What I am about to say may jeopardize my account, but it REALLY has to be said NOW, because we are wasting valuable time... And I decided to have the courage to post this even though I know how excited we all were (me too, I initially)... Well, here it is...
Shiva's premise is sadly unsalvageably flawed.
His premise:
- If 80% of a district's Straight Party Votes are for the Republicans...
- ... Then 80% of a district's Individual Candidate (party breakaway / party protest) Votes must be for the Republican Presidential nominee.
Anyone who thinks logically realizes quickly how wrong that premise is, on so many levels.
Think again: Most people in America vote straight party; either all D or all R. What is the only reason for using individual choices? Well, it's what you do if you like your party but STRONGLY dislike certain candidates within your own party and would RATHER prefer the opposing party's candidate in that spot. An "Individual Candidate Choice" ballot is a PROTEST VOTE AGAINST (aspects of) YOUR OWN PARTY.
So Shiva assumes that anyone who BREAKS AWAY from the Republican Party (by refusing to vote Straight Party) MUST still vote for Trump (R) on their "Individual Choice" ballot. He does not consider WHATSOEVER that the breakaway vote COULD be against Trump.
In other words: Contrary to his claims, there is actually NO correlation between Straight Party Vote % for Party X and Individual Candidate Vote % for Party X's President. The whole purpose of going Individual Candidate is to BREAK AWAY from your party in some way. Therefore, trying to plot that difference on a line "as if it means anything" is Democrat-level-retardation (sad to say it).
For him to be right, the following statement would HAVE TO be true (which it's clearly not): "People who break away from the Straight Republican Party vote, do that because they want to vote FOR Trump for President BUT want to pick Democrats for House/Senate". As batshit insane as that statement sounds, that's EXACTLY what would have to be true for Shiva to be correct in his claim that "The breakaway vote (individual choice vote) percentage for President Trump MUST match the overall percentage of Straight Party Republican votes".
Next up is the other, fatal flaw which other people on TheDonald have already explained (and I individually created identical graphs earlier today before even seeing other people's posts, because I was stunned by the silliness of Shiva's math too). The problem is: Shiva's X axis is "Republican straight party votes / All straight party votes". And his Y axis is "(Trump individual choice votes / All individual choice votes) MINUS the X axis (the Republican percentage of all Straight Party votes)". This does not plot what he thinks it does. He thinks it shows voter fraud. What it actually shows is basic mathematical reality: The more Republicans are in a district, the less Trump votes (as a pile-percentage) will be in the pile of voters that "broke away from their party", because a bigger share of the breakaway vote pile will be republican defectors in a heavily republican area. And the inverse is true too; the vast majority of the breakaway "individual choice" pile in a Democrat area will be overwhelmingly for Trump. In fact if there are zero Republicans in an area, then 100% of all "individual choice" (breakaway) votes will be for Trump; and the exact opposite happens in Republican areas. Therefore we end up with the diagonal line showing declining Trump share aka "bigger pile of Republican party breakaway votes" in heavily Republican areas, in Shiva's video.
In fact, using Shiva's exact math, but changing all the variables, swapping Trump to Biden and swapping Democrats to Republicans, and vice versa, would draw the EXACT SAME line graphs, showing "diagonally falling line of Biden support the more Democrat an area is"! Because Shiva is misunderstanding and misrepresenting very flawed math.
Again: The more a party controls a certain area, the more of the "breakaway vote pile" will be defectors from THAT party. You will ALWAYS get a diagonal line trending downwards, for BOTH parties!
A lot of people are misunderstanding Shiva's misleading, incorrect graphs, and think they mean "The more republican an area is, the less Trump votes". That is NOT what they mean. They are simply showing "The more republican an area is, the more of the overall party protest breakaway vote pile will be against Trump/the Republicsn party". Well DUH!
