I agree man and I tried to explain that to him, because at first I didn't even know that I got the timestamp wrong...I used the assumption that they were rounded to the nearest 3-digits...I followed up with a massive array of all possible 3-digit rounded values up to a (k+3)-th digit starting from 0.XX(X-1)4444...445 all the way up to 0.XXX4444...444 with (k+3) digit increments, and I multiplied that array against the raw vote total, which crucially has a (k+3) number of sig figs than 3, and I sorted the array in both ascending and descending with respect to the fractional part of the new partisan totals...I then compared the answers with the largest fractional part, which would look something like this A.XX(X-1)999...9999 to the answers with the smallest fractional part which looks something like this A.XXX000...0000 to find the best fit integer-ratio pair...if they didn't line up I used a comparator to find the terms that did line up to the (k)th digit...crucially speaking ALL of them ended up pointing at a simpler ratio, and we know numbers were always increasing in the NYT edison timeseries, as can be noted by the irregular time intervals.
I am going to check out your spreadsheet after I get home today because I am curious about your feedback...he didn't really get under my skin I just hadn't slept in 3 days up to that point...my point here is to only help.
ratios are the simplest mathematical tools to understand...so is the machinations of your mind apparently you moron
exactly...like how there were a few explanations for the switching of 5500 Trump votes to Biden in two counties in MI...and a few explanations to the switching of 93k votes in a county in VA on election night from - you guessed it - Trump to Biden...or the magical 450k votes across the country with no downballot votes, which is an inexplicable figure, all going straight to Biden...the data isn't noise, it is granular because the significant figures of the raw vote totals is high enough that greater resolution on regressing the non truncated values of the ratios can be achieved well beyond the 3rd digit place, and 16k has a 4-digit magnitude...Edison wont mutate the inputs algorithmically beyond low resolution time sampling for non ratios, and for high resolution time sampling for truncated ratios, which shifts liability for fraud beyond their purview...Dominion can claim ignorance of any fraud by blaming any anomalies on human operator error or cyberattacks, due to their lack of security protocols they were certain to place blame on...the door is wide open for a third party physical intrusion into the data feed...Hammer and Scorecard are real programs with real intrusive track records in clandestine operations...I am not saying that this is what is indicated here, I am saying that this is the granular information that we need in order to find out what really happened here.
lol, like how epstein killed himself
third party gains...like a voting shell game
Damnit I did it again...lol...I meant 11:08 PM
I checked it with simple integer-based math after fixing the 3-digit ratio imprecision...there is no other way to explain the simultaneous increases in vote totals and dilutions of the percentages that lead to wide variances in the higher resolution data, like the decreases that we see here...they don't happen everywhere but they do happen
working on that
already onto that part of the task
sorry dude...I have been a little busy with a few other projects to get this program translated...in a day or so I might have some free time to get that done, and thanks for the reminder
I see
you are an idiot
saying the same thing over and over again isn't going to change facts man
discovery bro...it is called discovery...once you have that, legally speaking, the next step is where the interesting stuff happens
I used revised ratios in my analysis...you can find it here: https://gofile.io/d/64wtx8
You are spot on about the appearance even in the truncated ratios!
trust me we are working on it...it will be YUGE
they already have it
this was from the raw data file
wtf...it said it in the source data
quick and dirty
Here is the correction:
The numbers highlighted in the marked data in the second screenshot is clear as day...I don't know what else to say but you are wrong about your conclusion, other than being right about the time stamp that you were referring to...I have the same analysis with multiple states and multiple datasets and I am currently cooperating with a team of independent researchers to verify everything...nothing has changed and I stand by my conclusion, with a caveat that I made a simple mistake and referenced the wrong timestamp in the last two screenshots
Dude calm down.
The answer is simple: You pointed out correctly that the last two screenshots are referencing a timestamp other than the one being highlighted in the second, third, and fourth screenshots.
The second, third, and fourth screenshots highlight the anomaly in the regressed dataset. The last two were an attempt to match the original json timestamps to the place I discovered the anomaly...as you pointed out they are the incorrect timestamp...that is all. The data is correct, but the timestamps weren't included in my original program so I didn't match them up correctly.
The point is that the anomaly occurred, and you can see this in the second, third, and fourth screenshots, whereas the first screenshot just gives you an outline of the headers. Ignore the last two screenshots, they were just included to give a guesstimate to when the anomaly occurred, and I messed that up...it was a simple mistake and it wasn't really all that relevant.
I can give you all of the resources I used and you can generate the graphs yourself to find the same anomalies I found, I am currently updating my program to include timestamps.
I am not the best communicator so if anything that I said is confusing you that is understandable, I am just here to open some wavering eyes on the issue of fraud.
I just updated it and I am about to post it...started catching up on sleep last night so I'm a little lagging behind...thanks for the kind words
we have better data now
Here is a link to the data that I compiled for every state and DC:
Nice eye...I put the arrow in the wrong place...the real change was a few frames back...I took that screenshot while I was half asleep...the arrow is pointing to the wrong value in the JSON that is all...my mistake.
These are real number though...all I did to calculate the new ratios was find the best-fit ratio for a denominator containing the raw vote total that produces an integer numerator raw partisan total, of which matches the truncated 3-digit ratios provided...the raw vote total is almost always unique enough that the real ratios can be derived within a relatively high precision, less than the number of significant figures in the raw vote total.
might be a 25 day old account...I and my analytical skills are neither 25 days old or 25 years old