Elon Musk modified the timeline to ensure his tweets came up first. Musk Response. UPDATE with Data to prove it.

[Uttamattamakin]

48 minutes ago, Bitter said:

I feel we're getting closer and closer to Godwin's Law and kind of surprised we haven't hit it yet.

Well we wouldn't want to get political which you know mentioning Obama isn't but mentioning certain other things that were already removed by a mod is so... Yeah no Godwins laws. here. 

 

 

[HenrySalayne]

9 minutes ago, wanderingfool2 said:

Under your logic of "deduction" I can claim that Obama must have been buffed;

No, you are talking about induction, the opposite.

 

10 minutes ago, wanderingfool2 said:

You literally don't know where to find the data...when it's a topic about Twitter, and I mention looking at impressions.  How about going to twitter, and the account and the date and lets see...looking at the impressions.

I honestly thought you were talking about a more sophisticated procedure than manually scrolling through a Twitter page and searching for these posts. Have you at least put all the data into a spreadsheet to make some graphs? Like this one?

 

Spoiler

Don't klick this spoiler, you can find it yourself.

Spoiler

I'm serious, this is easy to find.

Spoiler

Really?!? You literally don't know where to find the data?

Spoiler

Ok, fine, you hypocrite.

Source: socialblade.com

 

[wanderingfool2]

9 minutes ago, HenrySalayne said:

No, you are talking about induction, the opposite.

No, the induction part is assuming that if you are claiming "deduction" is how you specidied that I could use "deduction" to make the exact same claim.

i.e. You are claiming if p then q, where you show q (which is still 100% faulty logic).

The inductive part is that assuming "if p then q, showing q is deductive reasoning"

So I present that Person A was boosted, and give "proof" by showing the increase.  So it's still showing "deductive" reasoning how you define it as being proof.  So it uses inductive, but the claim is that it's the same "deductive" reasoning.  It's literally me essentially showing the contrapositive. (by making light that under your reasoning I could make the same claim on another account, which clearly doesn't have the same "boost").

 

15 minutes ago, HenrySalayne said:

I honestly thought you were talking about a more sophisticated procedure than manually scrolling through a Twitter page and searching for these posts. Have you at least put all the data into a spreadsheet to make some graphs? Like this one?

I see no need to waste my time to do that, it shows enough that there are spikes in the past.

[Uttamattamakin]

12 minutes ago, HenrySalayne said:

No, you are talking about induction, the opposite.

 

I honestly thought you were talking about a more sophisticated procedure than manually scrolling through a Twitter page and searching for these posts. Have you at least put all the data into a spreadsheet to make some graphs? Like this one?

 

  Reveal hidden contents

Don't klick this spoiler, you can find it yourself.

  Reveal hidden contents

I'm serious, this is easy to find.

  Reveal hidden contents

Really?!? You literally don't know where to find the data?

  Reveal hidden contents

Ok, fine, you hypocrite.

Source: socialblade.com

 

Looks exactly like what one would get by adding a step function to a random fluctuation.  The natural data is what you saw before around... new years 2023.  Then we see that same noise being forced upward like a function shaped like this is being applied twice. 

Not exactly like this plot but the same idea.  (I am not saying this is the data I am saying this is a simple mathematical model that demonstrates the principle I am trying to explain here.  This loosely matches what we see in the data.)


Some smooth natural fluctuation then at some point in time a sharp discontinuity.  On a plot like this one a straight or nearly straight vertical line would also not be natural.  Perfectly horizontal or vertical lines are not something that would happen naturally in a graph like this.  They mean someone is either artificially boosting or suppressing a user. 

To visualize it @wanderingfool2 if Twitter suppressed a user then their graph would show a drop off that would be a step function going down.  They'd go from getting some random number of likes/ unit time to 0, instantly.  What would that look like on a graph like the above?  That would be a perfectly vertical line going down (not at much of any angle.) 

 

[Uttamattamakin]

8 minutes ago, wanderingfool2 said:

I see no need to waste my time to do that, it shows enough that there are spikes in the past.

The spikes are not the issue.  Having some spikes up and down is just the natural randomness of the situation.  Having the spikes be much higher 2x to 3x higher after that unnatural perfectly horizontal line where in Musk was being forced upon the feeds of people who do not follow him, or anything he tweets about, is unnatural.  

You can choose to not believe it.  This is a free world.  Just don't open an umbrella and tell the rest of us it's not raining. 