To clarify even more to help people understand: If you, for example, say that a FLAT 4% of all Republicans everywhere in America dislike Trump, and those people all do a "breakaway Individual Choice vote" against him, then the MORE Republicans are in a district, the MORE Anti-Trump breakaway votes will exist in the Individual Candidate pile. For example, 4% of 1,000 republicans is 40 anti-Trump "protest votes", but 4% of 1,000,000 republicans is 40,000 anti-Trump "protest votes". So, again: The MORE REPUBLICAN a district is, the LESS votes/support Trump will have in the breakaway "Individual Candidates" ballot pile. Therefore, as the Straight Republican Party % goes up, the Trump Individual Choice % goes down.
This problem is THE most severely flawed premise of Shiva's, because he assumes that "party breakaway" vote percentage FOR PRESIDENT must exactly match "straight party" vote PARTY percentage. But reality is the inverse of that. The whole point of breaking away is that you're voting AGAINST portions of the party. This diagonal downwards line holds true even if we assume that only 3-6% of republicans left Trump.
The final flaw of Shiva's presentation is that he tried to draw trendlines that went "horizontal until X=20%, and then diagonal downwards after 20%" on his plot, but that was not true at all. The data plot points were all clearly diagonal even from the beginning, as a logical result of the flawed math that I and others have described. In ONE of his graphs there seems to be a flatter start, but that's very easily explained by "sticky Democrat" demographics of that particular city. A flat start (the left hand side) in his graph means "there is an extremely low proportion of Straight Republican votes in that area (X axis = 0-20%), and there is also an extremely low amount of Democrats breaking away for Trump there; so the line is drawn as a flat line since breakaway votes from Democrats going for Trump are on the same low level as the low amount of Straight Republican votes in that city". Or, summarized more briefly: Any "slightly flat" starts on the graphs just mean that Democrats in heavily blue areas are refusing to cross over to Trump, and therefore the breakaway votes from Democrats switching to Trump are as low as the Straight Republican votes there.
If you've read this far, here's actually the biggest bombshell: It would actually have been weird and indicative of fraud if the lines WEREN'T going diagonally downwards for Trump. If they had been straight horizontal lines (or even upwards trajectory), as Shiva believes they were meant to be, then THAT would have shown election fraud BY Trump to swap "Republican defector / protest votes" in Republican areas back to himself to maintain a flat/positive line... So thank GOD that Shiva's line WASN'T a flat horizontal line! It would have been very clear evidence of election fraud by OUR side! (As I said earlier, a STRAIGHT line would require that the people who are defecting away from "Straight Republican" do so because they want to vote TRUMP for President AND DEMOCRATS for House/Senate, which would be batshit insane!)
A really big "red flag" during Shiva's movie-length presentation was that he could never clearly elucidate exactly what he was showing. He confused himself many times. He referred to his X/Y axes inconsistently throughout the video and struggled to speak clearly and confidently. It looked as if he was going over everything himself and still trying to understand what exactly he was looking at. Heck, he rambled so much that the two guests only got about 10 words in total in the 80 minute long presentation. That is because the premise and math was a mess. I am VERY sad to say it. Shiva means very well, but his math is fatally incorrect and is wasting our time that we SHOULD spend searching for actual evidence. :-(
I wanted to contact Shiva privately but he does not seem to list any contact email anywhere on his online presence (a little bit amusing for a man who claimed to have invented email, heh)...
Anyway, please carry on the amazing work everyone. The crowdsourced work being done here is incredible! I am excited to see what else we uncover in the coming days and weeks! I am as energized as I was in 2016 against Crooked Hillary! Let's do this! Let's burn down the establishment swamp!
Edit: Despite my best efforts to be abundantly clear and easy to follow (hence the length and repetition of my post), a few people are still having trouble with the math. So I put together a comment to attempt one more time to explain it: https://thedonald.win/p/11Q8XO2YxF/x/c/1BjFNXpNSu
And one more thing to keep in mind here: What I have said here is the actual truth, and it will come out no matter what some diehards want to believe. The question you have to ask yourself is; Do you want to fix this Shiva situation NOW? Or do you want to get hit by the shrapnel from that grenade when it explodes and his clownish math gets ridiculed? We really only have those two choices as a community. This is an absolute black and white situation without any shades of gray. His math is awfully broken and we have to decide how we proceed forward now. Do we continue hitching our horses, wasting time and getting ridiculed? Or do we get back on track and focus where our help IS needed?
