The MSM shows data in a way to fit their point of view. This can be by not showing everything, leaving out important information, changing the parameters, etc.
Showing a picture of a single vote that's "split" between candidates is doing the exact same thing. Another is showing only the times Trump goes down 0.1% or Biden up 0.1%. You could easily create the narrative that the machines were giving/taking a set percentage (magically 0.1% of total) of votes.
Since I seem to have to keep repeating myself, I'm not talking about large shifts outside the margin of error from rounding. I'm talking about the changes likely caused by the rounding. The large shifts should be getting attention, but people are too focused on the others.
I think you misunderstood. The instances I'm talking about are the +/- 0.1% changes that people are claiming are evidence of fraud or votes being stolen. When people do that, they're overshadowing actual instances where votes appear to be swapped.
I wrote a program to grab all instances of vote losses exceeding 0.1%. The swaps aren't as commo, but the fact they are there at all is a major issue. PA was so bad that I don't think they even know how to count. NJ has a near 80k exact vote swap. You've also got some smaller ones in other states. Then there's the vote dumps you can find in the data.
This thread about 1 vote being split between multiple candidates is a great example of not understanding the original data. It's about as bad as what the MSM does when presenting their data.
Total vote count is "exact" (some states have some odd counts). You are then given a percentage for each candidate. We don't know the exact vote count per candidate, only a percentage that is rounded which is used to estimate. When working with 1m votes, that rounding can cause the candidate estimates to be off by 1k.
A lot of these that are posted are likely due to people not knowing they are estimates. That's from the original source not providing all the information on how the numbers are gotten. Sound familiar?
Yeah, definitely. I'd just hope people stop using the rounding errors as evidence of fraud. There's other inconsistencies that can't be easily explained without the generic "it was human error during 'x'." I'm still curious about that exact vote switch in New Jersey that was nearly 80k.
A little too convoluted. This is simply the data giving a total votes and a percentage for each candidate. The votes for each candidate are estimated and not the actual values. Too many people get focused on that and forgot about the other inconsistencies.
No idea what they did on previous elections, but this may be their normal way of doing it. Completely stupid to have 0.1% rounding errors when dealing with millions, but what can you expect from them?
Limitations with the live data. All percentages are rounded to 0.1%.
Let's say there's 10000 votes. 50.6% for Biden and 49.4% for Trump. Add 1 vote for Trump to make it 5060-4941. The percentages are still 50.6% (50.595%) and 49.4% (49.405%) when rounded.
10001*50.6%=5060.506
10001*49.4%=4940.494
We now have a fractional vote that appears in the live data with Biden gaining more of the vote than Trump. If this was enough to cause the rounding to increase, we'd see an increase of 0.1% of the total vote to Trump and 0.1% decrease for Biden. That's why it looks like there's increases in a specific multiple (0.1% of total) when you take all the rounding errors and put them in a row.
When looking at the live data, you have to remove any possible rounding errors. That can remove some fraud, but it can't be proven. I believe someone on here did that months ago and still found many unexplainable instances.>
It likely died as it's easily explained. The vote counts are purely estimates, and it's not even that accurate. We have changes of 0.1%. If you take the total votes and multiply it by 0.1%, that's the "multiple value" they're claiming.
Checked a few of them. The 0.1% changes were when Biden gained 0.1% and 3rd party lost 0.1% or Trump lost 0.1% with 3rd party gaining 0.1%. Completely expected, if you know that the counts are estimates.
You should separate out any lost/switched votes when there's a 0.1% change. That can simply be a 0.0499% vs 0.05% difference when rounding. Could even go as far as separating out anything that might be possible. Without more accurate data, it's impossible to prove they aren't rounding errors.
Another issue could be that votes were switched on top of new votes being counted. For the rounding issue, ignoring any 0.1% changes can clean up the data.
Michigan: 3 instances Last one is most likely a rounding error, if that group was almost completely for Biden.
Pennsylvania: Complete mess Some explanations for the massive vote drops would be needed.
Georgia: Only 1
New York: Pretty interesting Has a massive amount for Biden then removes most of those a little later. Then has another drop right after.
New Jersey: Explanation required
Some of these definitely need to be looked into, but I believe most of the "switched votes" in the OP occurred due to rounding errors.
Yeah, I know. The issue is that many sources don't disclose that, nor even know. Without taking it into account, you can come up with weird conclusions and completely ignore the actual suspicious instances.
The rounding may not have been done with bad intentions. It's easy to forget about the consequences of your actions when programming. What we'd need to find is how the previous elections did it. If those used exact counts or more digits, then I'd have to agree with you on it being intentional.
I'm guessing here, but they should still have the exact counts on their system. Looks like they should have it by county too. With that, we'd be able to find any irregularity on a per county basis.