No. You need to understand what this graph shows. It does NOT show Trump's performance among Republicans, it shows his performance against ticket-splitting voters.
By definition, these are either:
Republicans who voted for Biden
Moderates who voted split ticket
Democrats who voted for Trump
So, because there are more Republicans in heavy R districts, there WILL be more Republicans in those districts who voted for Biden than there will be R's in heavy D precincts who voted for Biden.
So OF COURSE Trump will UNDERPERFORM with ticket splitters in heavy R districts COMPARED to straight ticket, hardcore Republican voters. Just like Biden underperforms with ticket splitters in heavy D districts. That's almost what this chart shows. Except it doesn't. It shows Biden doing much better than Trump in Heavy D. districts among ticket splitters (look at the left of the graph - Biden's dots there should be lower, in line with the rest of the blue dots, but they are higher, and Trump's is lower).
BUT it also shows that Trump UNDERPERFORMED in heavy Democrat districts among large numbers of Democrat ticket splitters who by definition would have voted for Trump, and among moderates who split their ticket.
Trump's performance along the LEFT side of the chart should be higher (more in line with the rest of the normal slope), but it's not. And its only in Heavy D districts where he's getting screwed.
It is far more plausible to me that heavy Democrat districts stuffed ballots for Biden in those areas, which would hurt Trump's score and enhance Biden's and that's exactly what we see here.
If you saw the data analysis from a few days ago by Dr. Shiva, MIT PhD, this plot is showing the same results as his, just with an additional scatter on top of Shiva's. If you didn't check it out, I highly recommend it.
Shiva found that, while examining Michigan's top 4 counties, that as a precinct skewed more %republican in it's straight-party-votes, Trump received LESS votes, which defies logic. The new plot added on top of that (represented by the blue dots for Biden's performance) shows that as precincts got more %republican, they casted more votes for Biden. None of it makes any sense, considering Trumps very-high approval rating amongst republicans.
Shiva concluded that it is 99% likely caused by an algorithm giving votes to Biden.
I'm sorry, bro, but it's Shiva who is mistaken. He's reached the wrong conclusion, an implausible one, but I think he stumbled into the real fuckery.
Read it carefully. Grasp what I'm saying. Many have already pointed out Shiva's flawed analysis. I point it out too. A negative slope (left and down) is NORMAL and EXPECTED when plotting this data.
What ISN'T normal or expected is how it actually flattens out in heavy Democrat districts.
It's the heavy D districts where the actual fuckery is going on.
How is this different? Trump is shown in red and still shows a decrease in larger percentage republican districts. Biden in blue is being shown to way over-perform in Republican districts. Also Trump is over-performing in Dem districts until the algorithm kicks in.
You're missing a key point - this chart shows Trump's relative strength with ticket-splitting voters measured against his strength in the precinct overall.
Ticket splitters are:
Moderates
Republicans voting for Biden
Democrats voting for Trump
The chart takes the % of ticket splitters for Trump and plots it against the % Trump strength in the district as measured by straight party voters.
In a heavy Republican district, there will be MORE Republicans who vote for Biden than there will be Democrats for Trump. And vice versa: in a heavy Democrat district, there will be MORE Democrats for Trump than there will be Republicans for Biden.
So, in a case with a heavy R district, Trump should be expected to underperform among moderates and registered party ticket splitters than he would compared to hard core straight party Republicans, right?
Of course.
So Shiva is wrong. The negative slope with Trump doing poorly among ticket splitters in heavy R districts is totally normal. It's expected.
What isn't expected is how his performance in heavy Democrat precincts is MUCH LOWER than the trendline, while Biden's performance is MUCH HIGHER than the trendline.
I was under the assumption that it was straight party. But you are saying it was split party ticket? If that's the case then would it be that they would have to take votes away from one and add them to the other? Otherwise you would end up with a negative percentage which can't happen? So then the graph with all of the clumped voting would be the one that is the problem, not the downward (negative) sloping one? Is that correct? I am just asking, not trying to be an ass.
The RED dots on Dr. Shiva's plot (that's the image above, it's not mine), show the relative strength of Trump among ticket splitters vs. straight party voters.
The farther right the dot, the stronger the precinct was overall for Trump, as measured by the % of straight party voters who cast ballots for Trump.
