If this is your first visit, be sure to
check out the FAQ by clicking the
link above. You may have to register
before you can post: click the register link above to proceed. To start viewing messages,
select the forum that you want to visit from the selection below.
OP, I have not seen any evidence that COVID is any worse than the flue...as a matter of fact, I haven't even heard anything about the flue this year. They must be counting flue deaths as COVID. Don't be scared of that dark thing following you around...it's just your shadow.
I am not alone, even 12% of Democrats believe the is election is a fraud. Do you know why 12% of the Democrats believe that Democrats cheated?...because they did it themselves.
80% of Republicans think the election was stolen and the media is the enemy
"Donald Trump’s message of election fraud has reached millions of people even though Twitter, Facebook, and the media have said it was baseless every ten minutes since election day. Word is spreading so fast that even 12% of Democrats think the election was stolen.
…Politico conducted its 2020 Voter Priorities Survey and results show that a vast majority of Trump voters believe the election results are not valid and that illegal voting and fraud took place during the election. Twenty-one percent of Trump voters believe the results are valid and 79 percent believe the election was stolen.
This is a crisis for any democratic nation. The way to solve it is with a full and fair investigation which either cleans up the count or puts people in jail. Possibly nothing short of a new election (with paper ballots) will work."
Anomalies in Vote Counts and Their Effects on Election 2020
A Quantitative Analysis of Decisive Vote Updates in Michigan, Wisconsin, and Georgia on and after Election Night
Executive Summary
In the early hours of November 4th, 2020, Democratic candidate Joe Biden received several major “vote spikes” that substantially — and decisively — improved his electoral position in Michigan, Wisconsin, and Georgia. Much skepticism and uncertainty surrounds these “vote spikes.” Critics point to suspicious vote counting practices, extreme differences between the two major candidates’ vote counts, and the timing of the vote updates, among other factors, to cast doubt on the legitimacy of some of these spikes. While data analysis cannot on its own demonstrate fraud or systemic issues, it can point us to statistically anomalous cases that invite further scrutiny. This is one such case: Our analysis finds that a few key vote updates in competitive states were unusually large in size and had an unusually high Biden-to-Trump ratio. We demonstrate the results differ enough from expected results to be cause for concern.
With this report, we rely only on publicly available data from the New York Times to identify and analyze statistical anomalies in key states. Looking at 8,954 individual vote updates (differences in vote totals for each candidate between successive changes to the running vote totals, colloquially also referred to as “dumps” or “batches”), we discover a remarkably consistent mathematical property: there is a clear inverse relationship between difference in candidates’ vote counts and and the ratio of the vote counts. (In other words, it's not surprising to see vote updates with large margins, and it's not surprising to see vote updates with very large ratios of support between the candidates, but it is surprising to see vote updates which are both).
The significance of this property will be further explained in later sections of this report. Nearly every vote update, across states of all sizes and political leanings follow this statistical pattern. A very small number, however, are especially aberrant. Of the seven vote updates which follow the pattern the least, four individual vote updates — two in Michigan, one in Wisconsin, and one in Georgia — were particularly anomalous and influential with respect to this property and all occurred within the same five hour window.
In particular, we are able to quantify the extent of compliance with this property and discover that, of the 8,954 vote updates used in the analysis, these four decisive updates were the 1st, 2nd, 4th, and 7th most anomalous updates in the entire data set. Not only does each of these vote updates not follow the generally observed pattern, but the anomalous behavior of these updates is particularly extreme. That is, these vote updates are outliers of the outliers.
The four vote updates in question are:
An update in Michigan listed as of 6:31AM Eastern Time on November 4th, 2020, which shows 141,258 votes for Joe Biden and 5,968 votes for Donald Trump
An update in Wisconsin listed as 3:42AM Central Time on November 4th, 2020, which shows 143,379 votes for Joe Biden and 25,163 votes for Donald Trump
A vote update in Georgia listed at 1:34AM Eastern Time on November 4th, 2020, which shows 136,155 votes for Joe Biden and 29,115 votes for Donald Trump
An update in Michigan listed as of 3:50AM Eastern Time on November 4th, 2020, which shows 54,497 votes for Joe Biden and 4,718 votes for Donald Trump
This report predicts what these vote updates would have looked like, had they followed the same pattern as the vast majority of the 8,950 others. We find that the extents of the respective anomalies here are more than the margin of victory in all three states — Michigan, Wisconsin, and Georgia — which collectively represent forty-two electoral votes.
Extensive mathematical detail is provided and the data and the code (for the data-curation, data transformation, plotting, and modeling) are all attached in the appendix to this document[1]. Background
Late on Election Night 2020, President Donald J. Trump had a lead of around 100,000 votes in Wisconsin, a lead of around 300,000 votes in Michigan, and a lead of around 700,000 votes in Pennsylvania. Back-of-the-envelope calculations showed that in order to overtake President Trump, Joe Biden would have to substantially improve his performance in the remaining precincts — many of which were in heavily blue areas like Detroit, Milwaukee, and Philadelphia.
On Election Night, conflicting news reports came in that various precincts were stopping their count for the evening, sending election officials home, or re-starting their counts. There remains a large amount of confusion to this day about the extent to which various precincts stopped counting, as well as the extent to which any state election laws or rules were broken by sending election officials home prematurely. Whatever the case is, various precincts in Wisconsin, Michigan, and Pennsylvania continued to report numbers throughout the night.
By the early hours of the following morning, Wisconsin had flipped blue, as did Michigan soon after. A few days later, Georgia and Pennsylvania followed suit. Given the uncertain context, many American observers and commentators were immediately uncomfortable or skeptical of these trends.
For context, using publicly available data from the New York Times, here is a visualization of the number of votes by candidate in Michigan from the beginning of election night to 7pm Eastern Standard Time (EST) on November 4th, 2020: Fig. 1. X-axis is the Month-Year Hour of the time, Y-axis is the number of votes as of that time, expressed in millions of votes. The red series is the running number of votes for Donald Trump, and the blue series the running number of votes for Joe Biden.
As this graph shows, Joe Biden overtook President Trump’s lead through a small number of vote updates which broke overwhelmingly for Biden in Michigan in the early hours of the morning of November 4th.
The situation in Wisconsin is even more stark: a single update to the vote count brought Biden from trailing by over 100,000 votes into the lead. Here is the comparable graph, over the same time range, for Wisconsin, with the x-axis (time) expressed in Central Standard Time (CST): Fig. 2. X-axis is the Month-Year Hour of the time, Y-axis is the number of votes as of that time, expressed in millions of votes. The red series is the running number of votes for Donald Trump, and the blue series the running number of votes for Joe Biden.
Various versions of these graphs spurred online discourse. While some commentators provided relatively partisan analysis, others merely expressed surprise at the near-vertical leaps in some of these vote updates. Is it likely this phenomenon would arise organically? In an attempt to address this question, this report assesses how extreme and unusual these spikes are with respect to both other vote updates in the states of Michigan, Wisconsin, and Georgia, as well as those around the nation.
