Each EIRS hotline report represents over 4,000 similar not reported- based on known data for

[philb]

most of the types of incidents that have known incidence data

It is known there were about 1 million RoboCalls
5 EIRS reports that reported RoboCalls and 3 more that might have been
(there were a lot more reported to News Media than to EIRS)

So over 100,000 RoboCalls per each report to EIRS

none reported about warning letter in Orange Co. Calif. but 14,000 letters known to be sent out

Sarasota Dist 13 disappearing votes --- 15,000 excess undervotes

3 reported to EIRS hotline didn’t see the race on screen, and 4 reported disappearing votes
(but mentioned they noted it happened to others)

At least 2000 didn’t report for each one who did; actually its known a good many people observed the problem on the review screen and fixed the problem so didn’t show up as undervote; so the ratio of not reported to reported is likely at least 4,000 to 1


Similar ratio in New Jersey Congress Dist where over 10,000 undervotes and 2 reported disappearing votes
5,000 for each EIRS report


Robo-Call data
In all, a total of 1 million calls were spread over 53 House races; this means that an average of 20,000 calls were made in each district, each of which contained about 200,000 votes. The calls could potentially have reached one out of every 10 voters in the targeted races.
http://www.salon.com/news/feature/2006/11/21/cheat_sheet/index1.html

Orange County warning letter which caused displinary action against sender:
This October, nearly 14,000 Orange County, Calif., residents received a letter stating that if they attempted to vote on Nov. 7, they could face jail time or deportation. The recipients -- all of whom had Latino last names and were registered as Democrats -- were discouraged in formal Spanish laden with grammatical errors from voting in the midterm elections.



Sarasota Dist 13 disappearing votes

15,000 excess undervotes


3 reported didn’t see the race on screen and 4 reported disappearing votes

At least 2000 didn’t report for each one who did; actually its known a good many people observed the problem on the review screen and fixed the problem so didn’t show up as undervote; so the ratio of not reported to reported is likely at least 4,000 to 1


Likely, some things are likely to be noticed and reported more than others

2004 had a bigger EIRS effort, and likely a lower ratio of unreported to reported

[bleever]

1. Very important point. Actual reported incidents represent many, many more.

K&R.

[philb]

2. There were a lot of precincts with broken machines/long lines but only 1 or 2

in those precincts knew how to report and bothered to do so.

Similar for scanners that would not read the ballots so ballots had to be put in a box to read later

similar for most of the things that happened to the whole precinct but only 1 or 2 in the precinct reported.

The EIRS reports indicate the kinds of irregularites that happened in the different counties throughout the U.S.
But the hotline was more highly publicized in urban areas in swing states from 2004, when a big EIRS effort was done for the first time.
So more knew about it in states like Ohio and Florida and Pennyslvania that were swing states in 2004
and had lots of irregularities reported then. But from the things that are known to have happened to large numbers who would have known about it, its clear only 1 or 2 in a precinct or a few in a county reported them.
So its clear that if someone reported a long line in a precinct, it affected a lot more people.

Likewise for things that affected counties or districts(like Robo-Calls)(large numbers of calls,a few reports)
Each report represented things that happened to a lot more people, for most of the types of irregularities