worse than ever. For example, this year we had the last name Elliot split, with one Elliot person in one book and another Elliot person at the same address in the following book and in a different line. Similarly the last name Rogers was split across two books and lines. It makes the RT letter signs less useful -- e.g., which line does an Elliot person know to stand in? A problem we had in November 2018 and in other elections was that the letter-division signs at the Registration Table were hard to read from a distance. Because we have just a single door for entry/exit and other configuration constraints, incoming and outgoing voters must cross paths in front of the Registration Table. To provide room for this, the RT waiting lines have to be about ten feet from the table and signs. The provided signs are small and the writing must be crowded because of the pollbook letter splitting problem. All this makes the signs hard to read from where the RT voters must line up. In this election we solved this by putting a manila file jacket over the sign at the top of each stand; the letters were then written on a standard 8.5 x 11 sheet of paper taped on the front of each jacket. This let us write much larger letters and seemed to solve the problem. (That is, it solves the problem except for people with the same last name, but who must be in different lines. We would have to put first names on the signs as well as last names; that's much more than even our big signs can contain. The WCBOE needs to finally address this letter splitting problem. It can't be that hard for the WCBOE or print company to program.) From the start in this election we put tape markers on the floor to better separate the RT waiting lines. We had started doing this at the end of the last election when the lines grew large and the waiting area was crowded. The marking worked well in this election and will be done in 2020. As in prior elections we put masking tape arrows on the floor and signs on some posts to help guide voters from the polling booths in the back of the large room to the Tabulator by the door in the front. The Tabulator isn't visible from the voting booths. The path to it goes in front of the RT waiting lines and then angles to the left. See http://jgkhome.name/WakeBOE/Sanford_Creek_Elementary_Layout.pdf. After handing out a ballot our BT Officials always pointed the voters both to the booths and to the Tabulator, but many voters did not pay attention to the last part. The arrows and signs seemed to help, but still many people ignored them. Our Tabulator Monitor had to wave his hands to signal the confused voters in the back of the room; our BT and HT officials had to keep saying, "Follow the arrows!" when exiting people seemed to be lost. In one election we had tied a big helium balloon on the Tabulator, but that didn't work either. In 2020 we will try some bigger and better signs. We will need more sign stands for this. > CURBSIDE PARKING < ======== ======= Another problem we had in previous elections was how to clearly designate the entry to the curbside parking area and how to show which direction cars must face. (Getting to curbside parking can be pretty complicated, given the SCES lane configuration for carpools. See http://jgkhome.name/WakeBOE/Sanford_Creek_Elementary_Parking_Area_Traffic_Flow.pdf.) This election we tried to better designate the Curbside parking area entrance by putting some tan masking tape arrows on the asphalt. Nevertheless some people seemed to have a problem understanding the purpose of (or seeing) the doorbell stand and instead parked some distance from it. We subsequently made some additional signs to direct people to the proper side of the parking area lane so they would use the doorbell. We did not get enough curbside users after this addition to see how well it worked. We try to warn people before each election to avoid the carpool times; they are particularly a problem for someone trying to reach the curbside area. I asked our curbside people to give a slip to any curbside voter. It references http://jgkhome.name/WakeBOE/19-19_2019-11-05.htm which lists the main times to avoid and shows how to access the curbside parking area. This election we saw there was another such carpool line for "ball handlers"; it started at around 5:00 PM. Fortunately it was much shorter than the main afternoon line. > CURBSIDE PROCESSING < ======== ========== Our final voter of the day, right at poll closing time, was a curbside voter who required Help Table assistance and eventually had to complete a provisional envelope. A perfect storm of what you do not want to happen when you need to get packing done ASAP to not miss the 9:00 delivery deadline. Our Ballot Table officials pointed out it would be helpful to have the BT documentation explicitly point out the special things they must do for curbside processing. Currently most of the curbside ATV processing is described only in the curbside documents. The BT officials must depend on the curbside official to tell them what to do, which might be a problem when inexperienced people are involved. For a curbside voter the BT officials must provide a ballot when requested by the curbside official, put the ballot style on the ATV at that time, but NOT number the ATV until the ATV is later returned by the curbside official, then add to the BT curbside count at some point. (Note that the ATV might not come back to the BT, e.g., if a provisional envelope is required; so when the count add is done is a question; the BT count may end up smaller than the true curbside number. This is probably not a big deal given the purpose of the count and that provisional