[wanderingfool2]

6 minutes ago, Uttamattamakin said:

Perfectly horizontal or vertical lines are not something that would happen naturally in a graph like this

Are you trying to ignore the fact they only have 2 data points collected during that horizontal stretch.  Guess what, you get straight lines in graphs like that when you only use 2 points of data over a 12+ hour stretch.

 

YOU are assuming perfect 1 hour interval data points.  It clearly isn't the case.  Actually it looks more along the lines in relation to when he was tweeting when the datapoints were generated.  So YES you expect straight lines when you have a variable sampling rate and when you sample between 2 points quite far apart temporally.  Also, artificially boosted would see a spike, but it still would see a spike in relation to the amount of users online...so again you wouldn't get a straight light.

 

I've already gave a 100% good reasoning why that graph you showed of impressions is garbage.  But if you want to keep going on about it then address these CLEAR, issues with the graph.

 

Their claimed impressions on 1 hour intervals: has at least 6 sections where there is are straight lines for a 4+ hour period.  Simplest reasoning the dataset uses variable sampling.

 

They claim a starting point at 2:30 and another point at 10:30.  That is 8 hours, yet based on the bottom axis it is at least 12+ hours.  How do you explain that inconsistency?  If they had the raw data, why mark the 2 points in time so badly...guess what, it's because they have terrible data or trying to fit their narrative.

 

You don't get to have things both ways, that sometimes there are natural spikes and when it fits your logic those spikes aren't "natural".  The simple fact is, the data is garbage data that lacks granularity in time.  It's like having a ruler with only cm markings, and then trying to guess the mm and presenting your data as though it has mm accuracy.  Data collection just doesn't work like that.  If you collect data at non-fixed intervals you can't make conclusions about "weirdness" between the two sample points.

 

6 minutes ago, Uttamattamakin said:

You can choose to not believe it.  This is a free world.  Just don't open an umbrella and tell the rest of us it's not raining. 

Says the person stating they are right without showing any proof other than "oh the journalist said so".  You don't get dictate what I say.

[HenrySalayne]

54 minutes ago, wanderingfool2 said:

No, the induction part is assuming that if you are claiming "deduction" is how you specidied that I could use "deduction" to make the exact same claim.

i.e. You are claiming if p then q, where you show q (which is still 100% faulty logic).

The inductive part is that assuming "if p then q, showing q is deductive reasoning"

So I present that Person A was boosted, and give "proof" by showing the increase.  So it's still showing "deductive" reasoning how you define it as being proof.  So it uses inductive, but the claim is that it's the same "deductive" reasoning.  It's literally me essentially showing the contrapositive. (by making light that under your reasoning I could make the same claim on another account, which clearly doesn't have the same "boost").

No, you ignore the temporal link.

We had the report about Elon demanding a change (the theory) days (and in a vacuum) before we had any data (the observation). 

In your example it is inverted, starting with the observation and following it up with a theory. Despite me presenting it as straight forward, It is actually a continuous process, just the starting point differentiates between deduction and induction and one will eventually follow the other.

 

In this particular case there is also some confusion about qualitative and quantitative statements (you actually mention this in form of a bad example twice) but somehow you pushed a quantitative "example" into the arena to fight a qualitative claim. This is basically the last two pages of discussion in a nutshell. ¯\_(ツ)_/¯

What you really want to do is find a graph showing the interactions for several accounts and take a look how they developed over the past few weeks. If other accounts show similar rises in the mean, this would at least show that the changes are not unilateral favouring one account. I mean at this point it's IMHO pretty unfounded to argue that nothing has changed and that there Musk's account has not benefited.

[wanderingfool2]

1 hour ago, HenrySalayne said:

No, you ignore the temporal link.

We had the report about Elon demanding a change (the theory) days (and in a vacuum) before we had any data (the observation). 

In your example it is inverted, starting with the observation and following it up with a theory. Despite me presenting it as straight forward, It is actually a continuous process, just the starting point differentiates between deduction and induction and one will eventually follow the other.

I'll say this again.  CORRELATION DOES NOT IMPLY CAUSATION.  You are applying a logical fallacy by claiming you proved something.  Again, IF P THEN Q you CANNOT prove P by showing Q.

 

Guess what, the article was written after the observation.  I already presented just as plausible explanations for a spike in traffic, including but not limited to the fact that the block list was based on absolute count instead of % count.

 

1 hour ago, HenrySalayne said:

What you really want to do is find a graph showing the interactions for several accounts and take a look how they developed over the past few weeks.

