Analysis

25 September 2020 #BetweenTheLilneDotVote Analysis ANNOUNCING our 2nd Tranche of Data! In a word, as it stands (more on that in a bit) our BTL data is 100% the opposite of all the published polling you see in the news. If our data is right, everyone else is wrong.

2) Before we dive into the data above, allow me to share, again, our 1st tranche of data. I'll briefly explain the difference between the two below.

3) The BTL Algorithm is the difference. Our 2nd tranche benefits from it, our 1st did not. Here's the quick story. By July 29, we had collected enough data that I felt it was important to share. There was, however, a severe problem with the data. It was sample-bias skewed.


4) Our data was SO Trump leaning, that it was simply NOT possible to believe in it. My theory was that, as I am such an ardent Trump supporter, the people participating in our poll were influenced by and accessible to me, personally. So, obviously, a massive Trump bias, right?


5) Still, I couldn't throw my data away. Not only was it too hard fought in order to gain, I refused to believe it didn't have value. So, I worked and worked. At first, I broke my brain over the math, and I got nowhere. Here's what I did...


6) I used my intuition to explode Biden support and reduce Trump support with the following goal. What I wanted was an equal influence from all 3 groups, Democrats, Independents, and Republicans. The logic is simple.


7) If you have too many Republicans participating in your poll, as we at BTL do, you have to INCREASE the impact of those Democrats who have been so kind as to offer their votes. Hoping to give their voice an equal impact on our poll's outcomes, I simply increased its count.


8) The result was the chart you see above. I did NOT claim it to have the slightest scientific basis, and I don't claim that, for it, now. It was just the data I had, and I did the best with it that I could. What arose, though, had a serious impact we must turn to now.


9) The BTL model, our data model, couldn't be more simple. We work with our data to effect equality among three groups, Democrats, Independents, and Republicans. Simply 33% for each. We intend, in the future, to alter this model, once we build our own model of America's voters.


10) I hope you'll work with me on this. We're starting out with simple equality of impact. To do this, we have to expand the impact of those whose numbers we have the least. Since we have the least Democrats, their vote must have the largest increase in impact.


11) In our 1st tranche, I simply did that by instinct. What that means is this. I looked at the numbers and I - with no other mechanism than my own personal guess work - gave the most impact to Democrats, more impact to Independents, and I left the Republican count alone.


12) After publishing our 1st tranche, the game was on. What I knew I'd need was an algorithm. I started out with nothing, and built the algorithm from scratch. It's taken me these many weeks to do so. My algorithm works. It is proprietary, so I won't tell you much about it.


13) I am both happy and honored to tell you this much. What I did, instinctively, with the 1st tranche of data, my algorithm now performs with pure mathematical simplicity over our 2nd tranche, and will do so for all our work moving forward. It equalizes impact.


14) The algorithm is sensitive and easily can accommodate various different models. We can easily change the parameters for the model, off its current equal 33% basis. So, if we come to believe there are more Democrats, fewer Republicans, etc., we can change it like so...


15) Imagine our data informs us that America really does have 39% Democrats, 42% Independents, and only 19% Republicans, we can easily shift our parameters to reflect that count. One thing that we PROMISE, is that our model will ALWAYS be published.


16) And by "published," I do NOT mean in the fine print somewhere that you have to be dedicated to find. Right now, the model is 33% each for Ds, Is, and Rs. No, that is NOT our actual sample. More about that below. But it IS the actual impact on our projections.


17) I know. Promises are cheap. On that very note, I report - not promise - that our data is presented with the actuality of equalized impact by the three political identifications, and we have the algorithm to ensure it.


18) Now we return to our 2nd tranche, published for the first time above. Here's that chart, again. A question jumps to mind. If we've equalized the impact of our smallest contingent, Democrats, and our next smallest, Independents, then why such a Trump dominance?

19) Truth is, I was SHOCKED when our algorithm spat out data that looked almost exactly as my instinctively processed data had. I truly expected the two campaigns to be far closer together. I was, in fact, disappointed when Biden's numbers showed up so low.