The higher the dot, the more split ticket votes Trump received, expressed as a percentage of Trump votes MINUS the relative strength of the precinct.
My point is Shiva's analysis has a flaw and he focused on the wrong part of the graph.
What I'm trying to call attention to is how the trend line is bends, the more you get to Democrat precinct, and this should NOT be the case, because of course Trump does better among moderates who split their ticket in a heavy Democrat precinct than he does vs. the precinct as a whole.
The problem is, he's not doing as well as he should be doing...which indicates something is wrong in those precincts.
I don't think he was incorrect, this is just additional data with different X/Y scales. He is showing the blue dots as Biden, showing the relationship between the two along the same x-axis, with the y-axis now being "candidates performance amongst own party"
This IS Shiva's plot. His plot is just that. It's perfectly accurate because it's just data. It's his analysis that was already debunked at T_D.win and elsewhere.
I'm arguing that he misinterpreted the data and the REAL fraud is on the LEFT side of the graph, which is HEAVY Democrat precincts.
What if Shiva is wrong, but he just missed the fraud?
I believe Shiva overcomplicated his analysis as follows:
He subtracted the % of Trump precinct strength from the % of his votes among ticket splitters. He did not need to do this, and this is what created the downward sloping line, which he then thought was unexpected and zeroed in on the WRONG END of the graph (the Republican end).
He should have simply plotted the % of Trump votes among ticket splitters without subtracting the precinct strength. That would have produced the expected flat plot of precincts along the x-axis.
But it ALSO would have produced a VERY ODD downward bend in Democrat precincts (everything left of the 30% line in this county) which surely would have gotten his attention (remember, he mentioned he expected the dots to be flat along the x-axis)...
This revised plot, if someone can do it, WILL show the actual Democrat fraud, and pinpoint the exact precincts where it occurred.
Dr. Shiva's analysis (and that of a 4chan anon, above) seem to think that the downward slope of Trump's ticket splitting voters (charted in red, above) suggests votes being stolen from him in strong Republican precincts in Oakland County.
For various reasons, I believe that to be incorrect, and the distribution along the right side of the slope is totally normal.
What looks really screwy to me, is how the slope is breaks downward around the 30% and over mark, but in precincts below 30%, which are heavily Democratic, Biden quite clearly overperforms among ticket splitters (compared to the rest of the slope) while Trump clearly underperforms (the left side of the red dots plot should be higher, more in line with the rest, same with the left hand blue dot plots for Biden, which should be lower).
Precincts that are 70% or higher Democrat precincts are, in my view, FAR more likely to feature exclusively Democrat poll workers, which in turn means a higher likelihood of ballot shenanigans.
Some quick background on this plot and underlying analysis:
The x-axis of the chart is a measurement of precinct strength for Republicans, expressed as a percentage of GOP straight ticket voters. The further to the right we look, the higher the percentage of GOP straight ticket voters are found in that precinct.
The y-axis measures ticket-splitter percentage over or under the strength of straight party ticket voters (with red being a measurement of Biden's performance in the precinct, and blue being a measurement of Trump's performance in a precinct). So the two left-most red and blue dots are plots for both Biden and Trump in a precinct that is almost 95% Democrat straight ticket voters. Because the dots are both at 0% on the y-axis (vertical), that means that Biden scored ~95% among ticket splitters, and Trump scored ~5%.
This seems highly unlikely to me, and here's why:
In a precinct that is so overwhelmingly Democrat, where 95% of all registered party voters are Democrats, to which political party do the ticket splitters MOSTLY belong to?
That's right. Democrats. In fact, 95 out of 100 registered party ticket splitting voters will be Democrats. And by definition, which presidential candidate does a Democrat have to vote for if he or she is a ticket splitter?
That's right. Donald J. Trump.
Now you're starting to catch on.
Who else might be ticket splitters in any precinct? That's right: moderates. They are, by definition, middle of the road types. So they will, generally speaking, vote 50-50 Trump or Biden, give or take 15 points either way, but it all averages out.
So if we're going to plot Donald Trump's strength among ticket splitting voters in a district where Trump is only expected to win 5% of straight party voters, should we expect ticket splitters who are 50/50 moderates and the rest are overwhelmingingly (95%) Democrats who voted for Trump, do you think Trump will do better or worse than his performance with straight party voters?
If you said "Better," you're absolutely right.
