On election night we are accustom to seeing a graph with two lines, one for each candidate.  This is the raw election data as it comes in.  That is not what we are talking about here.  

What we are looking at in the CVR analysis is one line representing the total percentage of one candidate. 

The CVR is the sequence of votes as they are counted. "This allows a data analyst to “replay” the election ballot by ballot to spot unnatural patterns of voting ".  You can think of the CVR as an approximation of time. 

Trump's percentage is the number of Trump mail-in votes at any given point, divided by the number of Trump mail-in votes plus the number of Biden mail-in votes. This is then multiplied by 100.

The reason is because it is a ratio or percentage, not raw counts coming in. You would expect to see some variation at the beginning of counting, but after a large number of votes have been counted (over about 1000) that line should settle down to a flat line vertical to the x-axis in a fair election (see green line in above graph) .  

This is due to the law of large numbers or the coin-flip law.  At 20K votes if one more vote comes in it has only 1/20,000 of an effect on the final outcome.

Think of how easy it is to change your bowling average at the beginning of the season, and then think of how hard it is to change your bowling average at the end of the season.

The fact that the total percentage continues to rise despite thousands of votes could only be due to computer manipulation.

There are numerous opportunities to cheat using mail-in voting (2000 mules). In this case the CVR record of mail-in votes presented a rare opportunity to analyze what should be random data. 

Mail-in ballots are mailed out to the public at random, they are filled out by the voter at random, and mailed in at random.

The irony is that were were told democrat votes showed up last in 2020, the CVR data shows just the opposite. We had a front loading of votes for democrats.

The Mesa Pattern of fraud was discovered by Jeffrey O'Donnell when he was looking at both the CVR data and the forensic copy of the Mesa County, Colorado hard drive.  

What he discovered from the forensic copy was there was an unauthorized creation of a second database which happened after ~25K votes had been cast in the Dominion machine (Mesa Report #3). The second database did not contain the same composition of votes as the first.  The first database then disappeared never to be seen again.

"initial ballots were either secretly reprocessed inside the computer software into a second database or left behind in the initial database, which was then hidden from view from the clerks"-Jeff O'Donnell

He then matched this to the Mesa Pattern of fraud in the CVR graphs. He later observed this same pattern all across the country, not just in Mesa County, Colorado (FingerprintsofFraud.com).

 The CVR analysis shows a front-loading of votes for the democrat candidate, or a removal of valid republican votes, or both. 

The red cone represents the highest and lowest limits within the range of possible votes for a fair election, and is based on the published final outcome of an election.

Any line that falls outside the red cone is evidence of computer manipulation.  This is what Jeffrey O'Donnell has termed the Fingerprint of Fraud.

Video explanation of the Mesa Pattern of fraud.  It's recommended that you watch both videos here, and here (2 min each) 

Trump's Total Percentage vs. CVR (time). Trump won Shasta County by 65.4%.  Nevertheless, a clear pattern of fraud was observed.  Trump's percentage did not cross into the expected range until 58K votes were counted.  This indicates a front loading of votes for Biden or a removal of valid Trump votes. 

Rob Bernosky (R) Cumulative Percentage vs. CVR (time). Mail-in votes. 

 Mesa Pattern was observed indicating a front loading of votes for Shirly Webber (D) or a removal of valid Rob Bernosky (R) votes.  Cheating in this fashion keeps the republican cumulative percentage artificially deflated until the very end of the election.  

Brian Dahle (R) Cumulative Percentage vs. CVR (time). Mail-In Votes 

There was a front loading of votes for Gavin Newsom or a removal of valid Brian Dahle votes.   

This same pattern of fraud can be observed in the Nov 2022 election between Newsom and Dahle (see below)

Brian Dahle Cumulative (R) Percentage vs. CVR (time). Mail-In Votes 

There was a front loading of votes for Gavin Newsom (D) or a removal of valid Brian Dahle (R) votes. .  

Now look at the shape of the graph for the SOS race Nov 2022 (see below).  Do you see the sharp decline at the beginning followed by a little nipple up and a then a gradual incline to the finish?  Video (3 min)

They used the SAME EXACT algorithm to steal votes in both the SOS and Governor's race of Nov 2022!