DO EXTRAORDINARY CLAIMS REQUIRE EXTRAORDINARY EVIDENCE?
No, absolutely not.
First, there’s a definitional issue. What’s an “extraordinary claim” and what’s “extraordinary evidence?” Those terms are almost never defined by the proponents of the slogan, which means they’re free to adjust the meaning at anytime and push the burden of proof higher and higher.
Second, if the slogan were true, we could hardly ever have confidence in a whole bevy of extraordinarily improbable events, like Neil Armstrong walking on the moon.
Or consider a person winning the lottery, who we’ll call “Don.” Let’s say Don tells you he won the mega million lottery. Is that an extraordinary claim? Yes, certainly from a probability standpoint it’s very unlikely Don won the million dollar lottery. But does that mean you need “extraordinary evidence” to believe that it is likely that Don is telling the truth? No, of course not. Regular evidence, like Don’s winning lottery ticket, would suffice.
Third, the fatal flaw is that although the slogan is catchy, it fails to appreciate all of the factors needed to asses the probability that an event occurred. One factor forgotten by the slogan is the likelihood that if the extraordinary event had not occurred, what’s the probability that we’d have the evidence that we currently do suggesting the unlikely event’s occurrence?
So, let’s go back to the lottery for a moment. Consider a pick in the Mega Ball Million, for which the odds are 300 million to one. If the slogan were absolutely true, the evidence presented by the nightly news claiming to have the winning number would be swamped by the improbability that the reported pick was in fact the winning number.
But in assessing the likelihood that the news reported the winning number correctly, one question to ask is, what’s the likelihood that the news would’ve announced that particular number if it were not in fact the winning ticket? If that probability is sufficiently low, it can counterbalance any intrinsic improbability in the reported number itself.
So the evidence that it takes to counteract the low probability of a reported winning lottery number needn’t be enormous or unusual at all, which is why you’ve probably never questioned the reported lottery pick. Rational thinking tells you it just needs to be more probable given the truth of the hypothesis than its falsehood.
So, let’s apply this to the election. One question to ask in assessing the claims of a stolen election is what’s the likelihood that in all contested states, there would be dozens of eyewitnesses attesting to fraud and numerous statistical irregularities strongly indicative of fraud, if election fraud (extraordinary event) did not occur?
Like with the lottery, if the probability is sufficiently low that we would not have dozens of witnesses swearing to fraud and numerous statistical indicators of fraud unless fraud had in fact occurred — then normal evidence of fraud is sufficient to counteract the intrinsic improbability of a stolen presidential election.
Recap: the phrase “extraordinary claims requires extraordinary evidence” is pithy, but logically problematic. In my experience, it’s often used to deceive people during a debate by allowing the proponent to shift the burden of evidence higher and higher without ever fully considering all probabilities involved in assessing the truth of the claim based on the known evidence.
Categories: Election
Here is the germ of an idea for nailing down verifiable election integrity.
1. Put a random, unique serial number on each ballot.
A. Ensure that the serial number does not contain information that can signal who, what, when, or where with regards to the ballot. Unique identification, not categorization.
B. Ensure that information that could be used to identify the voter with the ballot is not stored.
C. Allow the voter to copy down the serial number so they can later go to a web site and enter that serial number to see what the contents of the information stored about their specific ballot actually is.
2. Store the selections of each ballot.
A. Different identifier “number” for different types of ballots (County/City/etc.).
B. Numbers indicating the selections for each scanned ballot.
C. Index the stored data by the unique ballot serial number.
D. Leave field open for ballot set check sum to be filled later.
3. Take sets of 100 ballots (100 is just a SWAG) and get a checksum of all their stored data and give the set a unique identifier.
4. Store the ballot set ID in each of the ballots field set aside for it.
All the above will enable easy identification of changes to or missing ballot information. The nature of a proper check sum (or even cryptographic digital signature) can mathematically identify even the slightest change in the data.
Also, a voter can verify the integrity of their own ballot without identifying themselves, maintaining anonymity.
Identifying number could be in two parts separated by a dash. First part includes Two-letter state abbreviation and 8-digit date; Second part is the random ID number. For example:
MS/11/03/2020-1235678
If this is printed TWICE at the BOTTOM of the ballot, then the perforated one part could be folded and torn off to give to the voter as his “receipt of voting”
I would like to be able to check my ballot to be sure my vote was properly calculated. Today, just for piece of mind, I would like to know if my vote was cast properly. If so, great. If not, I would need to know who or how to contact to correct the error.