Through several investigative mechanisms, we find these four vote updates to be extraordinarily anomalous. While these alone do not prove the existence of fraud or systemic issue, it invites further scrutiny. The Concept, the Intuition, and the Measurement
Data analysis relies on recognizing and evaluating patterns in data. When we find anomalous data, that is often an indication of underlying differences. This is why in this report we focus on these four vote updates.
There are also a number of general intuitions upon which we draw to direct our research. In general, the larger the sample size, the smaller we expect the deviation from the population average to be. While anomalous vote ratios may occur, the statistical chance of anomalous margins goes down as the size of the sample (or vote update) goes up.
The basic intuition is: big margins are one thing, and so are super-skewed results, but it’s weird to have them both at the same time, as they generally become inversely related as either value increases.
We will demonstrate below that the data overwhelmingly follow this intuition, but that four key vote updates identified by this report cut against this intuition.
In particular, we will show the existence of a very strong inverse relationship within vote updates, across all states and times, between the difference of votes for Joe Biden and Donald Trump (often referred to as the “Biden-Trump margin”) and the the ratio of Joe Biden’s votes to Donald Trump’s votes (often referred to as the “Biden:Trump ratio”). As described in more detail in the next section, we take the natural logarithm of the ratios so that they are symmetric, i.e. so that we are not treating the two candidates differently when graphing and analyzing. These values are often referred to as “Biden:Trump log-ratio.” Since the logarithm is an order-preserving transformation — i.e. if x is bigger than y, then log(x) will be bigger than log(y), and vice versa — we sometimes use them interchangeably when precision is not required.
At any geographical level, we can test the assumption of an inverse relationship between vote update size and the extremity of the ratio between the candidates’ votes, and, as we will see here, the relationship is extremely strong. Across states red and blue, where turnout is high and low, there is an obvious inverse relationship between the two. Measuring This Relationship Between The Candidate’s Margin and their Ratio
Let us now attempt to quantify the nature of the inverse relationship in the context of a particular state. First we take our data set of running vote totals[2] for each state, and, for each state, calculate the vote differential for each candidate between updates. This produces a sequence of vote differences, the sum of which, within any given state, is the total.
To begin, we consider each sequential update in the state of Michigan where the vote totals for both Trump and Biden are greater than zero[3]. For each of these, we compute two values:
The difference between the number of votes for Biden and the number of votes for Trump — the “margin”
The logarithm[4] of the ratio between the number of votes for Biden and the number of votes for Trump — the “log-ratio”
Note: both of these metrics are symmetrical. If we let f1 be the first metric and f2 the second, the reader will note that, for any positive numbers (X, Y):
And that:
In other words, given X for Biden and Y for Trump, either metric will produce a score which is the opposite of what it would produce if the update instead had Y votes for Biden and X for Trump. This property is extremely useful, and will come in handy during the statistical analysis.
Readers might ask: Why are you measuring the ratio? Why not measure the difference between the vote proportions (or, equivalently, their percentages). The answer to this lies in what we are looking for, i.e. evidence of fraud or foul play which manifests in extremely unusual outcomes. In particular, ratios are almost never used in expressing vote counts (one typically hears of percentages or, when a race is close, numbers) and so anyone committing fraud and looking to “cover their tracks” is more likely to be “gaming” the metrics they’re used to, and much more likely to leave tells in metrics they’re not considering.
This obscures critical differences between the two statistics.
Ratios demonstrate an important property: the farther ahead a candidate is, the harder it is to move the next 1 percent ahead. They reflect the relative difficulty of each marginal vote as the pool of remaining votes decreases.As a candidate approaches 0% or 100% of the vote, the rates at which the ratio of that candidate’s votes to the other candidate’s votes converge to zero or infinity are very different.
Ratios allow us to spot a potential sign of fraud: unusually low ratios between the losing (major) candidate and other, less well-known candidates. Because those who watch and participate in elections tend not to think in these terms, if there is fraud, they’re much less likely to have covered their tracks in this respect. A tin-pot-dictator style election where the favored candidate gets 99% of the vote is obviously suspect, but less attention is often paid to details like whether the ratio between the most popular losing candidate and long-shot third-party candidates actually makes sense[5]. Looking at metrics which are less popular in practical use will be tremendously helpful here, as we will see.
To illustrate this, let us consider a sequence of two hypothetical elections between Tom and Harry. Imagine that the first time around, Tom wins with 55% of the vote to Harry’s 45%. Four year later, Harry is the challenger and Tom improves his margin to 60% of the vote. There are many ways that this can happen; winning over new voters, Harry’s previous supporters no longer voting, Harry’s supporters switching to Tom, or some combination of any of the above. Let’s consider merely the last case for the moment. For Tom to get from 55% to 60%, he must convert one out of every nine, or just over 11%, of Harry’s supporters. This may not be easy, but is hardly outside the realm of possibility.
Now consider another hypothetical election in a heavily partisan electorate, between Alice and Bob. In the first election, Alice gets 90% and Bob gets 10%. In order for Alice to achieve the same absolute percentage increase as Tom, i.e. 5%, she must convert 5% among a population of 10%. In other words, she must convert one out of every two supporters of Bob. For reasons outside the scope of this paper, this may not be 4.5 times as difficult as a candidate getting from 55% to 60% of the total vote, but it is without question much harder. A useful example of this is this is San Francisco, CA, which, despite being one of the bluest cities in America numerically and culturally, is one where Democratic Presidential candidates consistently get about 90% of the vote but never seem to crack 95%. There are Republicans in San Francisco, however few of them, and converting half of them is a tall order. This makes ratios a useful tool in our arsenal for answering questions of the form “how much is too much”?. This allows us to assess the data in a way which we believe is qualitatively different — and qualitatively superior — to the common forms of assessment used by average individuals and the news media.
This election represents an extraordinary and unique opportunity for election integrity analysts and the application of statistical fraud detection research, as it is likely the first national election in American history, at the very least, where the general public has had access to time-series election data. Even well-respected academic papers which study election fraud in other countries[6] seem to mostly study after-the-fact information about final tallies; analysis is done on statistics about voter turnout, digit frequencies, and other information which is available in after-the-fact official numbers. After all, if reports of widespread fraud and corruption ordered from the top in elections in, e.g., Russia, Uganda, Ukraine, Iran, etc., are to be believed, then those governments, which tend to have much more control over what can and cannot be published than our government, are unlikely to want to increase the number of dimensions along which their claim to legitimacy can be audited. A Look at Michigan
Let us now calculate these two values for each vote update in Michigan where both Biden and Trump have positive values. If it follows the intuition that there as an inverse relationship between the margins of an update and its ratio, we should expect to see a large cluster of data with a few points above, below, to the left and right, and virtually no points in either the top right (which would represent a simultaneously extreme Biden-Trump margin and Biden:Trump ratio) or the bottom left (which is analogous but favorable to Trump).
Here is that distribution, presented as a scatter plot, with the numerical margins as the X-axis and the log-ratios as the Y-axis. Fig. 3. The X-axis is the difference between the number of Biden votes and the number of Trump votes in each vote update, and the Y-axis is the natural logarithm of the ratio of the two.