Yet this chart shows, at the very least, several precincts where he's far worse, and many other precincts to the left of the 30% line, at a much lower clip than is expected.
Something IS screwy here, and it's almost exclusively screwy in heavy Democrat districts.
If any data pedes have access to Oakland's precinct data, there are two key questions that might tease this out a bit more:
Can we tell how many "Biden only" votes came from these heavy D precincts compared with the % of "Biden only" votes from other places?
Are "Biden only" votes considered "ticket splitters" or "straight party" voters? If ticket splitters, we're on to something. Maybe big.
No. You have got your analysis completely backwards. Trump is red in this graph and Biden is blue. Trump is overperforming in Democrat precincts as indicated by the first 30% which actually has a linear trend in Trump's favor if you look at it. Suddenly that trend breaks down as the precinct gets more Republican, and the slope of the trend line gets bent like a transistor, as Dr. Shiva suggested.
The reason is simple: it's easier to steal votes in heavily Republican districts from Trump and have it go unnoticed. If a precinct normally goes 70/30 Republican but it only ends up 60/40 for Trump then no one really notices, Trump still wins the county he was expected to but Biden gets the phantom votes added to his statewide total.
Sorry, friend. Not trying to be an ass. But you're not understanding what Shiva's graph ACTUALLY shows.
I'm aware that Trump is red.
I'm saying that because of the way Shiva chose to plot this data, by subtracting Trump's % of precinct strength from his % of ticket splitter strength, it SHOULD yield a negative number in heavy Trump districts because in heavy R districts, Republicans outnumber moderate ticket splitters.
Example: In a 90% R precinct, let's say Trump gets 80% of ticket splitters to Biden's 20%. I think you would agree that's an impressive number, right? Trump ticket splitters are, by definition, moderate voters who do not cast an exclusively Republican ballot, or they are Democrats who voted for Trump. But in a heavy Republican district, there won't be very many Democrats in the first place, right? So most will just be moderates, who don't necessarily agree with their neighbors politically. So Trump getting 80% of moderates in a heavy R precinct is pretty amazing, I'd even argue it's highly unlikely. But I'll give it to you.
So under Shiva's math, we take 80% and subtract the 90% precinct strength, which yields -10 (80%-90%=10).
We then plot that on the chart at (90, -10)
That's the lower right, WELL below the x-axis.
And that's WITH Trump overperforming among moderate ticket splitters.
Hope that helps.
Honestly, there's no need to subtract the precinct strength. I think that's where Shiva got confused. If he hadn't done that, I think the resulting graph would show my point more clearly: a flat line (as expected), except on the left, which would slope sharply down to the left, indicating something odd happening in Democrat precincts.
These aren't ticket splitters, in Dr. shiva's dataset. These are precincts with x% of Republican voters. As the precinct gets more Republican (after the 20% or 30% mark), Trump is getting less of the overall vote for that precinct, less than the percentage of Republican voters. Go watch his video again he explains it very well.
Michigan voters have the option of voting straight ticket, either for every Republican or each Democrat on the ballot.
Ok, so they aren't necessarily registered Republicans but they have clearly indicated their desire to vote Republican by marking straight party on the ballot. What are the odds that as a precinct got more straight Republican that the percentage of Trump votes also dropped precipitously??? These aren't small amounts in Shiva's analysis, these are 20 and 30% differentials between Republican straight party and Trump. That's insane, buddy. Indicative of fraud. Especially since the linear weighting kicks in at precisely 20% or 30%. Sorry, the President is the most important ticket. This is fraud.
I think it's insane that Michigan allows you to mark straight party and also individual races. It means these ballots can be tampered with after the fact. Straight party selection voting should be outlawed.
There is no data here indicating that the percentage of Trump votes dropped in strong R precincts.
The chart is measuring which way ticket splitters (moderates and squishy R's and D's) voted compared to the relative strength of the district.
In other words, moderate voters are typically 50-50 voters, they could go either way. Maybe they break a little for Trump in his stronger areas, but do you really think they would break STRONGER than hardcore Republicans would? Of course not. Yet that's what you're saying you expect to see: Trump getting just as many moderate voters as he got R's in a heavy R district. It's silly.
So let's do another more extreme example: 1,000 votes cast in a precinct.
995 are hardcore Republicans, and vote straight party tickets. That's 995 votes for Donald Trump.
