Its the same thing we've been saying all along. Biden vote counts are a statistical improbability but he went even further to identify 10 suspect (and 4 extremely suspect counties) where the cheating occurred.
This is just my first pass and I'm by no means a data expert, so someone else please verify that I interpreted the report correctly in my summary. Some areas confused me, like the 6th and 7th graphs, which even in my highly advanced 81 IQ mind was word salad.
First graph: They point out that it is rather interesting that contested Allegheny county "found" ~32k more ballots for Biden than any Dem has ever gotten, while securely blue Philadelphia county "lost" ~31k ballots. Almost like votes from "reliably won" areas were siphoned into "oh shit we're losing" areas.
Second and third graph: There's very little differential between this election and previous, in terms of voting patterns. For a "Biden blowout" you'd expect unprecedented voter turnout across the board, but curiously, the few counties that deviate within reason from the norm show lower-than-average, antipathic / unenthusiastic turnout for Biden compared to past Democrats.
Despite this, there are giant "spikes" where unusual outlier votes for Biden just suddenly show up. In essence, you'd expect a steady upward curve for Biden to have won, but the actual data is a downward curve with randomly placed, mountainous leaps straight up to offset the overall loser trend.
Fourth graph: From Obama to Hillary we saw a huge drop in blue voters, and given how much excitement there was for Obama (muh first clean and articulate black president, etc.) compared to Hillary, and then comparing how little excitement from Hillary to Biden (at least she actually TRIED to GOTV and campaign in a few states...), you'd expect a similar plummet.
Instead, the data shows that Biden was apparently VASTLY MORE POPULAR than Obama, despite generating nothing even in the same ballpark in terms of excitement and turnout.
Several of the contested counties had turnouts higher than the maximum increases in turnout that Obama saw -- phrased another way, every county that Obama blew out the vote and set new records never seen before? Biden beat him by a mile... but only in different counties which reliably never swing that dramatically.
(This also incidentally disservices the idea that the new bloc of Biden votes were allegedly "anybody-but-Trump" votes, because if so we would have seen some evidence of building up to that in Hillary voters.)
Fifth graph: In ALL counties, Biden won a vastly larger share of the vote than there are registered Democrats. It is unprecedented (by now you're probably tired of that word) for a Democrat to turn out that many more Democrat votes than there are Democrats to vote for him. A few hundred, maybe a thousand or so vote-flippers from Republican and Independent to Democrat, sure, but not only is that normally offset by a comparable number of reverse conversions, but it is unheard of for a net profit of tens of thousands (~38k) voters to suddenly abandon their core party principles and vote for the other guy like that.
Sixth and seventh graph: Essentially, just plotting out that approximately 70% of registered D voters voted for Biden, which is unusual, but not outlandish -- except in the contested counties, which were 76%+. It may not seem like much, but a +6% increase is HUGE DEVIATION in terms of realistic turnout.
It's fascinating to note that they had to drop Philly and Allegheny to keep the chart "unbiased", otherwise the chart would get ridiculously fucked up by including those outliers. You should NEVER have to drop data to fit a graph -- when that happens, either your sampling method is fucked up (unlikely, since these are all counties in the same area), or the data itself has been massaged (aka fraud.)
In relatable terms, it'd be like trying to plot a 1000-point graph of average annual income in America, and having for your sample 998 regular guys making $20-$100k a year, and then just for giggles, some asshole decided to include Jeff Bezos and Warren Buffet as the last two data points in the "blind" sample. (So now the average annual income for guys like you and me is several hundred million a year, woohoo!)
To directly quote the issue here: "The conclusion is that in these ten suspect counties, 76% of PA registered Democratic voters supposedly voted. We have something of a contradiction. In counties that delivered increase votes for Biden a higher proportion of registered Democrat voters voted than otherwise in the state (which was 70%±)."
Eighth chart: Benford's Law. Buckle up and obey the number limit, it's the Law.
It's a mathematical certainty used in accounting and data entry to check for fraud and make sure nobody is fucking around and just entering random blurbs to populate the data. In essence, hard disciplines that deal with numbers regularly are "biased" towards certain numbers, like "1" being overwhelmingly the leading digit of any column (it's weird but makes sense, the majority of large numbers will on average start with 1 as the "overflow digit"), and this is so reliable that auditors can always tell when random number generation or manual manipulation is involved because it DOESN'T follow Benford's Law.
Which... Biden's vote tallies don't. If the IRS were looking at Biden's votes, they'd be flagging that shit for a full audit because it doesn't pass the sniff test. (The vote totals, not championship sniffer Biden himself.)
Benford test can easily fail just from the polling locations being mostly similarly sized. It's not a good test, it won't detect systemic cheating like +100 in every precinct across the board, or a counter throwing out all the Trump ballots, or a single huge fraud in one field.
The other report is much better especially the chart with Chester county. A county with more republicans that went big for Biden - no way. I bet they just swapped the Biden/Trump results so Biden got Trump's votes.
I have an IQ of 80, man. Can you break down what all those graphs mean?