curbside voters shouldn't happen often.) One of our curbside officials asked if steps 3 and 4 of the process on Manual page 71 should be reversed to better match the normal RT processing -- get the voter to state their name and say it back before the voter signs the ATV. > AUDITS OF ATVS AND PROVISIONAL ENVELOPES < ====== == ==== === =========== ========= The Registration Table officials did an excellent job. There were *zero* errors in the ATVs they created. It was very interesting that there were no "Verify" voters in the 764 ATVs; usually there would be quite a lot. I found one ATV processed by the Help Table in which a voter had been allowed to vote in our precinct when she should have been transferred to another. However the same ballot style applied in both precincts. I was able to check only two of the three provisional envelopes. The third one was for a curbside voter that was still being processed when the polls closed. For the other two envelopes, one was missing a birth date. The other was good and arose because a Help Table in another precinct had transferred a voter to our precinct without initialing the ATV and had also not completed another field. The Help Line told us to have the voter vote provisionally. For the mistakes made by our people, I again emphasized the need to follow the flowcharts. Doing that would have prevented both of their problems. > RECONCILIATION FORM < ============== ==== It took longer than necessary to complete the Reconciliation Form. The ATV vs. Tabulator counts matched right away -- the important check. The Ballot Table officials had done a good job all day making sure those numbers stayed in sync and quickly corrected an error they once found earlier in the day. However the ballot inventory check did not work. The total unused ballots had not been calculated correctly. After a couple of reworks, the error was found and everything balanced. We found that none of our pens wrote well on the yellow fields on the Reconciliation Form. Eventually we just wrote in the white area off to the side. > CAMPAIGNERS < =========== We had almost no campaigner problems. As usual they wanted to use the restroom in our break room. We pointed out the WCBOE restriction on this since using that restroom requires going thru the voting enclosure. I also checked with the school's Assistant Principal to see if campaigners could use other school restrooms accessed via the SCES main entrance. I expected a "No" answer and that's what I got; it would definitely lead to security problems. He said there was a porta potty at the nearby Rolesville park gravel lot. I later conveyed all this to the campaigners up on the hill where they usually set up shop. Hopefully campaign officials are warning their volunteers about this 19-19 constraint. Maybe the WCBOE could figure to some way to spread the word. > VOTER COMPLAINTS/DIFFICULTIES < ===== ======================= I didn't hear any complaints myself. One official told me that a voter said she had contacted some Rolesville employee about where to vote and was told to vote at our polling place. However in fact it was the wrong polling place for that voter. Our Help Table officials pointed her in the right direction. In this election most of the voter redirects by our HT were to 19-09 as in the last election. Several years ago they were to 19-18, which was expected immediately after the 19-10 split into 19-18 and 19-19. Several unincorporated voters arrived at the Help Table and had to be told why they could not vote. Our Help Table Officials pointed out that the "No Ballot Style" wording in VAST was confusing to them. Couldn't it or the flowchart be more explicit about the cause -- the voter lives in an unincorporated area, so is ineligible to vote in a town election? > SPOILED BALLOTS < ======= ======= In the first few hours of the day we had eight spoiled ballots, which seemed very unusual for such a simple election. It turned out that all of them were ballots for the Rolesville election and had been rejected because the voter had marked three Commissioners instead of the two that were allowed. (In contrast the Wake Forest race allowed three.) When we figured out the problem, I put a note on the entry door to emphasize that Rolesville voters could vote for only two commissioners. We did not have a single Rolesville spoiled ballot for the rest of the day. > SUPPLIES < ======== 1) There were no inventory labels on various kit bags as was done in a prior election. Having those bag labels made inventory checking much easier than having to share the single gray bin supply list among a bunch of people. Having no bag labels slowed down setup. 2) The cardboard bins for the booths and signs were packed VERY tightly and at least one was hard to unload. When it came time to put the toothpaste back in the tube at the end of the day, it would not fit. We left some signs beside the bins instead of in the bins. 3) The initial (Saturday pickup) ballot box seals still pop off easily, e.g., when the boxes are being removed from a car trunk. Not very secure. Similarly at pack-up we found the "Not Voted" stickers did not stick well; the "Voted" stickers were better. 4) For our polling place, which has a hard-to-get-to curbside parking area, we would like some simple curbside signs with arrows. Currently we have to put three signs in a row to convey the same information and we use four of those three-sign combinations. 5) All the table sign stands this time did not wobble vs. in past elections. We could use a couple more to hold our special signs (see above). 