The remaining 5 voters are the precinct's only moderates, and 4 of them split their ticket but vote Trump. The last guy splits his ticket but votes for Biden. That's 80% for Trump among ticket splitters.
This just means that Trump's percentage of votes among ticket splitters was 19.5% below his percentage of hardcore voters. There's NOTHING unusual about that.
Trump actually WAY outperformed Biden 999 votes to 1, but we're measuring how he performed among ticket splitting voters compared to hardcore voters.
You tracking with me yet?
I hope so, because the ultimate point I am urging you to take note of is the relatively flat precincts on the left, which are precincts in heavy Democrat areas.
In heavy D areas, moderates, being 50-50, so they are likely to vote for Trump by a margin greater than the hardcore Democrats are voting for him, so Trump's plot there SHOULD be pretty stinking high. But it's not. It's BENT. Downward.
WTF. Poll workers in those Dem precincts are the ones that screwed us, probably with ballot stuffing for Biden, not some B.S. algorithm in R precincts.
No, that is not the Y axis on Dr. Shiva's chart. He's not even looking at so-called ticket splitters directly. He's looking at (Trump - Republican Straight Party) as a vote %. You have to subtract the Republican straight party votes from Trump's totals and then measure that percentage as a a fraction of the overall vote for that district.
It shouldn't be negative unless Trump is unpopular among Republican voters (not the case in the least).
That would indicate that people are going into the polls, and ticking both the Straight Party Republican box and Pedo Joe Biden in massive numbers, to the tune of 20-30% in some cases. It ain't happening. This is software fraud.
as you go to the right, you are getting into predominantly more party affiliated districts. so, the farther right you go, that district would be more republican (red) or DEM (blue). obviously one would agree that the more politically affiliated a district is, the more likely that district would vote for the president. there is zero chance that the more republican a district got ( especially rep) that they wouldn't vote for trump. this is evident as you go more right, you go up or at least have a higher average line.
see how in less rep districts, they are less likely to vote trump , right? hence why they are below 0%
this graph literally states that the more republican a district is, the more they voted for Biden, highly unlikely- and the opposite, the more democratic district, the less likely they were to vote bide.
makes no common sense
Imagine in a district with 1000 voters, 990 of them vote a Republican straight ticket.
This would score 99% on the x-axis (far right) on the chart.
Now imagine the last 10 voters are ALL moderate, undecided voters who split their ticket for both Republicans AND Democrats. Let's say 8 vote for Trump and a Democrat congressman, and the other 2 vote for Joe Biden and a Republican congressman. That's what makes them "ticket splitters."
That means among ticket splitters, Trump got 80% of them. Wow. This is great. Moderates breaking HEAVY for Trump right?
You'd argue that Trump is overperforming among moderates and this plot should be on the upper right side of the graph.
That's wrong.
Because Dr. Shiva takes the precinct score (99%) and subtracts it from the ticket splitter score (80%) which then yields a score of -18 (80-99=-18) and he plots it on the y-axis at the 99 percent mark (lower right!).
This means that even in districts where moderates break heavily for Trump, unless they break at a greater rate than Trump's straight ticket hard core voters (right side of the x-axis), they will ALWAYS score below the trendline. This also means the opposite is true, in weak Republican areas (left side of the x-axis), moderates breaking heavily (or even just barely better than the district strength for Trump) will still outperform Trump's straight ticket voters (meaning ABOVE the 0% x-axis line).
That means that the downward left to right slope of Dr. Shiva's graph was PERFECTLY NORMAL, but his analysis is wrong.
But his graph ALSO shows something quite strange: a flattening of the trend line in heavy Democrat precincts.
And THAT's where I'm saying the fraud occurred, mostly likely through plain old ballot stuffing.
Incorrect, UNLESS you're suggesting the scale changes at the 30% mark.
But where is your evidence for that? The graph scale appears to be consistent.
Because the fraud occurs in heavy Democrat districts to the far left of the chart. Where it appears flat and out of line with the rest of the slope.
In the place where 100% of the poll workers are Democrats, too.
exactly the same... the more republican the worse trump does and it flips to biden
https://www.youtube.com/watch?v=C0-eJfOg3nc
No. You need to understand what this graph shows. It does NOT show Trump's performance among Republicans, it shows his performance against ticket-splitting voters.