20) A theory was born. Republicans are virtually united behind Trump. That's not news. In our data at BTL, Independents are breaking for Trump about 8 out of 10. If we're reflective of the nation's Independents, that should be - but won't be - news.


21) Imagine. What if our data is correct? I'm not saying it is. I'm only asking the hypothetical. What if our data is right, as to the break in the Independents' vote? If 8 out of 10 Independents break for Trump, he wins in a landslide. 80%? Can that be possible?


22) Let's talk about the art of the possible in polling for a moment. It explodes my brain every single time a reporter says: This X percentage of Americans feel thusly... No. This X percentage of those polled here, feel thusly... Do you see the difference?


23) A poll is a guess. If we've sampled our people correctly, and if their opinions as stated reflect all of America correctly, then, our poll indicates that this percentage of Americans feel like so. You don't have to say those words. You do have to have that humility.

Interim. I need a break for this reason. My data truly does come in the context of inspiration from 2016's Dornsife Poll. I never fail to give them recognition. We'll be making that tie next, and I have to meditate on that for a bit. Back soon...


24) An historical approach always grounds us. Why did Dornsife correctly call the 2012, and 2016 elections? I believe the answer is to found in the method they used, back then, but have now rejected. Back then, they used ONLY 3 questions.


23) Their 3 questions were: 1) Will you vote? 2) For whom? 3) Who will win? Forgetting their sample - it was flawless back then! - and focusing solely on these 3 questions, you see perfect genius at play. I will never forget my emotion when I discovered those questions in 2016.


24) That Dornsife has abandoned its previous format breaks my heart. In 2012 and 2016, they were by far the strongest polling outfit in the world. Those 3 questions were the very basis of that domination.


25) I'm about to show 5 of their charts. You can't know about these charts unless you dive deeply into the content they provide. They're hidden in plain view. They are: 1) Democrats 2) Lean Democratic 3) Independents (No party affiliation) 4) Lean Republican 5) Republican

26) These 5 charts answer a single question: Who, in your state, do you think people will vote for? That's a question! It follows perfectly from the Flynn Doctrine: Who the people think will win, will win. If you follow the Flynn Doctrine, then predicting winners is the job.


27) Dornsife's pure Democrats poll...

28) Dornsife's Lean Democrat poll:

29) Dornsife's Independent/No Party Affiliation poll: (This the one that matters!)

30) Here's the Dornsife's Lean Republican numbers:

31) And here's Dornsife's break out for those who identify as Republicans:

32) The slightest examination reveals that even at Dornsife, they seek only the truth. I publish my advocacy, boldly. In polling, however, truth alone is the goal. If American Democrats have won the day, then we conservatives must bear the brunt.


32) If there's nothing else we ascertain from the above charts, we realize this. Men interpret charts of data, the same way their illusions inform.


33) But perhaps we may glean a bit more. Perhaps, somewhere in the data, the truth lies. That is my prayer. That is my prey. I hunt the truth and will not stop until I bag it.


34) Please head over to our site, now. We ask the same questions Dornsife used to ask, but add two of our own. It appears we're the only out fit out there employing this method, this year. You'll find our 5 questions here:

35) And you can count on our honesty, 100%, always. We state our position. We account for it. We fight the bias all human enterprises suffer. We seek the truth. Please join us!


Thread ends at #35.


September 2020 #BetweenTheLinesDotVote Analysis Dornsife Predicts Trump Victory! Okay, not really, but this chart - one of 8 now available - actually does lean Trump's way. Check it out.

2) I'll be short right now. Here's why. I have decided to extract all the data and show it in my own way. I do consider this important work or I'd not waste my time, I promise.


3) I think the tactic involved here is called an Information Dump. You've seen it in movies about law cases, when one side requests specific information and the other sends truck loads of boxes inside of which is the information, but there's almost no way to find it.


4) As of right now, I can't tell you how many new charts Dornsife has added. Being me I'm kind of in data heaven, bopping about all over the place at their website. It's a good site, for real. Within the vast number of new data presented, I believe I've found what they're hiding.