As we can see, most observations follow the basic contour of our hypothesis, i.e. the more extreme an update is in one respect, the less extreme it is in another.
For example, the update at (-3,622, -6.449), has a fairly extreme ratio of Biden:Trump votes — about 1:632 — but is not very large, producing only a margin of -3,622 votes for Biden, which, as we can see, is not terribly extreme in the context of this distribution. Similarly, the point all the way to the left, (-66,456, -0.816), is one where Biden’s margin is a significant -66,546, but where the ratio, of about 1:2.26, is not particularly unusual for a vote update which favors Trump.
We can see this pattern as well in almost every Biden-favoring update as well. For example, the update with the 3rd greatest margin for Biden, at (34,450, 0.296), is 134,326 Biden votes to 99,867 Trump votes, and only has a Biden:Trump ratio of 1.34:1. And the update with the 3rd greatest Biden:Trump ratio, at (6,091, 2.184), in which Biden received 6,863 votes and Trump received 773 votes, has a fairly extreme ratio of 8.884 but only nets Biden 6,091 votes, a relatively small amount compared to what we will examine next.
Two points stand out.
Let us first consider the less extreme of these, i.e. the point at (49,779, 2.447). This point, representing a vote update which went 54,497 for Biden and 4,718 for Trump and arrived at 3:50am ET on November 4th 2020, is both the second-largest vote margin of Biden’s, at 49,779, and also has the second largest Biden:Trump ratio at 11.55:1. As we can see and as was described above, the update with the next largest margin was an update with merely 7,776 votes, while this update had over 7 times as many votes and broke more heavily for Biden.
The oddness of the update described above pales in comparison to that of the update in the top right corner, however. That update, at (135,290, 3.164), represents the vote update described at the top of this report, and is responsible for the extremely noticeable spike which nearly eliminated Trump’s lead in one shot. It arrived at 6:31am ET on November 4th, and went 141,258 for Biden to 5,968 for Trump — representing both the largest vote margin for Biden of any of the 502 updates we have here, at 135,290, while also representing, by a factor of more than 2, the largest Biden:Trump ratio, at a whopping 23.67:1 (the log of which is 3.16). As we will see when comparing with other states, by our metric this is the single most anomalous point in the nation.
This update is also particularly interesting for another reason: there are 2,546 non-two-party votes, while Donald Trump only has 5,968. Here is a histogram of vote-total-weighted Other:Trump ratios[7]: Fig. 4. The x-axis is, for each vote update, the ratio of other (non-2-party) votes to votes for Trump, multiplied by the number of total votes in that update. The y-axis is the number of vote updates in that “bin,” where each bin has a range of 500.
As we see, when we weight by the number of votes in any given update, this update is particularly anomalous. The next closest vote-weighted Other:Trump ratio is less than two-thirds of this one, and the median — 137.56 — is smaller by a factor of about 464.5. For such a large batch of votes to be counted while also showing such an exceptionally poor performance of Trump relative to the non-two-party vote is clearly very surprising.
In particular, it calls into serious question the veracity of this vote update, and is perhaps some of the strongest direct evidence of fraud in this entire report. Someone looking to fraudulently improve Joe Biden’s margins relative to Donald Trump is likely to be focused on covering their tracks by keeping Joe Biden’s share of the update at a reasonable value. 95% might seem plausible, but 99.9% at this scale becomes prima facie implausible to any honest observer. One effective way of achieving the desired goal of decreasing Donald Trump’s lead at this point would have been to suppress the Trump vote while artificially inflating the non-two-party vote in an attempt to disguise just how Biden-favoring this update actually was. Indeed, this is precisely the reason this report uses ratios -- because they are a metric virtually never used for any practical purpose in discussing election results, someone committing fraud is far less likely to consider how unusual a ratio might look. In particular, because the non-two-party candidates received far less media attention than in the 2016 Presidential election, and the Green Party candidate was even successfully sued off of the ballot in one or more states, it is hard to believe that this vote update only favored Trump over the non-two-party vote by less than a factor of 2.5, when the statewide ratio was over 31[8]. Absent a compelling explanation of why this particular update -- at such a crucial time, in a crucial state, which improved Biden’s standing in the state so dramatically -- also had non-two-party votes performing so unusually relative to Trump votes, it seems unlikely that this vote update reflects an honest accounting of the legitimate votes.
Subsequent sections of this report quantify how extreme it is in other respects and consider the implications if it had been slightly less extreme. A Look at Wisconsin
Here is the analogous graph for Wisconsin. Fig. 5. The X-axis is the difference between the number of Biden votes and the number of Trump votes in each vote update, and the Y-axis is the natural logarithm of the ratio of the two.
The patterns in this graph are somewhat more bizarre. The updates favoring Trump (i.e. those to the left of zero on the x-axis) exhibit an inverse relationship between the margin of victory in a Trump-favoring update and the ratio between Trump and Biden votes. For example, the update at, (-5,433, -4.564), which is the most extreme in the state in terms of ratio, is from an unusually Trump-favoring batch of ballots which went 5,490 for Trump to 57 for Biden, i.e. a Trump:Biden ratio of about 96:1 for Trump. This number itself is quite large, but, critically, it is not anomalous with respect to the shape of the distribution. The tell-tale sign of oddity here is not extremity with respect to either value, but co-extremity.
Biden’s distribution looks slightly odd here, but there is one point which especially stands out, i.e. the one in the top right, at (118,215, 1.74). This was the vote update which arrived at 3:42am CST on November 4th, and went 143,379 for Biden to 25,163 for Trump[9], giving a margin of 118,215 and a Biden:Trump ratio of about 5.7:1 — about 3 times larger than the update with the next largest margin (which was 39,499). At the same time, only one update — one with a mere 6,435 votes (i.e. about a factor of 18 fewer than the update in question) which went 3,037 for Biden to 495 for Trump — has a larger ratio, at around 6.14:1. A Look at Georgia:
Fig. 6. The X-axis is the difference between the number of Biden votes and the number of Trump votes in each vote update, and the Y-axis is the natural logarithm of the ratio of the two.
This one seems only slightly more anomalous than other such graphs, but, as we will see, actually contains two of the nine most anomalous vote updates in our combined distribution of 8.954 vote updates. In particular, the point at (136,155, 1.543), representing a vote update which arrived at 1:34am EST on November 4th, is the update with the largest margin of all of the updates in Georgia — it also has the 10th largest Biden:Trump ratio. There are a few smaller updates with more extreme ratios, but, as we will detail later in this report, this point is in fact unusual. A Short Survey of Other States
We now turn to other states, particularly those with similar characteristics (e.g. a swing or blue state where one or two urban cores offsets an otherwise very Republican population). These help us establish an initial baseline of what these distributions should look like within any state before we begin comparing updates directly across states. Pennsylvania:
Fig. 7. The X-axis is the difference between the number of Biden votes and the number of Trump votes in each vote update, and the Y-axis is the natural logarithm of the ratio of the two.
