[Date Prev][Date Next] [Chronological] [Thread] [Top]

RE: Multi-page ballots and turnout percentages

Just to get some bug numbers out there so people can follow along as we address some of these issues:


1969  Pima’s request that Rob is talking about here.

2129  Mark’s bug that Tab is working on.

1525  Another cards cast problem that is still outstanding.

2140  New RCR representing Mark’s request to show percents reporting by race in SOVC




-----Original Message-----
From: owner-support@dieboldes.com [mailto:owner-support@dieboldes.com] On Behalf Of Robert Chen
Friday, October 11, 2002 1:06 PM
To: support@dieboldes.com
Subject: RE: Multi-page ballots and turnout percentages


Ken is correct that we try to avoid multi-card elections like the plague but, just like the plague, despite whatever precautions you may take, you can still get it.  Jurisdictions like Pima County, Fresno and Alameda are probably lucky to not have multi-card ballots. 


Pima may go away from that list now that we have multi-lingual support.  Brian Crane from Pima did make a suggestion that might be interesting if this were to ever get resolved.  They suggested having the first card in a multi-ballot to be the "control" card and that cards cast results are based on that control card.  If a voter returns a ballot without the "control" card, then the county can get GEMS/AVOS to accept the ballot with a generic control card and it is assumed that the voter blank votes the first card.  If the voter returns the ballot without second card, then it is assumed that the voter blank votes the all the races on the second card. 


rob chen

-----Original Message-----
From: owner-support@dieboldes.com [mailto:owner-support@dieboldes.com]On Behalf Of Ken
Friday, October 11, 2002 12:19 PM
To: support@dieboldes.com
Subject: RE: Multi-page ballots and turnout percentages

-----Original Message-----
From: owner-support@dieboldes.com [mailto:owner-support@dieboldes.com] On Behalf Of Mark S Earley
Friday, October 11, 2002 9:09 AM
To: Support
Subject: Multi-page ballots and turnout percentages


How have people handled Election Night Reporting in accounts with multi-page ballots? The Turnout percentages are misleading as everyone knows. What have others done or expect to do in this situation?


I’d like to hear from all interested parties on this.  Here are a few comments of my own.


Multi-card ballots have always been weak in many ways in GEMS.  You should avoid them at nearly all costs.  Have a look at the lengths Rob Chen went to avoid them in California – attached.


This has an interesting feedback loop however.  People avoid multi-card elections at all costs because GEMS multi-card support is weak.  That makes multi-card elections very rare.  Since multi-card elections are very rare, improving GEMS is always a low priority relative to common pressing problems.  So GEMS multu-card support remains weak.  Rinse, repeat.


I am considering advising the counties to turn off the Cards Cast feature, and keeping the Times, Totals, and Registered Voters options turned on. This will put the county-wide races near the beginning of the reports, giving an accurate appraisal of the numbers of voters. This will work well on the Summary Reports where the Times Voted for each race gives a percentage.


Right.  You have reinvented the standard operating procedure.  A jurisdiction-wide race will always give you the numbers they want.


The SOVC Report does not have a comparable percentage stat on it within the races. Most users are used to the SOVC showing the Turnout percentage in the third column, and turning this feature off will likely upset them and the reporters they serve.


Hmmm, indeed.  SOVC should show the percent as well.  I will create a new RCR to that effect and get this assigned.


One workaround is to export the precinct results, import into Excel, massage the formatting and use the times counted for countywide races as the turnout. For one county this will be manageable. For many counties and remote support, this will be problematic.


Tab is working on fixing the cards cast problem.


Mark Earley

850 422-2100 - office/fax

850 322-3226 – cell