Its the same thing we've been saying all along. Biden vote counts are a statistical improbability but he went even further to identify 10 suspect (and 4 extremely suspect counties) where the cheating occurred.
It’s gonna take more than that to flip some of these things in court. I hope (and suspect) they’ve got more actual evidence
Those voters in those counties need to be polled by phone, and any trump voters need to asked if they'll go on record.
This is just my first pass and I'm by no means a data expert, so someone else please verify that I interpreted the report correctly in my summary. Some areas confused me, like the 6th and 7th graphs, which even in my highly advanced 81 IQ mind was word salad.
First graph: They point out that it is rather interesting that contested Allegheny county "found" ~32k more ballots for Biden than any Dem has ever gotten, while securely blue Philadelphia county "lost" ~31k ballots. Almost like votes from "reliably won" areas were siphoned into "oh shit we're losing" areas.
Second and third graph: There's very little differential between this election and previous, in terms of voting patterns. For a "Biden blowout" you'd expect unprecedented voter turnout across the board, but curiously, the few counties that deviate within reason from the norm show lower-than-average, antipathic / unenthusiastic turnout for Biden compared to past Democrats.
Despite this, there are giant "spikes" where unusual outlier votes for Biden just suddenly show up. In essence, you'd expect a steady upward curve for Biden to have won, but the actual data is a downward curve with randomly placed, mountainous leaps straight up to offset the overall loser trend.
Fourth graph: From Obama to Hillary we saw a huge drop in blue voters, and given how much excitement there was for Obama (muh first clean and articulate black president, etc.) compared to Hillary, and then comparing how little excitement from Hillary to Biden (at least she actually TRIED to GOTV and campaign in a few states...), you'd expect a similar plummet.
Instead, the data shows that Biden was apparently VASTLY MORE POPULAR than Obama, despite generating nothing even in the same ballpark in terms of excitement and turnout.
Several of the contested counties had turnouts higher than the maximum increases in turnout that Obama saw -- phrased another way, every county that Obama blew out the vote and set new records never seen before? Biden beat him by a mile... but only in different counties which reliably never swing that dramatically.
(This also incidentally disservices the idea that the new bloc of Biden votes were allegedly "anybody-but-Trump" votes, because if so we would have seen some evidence of building up to that in Hillary voters.)
Fifth graph: In ALL counties, Biden won a vastly larger share of the vote than there are registered Democrats. It is unprecedented (by now you're probably tired of that word) for a Democrat to turn out that many more Democrat votes than there are Democrats to vote for him. A few hundred, maybe a thousand or so vote-flippers from Republican and Independent to Democrat, sure, but not only is that normally offset by a comparable number of reverse conversions, but it is unheard of for a net profit of tens of thousands (~38k) voters to suddenly abandon their core party principles and vote for the other guy like that.
Sixth and seventh graph: Essentially, just plotting out that approximately 70% of registered D voters voted for Biden, which is unusual, but not outlandish -- except in the contested counties, which were 76%+. It may not seem like much, but a +6% increase is HUGE DEVIATION in terms of realistic turnout.
It's fascinating to note that they had to drop Philly and Allegheny to keep the chart "unbiased", otherwise the chart would get ridiculously fucked up by including those outliers. You should NEVER have to drop data to fit a graph -- when that happens, either your sampling method is fucked up (unlikely, since these are all counties in the same area), or the data itself has been massaged (aka fraud.)
In relatable terms, it'd be like trying to plot a 1000-point graph of average annual income in America, and having for your sample 998 regular guys making $20-$100k a year, and then just for giggles, some asshole decided to include Jeff Bezos and Warren Buffet as the last two data points in the "blind" sample. (So now the average annual income for guys like you and me is several hundred million a year, woohoo!)
To directly quote the issue here: "The conclusion is that in these ten suspect counties, 76% of PA registered Democratic voters supposedly voted. We have something of a contradiction. In counties that delivered increase votes for Biden a higher proportion of registered Democrat voters voted than otherwise in the state (which was 70%±)."
Eighth chart: Benford's Law. Buckle up and obey the number limit, it's the Law.
It's a mathematical certainty used in accounting and data entry to check for fraud and make sure nobody is fucking around and just entering random blurbs to populate the data. In essence, hard disciplines that deal with numbers regularly are "biased" towards certain numbers, like "1" being overwhelmingly the leading digit of any column (it's weird but makes sense, the majority of large numbers will on average start with 1 as the "overflow digit"), and this is so reliable that auditors can always tell when random number generation or manual manipulation is involved because it DOESN'T follow Benford's Law.
Which... Biden's vote tallies don't. If the IRS were looking at Biden's votes, they'd be flagging that shit for a full audit because it doesn't pass the sniff test. (The vote totals, not championship sniffer Biden himself.)
Benford test can easily fail just from the polling locations being mostly similarly sized. It's not a good test, it won't detect systemic cheating like +100 in every precinct across the board, or a counter throwing out all the Trump ballots, or a single huge fraud in one field.
The other report is much better especially the chart with Chester county. A county with more republicans that went big for Biden - no way. I bet they just swapped the Biden/Trump results so Biden got Trump's votes.
Do you need it in Somalian?