By definition, these are either:
So, because there are more Republicans in heavy R districts, there WILL be more Republicans in those districts who voted for Biden than there will be R's in heavy D precincts who voted for Biden.
So OF COURSE Trump will UNDERPERFORM with ticket splitters in heavy R districts COMPARED to straight ticket, hardcore Republican voters. Just like Biden underperforms with ticket splitters in heavy D districts. That's almost what this chart shows. Except it doesn't. It shows Biden doing much better than Trump in Heavy D. districts among ticket splitters (look at the left of the graph - Biden's dots there should be lower, in line with the rest of the blue dots, but they are higher, and Trump's is lower).
BUT it also shows that Trump UNDERPERFORMED in heavy Democrat districts among large numbers of Democrat ticket splitters who by definition would have voted for Trump, and among moderates who split their ticket.
Trump's performance along the LEFT side of the chart should be higher (more in line with the rest of the normal slope), but it's not. And its only in Heavy D districts where he's getting screwed.
It is far more plausible to me that heavy Democrat districts stuffed ballots for Biden in those areas, which would hurt Trump's score and enhance Biden's and that's exactly what we see here.
What does this scatter plot show
If you saw the data analysis from a few days ago by Dr. Shiva, MIT PhD, this plot is showing the same results as his, just with an additional scatter on top of Shiva's. If you didn't check it out, I highly recommend it.
Shiva found that, while examining Michigan's top 4 counties, that as a precinct skewed more %republican in it's straight-party-votes, Trump received LESS votes, which defies logic. The new plot added on top of that (represented by the blue dots for Biden's performance) shows that as precincts got more %republican, they casted more votes for Biden. None of it makes any sense, considering Trumps very-high approval rating amongst republicans.
Shiva concluded that it is 99% likely caused by an algorithm giving votes to Biden.
Whoa sounds suspect thanks for the explanation
I'm sorry, bro, but it's Shiva who is mistaken. He's reached the wrong conclusion, an implausible one, but I think he stumbled into the real fuckery.
Read it carefully. Grasp what I'm saying. Many have already pointed out Shiva's flawed analysis. I point it out too. A negative slope (left and down) is NORMAL and EXPECTED when plotting this data.
What ISN'T normal or expected is how it actually flattens out in heavy Democrat districts.
It's the heavy D districts where the actual fuckery is going on.
How is this different? Trump is shown in red and still shows a decrease in larger percentage republican districts. Biden in blue is being shown to way over-perform in Republican districts. Also Trump is over-performing in Dem districts until the algorithm kicks in.
You're missing a key point - this chart shows Trump's relative strength with ticket-splitting voters measured against his strength in the precinct overall.
Ticket splitters are:
The chart takes the % of ticket splitters for Trump and plots it against the % Trump strength in the district as measured by straight party voters.
In a heavy Republican district, there will be MORE Republicans who vote for Biden than there will be Democrats for Trump. And vice versa: in a heavy Democrat district, there will be MORE Democrats for Trump than there will be Republicans for Biden.
So, in a case with a heavy R district, Trump should be expected to underperform among moderates and registered party ticket splitters than he would compared to hard core straight party Republicans, right?
Of course.
So Shiva is wrong. The negative slope with Trump doing poorly among ticket splitters in heavy R districts is totally normal. It's expected.
What isn't expected is how his performance in heavy Democrat precincts is MUCH LOWER than the trendline, while Biden's performance is MUCH HIGHER than the trendline.
I was under the assumption that it was straight party. But you are saying it was split party ticket? If that's the case then would it be that they would have to take votes away from one and add them to the other? Otherwise you would end up with a negative percentage which can't happen? So then the graph with all of the clumped voting would be the one that is the problem, not the downward (negative) sloping one? Is that correct? I am just asking, not trying to be an ass.
It's cool.
The RED dots on Dr. Shiva's plot (that's the image above, it's not mine), show the relative strength of Trump among ticket splitters vs. straight party voters.
The farther right the dot, the stronger the precinct was overall for Trump, as measured by the % of straight party voters who cast ballots for Trump.
The higher the dot, the more split ticket votes Trump received, expressed as a percentage of Trump votes MINUS the relative strength of the precinct.
My point is Shiva's analysis has a flaw and he focused on the wrong part of the graph.
