About a week ago, I was informed that five counties had changed all of their ID numbers. I asked for sample data from one of the counties, Allegany, so I could see for myself. It was true. Every county-assigned voter ID (CID) number, had changed. Not just a little, but a lot. For instance, CID 1,341 was changed to CID 92,298.

Changing the CID numbers could have totally disrupted algorithmically-imposed patterns built into the mapping of CID to state ID (SBOEID) numbers.

They didn’t. The Spiral remains in the new numbers. This implies that the original ranking was maintained. However, it wasn’t. This suggests the algorithm was important enough to preserve, and that another algorithm was devised to ensure that happened while the ranking was modified.

A close inspection of the original numbers shows there are gaps between sequential numbers. Referring to the illustration above, you will see the first such gap is between CID 13,214-13,220. Numbers 13,215-13,219 don’t exist. This is reflected in the ranking of the numbers 13,214 and 13,220, which is 8,375 and 8,376, respectively. The ranking shows the position of each number in the full list. In this case, 13,220 is only one higher than 13,214 because it is the next highest number.

With the new CID numbers, the record for CID 13,214 has become CID 102,671, and the record for 13,220 is 102,673. The original 5 missing numbers has been reduced to 1, because the other 4 numbers have now been assigned. By looking at the “Rank Diff” column, we see that the difference in the rank of these numbers is 3,170. Meaning, 3,170 new numbers have been assigned between previously existing CID numbers.

This by itself doesn’t have to be nefarious. New records have certainly been added over the past two years, perhaps 3,170 of them. If sorted differently than the original records when new CID numbers were assigned to every record in the county, this is exactly the result we would expect.

If sorted by SBOEID, as the numbers in the illustration at the top of this article are sorted, the Spiral algorithm remains embedded in the data, despite the change to the CID numbers. This would be expected if each CID number had the same value added to it. For instance, if the new numbers were the old number +10,000. That wouldn’t change the ranking at all, though the “in-between” numbers would no longer be in-between, but at the end. Unless, that is, they wanted to use those in-between numbers. Why that would be the case when all the numbers have been changed is something I am curious about.

The “CID Diff” column shows how much was added to the original CID to arrive at the new CID. Some are the same, but not all. The minimum difference between old and new CID numbers is -87,875. The maximum difference is 187,761. That is a range of 275,636. This means that the new CID numbers were not adjusted in the same way. And yet, the algorithm remains, as solid as it was before. Why?

The dataset I received didn’t have SBOEID numbers for new records. This prevented some analyses from being done. However, by sorting on CID numbers (Figure 2) it is easy to see where new CID numbers were inserted. Every blank cell in the State Voter ID column is new. And yet, the registration dates for many of these are old enough to predate the new numbering system and the introduction of the Spiral in 2007.

The SBOEID Gap field reveals the characteristic repunit pattern associated with the Spiral (gaps of 1111 or 1112, etc), despite the presence of blank cells between them. This suggests that the new numbers were assigned to a list where the Spiral partition numbers were treated differently from the rest.

To do this implies that the new algorithm was aware of the old algorithm and was designed to retain its structure. Maintaining the original algorithm while imposing another one is not a trivial task, and wouldn’t be done unless there is some value embedded in the original Spiral mapping of the numbers.

That said, I leave open the possibility that I’ve missed some detail that could cause this naturally. It isn’t looking that way so far, but the possibility shouldn’t be ignored.

I think we can say that when they know they’ve been found out, they’ll do what’s necessary to maintain the structure and its use while simultaneously further obfuscating its discovery and analysis by auditors.

That they changed CID numbers to torment you & team of sleuths, is also a possible "reason" at least in part. The changed CIDs is cause for new doubts and suspicions.