Edit: Well, all I can say is that I tried. For those who were able to understand the math: Thank you, and remember to pass it on to those willing and able to understand. It pains me to see ignorance in our community. These are the most important times of our lives, and a large portion of the community is being willfully ignorant and wasting our time with Shiva's clownish maths. It's very painful to see.
You're wrong. If a Trump protest vote was occurring, it would have registered going across, as they show in the video. They go over this. The declining slope is the indication of a compensatory fraud measure. It's like the data analysist says, "it's too perfect." We don't know what other people are voting for, nobody knows the results. For that line to exist, it's like, very very mysteriously, as straight party votes for Trump came in then, at the same time, juuuuust the right amount of one candidate (for biden) ballots came in. It's too perfect. You're wrong. The "diagonal," the downward slope, should simply not be there. The lead was compensated against Trump in real time.
Please stop spreading fake news. You are not helping us win. You are helping waste our time and are literally regurgitating the video verbatim. And just proving that wishful upvotes don't have any clue. Please read the post. I already went over everything. Literally everything; all the math and every premise; of Shiva's video are incorrect. And there is widespread misunderstanding of those incorrectly calculated graphs. If the slope was going straight across horizontally (as Shiva wrongly believes it should) then that would show election fraud BY Trump instead. And we certainly don't want that. I even wondered if Shiva is a Democrat operative. At least one of his guests in the video was involved in trying to help Democrat Al Gore win against Bush in the 2003 recount.
But to clarify yet again:
Around 3-6% of Republicans are estimated to be never trumpers who find him crass and don't want him as President.
Now let's do some math...
Assume that we always have a city of 100,000 voters, just to make the percentages easier to visualize mentally.
3% of Republicans universally defect and vote against Trump (this is lowballing it using the lowest estimate).
Let's say that 4% of Democrats vote for Trump everywhere.
Alright. Let's go:
In a Dem district with 15% Republicans (15,000 people) and 85% Democrats (85,000 people), you would have 0.03 * 15000 = 450 Republicans voting "individual choice: No Trump, but Yes other Republicans in House/Senate", and you would have 0.04 * 85000 = 3400 Democrat defectors using "individual choice: Trump". This means that Trump got 3400 / (3400 + 450) = 88% of the "individual choice" vote in a heavy dem district.
In a Mixed district with 45% Republicans (45,000 people) and 65% Democrats (65,000 people), you would have 0.03 * 45000 = 1350 Republicans voting "no Trump", and 0.04 * 65000 = 2600 Democrats voting "yes Trump". This means that Trump got 2600 / (2600 + 1350) = 66% of the "individual choice" vote in a mixed district.
In a heavy Rep district with 85% Republicans (85,000 people) and 15% Democrats (15,000 people), you would have 0.03 * 85000 = 2550 Republicans voting "no Trump", and 0.04 * 15000 = 600 Democrats voting "yes Trump". This means that Trump got 600 / (600 + 2550) = 19% of the "individual choice" vote in a mixed district.
This is the exact diagonal downward line that Shiva got because he and the others do not understand math and statistics.
Again, as I tried to explain in detail: Even though you have a Flat 3% Never Trumper defectors everywhere, Trump wins less and less of the "party protest" votes the more Republican a district is, because more and more of the total number of defectors are Republicans!
This kind of plot line shows up on both parties identically. Shiva and the two other dudes don't understand what they are doing and they never bothered using the same clownish formula on the Democrats, but it's explained in my post and in all of my followup comments. The shortest summary version possible is: The more one party controls a district, the more of the defectors will be from THAT party. Shiva then cluelessly takes the percentage of defectors and plots that as a crazy Y axis. Well DUH of course the deficit grows when you get into more and more partisan districts where all defectors are from our/their party. So yes, both parties have downward diagonal lines. The more a party owns a district, the larger their percentage of the total defector vote, hence why Trump/Biden BOTH trend downwards completely diagonally using Shiva's clownish math. And if they DIDN'T have diagonal downward lines, then THAT would be fraud.