What I'm trying to call attention to is how the trend line is bends, the more you get to Democrat precinct, and this should NOT be the case, because of course Trump does better among moderates who split their ticket in a heavy Democrat precinct than he does vs. the precinct as a whole.
The problem is, he's not doing as well as he should be doing...which indicates something is wrong in those precincts.
Ok, good deal. Nice work. Honestly whatever helps us win is fine by me.
I don't think he was incorrect, this is just additional data with different X/Y scales. He is showing the blue dots as Biden, showing the relationship between the two along the same x-axis, with the y-axis now being "candidates performance amongst own party"
Dude, you're not understanding.
This IS Shiva's plot. His plot is just that. It's perfectly accurate because it's just data. It's his analysis that was already debunked at T_D.win and elsewhere.
I'm arguing that he misinterpreted the data and the REAL fraud is on the LEFT side of the graph, which is HEAVY Democrat precincts.
What are the credentials of the person posting this? We know Shiva, who is this?
Ask yourself this question:
What if Shiva is wrong, but he just missed the fraud?
I believe Shiva overcomplicated his analysis as follows:
He subtracted the % of Trump precinct strength from the % of his votes among ticket splitters. He did not need to do this, and this is what created the downward sloping line, which he then thought was unexpected and zeroed in on the WRONG END of the graph (the Republican end).
He should have simply plotted the % of Trump votes among ticket splitters without subtracting the precinct strength. That would have produced the expected flat plot of precincts along the x-axis.
But it ALSO would have produced a VERY ODD downward bend in Democrat precincts (everything left of the 30% line in this county) which surely would have gotten his attention (remember, he mentioned he expected the dots to be flat along the x-axis)...
This revised plot, if someone can do it, WILL show the actual Democrat fraud, and pinpoint the exact precincts where it occurred.
That's data we can actually use.
Dr. Shiva's analysis (and that of a 4chan anon, above) seem to think that the downward slope of Trump's ticket splitting voters (charted in red, above) suggests votes being stolen from him in strong Republican precincts in Oakland County.
For various reasons, I believe that to be incorrect, and the distribution along the right side of the slope is totally normal.
What looks really screwy to me, is how the slope is breaks downward around the 30% and over mark, but in precincts below 30%, which are heavily Democratic, Biden quite clearly overperforms among ticket splitters (compared to the rest of the slope) while Trump clearly underperforms (the left side of the red dots plot should be higher, more in line with the rest, same with the left hand blue dot plots for Biden, which should be lower).
Precincts that are 70% or higher Democrat precincts are, in my view, FAR more likely to feature exclusively Democrat poll workers, which in turn means a higher likelihood of ballot shenanigans.
Some quick background on this plot and underlying analysis:
The x-axis of the chart is a measurement of precinct strength for Republicans, expressed as a percentage of GOP straight ticket voters. The further to the right we look, the higher the percentage of GOP straight ticket voters are found in that precinct.
The y-axis measures ticket-splitter percentage over or under the strength of straight party ticket voters (with red being a measurement of Biden's performance in the precinct, and blue being a measurement of Trump's performance in a precinct). So the two left-most red and blue dots are plots for both Biden and Trump in a precinct that is almost 95% Democrat straight ticket voters. Because the dots are both at 0% on the y-axis (vertical), that means that Biden scored ~95% among ticket splitters, and Trump scored ~5%.
This seems highly unlikely to me, and here's why:
In a precinct that is so overwhelmingly Democrat, where 95% of all registered party voters are Democrats, to which political party do the ticket splitters MOSTLY belong to?
That's right. Democrats. In fact, 95 out of 100 registered party ticket splitting voters will be Democrats. And by definition, which presidential candidate does a Democrat have to vote for if he or she is a ticket splitter?
That's right. Donald J. Trump.
Now you're starting to catch on.
Who else might be ticket splitters in any precinct? That's right: moderates. They are, by definition, middle of the road types. So they will, generally speaking, vote 50-50 Trump or Biden, give or take 15 points either way, but it all averages out.
So if we're going to plot Donald Trump's strength among ticket splitting voters in a district where Trump is only expected to win 5% of straight party voters, should we expect ticket splitters who are 50/50 moderates and the rest are overwhelmingingly (95%) Democrats who voted for Trump, do you think Trump will do better or worse than his performance with straight party voters?
If you said "Better," you're absolutely right.