The inverse relationship is immediately visible here. We have points near the bottom (representing high Trump:Biden vote ratios), a few points far to the left (representing high Trump - Biden values), and a couple (much farther) off to the right, representing a high Biden-Trump margin, but which are not particularly extreme in terms of their Biden:Trump ratio. Minnesota:
Fig. 8. The X-axis is the difference between the number of Biden votes and the number of Trump votes in each vote update, and the Y-axis is the natural logarithm of the ratio of the two.
While there is an update which is more extreme in terms of how large the Trump:Biden ratio is, and several updates with extremely large Trump-Biden margins, we see the basic shape remains the same. New York:
Fig. 9. The X-axis is the difference between the number of Biden votes and the number of Trump votes in each vote update, and the Y-axis is the natural logarithm of the ratio of the two.
The vote margins for each update are clustered fairly heavily around zero, while the few updates which have exceptionally large margins for either candidate have ratios which are not nearly as extreme as those of many other updates. Consolidating, Comparing, and Measuring
Having taken a brief tour of states with similar characteristics, i.e. where Joe Biden is currently in the lead and the Democratic vote comes overwhelmingly from a single urban area (or perhaps two, in the case of Pennsylvania), we can see that the Michigan and Wisconsin graphs both look unusual. In order to more rigorously assess the extent to which this is actually anomalous, it is necessary to accommodate the reality that the typical Biden-Trump margin and Biden:Trump ratio will vary substantially between states. If we merely take these values as they are, then most of the differences between, e.g., Alabama and California would likely just be artifacts of the massive discrepancies between how the candidates each performed in these states.
To achieve this, we can use a data transformation process called standardization. This is a process by which, for a series of numerical data, the mean of the data is subtracted from each point, and then the result is divided by the standard deviation. This will produce a series of distributions which permit an apples-to-apples comparison of these values (i.e. per-vote-update Biden-Trump margin and Biden:Trump log-ratio) between states which are both very different in size and lean very differently, politically. Data standardization is a very common technique in machine learning for training models on data sets with very different numerical magnitudes and means[10], as it provides precisely the functionality we need here.
We can thus standardize each individual (margin, log-ratio) point within its state[11], and plot it as we did before. Here is what that graph looks like. The values for Michigan are in red, those for Wisconsin are green, and the values for all other states are blue: Fig. 10. The X-axis is the difference between the number of Biden votes and the number of Trump votes in each update, standardized by the distribution of such values of its state. The Y-axis is the log-ratio of Biden votes to Trump votes in each update, again standardized by the distribution of such values in its state.
Out of these 8,954 vote updates across the country, we can see how overwhelming the pattern is. In particular, we see that — with a few notable exceptions — as one value grows more extreme in any direction, the other tends to become less extreme. This brings us to the visually identifiable exceptions.
Directing our attention to the points on the far right end of the distribution, i.e. those which have the most extreme Biden-Trump margin with respect to their state, we immediately see one point from Michigan, which is quite far above where the shape of the plot would otherwise predict it being. This, the point at (15.494, 4.989), is the vote update which arrived at 6:31am EST on November 4th, went 141,257 to Biden and 5,968 to Trump. Recall: this update had both the largest margin (135,290) of any of the 574 updates[12] in Michigan, by about 85,000 votes and a factor of about 2.7 over that of the update with the next-largest update, (5.679, 3.912) — which, critically (and surprisingly, vis a vis what this distribution shows), was both the second largest in terms of Biden-Trump margin and Biden:Trump ratio[13]. It also had the largest Biden:Trump ratio (roughly 23.69:1), by more than a factor of 2 over that of the update with the next-largest Biden:Trump ratio. The visual discrepancy between that update and the overwhelming pattern followed by the other updates is glaring, and we will shortly quantify just how extreme it is.
Next, consider the green dot just slightly down and to the left of the red outlier. This is the vote update in Wisconsin which arrived at 3:42am CST on November 4th, which went 143,379 for Biden and 25,163 for Trump, for a margin of 118,215[14]. It was the update with the largest Biden - Trump margin in Wisconsin by a large distance[15] and, in Biden:Trump ratio, was second largest -- second only to an update which was 26 times smaller and yet only slightly more extreme in its ratio[16].
We also see a red dot at (5.679, 3.912), which corresponds to the vote update which arrived at 3:50am EST on November 4th and went 54,497 for Biden to 4,718 for Trump, for a margin of 49,779 and a ratio of 11.55:1. It is worth noting that, while not nearly as anomalous as the 6:31am EST update, this one was very extreme along both dimensions in its own right. As we will see, however, it ends up being the seventh most extreme value in terms of its non-adherence with the distribution as a whole.
While both of these points would be unusual on their own, it is exceptionally unlikely that both of them would have come from the same state, critical to the election, less than three hours apart during an overnight counting process — a process subject to great controversy and where there remain, nearly three weeks from election day, many unknowns. Together, these two vote updates provided Joe Biden with the votes required to deliver him the lead in the state. Quantifying the Extremity
Having demonstrated visually how anomalous the four key vote updates are, we can now proceed to attempt to quantify how unusual it is that these three points exist at once and that two of them are from the same state.
The below graph has two particularly interesting visual properties:
The graph is presented two-dimensionally, but it’s really three-dimensional. It’s visibly much denser in the center, has what appear to be something like two normal distributions, and as you move farther from the origin along a positive-sloping line which runs through the origin, the lower the density you can expect.
The outer “edges” of the graph, in the top-right and bottom-left quadrants, closely resemble the shape of the line y = 1 / x.
We similarly expect points to be in both the top-right and bottom-left quadrants, and between an outer line which has the shape of y = 1 / x and the origin. Since these values will thus mostly be either both negative or both positive, we can see that multiplying each point’s x-coordinate with its y-coordinate is a useful way of assessing the extent to which it follows this sort of distribution. Since there are more points near the origin than there are on the visible “boundary lines” (i.e. the sequences of points on the outer edges in the first and third quadrants which visibly form these lines which look like a graph, if perhaps scaled, of y = 1/x).
We thus, for each (again, both standardized by state) coordinate pair of Biden-Trump margin and the log-ratio of Biden to Trump votes, can multiply these values and examine the distribution of the resulting products. Here, the larger a value is in magnitude, the less it follows the non-co-extremity. Plotting these products gives us: Fig. 11. Histogram of products of x and y values for each coordinate pair in Fig. 10
As we can see, the values are overwhelmingly concentrated near the median, and the graph is profoundly right-skewed — otherwise, the x-axis would not need to stretch all the way to 80. All but 60 out of 8,954 unique updates have values less than 10, and all but 10 have values less than 20. In other words, an overwhelming share of updates seem to track this rule pretty closely, but a small number of updates are truly extreme outliers.
A quick dive into these ten points reveals data which, by this point in the report, will be very familiar to the reader:
As we can see, four of the seven most anomalous vote updates — which is to say, updates in which the margin and ratio are co-extreme — are in election-critical states and occurred during the same five hour period where the circumstances on the ground were (and remain) contested and highly suspicious.
