The data provided by Elections Ontario is for the 2014 provincial elections, which as best we could determine were ALL "paper based" non-electronic votes, that allowed candidates (by referee) to observe the counting process and where necessary undertake various processes when the counting of votes appeared to require further review. The data also breaks the "invalid votes" into two reported values, "Rejected" and "Unmarked".

The data provided by the Township of Clearview is for the 2014 municipal elections that used a totally electronic, internet based system that appears to only provide a single broad based description of "spoiled ballot" for votes that were either not counted or were "spoiled".

The data from the Township of Clearview does not provide any information on what criteria the "secret" software uses to count votes or to determine what constitutes a "spoiled ballot" so candidates have no opportunity to validate the integrity of the voting process or review the accuracy of results until AFTER the returning officer received the results from the voting system provider and has announced the results! The data also fails to break down "invalid votes" into either "Rejected" or "Unmarked" ballots.

In 2014 the Township refused to require as part of their RFQ for the voting system that a copy of the source code used in that process is lodged with a  third party (such as the Ontario Privacy Commissioner) do that in the event of the types of data difference dealt with in this paper can be reviewed by the courts under  a judicial review of what appears to be a flawed process not tested nor approved by the Province of Ontario.

Provincial Voting statistics 2014

Bad vote average is the number of total uncountable votes /total cast votes expressed as a percentage of votes cast.
Bad vote average for all Ontario Provincial elections in 2014
Average for all races = 0.722562% (Median value 0.694317%)
(Hi = 028 HALDIMAND-NORFOLK - 1.854905%, Lo = 039 KITCHENER-WATERLOO - 0.459600% with only 5 of the races above 1%)

Clearview voting statistics 2014

Bad vote average is the number of total uncountable votes /total cast votes expressed as a percentage of votes cast.
268 votes cast in Clearview 2014 were rejected by the secret on-line counting system as being spoiled votes and as best we can determine there were another 11 totally non-attributable votes in the "honeypot", meaning that
279 votes were rejected by the SOFTWARE used for counting!

For the various races the number of "spoiled votes" were as follows
What's wrong with these numbers?
In the seven races for the 2014 election that were not acclaimed there is a HUGE variation in the Bad vote average for the Township of Clearview. In 4 of the races the bad vote average was over twice the provincial average and the race for Deputy Mayor had a bad vote average over 4 times the provincial average and twice that for the Mayor's race.

Each voter in Clearview has the opportunity to vote on 1 ward race and the 2 at-large races with the same single voter code as issued by mail. The system approved by the council is fully electronic and used composite voting controls, so if a voter successfully "logs on" to the voting system and places their votes for the 3 races it should be expected that their votes in each individual race SHOULD ALL be deemed a valid vote. If a voter chooses not to vote in any of the three races set out in their "voting package"  their vote should not be deemed to be "spoiled", it should simply not be counted either as a vote or an "offer to vote".

A separate ballot is designed to be used for only one office, by-law or question, as the case may be; a composite ballot is one that combines the contents of two or more separate ballots. 1996, c. 32, Sched., s. 41 (5)
Q - Did the returning officer and elected councilors fail to recognize these disparities and if so what steps will be undertaken from this letter to validate the integrity of the chosen voting system for the voters of Clearview Township?

Q - How can there be such a difference in the bad vote average for the 3 races with common voter participation?
The process would appear to rule out the electronic communication infrastructure (integrity of the internet connection) as being a factor in just 1 or more of the three votes being deemed by the SECRET software used to count votes to be "spoiled" since it is highly unlikely (and maybe not even possible) for each of the three votes to take place at different times from three different internet access points.
Without  a court having free access to source code that issue could never be resolved.

In the case of the two at-large votes the number of votes cast were almost identical (5,111 to 5,095, a
difference of just 16 votes cast) yet the number of "spoiled" votes reported by the voting system for the Deputy Mayor role (158) was 2.3 times that for the Mayor (68) and was far above the difference of 16 cast votes between polls.

In a single voter "package" with three votes to be taken, would not all of the votes not be counted if one of the votes from the three available was found defective?

In that scenario it would be expected that the number of defective votes would be equal across all three races.
Questions for council, and the returning officer:

1. Why did these disparities in data not "trigger" some concern by the returning officer for the integrity of the system council had approved?

2. If these results did "trigger" any form of analysis and investigation by either the returning officer or council, where can the public find the results of that "investigation" or review  and why was this process kept from the public?

3. What are the grounds for a vote being considered as "spoiled" by the voting system approved by the Council for the Township of Clearview?

4. What are the criteria regarding the impact of votes that end up in the "honey pot" confirmed by the software creator to be a place where any votes that cannot be determined to be from a  registered voter, or from an "approved" internet access point (something else for which there does not appear to be a public definition see additional note C) below on that concern ) and cannot be "traced" by any means to a specific voter number, end up?
Clarifying notes:
A. The Township of Clearview was advised long before the last election of the problems with the visibility of voting codes and the returning officer determined that no action would be taken by the Township on that clearly identified flaw in the process that may well also have led to the re-issuance of voting credentials (maybe ask the returning officer if any credentials had to be re-issued and upon what grounds, together with the steps taken to ensure that the request was for valid reason and that the software was then changed to reject or deny a vote based upon the original voting credentials.

B. The Township of Clearview and all associated levels of the Provincial Government and Elections Ontario have been
previously advised in 2013 of these problems with this particular voting system and ALL have failed to act or even acknowledge receipt of such concerns. (Letters available upon request).

C. For the 2014 election at least one web site being run by a candidate had a link to the "official" voting site from that privately owned web site. Without knowing the specific routines employed by the software (and that would be available for review if a copy was placed in escrow) it is difficult to determine if votes placed through that link (created without any directive base of linkages specific to and approved by the voting software supplier) might well have all ended up in the "honey pot". That would have meant that the supporters of that particular candidate might well have "added" to her loss. I fully understand that "conjecture" might be inappropriate analysis tool, but when the full process that voters are being required to use is not "open" with regard to validity and integrity testing, conjecture is the only tool upon which concerns can be based.
The differences between the Provincial and the Township of Clearview voting averages.
UPDATE - August 10, 2018

We have now filed two Freedom of Information requests trying to find out WHY so many votes were rejected in 2014.

The first was deemed a vexatious request by the "illegally delegated Head for FOI" who is also the returning officer!

The second, delivered to each member of council under MFIPPA section 3 has been rejected and sent back to us by CAO Sage, who as best we can determine has absolutely no standing in regard to FOI matters.

Both of these responses are now under appeal with the Privacy Commissioner but we do not expect any resolution until long after the 2018 election, so the Sage/Vanderkruys/Burton administration will once again have succeeded in FORCING voters to vote using a flawed system that is proven not to produce representative results.