The more one party controls a district, the more of the defectors (numerically) will be FROM THAT party, even though the percentage of defectors (3%) never changes.
Or, in Common Core terms:
3% of 100 Republicans is 3 defectors.
3% of 1,000,000 Republicans is 30,000 defectors.
So yeah, the more Republican a district is the more defectors. And Shiva the clueless time waster counts the number of defectors.
The problem is that the massive downward swing you are seeing, with a "-30% swing against Trump", is a fabricated lie by Shiva's clownish math. His Y axis is NOT the total vote percentage lost by Trump. It is the protest vote percentage AGAINST Trump, MINUS the Republican Straight Party percent. For example, if his dataset was 4 Individual Choice ballots (3 never Trumpers, and 1 Dem that voted for Trump), then Trump only got 25% of those Individual Choice ballots. He then does 25% - RSP%. The republican straight party percent, such as 50% of all straight party votes being for Republicans. The result then is 25-50 = - 25%.
So his line would say (and does indeed say): Three people defected against Trump. Trump got -25% on the Y axis...
Shiva's biggest mistake, and the reason for all of his bad math, is that he believes that the Republican defector votes should ALL be FOR Trump BUT AGAINST the OTHER REPUBLICANS, which is the ONLY way to get a flat horizontal line all the way with his clownish formula.
But it actually gets even worse for Shiva's unfixably broken math: He actually subtracts the "straight republican vote %" from the share % of individual choice votes that Trump got, and then demands that the line should be straight, which is completely nonsensical and makes the assumption that every Republican who decides to cast an individual choice ballot (instead of a straight party vote) will ALWAYS be putting Trump as President and DEMOCRATS as House/Senate, for Shiva to be correct. Hint: He is not correct.
Again: The more Republican a district is, the more Republican voters exist overall in the voterbase, and therefore most of the defectors in the district (in raw numbers) will be Republicans, meaning that less and less of the "party protest pile" will be "Trump votes". So the more Republican a district is, the less pro-Trump votes in the PROTEST PILE, as a natural consequence of the population size. Shiva's entire mistake is that he looks at that PROTEST pile and DEMANDS that the protest pile's PRESIDENT TRUMP votes should always match the straight party republican votes. But the ONLY way for that to be true would be if the protest is "I refuse to vote with the straight republican ticket method; I will now vote TRUMP for President and DEMOCRATS for House/senate". Which would obviously be insane and is not how people think. Therefore we can conclude that defectors from the Republican party are voting against Trump, but still FOR House/Senate Republicans. It really blows my mind that Shiva doesn't understand this. But then again, he was so confused by his own graphs in the video, that he probably never analyzed it deeply enough to think about it.
If people still don't understand what I've explained, then Common Core has failed us. But I see that many people understand already.
I am sure that even Shiva will understand after this is pointed out to him. He is clearly an intelligent man. Just bad at math, like the other two hobbyists in that video. All three of them are lying about being data analysts, since they didn't even do basic data analysis requirements such as fitting the chart to a linear regression, or determining the colinearity or variance. Or even testing it for the opponent (where their shitty, clownish formula would just show the exact same, natural result, due to how these 3 clowns designed their idiotic formula).
Their formula and graphs LITERALLY ONLY tell us "The more Republicans in a district, the more Republicans in the district's party-dissent pile." Which is a total DUH, OF COURSE, and is the EXPECTED result.
I knew very well that the facts would upset a lot of people, but facts do not care about your feelings. It's very embarrassing that some people in our community are acting this way.
Alright, now let's get back to real leads and investigations! :-)
I understand what you're saying, I'll stop pushing it, but showing, 88-15= 77, 66-45= 21, 19-85= -66, which are not good estimates according to the graph, would have given better clarity to us dimwits in your "overexplained post."