5) I have been threatening to share my own data with you, and I absolutely WILL DO SO, tomorrow. I am MASSIVELY excited about that. There's a strange connection here. One of the charts I've found at Dornsife looks a great deal like my data. Wow!

6) There's so much this chart tells me. It reaffirms my strong suspicion that the current sample - whose data shows up just like this, but inverted - is heavily weighted with a Democrat leaning.


7) I admit, I am pleasantly surprised that Dornsife, in this Information Dump, has allowed such data to show up. I remain profoundly disappointed that they bury this data so deep that, well, I may be the one and only guy looking at it.


8) Speaking of being the only guy in the world to do something, I have a wooden sword story to share with you about this remarkable weapon. Look at the phenomenal craftsmanship and beauty. And I can tell you, it holds to your hand with power and grace.


9) I bought this incredible work of martial art around 2007 or so, for $125. It was advertised at the swordmaker's website and I was stunned the moment I saw it. It may well be the greatest bargain in my entire sword collection. You won't believe what I discovered later.


10) My Claymore is 1 of 2, and that's it. The other is owned by the sword maker. And, for that reason, she discontinued to offering. So, outside of herself, I am the one and only person to own this beauty, and always will be, since it's not on offer. How about that?


11) When my sword maker friend discussed this with me, she explained that she had some super duper extra phenomenal lathe thing, that was required for this kind of fine work. And, yeah, you guessed it, it broke and she couldn't justify the expense of replacing it.


12) Life is filled with heart breaks, many of them small, or even not noticed. I noticed that one. It made me proud to be the one person to pay the damned $125 for this work of art. It made me sad that the means of production was no longer under the control of its maker.


13) There is truth within the data, friends. Precisely as I found and to this day - I just took those shots right now - own that wonderful sword, I will find the truth within the data, even if I'm the only guy in the world to do so.


14) I'm going to go make some espresso now, and then I'll come back and chart out Dornsife's strongest, best chart. This will be fun!


Thread pauses temporarily at #14.


P.S. It's amazing what you learn WHILE writing up a Twitter thread. My first post references 8 charts. While working on the thread, I realized there are MANY more than just 8. Thus, data dump. I really am in data heaven right now!

Interim Post... So, here I am innocently working on my data, and what do I find? The chart I posted above was of the ENTIRE sample, and it gently calls for Trump. Well...check this out! This is the Independents!!! For my own call, It won't be Ds or Rs, but Is who determine 2020.

New images, but no lines drawn... I'm working with my tech people. My printer/scanner is, for some inscrutable reason, off line. So, I can't draw my lines. Here, however, is the data in my Excel format. It's an amazing chart. Biden and Trump will follow.

15) After many technical challenges faced - and some NOT yet overcome - I'm ready to finish up today's work. Technology is amazing, and I will learn to use it. Boy do I struggle, as you'll see. Let's return to this chart. Within the too close to call, still Trump leads. Amazing.

16) What exactly is this chart tracking? The Dornsife people ask, for this chart, just this one question: In your state, who do you think others will most vote for? That makes it essentially a "Who will win?" question. It sits easily within the Flynn Doctrine.

17) The Flynn Doctrine is: Who The People Think Will Win, Will Win Buried deep within their website, this is their poll that most strongly calls for Trump. Isn't that interesting?


18) Now we turn to technology. What I do is enter data into an Excel spreadsheet. The data is formatted into a lovely chart. I print out the chart, and with pen and straight edge, draw my support and resistance lines by hand. But, my printer's out today!


19) So, never to be stopped, I started learning how to use a snipping tool. As you'll see, I have a lot to learn. But, tools are tools, and we must learn them at need, right? This is the Biden form from the same data.

20) As we've seen elsewhere, the trend is clearly downward. And, when you place that over the Trump line, as you saw above, Trump is pulling away, and Biden's channel is clearly downward. How about that? You won't see that chart anywhere else.


21) And here, along with my terrible handwriting, made worse by using my mouse, is the Trump chart from the same data.