Yet this chart shows, at the very least, several precincts where he's far worse, and many other precincts to the left of the 30% line, at a much lower clip than is expected.
Something IS screwy here, and it's almost exclusively screwy in heavy Democrat districts.
If any data pedes have access to Oakland's precinct data, there are two key questions that might tease this out a bit more:
Can we tell how many "Biden only" votes came from these heavy D precincts compared with the % of "Biden only" votes from other places?
Are "Biden only" votes considered "ticket splitters" or "straight party" voters? If ticket splitters, we're on to something. Maybe big.
No. You have got your analysis completely backwards. Trump is red in this graph and Biden is blue. Trump is overperforming in Democrat precincts as indicated by the first 30% which actually has a linear trend in Trump's favor if you look at it. Suddenly that trend breaks down as the precinct gets more Republican, and the slope of the trend line gets bent like a transistor, as Dr. Shiva suggested.
The reason is simple: it's easier to steal votes in heavily Republican districts from Trump and have it go unnoticed. If a precinct normally goes 70/30 Republican but it only ends up 60/40 for Trump then no one really notices, Trump still wins the county he was expected to but Biden gets the phantom votes added to his statewide total.
Sorry, friend. Not trying to be an ass. But you're not understanding what Shiva's graph ACTUALLY shows.
I'm aware that Trump is red.
I'm saying that because of the way Shiva chose to plot this data, by subtracting Trump's % of precinct strength from his % of ticket splitter strength, it SHOULD yield a negative number in heavy Trump districts because in heavy R districts, Republicans outnumber moderate ticket splitters.
Example: In a 90% R precinct, let's say Trump gets 80% of ticket splitters to Biden's 20%. I think you would agree that's an impressive number, right? Trump ticket splitters are, by definition, moderate voters who do not cast an exclusively Republican ballot, or they are Democrats who voted for Trump. But in a heavy Republican district, there won't be very many Democrats in the first place, right? So most will just be moderates, who don't necessarily agree with their neighbors politically. So Trump getting 80% of moderates in a heavy R precinct is pretty amazing, I'd even argue it's highly unlikely. But I'll give it to you.
So under Shiva's math, we take 80% and subtract the 90% precinct strength, which yields -10 (80%-90%=10).
We then plot that on the chart at (90, -10)
That's the lower right, WELL below the x-axis.
And that's WITH Trump overperforming among moderate ticket splitters.
Hope that helps.
Honestly, there's no need to subtract the precinct strength. I think that's where Shiva got confused. If he hadn't done that, I think the resulting graph would show my point more clearly: a flat line (as expected), except on the left, which would slope sharply down to the left, indicating something odd happening in Democrat precincts.
Check my math, you'll see I'm right.
These aren't ticket splitters, in Dr. shiva's dataset. These are precincts with x% of Republican voters. As the precinct gets more Republican (after the 20% or 30% mark), Trump is getting less of the overall vote for that precinct, less than the percentage of Republican voters. Go watch his video again he explains it very well.
No, friend. That's incorrect.
At 20:36 he lays out the exact equation. Here it is, verbatim from his video:
X-axis: Republican Straight Party (RSP) Vote percentages %
Y-axis: Difference of %Trump Individual Candidate Votes minus %RSP votes
It's right here if you want to verify: https://youtu.be/Ztu5Y5obWPk?t=1236
Michigan voters have the option of voting straight ticket, either for every Republican or each Democrat on the ballot.
Ok, so they aren't necessarily registered Republicans but they have clearly indicated their desire to vote Republican by marking straight party on the ballot. What are the odds that as a precinct got more straight Republican that the percentage of Trump votes also dropped precipitously??? These aren't small amounts in Shiva's analysis, these are 20 and 30% differentials between Republican straight party and Trump. That's insane, buddy. Indicative of fraud. Especially since the linear weighting kicks in at precisely 20% or 30%. Sorry, the President is the most important ticket. This is fraud.
I think it's insane that Michigan allows you to mark straight party and also individual races. It means these ballots can be tampered with after the fact. Straight party selection voting should be outlawed.
There is no data here indicating that the percentage of Trump votes dropped in strong R precincts.
The chart is measuring which way ticket splitters (moderates and squishy R's and D's) voted compared to the relative strength of the district.