It is worth noting here that roughly 15% of the vote updates in the data set of 8,954 were from these three states. If we assumed it equally likely that any particular state should end up at any of these extreme points, there would be about a 1.2% chance that three states are represented in three out of the top four or four out of the top seven spots, and about a 0.99% chance that these three states would occupy five out of the top seven spots. It is thus very surprising to see the states in question be so disproportionately represented in the top 0.11% of the distribution of co-extremity[17]. Predicting More Typical Results and Assessing Their Implications
We now proceed to ask: How extreme did these vote updates need to be for Biden to win these states?
To do this, we consider “level sets”[18] of the products of the x and y values of the coordinates we are plotting, and consider the percentiles of these (with respect to the values graphed in Fig. 10). Each level set is a point in that distribution, and has a corresponding percentile. For example, the 99th percentile of products is about 6.6 — much smaller than the values of 77.30, 46.45, 33.23, and 22.22 which we see for these four updates. We can now determine what each of these updates might have looked like if they were only at the 99th (or other) percentile of co-extremity. In deciding how to do that, we must consider — what makes more sense? Holding the margin constant and seeing what the ratio would look like, or holding the ratio constant and seeing what the margin would look like? The latter makes far more sense in this scenario, since the former suggests an equal number of ballots for both candidates may have been held back improperly, while the latter likely suggests that an excess number of ballots for the winning candidate were produced. We are interested in testing for the latter scenario.
Since we are using ratios to predict margins, it makes sense to show what the graph in Fig.10 looks like when the axes are reversed, so that one can see how the margins vary with the ratio. Fig. 12. This is the same graph as Fig. 10, but with the axes flipped. The X-axis is the log-ratio of Biden votes to Trump votes in each update, again standardized by the distribution of such values in its state. The Y-axis is the difference between the number of Biden votes and the number of Trump votes in each update, standardized by the distribution of such values of its state.
This shows the same data as shown in Fig. 10, but is a more natural presentation for using ratios to predict margins. The pattern becomes somewhat more clear when we simply look at absolute values, as our subsequent examinations rely on metrics which treat pro-Biden and pro-Trump vote updates symmetrically.
We can also consider the “level set” of (margin, ratio) combinations which form a particular percentile of co-extremity. Here, we show the absolute values of the (standardized) log ratio and margin, with level-set annotations for the 95th, 99th, and 99.5th percentile: Fig. 13. This is the same graph as Fig. 12, but where the absolute value of the coordinates of both points is taken first, so as to present a consolidated view. The x-axis is the absolute value of the (standardized) log-ratio of Biden:Trump votes in each update, and the y-axis is the absolute value of the (standardized) Biden-Trump margin in each update.
This allows us to clearly see how extreme vote updates are, with respect to the generally observed property of them being bounded by an inverse curve[19]. The solid black line represents the 95th percentile -- i.e. 95% of vote updates are inside of this curve (i.e. have less co-extreme margins and ratios). The middle black line, with dashes and dots, represents the 99th percentile, i.e. 99% of the 8,954 vote updates are less co-extreme than any of the points on this line. And the highest-up line (dotted black), represents the 99.5th percentile, i.e. 99.5% of the 8,954 vote updates are less co-extreme than any of the points on this line. As we can see, all four of the vote updates in question (the two red points, the green points well above this line, and the farther-up yellow point), are well above even this line. Indeed, the least extreme of these points, represented by the lower red dot which is above the 99.5th percentile curve, is the 7th most co-extreme point out of all 8,954 vote updates, and represents the 99.92nd percentile. This raises the obvious question: what might these vote updates look like if they were less extreme?
Graphically, this would involve moving them some combination of down (representing a lower margin) and to the left (representing a lower ratio). In theory, we would simply calculate the shortest distance to any particular percentile level-set curve and choose that particular (margin, ratio) combination. Doing so would ignore a crucial aspect of the nature of the data, however. In particular, decreasing the ratio at any given margin implies the total number of votes in the update would go up. In particular, given the scale of the anomalies here, this would imply a scenario in which a large number of votes -- possibly hundreds of thousands -- for both candidates were somehow withheld. While it is possible that this is the case, it would almost certainly represent a data-entry error on the scale of hundreds of thousands of votes which affected both candidates equally or nearly-equally. Since the margin is the metric which matters for the result, if there was foul play, it is much more likely that votes are being subtracted from one of the candidates while also added for another.
Since we bring no a priori assumption about what these updates should like, it is worth considering what they would look like if these ratios are accurate and they merely represented the 99th percentile of co-extremity. Graphically, this represents taking the four points in question and “dragging” them down to the middle of the three black lines plotted. If this were done, these vote updates would have staggeringly smaller margins but would still be more co-extreme than 99% of the 8,954 vote updates studied. We have no affirmative reason to believe that this was precisely the case. Indeed, we cannot, with the data available to us, affirmatively make a case for any particular outcome. It is merely useful to consider what updates with these ratios would have looked like if they were more co-extreme than only 99% of the 8,954 vote updates studied, as opposed to 99.92%.
If these results seem unrealistic or implausible, this is a result of how bizarre these vote updates are with respect to the rest of the distribution.
First, let us consider the MI update at 6:31AM EST on 11/4. Its product is 77.3, and its (margin, log-ratio) coordinate pair was (15.494, 4.989)[20]. If it were to only be at the 99th percentile of co-extremity, then its product would only be 6.600. So, if the (standardized) log-ratio value of 4.989 is held constant, the (standardized) margin value would be a mere 1.323, as opposed to 15.494.
At this point, we have to undo the standardization process which allows us to fairly compare values between states. Since these were standardized with respect to the vote updates in MI, we can determine the actual margin value corresponding to a z-score[21] of 1.323. Looking up the numerical margin which corresponds in Michigan to a z-score of 1.323[22], we see it is about 11,834, while the z-score of this actual observation was 15.494, corresponding to a margin of 135,424. In other words, if we hold the ratio constant, and this vote update were only at the 99th percentile of co-extremity, the margin in this vote update would have been 123,590 votes fewer.
Let us now consider the WI update at 3:42 AM CST on 11/4. Its product is 46.452 and its (margin, log-ratio) coordinate pair was (14.427, 3.220). If it were to only be at the 99th percentile of co-extremity, then its product would be 6.600. So, if the (standardized) log-ratio value of 3.220 is held constant, the (standardized) margin value would be only 2.050, as opposed to 14.427. Looking up the numerical margin which corresponds in Wisconsin to a z-score of 2.050[23], we see it is about 16,938, while the z-score of this actual observation was 14.427, corresponding to a margin of 118,396. In other words, if we hold the ratio constant, and this vote update were only at the 99th percentile of co-extremity, the margin in this vote update would have been 101,459 votes fewer.