6) The collapsible cones worked very well for our curbside parking area and seemed to stand up well to the heavy wind in the 2018 election. They do take up much less room for transportation. A great solution. 7) We were happy to see two rolls of blue tape, which we use a great deal. See http://jgkhome.name/WakeBOE/BOE_Supplies.txt for some other supplies comments from past elections that still apply. See http://jgkhome.name/WakeBOE/BOE_Wish_List.txt for my current general wish list. > DOCUMENTATION < ============= General ------- See my pre-election comments: http://jgkhome.name/WakeBOE/Comments_on_2019.v.1_CJ_Pocket_Guide;_CJ_Supply_List.txt http://jgkhome.name/WakeBOE/Comments_on_2019.v.1_Documents.txt I had also sent emails to elections@ and to individual staff members on some specific items not listed here. 1) There should be a revision date by each item on the Precinct Officials website. See http://jgkhome.name/WakeBOE/WakeBOE_TOC.htm for a possible format. There should be a change date by any added/changed FAQ. This was done a couple of years ago. For this election there was not even a global change date for the whole file. Same problem as in 2018. 2) There should be an errata file on the Precinct Officials website for key errors/omissions. Since there was none, I gave my officials a list of four problems and workarounds in my assignments email prior to Monday setup. 3) Avoid "duplicating" text; get it right in one place and refer to that place from others. If I don't know it is guaranteed to be identical, I feel required to review the "same" stuff found in multiple places. The CJPG and Tabulator User Guide have a lot of (almost) identical text. The AOQG and Manual also have a lot of duplicate material. Mobile Phone Instructions ------ ----- ------------ The Help Table QG says to set the phone volume to maximum, but does not say how. Our Help Table officials figured it out, but the phone document should probably provide the details as it does for other things. The phone kept downloading something at various times, which was troublesome when you needed to make (or get?) a call. I don't think the document said anything about this. Is this a known problem for all the phones, or just for the one we got? Ballot Table Quick Guide ------ ----- ----- ----- The BT QG does not say for Tuesday morning that BT officials should open each ballot box and verify that the box label and number/style of the contained pack are in sync. We verified with the WCBOE that this was an error and we did do the check. Videos ------ The current videos for the DS200 are very good. There should also be one of normal DS200 processing for a voter. I had previously emailed to the WCBOE a link to a video done by some Virginia election officials. > MISCELLANEOUS ITEMS < ============= ===== 1) When they delivered the replacement AutoMARK, they also delivered a second laptop. We had not reported any problem with the original laptop and had not requested a replacement. The lady replacing the AutoMARK did not know why she had also been told to deliver a laptop. We reached the Help Line and they said it was just an extra we could use. Why wasn't the delivery person told that so she could tell us, or why weren't we notified some other way? We thought it might have been a mistake or somehow related to the AutoMARK replacement. As it turned out, we had very low HT activity and had no problems using a single laptop all day for normal business and for practice. However two laptops would probably be very good for the 2020 election. It was just confusing that we got two this time and in this way since before we had always gotten just one. 2) As usual the people doing setup did not know to use the extenders for the accessible booth. I caught sight of the incorrect setup later and showed how to fix it. The relevant part of the Manual (page 26) is apparently ignored each year when the Manual is read and I guess the extenders are not mentioned in training. 3) When we were discussing cell phone restrictions at Monday setup, one official suggested that having a selfie station somewhere outside the voting enclosure (50+ feet away) might help handle those who want to take a picture of the voting experience. In fact we did have one of those people on Tuesday and I had him go up on the hill to get a picture of him, the Media Center, and the signs, etc. we have out front. 4) We had a problem with the room temperature. It felt pretty cool to a number of people. I asked the Assistant Principal to see if it could be fixed, a harder problem than people might appreciate since all Wake County school temperatures are controlled from some central location in the county, not at the school itself. He tried to do an override, but it did not seem to work. 5) There are many floor outlets in the Media Center, each with a big cover plate. The one in front of our Tabulator location is partially sunk and is a tripping hazard, so I bring a small rug to cover it. The rug was being moved around this time as people voted; I eventually taped it down with masking tape. > PICTURES < ======== For some pictures and explanatory text about the voting enclosure, voting equipment, and the parking area, see https://photos.app.goo.gl/9Wo9XSD6ppoUxvUJA. If you are unfamiliar with viewing Google Photo albums, http://jgkhome.name/Misc/Google_Photos.htm gives some hints. In particular, note how to view the full text beside each picture. > PROPOSED SIGNS < ======== ===== I created a set of proposed signs to use in place of the handwritten signs we used in the 11/5/19 election. See the http://jgkhome.name/WakeBOE/WakeBOE_TOC.htm#proposed-signs section of my website. Report written by Jeff Knauth, Chief Judge for 19-19 with input from the other officials for this precinct