* *** POSTMORTEM_2012-07-17 TXT - 20 Jul 2012 17:13:50 - JKNAUTH Postmortem on 7/17/12 Election at Precinct 01-43 ========== == ======= ======== == ======== ===== We didn't have any significant problems in the second primary election. As expected, very few people voted (about 5%). Essentially all ATVs were "perfect scenario". The Help Table had absolutely no traffic -- no provisional voters, no unreported moves, no transfers in or out, no voter disputes, not even any "where is my precinct?" queries. The Help Table judges were thus able take some time to demonstrate the laptop capabilities to all the other officials and let those others become familiar with the laptop operation. The Help Table judges in turn learned what the Ballot Table people did, etc. This took only a small part of the day, however. It certainly would seem better if the BOE could come up with some useful work that precinct officials could do during this dead time. It seems to occur every couple of years; everyone has long known that the second primary was going to be like this. Reviewing the manual all day isn't very productive. I have no specific suggestions about what work the BOE might have available that the precinct officials could do (maybe some type of paper processing currently backlogged at the BOE?), but the current situation seems very wasteful of a good resource. See my postmortem on the 5/8/12 election, http://jgkhome.name/WakeBOE/Postmortem_2012-05-08.txt. The comments there still apply, e.g., the comments on documentation errors, putting version dates on Quick Guides, confusing writeup on incident reporting, etc. Some Negative Points ---- -------- ------ 1) We opened only two ballot packs. One had 100 ballots; the other had only 99. Because we had counted the ballots when opening each pack, we were able to reconcile exactly at the end of the day and were not thrown off by the short pack. See http://jgkhome.name/WakeBOE/Ballot_Count_Reconciliation_2012-07-17.txt. 2) The M100 case contained only the pink Quick Guide. The AutoMARK case contained a detailed set-up document, similar to Chapter 10 of the old manual. The M100 case should contain a detailed "Chapter 10"-like writeup for the M100 and steel ballot bin. We brought the old Chapter 10 to the polling place and used it as the setup guide. 3) In the Chief Judge packet we received several months ago, the Monday poll book pickup time was listed as starting at 4:00 PM. More recent communications listed it correctly as 4:30. The same thing happened in the May election. Since the pickup time constricts when Monday setup can be started, the correct pickup time should be listed in all BOE material, including the initial material sent to CJs. 4) We tried to transmit the M100 data 10-15 minutes after the polls closed. The system kept retrying. Then after about 20 of those retries, one of the officials accidentally disconnected the phone line from the wall jack, which is about 50 feet away from our M100 location and around a corner. When we realized this, we plugged the connector back in the wall jack and the retries then continued. Meanwhile we worked on other poll closing tasks. After about try #32 we started to call the BOE on a different phone to get advice. Of course as soon as the BOE number was dialed and before the call was answered, we got "Modem Operation Complete" at the M100. Maybe the M100 realized the big guys were about to get involved and that the jig was up. Normally for such a short ballot we would have expected the operation to complete after only a few tries; in the past, for us completion usually occurred after around 10 to 20 tries, even for a large ballot. However this time it was still going at 20 before we made the unplugging mistake, compounding whatever the initial hangup was. Anyway, eventually it worked and the only downside was that we didn't notify our coordinator at the usual time since the M100 was still retrying. In fact our coordinator stopped by the polling place shortly after the upload completed to see what was going on, so we never made the call to her. 5) Some of the documentation used for the first primary did not exactly apply for the second primary. For example, there were no write-ins possible for the second primary, yet the check list implied there was a write-in plastic bag, etc., although none was supplied for the second primary. Some Positive Points ---- -------- ------ 1) The right gray bin was delivered this time and Monday set-up went much more smoothly than in May. 2) A set of phone line couplers was included this time. That was exactly what we needed and worked fine. 3) The A-K (three books) and L-Z (three more books) split of the poll books worked very well for us. We had two Registration Table people and set up only two pole signs. With the light traffic this election, that presented no problems. Of course in the November election, we expect very heavy traffic. As expressed in May, we still want to see the poll books not split a letter across two books. We expect we will need four Registration Table people for the November election; splitting letters makes work distribution difficult and makes lines at the Registration Table confusing. 4) We did not have any ballot curl problems. 5) The new format of the ATVs and poll books continued to work well. Written by Jeff Knauth, Chief Judge