Let us now consider the GA update at 1:34AM EST on 11/4. Its product is 33.233 and its (margin, log-ratio) coordinate pair was (12.836, 2.589). If it were to only be at the 99th percentile of co-extremity, then its product would be 6.600. So, if the (standardized) log-ratio value of 2.589 is held constant, the (standardized) margin value would be only 2.549, as opposed to 12.836. Looking up the numerical margin which corresponds in Georgia to a z-score of 2.549[24], we see it is about 21,250, while the z-score of this actual observation was 12.836, corresponding to a margin of 107,143. In other words, if we hold the ratio constant, and this vote update were only at the 99th percentile of co-extremity, the margin in this vote update would have been 85,892 votes fewer[25].
Lastly, let us consider the MI update at 3:50AM EST on 11/4. Its product is 22.219 and its (margin, log-ratio) coordinate pair was (5.679, 3.912) If it were to only be at the 99th percentile of co-extremity, then its product would be 6.600. So, if the (standardized) log-ratio value of 3.912 is held constant, the (standardized) margin value would be only 1.687, as opposed to 5.679. Looking up the numerical margin which corresponds in Michigan to a z-score of 1.687[26], we see it is about 15,009, while the z-score of this actual observation was 5.679, corresponding to a margin of about 49,8929. In other words, if we hold the ratio constant, and this vote update were only at the 99th percentile of co-extremity, the margin in this vote update would have been 34,819 votes fewer[27]. Putting this all together, we see that if all four of these vote updates were extreme — but not as extreme — that the difference in margin would be greater than the margin of victory in all three states.
At the very least, it is possible to definitively say that Joe Biden’s victory in all three of these states relied on four of the seven most co-extreme vote updates in the entire data set of 8,954 vote updates.
Important Considerations
It is important to note one crucial indicator of why these results are bizarre. In virtually all other cases, areas which are very pro-Biden or pro-Trump have vote updates of varying size, and so a large vote update heavily favoring one candidate is almost always accompanied by even smaller updates which have a higher variance in the ratio, and at least some of them will favor the candidate who won the largest batch.
In particular, to accept the results as seen in Michigan as legitimate, one would need to believe that the one or two most possible pro-Biden areas of the state were somehow each counted their ballots entirely in one or two vote updates. If they were counted incrementally and released in smaller batches, as is typically the case, we would expect to see smaller updates with higher variance in outcome, and would almost certainly see updates with a higher Biden:Trump ratio than the two Michigan updates discussed in this report.
Indeed, if it is subsequently discovered that these did not comprise the entire count (for either mail-in votes or all votes) in these areas, then these results should be regarded with extreme suspicion. While vote counts are by no means a random sample nationally, given a small enough sub-population at which votes are counted, they eventually are. If it can be shown by those with access to time-series county-level (or precinct-level) data that, for whichever counties or precincts reported in this update, that there were other updates (or other updates with mail-in ballots), then these results become almost impossible to believe. This is to say, the believability of these updates relies on the premise that the one or two most Biden-favoring parts of the state (perhaps by ballot type) were counted entirely in these two batches. If it cannot be shown that the ballots counted during these spikes were qualitatively different from all other vote updates in Michigan, then the results are likely too extreme along multiple dimensions to be accepted at face value.
One would also need to believe that mail-in ballots, which have generally been understood to be more pro-Biden, sometimes substantially so, were counted in their entirety in these regions. While this data set does not provide breakdowns of how many votes in each update came from different types of votes, it is extremely surprising that we do not see smaller vote updates with mail-in votes which favor Biden more heavily.
This is also the case in Wisconsin, there the update discussed in this report, which had the largest Biden margin by far, also had the second-highest Biden:Trump ratio, by only a small amount. Accepting this at face value requires the belief that the most pro-Biden subset of the votes -- by geography and vote type -- was counted entirely in one batch. It would be extremely surprising if all mail-in ballots in the two most favoring Biden counties in the state, Dane and Milwaukee County, were entirely contained in this batch, and so it raises the question as to why we didn’t see even more pro-Biden updates in smaller, higher-variance vote updates in these heavily Democratic areas. If we are to accept that these votes were counted entirely in one batch, this raises serious questions as well. In particular, given the ambiguity -- to this day -- about where the vote-tabulation process was stopped and why, it makes little sense why these votes would be released in such an unusually large batch.
All of this is especially surprising when viewed in contrast to the prevailing analysis of the election, i.e. that Joe Biden’s victory was the result of improved performance in suburban areas. Looking at a map of final results by county, it is highly likely that these vote updates came from more densely-populated urban counties where Biden’s ratios were much higher. The findings here call that into question, however, as we can see that he relied heavily on four extremely aberrant vote updates which were almost certainly in heavily pro-Biden urban areas to provide a much-needed boost in the early hours of November 4th. Conclusion
This report studies 8,954 individual updates to the vote totals in all 50 states and finds that four individual updates — two of which were widely noticed on the internet, including by the President — are profoundly anomalous; they deviate from a pattern which is otherwise found in the vast majority of the remaining 8,950 vote updates. The findings presented by this report [28]suggest that four vote count updates — which collectively were decisive in Michigan, Wisconsin, and Georgia, and thus decisive of a critical forty-two electoral votes — are especially anomalous and merit further investigation.
In particular, the finding that the broader data follows general patterns and our ability to measure just how much any individual vote update does — or doesn’t — follow this pattern allows us to make concrete claims about both how extreme any given vote update is and about what any particular vote update might have looked like, had it been less extreme one one axis or another.
We further find that if these updates were only more extreme than 99% of all updates nationally in terms of their deviation from this generally-observed pattern, that, holding all else equal, Joe Biden may very well have lost the states of Michigan, Wisconsin, and Georgia, and that he would have 42 fewer Electoral votes — putting Biden below the number required to win the Presidency. Either way, it is indisputable that his margin of victory in these three states relies on four most anomalous vote updates identified by the metric developed in this report.
We once again note that this analysis is largely restricted to four individual vote updates out of a sample of nearly 9,000. This report by no means suggests stopping investigations in Michigan, Wisconsin, Pennsylvania, Georgia, or elsewhere; it is merely that these four key ballot updates are both profoundly anomalous with respect to a metric which removes any component of different states having different partisan leanings or a different number of voters. Furthermore, this analysis does not require that we regard the final vote totals in any of these states (or counties thereof) as suspicious, nor, critically, does it require that we accept that the observed data should follow any particular distribution a priori. We merely show that the data, adjusted appropriately to remove differences in size and political leaning between states, does follow a certain pattern, and that four key vote updates deviate profoundly from that pattern.
It is our belief that the extraordinarily anomalous nature of the studied vote updates here, combined with the staggering political implications, demands immediate and thorough investigation.
Democrats Are Currently Stealing a House Seat in New York
While most of the focus has been on the fraud that happened in the presidential election, it’s also worth nothing that other Republicans are also getting screwed by sketchy, late rushes of Democrat votes counted well after the election had ended. One case has become especially egregious, and it’s taking place in New York. The…
Post author
Post date
While most of the focus has been on the fraud that happened in the presidential election, it’s also worth nothing that other Republicans are also getting screwed by sketchy, late rushes of Democrat votes counted well after the election had ended. One case has become especially egregious, and it’s taking place in New York.