In other words, moderate voters are typically 50-50 voters, they could go either way. Maybe they break a little for Trump in his stronger areas, but do you really think they would break STRONGER than hardcore Republicans would? Of course not. Yet that's what you're saying you expect to see: Trump getting just as many moderate voters as he got R's in a heavy R district. It's silly.
So let's do another more extreme example: 1,000 votes cast in a precinct.
995 are hardcore Republicans, and vote straight party tickets. That's 995 votes for Donald Trump.
The remaining 5 voters are the precinct's only moderates, and 4 of them split their ticket but vote Trump. The last guy splits his ticket but votes for Biden. That's 80% for Trump among ticket splitters.
Using Shiva's math: % Trump Ticket Splitters - % Trump precinct strength = Y-axis plot.
% Trump ticket splitters = 80
% Trump precinct strength = 99.5
80 - 99.5 = -19.5
This just means that Trump's percentage of votes among ticket splitters was 19.5% below his percentage of hardcore voters. There's NOTHING unusual about that.
Trump actually WAY outperformed Biden 999 votes to 1, but we're measuring how he performed among ticket splitting voters compared to hardcore voters.
You tracking with me yet?
I hope so, because the ultimate point I am urging you to take note of is the relatively flat precincts on the left, which are precincts in heavy Democrat areas.
In heavy D areas, moderates, being 50-50, so they are likely to vote for Trump by a margin greater than the hardcore Democrats are voting for him, so Trump's plot there SHOULD be pretty stinking high. But it's not. It's BENT. Downward.
WTF. Poll workers in those Dem precincts are the ones that screwed us, probably with ballot stuffing for Biden, not some B.S. algorithm in R precincts.
No, that is not the Y axis on Dr. Shiva's chart. He's not even looking at so-called ticket splitters directly. He's looking at (Trump - Republican Straight Party) as a vote %. You have to subtract the Republican straight party votes from Trump's totals and then measure that percentage as a a fraction of the overall vote for that district.
It shouldn't be negative unless Trump is unpopular among Republican voters (not the case in the least).
That would indicate that people are going into the polls, and ticking both the Straight Party Republican box and Pedo Joe Biden in massive numbers, to the tune of 20-30% in some cases. It ain't happening. This is software fraud.
as you go to the right, you are getting into predominantly more party affiliated districts. so, the farther right you go, that district would be more republican (red) or DEM (blue). obviously one would agree that the more politically affiliated a district is, the more likely that district would vote for the president. there is zero chance that the more republican a district got ( especially rep) that they wouldn't vote for trump. this is evident as you go more right, you go up or at least have a higher average line.
see how in less rep districts, they are less likely to vote trump , right? hence why they are below 0% this graph literally states that the more republican a district is, the more they voted for Biden, highly unlikely- and the opposite, the more democratic district, the less likely they were to vote bide. makes no common sense
OK buddy, seriously, listen carefully.
Imagine in a district with 1000 voters, 990 of them vote a Republican straight ticket.
This would score 99% on the x-axis (far right) on the chart.
Now imagine the last 10 voters are ALL moderate, undecided voters who split their ticket for both Republicans AND Democrats. Let's say 8 vote for Trump and a Democrat congressman, and the other 2 vote for Joe Biden and a Republican congressman. That's what makes them "ticket splitters."
That means among ticket splitters, Trump got 80% of them. Wow. This is great. Moderates breaking HEAVY for Trump right?
You'd argue that Trump is overperforming among moderates and this plot should be on the upper right side of the graph.
That's wrong.
Because Dr. Shiva takes the precinct score (99%) and subtracts it from the ticket splitter score (80%) which then yields a score of -18 (80-99=-18) and he plots it on the y-axis at the 99 percent mark (lower right!).
This means that even in districts where moderates break heavily for Trump, unless they break at a greater rate than Trump's straight ticket hard core voters (right side of the x-axis), they will ALWAYS score below the trendline. This also means the opposite is true, in weak Republican areas (left side of the x-axis), moderates breaking heavily (or even just barely better than the district strength for Trump) will still outperform Trump's straight ticket voters (meaning ABOVE the 0% x-axis line).
That means that the downward left to right slope of Dr. Shiva's graph was PERFECTLY NORMAL, but his analysis is wrong.
But his graph ALSO shows something quite strange: a flattening of the trend line in heavy Democrat precincts.
And THAT's where I'm saying the fraud occurred, mostly likely through plain old ballot stuffing.