There really isn't one? TDW can use individual metrics like how many daily page views it's getting, what percentage of users have been active in the past day/week/month, that sort of thing. It doesn't need to compete with other sites.
You don't need the whole datapoints, or the exact ones for that matter. Find any exponential regression online, this one for example: http://www.xuru.org/rt/ExpR.asp#CopyPaste
Set a few points. I use these:
0 108
60 50
120 27
180 18
0 means Dec 3 of 2019, +60 is Feb 3 of 2020 and so on.
108 is about what I can see in the Y axys for that day, 50 for Feb 3, and so on.
Click calculate in that site, or any similar site, and it will give you a formula and a R^2 value. If your R^2 > 0.9 your formula is good enough. In this case is 0.98 which is very good. (you can't get greater than 1)
The formula you get in this case is y = 104.23 * e^-0.0016x
Put in the x the day you want, in our case day +335 (aprox nov 3) and that is.
However this relies on the model not changing from today to Nov 3. As time advance and election day draws near people will take more interest on politics and TDW will become more popular than the current trending.
If I have to guess, I will put it somewhere in between 2000 and 2500.
I'm still pretty sure Alexa is sketchy. It bases ratings on who installs the Alexa Toolbar (basically adware.)
There really isn't one? TDW can use individual metrics like how many daily page views it's getting, what percentage of users have been active in the past day/week/month, that sort of thing. It doesn't need to compete with other sites.
You can just copypaste the csv text into https://pastebin.com/
I'm signing off for the day, but tomorrow I will make you something more precise than my other comment.
Here is what I got:
https://imgur.com/a/LHKZjqY
By election day rank 2209.
I took away the first 25 days of the data because they were too unstable and wonky.
And if anybody care about the details, here is the complete summary:
Call: lm(formula = log(dataRead$rank) ~ dataRead$day) Residuals: Min 1Q Median 3Q Max -0.09139 -0.06494 -0.03192 0.02130 0.31735 Coefficients: Estimate Std. Error t value Pr(>|t|)
(Intercept) 11.4913622 0.0172111 667.67 <2e-16 *** dataRead$day -0.0114873 0.0001332 -86.22 <2e-16 ***
Signif. codes: 0 '' 0.001 '' 0.01 '' 0.05 '.' 0.1 ‘ ’ 1 Residual standard error: 0.09756 on 184 degrees of freedom Multiple R-squared: 0.9758, Adjusted R-squared: 0.9757 F-statistic: 7434 on 1 and 184 DF, p-value: < 2.2e-16
About 3000 by election day.
You don't need the whole datapoints, or the exact ones for that matter. Find any exponential regression online, this one for example: http://www.xuru.org/rt/ExpR.asp#CopyPaste
Set a few points. I use these:
0 means Dec 3 of 2019, +60 is Feb 3 of 2020 and so on.
108 is about what I can see in the Y axys for that day, 50 for Feb 3, and so on.
Click calculate in that site, or any similar site, and it will give you a formula and a R^2 value. If your R^2 > 0.9 your formula is good enough. In this case is 0.98 which is very good. (you can't get greater than 1)
The formula you get in this case is y = 104.23 * e^-0.0016x
Put in the x the day you want, in our case day +335 (aprox nov 3) and that is.
However this relies on the model not changing from today to Nov 3. As time advance and election day draws near people will take more interest on politics and TDW will become more popular than the current trending.
If I have to guess, I will put it somewhere in between 2000 and 2500.
this is a great site, the creators have done a hell of a job.