The race for NY-22 has suddenly flipped to the Democrat by 13 votes and the story behind how that happened will have you punching walls or facepalming, whichever you prefer.
The judge miraculously “found” enough ballots that he felt were okay and the incumbent now “leads” by 13 votes.https://t.co/miAlX6bZpD
After trailing in the polls by over 28,000 votes on election night, upstate Democratic Rep. Anthony Brindisi claimed victory Wednesday in the 22nd district over his Republican challenger — celebrating a slim 13-vote lead.
Yes, a Democrat that was losing by 28,000 votes in a district race (i.e. that’s a huge margin) has now claimed victory by 13 votes because a Democrat judge stepped in and put his stamp of approval on enough rejected ballots to make it happen. The implications here are just insane. If a single judge can decide he gets to be the arbiter of which ballots are accepted even after they were rightly rejected under the law, that opens nearly every close race up to judicial meddling.
In the grand scheme of things, assuming the Republican ends up losing here, one House seat isn’t a catastrophe. The GOP will still be well positioned to retake the chamber in 2022. But it sends a message that this kind of tampering is acceptable. Hopefully Claudia Tenney has some other legal recourse here, perhaps appealing this in federal court. I wouldn’t begin to assert that I know all her options at this point.
Regardless, this is the danger of mail in ballots and judges being allowed to nullify election law at their whim. We saw the same thing on a larger scale in Pennsylvania, where a court unilaterally decided that the election laws simply no longer applied in that state, allowing thousands of additional ballots to come in after the set deadline. Mass mail in ballots in general, at least when done in a way where they aren’t required to be in some days before the election, are also awful for the credibility of our system. How can anyone feel good or confident in the integrity of our elections when huge shifts in votes happen weeks after the fact due to late arriving and/or counted ballots?
Republicans control a lot of state levers of power, from governorships to state houses. They need to be wielding that power to fix this stuff. If they don’t, they can expect to be slowly weeded out by the very fraud they’ve ignored.
Johns Hopkins Study Saying COVID-19 Has 'Relatively No Effect on Deaths' in U.S. Spiked After Publication
BY MATT MARGOLIS NOV 27, 2020 11:24 AM ET
Share Tweet Share (Image: cdc.gov)Conventional wisdom is that COVID-19 has caused thousands of deaths in the United States and nearly 1.5 million worldwide. This perception has been directly challenged by a study published by Johns Hopkins University on November 22.
Genevieve Briand, assistant program director of the Applied Economics master’s degree program at Johns Hopkins University, critically analyzed the impact that COVID-19 had on U.S. deaths. According to Briand, the impact of COVID-19 on deaths in the United States can be fully understood by comparing it to the number of total deaths in the country.
According to the study, “in contrast to most people’s assumptions, the number of deaths by COVID-19 is not alarming. In fact, it has relatively no effect on deaths in the United States.”
Wait, what? Really?
That’s what it says. And, it should come as no surprise that the study was deleted within days.
So, how exactly did the study conclude that COVID-19 has had “relatively no effect on deaths”? Here’s how the study made this determination:
After retrieving data on the CDC website, Briand compiled a graph representing percentages of total deaths per age category from early February to early September, which includes the period from before COVID-19 was detected in the U.S. to after infection rates soared.
Surprisingly, the deaths of older people stayed the same before and after COVID-19. Since COVID-19 mainly affects the elderly, experts expected an increase in the percentage of deaths in older age groups. However, this increase is not seen from the CDC data. In fact, the percentages of deaths among all age groups remain relatively the same.
According to Briand, “The reason we have a higher number of reported COVID-19 deaths among older individuals than younger individuals is simply because every day in the U.S. older individuals die in higher numbers than younger individuals.”
Briand’s analysis found that the range of deaths amongst the older population has remained within the range of past years.
So, if COVID-19 has actually had no significant impact on U.S. deaths, why does it not appear that way?
To answer that question, Briand shifted her focus to the deaths per causes ranging from 2014 to 2020. There is a sudden increase in deaths in 2020 due to COVID-19. This is no surprise because COVID-19 emerged in the U.S. in early 2020, and thus COVID-19-related deaths increased drastically afterward.
Analysis of deaths per cause in 2018 revealed that the pattern of seasonal increase in the total number of deaths is a result of the rise in deaths by all causes, with the top three being heart disease, respiratory diseases, influenza and pneumonia.
“This is true every year,” explained Briand. “Every year in the U.S. when we observe the seasonal ups and downs, we have an increase of deaths due to all causes.”
Here’s where things get interesting.
When Briand looked at the 2020 data during that seasonal period, COVID-19-related deaths exceeded deaths from heart diseases. This was highly unusual since heart disease has always prevailed as the leading cause of deaths. However, when taking a closer look at the death numbers, she noted something strange. As Briand compared the number of deaths per cause during that period in 2020 to 2018, she noticed that instead of the expected drastic increase across all causes, there was a significant decrease in deaths due to heart disease. Even more surprising, as seen in the graph below, this sudden decline in deaths is observed for all other causes.
The study found that “This trend is completely contrary to the pattern observed in all previous years.” In fact, “the total decrease in deaths by other causes almost exactly equals the increase in deaths by COVID-19.”
Briand concludes that the COVID-19 death toll in the United States is misleading and that deaths from other diseases are being categorized as COVID-19 deaths.
On Thursday, Johns Hopkins University explained that they deleted the article on the study because it “was being used to support false and dangerous inaccuracies about the impact of the pandemic.”
Though making clear the need for further research, the article was being used to support false and dangerous inaccuracies about the impact of the pandemic. We regret that this article may have contributed to the spread of misinformation about COVID-19.
Marvis Gutierrez and Ariella Shua, the managing editors of the JHU newsletter, assured Lead Stories that the article/study was not censored. “The article in question was retracted last night, as it was being used to spread misinformation about the pandemic. We have preserved the article as a PDF and posted an Editor’s Note: with full clarification about our decision, highlighting the inaccuracies of the study…We were not censored, but decided to retract the article based on the reasons outlined in the Editor’s Note.””
The author of the article on Briand’s study, Yanni Gu, responding to the article being pulled, posted the following on LinkedIn:
Today, on November 27th, The News-Letter officially posted their reason for retracting the article, stating inaccuracies in the analysis. I am frustrated at the explanation, and I think it is disrespectful to Dr. Briand’s hard work putting data together and doing an honest analysis. If her analysis was to be contradicted, then at least an equal-level analysis should be done to provide more data and thus a new conclusion. Dr. Briand and her work deserve such respect.
I have received many messages asking the reason for taking the article down, and so I would like to officially express my opinions here. I even got emails saying that thanks to me, people now will not be wearing masks or practicing social distancing. They called me “a COVID denier and a minimizer” and that I have no idea the damage and the lives cost in me writing such an article. I was devastated to receive such accusations, but I stand my ground. The goal is never to undermine the effects of COVID-19 but to suggest a possible over-exaggeration in death numbers due to the pandemic.