22) I tried, honest I did. But, I simply wasn't able to spell "Resistance" with my mouse. So, 1st & 2nd R are simply two interesting resistance lines. I know it looks like a bad student in 1st grade, but I was able to spell support.


23) Please consider that they at Dornsife have hidden this data deep within their recently published overflow of data. Were I a Democrat, I would 100% want to know about this. What if this is the truest data, as it was for Clinton2 in 2016? Winning requires access to truth.


24) As I said above, tomorrow, I'll be providing my own data. I did once before. What I did then was attempt to correct for my overwhelmingly stronger access to Trump supporters than Biden's. This is called sample bias. We at BTL work everyday to overcome this.


25) In the meantime, and as I always say, do help us out, and most especially bring your Biden supporting friends to our honest polling site:

Thread ends at #25.


23 September 2020 #BetweenTheLinesDotVote Analysis Dornsife's TWO COMPLETELY DIFFERENT DATA SETS Part 1 of 2 I'll offer some short thoughts below. Here, I charge you, just look at these two charts and let me know what you see, will you?


2) In case you're not accustomed to polling data chart analysis, I'll give you a few tips on what to look at. First, note that it is the people being polled, but that the first is a 7-day window, and the second a 14-day. In fact, looks a little bit more closely.


3) Here are the two different data sets for today. Note, the data employs the date 22 Sep. This is completely correct. You cannot look at all of yesterday's data until today. As you look, note especially these two points: (N: 2575) on the 7-day, and (N: 5440) on the 14-day.


4) As an untested innovation this year, Dornsife increased its sample size from roughly 3,100 in 2016, to roughly 6,000 here in 2020. Thus, their 7-day data - what they've been publishing until yesterday - has about half the number of participants, at 2,575. The 14-day has 5,440.


5) Let's pause, and let go of data for a moment. I never expect anyone else to be as obsessive about details as I am. I always imagine you, my dear readers, as suffering under the yeah-alright-so-what problem when you attempt to follow my impassioned data ravings.


6) If possible, later today, I'll be posting my second attack on this 2-data sets problem. If not, then I surely will tomorrow. I hope to show you with charts how significant these changes really are, and offer substantive reasons that may be motivating this change.


7) But, if you simply look at the images I've posted above you can see, obviously, that not only are we dealing now with TWO DATA sets, they are each absolutely different. Please take a moment to consider the far-reaching implications of that.


8) Two data sets. Hmm. Let's turn positive first. Okay, we can slice and dice the data and openly present our onion layers to the world, and let the world decide which makes more sense of the two for itself. Or, more likely, the world will just see the general trend and not care.


9) Here's the counter-case to that. Dornsife now posts on its face, the 14-day Window. Before it was the 7-day. Why switch? To me, the answer is obvious. I'll show you the simple math below.


10) Here is the simple math for their Insignificance Range. 7- day: 49.37 - 44.05 = 5.32 14-day: 48.65 - 44.71 = 3.94 Play with the math and you'll see, they have reduced their Insignificance range by about 27%. Wow!


11) If you can, stay with me. In 2016, all of Dornsife's data was represented as inside or outside a 95% Confidence Range. They employed the same method in 2012, when they rightly predicted Obama over Romney, when most other polls leaned toward Romney.


12) How significant is the flip? Before, they used a Confidence ratio. Now, they use an Insignificance ratio. In case you can't quite visualize the difference, here again is their 2016 presentation of the data.

13) Here's the explanation, in case you can't read it: "The gray band is a "95-percent confidence interval." Figures lying outside the gray band mean that we are at least 95% confident that the candidate with the highest percentage will win the popular vote."


14) Funny thing, the poll was actually wrong. That is, Clinton2 won the popular vote, whereas the poll called for Trump's popular vote victory. No matter. He won the electoral college - NOT an analysis the data Dornsife provides - and calling for him in any way at all was epic.


15) Last point for now. Go again, and look at the 7- vs the 14-day windows. Not only do they drop their Insignificance range, they smooth out the data, reducing movement further. I assure you, this image of Biden calmly above, Trump sadly below, will be something we return to.


Thread ends at #15.