This **** is comical now. MAL you wreak of desperation, I know you even recognize it at this point. LOL can't wait until Jan 20th or even Dec 14th for that matter. ITS OVER, YOU LOST, better luck in 2024 you tin foil hat wearing loon
3:44 PM
To say out-loud that you find the results of the 2020 presidential election odd is to invite derision. You must be a crank or a conspiracy theorist. Mark me down as a crank, then. I am a pollster and I find this election to be deeply puzzling. I also think that the Trump campaign is still well within its rights to contest the tabulations. Something very strange happened in America’s democracy in the early hours of Wednesday November 4 and the days that followed. It’s reasonable for a lot of Americans to want to find out exactly what.
First, consider some facts. President Trump received more votes than any previous incumbent seeking reelection. He got 11 million more votes than in 2016, the third largest rise in support ever for an incumbent. By way of comparison, President Obama was comfortably reelected in 2012 with 3.5 million fewer votes than he received in 2008.
Trump’s vote increased so much because, according to exit polls, he performed far better with many key demographic groups. Ninety-five percent of Republicans voted for him. He did extraordinarily well with rural male working-class whites.
Trump grew his support among black voters by 50 percent over 2016. Nationally, Joe Biden’s black support fell well below 90 percent, the level below which Democratic presidential candidates usually lose.
Trump increased his share of the national Hispanic vote to 35 percent. With 60 percent or less of the national Hispanic vote, it is arithmetically impossible for a Democratic presidential candidate to win Florida, Arizona, Nevada, and New Mexico. Bellwether states swung further in Trump’s direction than in 2016. Florida, Ohio and Iowa each defied America’s media polls with huge wins for Trump. Since 1852, only Richard Nixon has lost the Electoral College after winning this trio, and that 1960 defeat to John F. Kennedy is still the subject of great suspicion.
Midwestern states Michigan, Pennsylvania, and Wisconsin always swing in the same direction as Ohio and Iowa, their regional peers. Ohio likewise swings with Florida. Current tallies show that, outside of a few cities, the Rust Belt swung in Trump’s direction. Yet, Biden leads in Michigan, Pennsylvania, and Wisconsin because of an apparent avalanche of black votes in Detroit, Philadelphia, and Milwaukee. Biden’s ‘winning’ margin was derived almost entirely from such voters in these cities, as coincidentally his black vote spiked only in exactly the locations necessary to secure victory. He did not receive comparable levels of support among comparable demographic groups in comparable states, which is highly unusual for the presidential victor.
We are told that Biden won more votes nationally than any presidential candidate in history. But he won a record low of 17 percent of counties; he only won 524 counties, as opposed to the 873 counties Obama won in 2008. Yet, Biden somehow outdid Obama in total votes.
Victorious presidential candidates, especially challengers, usually have down-ballot coattails; Biden did not. The Republicans held the Senate and enjoyed a ‘red wave’ in the House, where they gained a large number of seats while winning all 27 toss-up contests. Trump’s party did not lose a single state legislature and actually made gains at the state level.
Another anomaly is found in the comparison between the polls and non-polling metrics. The latter include: party registrations trends; the candidates’ respective primary votes; candidate enthusiasm; social media followings; broadcast and digital media ratings; online searches; the number of (especially small) donors; and the number of individuals betting on each candidate.
Despite poor recent performances, media and academic polls have an impressive 80 percent record predicting the winner during the modern era. But, when the polls err, non-polling metrics do not; the latter have a 100 percent record. Every non-polling metric forecast Trump’s reelection. For Trump to lose this election, the mainstream polls needed to be correct, which they were not. Furthermore, for Trump to lose, not only did one or more of these metrics have to be wrong for the first time ever, but every single one had to be wrong, and at the very same time; not an impossible outcome, but extremely unlikely nonetheless.
Atypical voting patterns married with misses by polling and non-polling metrics should give observers pause for thought. Adding to the mystery is a cascade of information about the bizarre manner in which so many ballots were accumulated and counted.
The following peculiarities also lack compelling explanations:
1. Late on election night, with Trump comfortably ahead, many swing states stopped counting ballots. In most cases, observers were removed from the counting facilities. Counting generally continued without the observers
2. Statistically abnormal vote counts were the new normal when counting resumed. They were unusually large in size (hundreds of thousands) and had an unusually high (90 percent and above) Biden-to-Trump ratio
3. Late arriving ballots were counted. In Pennsylvania, 23,000 absentee ballots have impossible postal return dates and another 86,000 have such extraordinary return dates they raise serious questions
4. The failure to match signatures on mail-in ballots. The destruction of mail-in ballot envelopes, which must contain signatures
5. Historically low absentee ballot rejection rates despite the massive expansion of mail voting. Such is Biden’s narrow margin that, as political analyst Robert Barnes observes, ‘If the states simply imposed the same absentee ballot rejection rate as recent cycles, then Trump wins the election’
6. Missing votes. In Delaware County, Pennsylvania, 50,000 votes held on 47 USB cards are missing
7. Non-resident voters. Matt Braynard’s Voter Integrity Project estimates that 20,312 people who no longer met residency requirements cast ballots in Georgia. Biden’s margin is 12,670 votes
8. Serious ‘chain of custody’ breakdowns. Invalid residential addresses. Record numbers of dead people voting. Ballots in pristine condition without creases, that is, they had not been mailed in envelopes as required by law
9. Statistical anomalies. In Georgia, Biden overtook Trump with 89 percent of the votes counted. For the next 53 batches of votes counted, Biden led Trump by the same exact 50.05 to 49.95 percent margin in every single batch. It is particularly perplexing that all statistical anomalies and tabulation abnormalities were in Biden’s favor. Whether the cause was simple human error or nefarious activity, or a combination, clearly something peculiar happened.
If you think that only weirdos have legitimate concerns about these findings and claims, maybe the weirdness lies in you.
No I think you can find anything on the interweb to support your narrative. All these posts and links and videos offer ZERO proof of wide spread election fraud in 2020. None, they explain how this could be possible or this might have happened or this software is vulnerable but again ZERO proof it actually happened. That's why fatso is laughed out of every court he and his crackpot team file in. Deeper and deeper down the rabbit hole MAL goes.
I just read something interesting one of the lawsuits mentions Edison County in Michigan. There is no Edison County in Michigan apparently, but that’s not what really surprised me. After hearing that the lawyers were submitting affidavits with misspelled words and improper format nothing would surprise me anymore. What surprises me is there’s no county in the United States of America named Edison! I would have bet the farm that there are multiple counties named after Thomas Edison! Or even some other Edison!
Always support your local wrestling tournament concession stands!!!
On another note, Bill Barr just told the Associated Press that the Department of Justice found no fraud. Guess that about wraps it up. Just have to roll up the mats and turn off the gym lights.
Sharpie gate...There is a photo of of the actual letter between election workers.
October 22 Maricopa County election official: “…we’ve heard you and we know you’re hearing issues and concerns about the Sharpie Markers. We NEED to use Markrs on Election Day, but for now and through 11/2 hand voters a Ballpoint Pen. You have plenty of pens in your supplies right now…Please message this to your inspectors and ensure they cascade this change down to everyone on the Board…”
Comment