I have already stated that that is what shows more.  And again, I already provided the counter to that by showing that spikes on the 13th-14th happened on a similar sized account.  Then you somehow pretended like I wasn't providing the source, despite the fact I've mentioned multiple times before what account I was talking about as the comparison.  I only switched to going fully after the "data" itself because people like you somehow thing that it's gospel truth.

 

1 hour ago, HenrySalayne said:

I mean at this point it's IMHO pretty unfounded to argue that nothing has changed and that there Musk's account has not benefited.

Based on a sample size of a singular account, and you are pretending that it is proof.  Literally what was posted lacks qualitative and quantitative.  So stop pretending as if that data is somehow proof that it exists, it lacks any sense of accuracy; including somehow stating that an 8 hour period is a 12 hour period. 

[HenrySalayne]

7 hours ago, wanderingfool2 said:

I'll say this again.  CORRELATION DOES NOT IMPLY CAUSATION.  You are applying a logical fallacy by claiming you proved something.  Again, IF P THEN Q you CANNOT prove P by showing Q.

I haven't proved something. I was pretty specific that the point made by OP was proven sufficiently to move forward in the discussion.

And yes, correlation does not automatically imply causation. But I would strongly suggest that catching someone with their hand still in the cookie jar is sufficient to call them a cookie thief.

But it seems more and more like there is nothing to move forward too, so instead everyone is turning around and try their best to somehow make the inconvenient data disappear. ¯\_(ツ)_/¯

 

7 hours ago, wanderingfool2 said:

Guess what, the article was written after the observation.  I already presented just as plausible explanations for a spike in traffic, including but not limited to the fact that the block list was based on absolute count instead of % count.

No, it was not. It was written after the event but before API data was published by a researcher showing the changes.

7 hours ago, wanderingfool2 said:

I have already stated that that is what shows more.  And again, I already provided the counter to that by showing that spikes on the 13th-14th happened on a similar sized account.  Then you somehow pretended like I wasn't providing the source, despite the fact I've mentioned multiple times before what account I was talking about as the comparison.  I only switched to going fully after the "data" itself because people like you somehow thing that it's gospel truth.

You countered "he is going way faster than before" with "but this other person also moved a distance". You can't just take the impressions on one post from one account and compare it to the total impressions over time for another account. It's incompatible data.

 

7 hours ago, wanderingfool2 said:

Based on a sample size of a singular account, and you are pretending that it is proof.  Literally what was posted lacks qualitative and quantitative.  So stop pretending as if that data is somehow proof that it exists, it lacks any sense of accuracy; including somehow stating that an 8 hour period is a 12 hour period. 

We have somebody hand-picking* a few posts to compare the data on the one side and Tom Graham simply pulling data from the API and put it into a graph on the other side. And you are telling me that Tom Graham is the one with the bad data?

 

*) of course not at all biased

 

 

 

And maybe put on some glasses because you could easily miss the intricate details, it's really subtle.

[mr moose]

20 hours ago, Uttamattamakin said:

The burden of proof is not static.  One side presents proof then to disprove it you need more or better data.  

You haven't proven your claim, you have shown that over a small data sample musk's impressions went up, this only proves that for a very small window his impressions went up.  It does NOT prove why, it doers NOT prove that they are still up and it does NOT prove that it was only musk's tweets that were improved.

 

These are things you have to prove if you want to make the claim that musk did this to improve HIS tweets.

 

 

 

16 hours ago, HenrySalayne said:

Ok, we can end the debate at this point. Your statement proves conclusively that you are unfit to distinguish between an online forum and a court room. I rest my case.

 

And all this post proves is that you can't/won't support your claims.

[mr moose]

4 hours ago, HenrySalayne said:

And maybe put on some glasses because you could easily miss the intricate details, it's really subtle.

And what about any of the other twitter users?  have they gone up also or is it only musk's?  Because for the OP's claim to be true only musk's tweets will have gone up.

You can't just keep ignoring the fact that the claim comes with specific attributes,  musk's tweet impressions going up is only one of them and the fact no one will post a recent graph, because that change occurred more than 3 weeks ago and I'd like to know if it is an ongoing trend or simply a one off, only compounds the fact that there is no data to support said claims.

[Fendrick]

The hero no one thought we needed.

[Uttamattamakin]

4 hours ago, mr moose said:

And what about any of the other twitter users?  have they gone up also or is it only musk's?  Because for the OP's claim to be true only musk's tweets will have gone up.

You can't just keep ignoring the fact that the claim comes with specific attributes,  musk's tweet impressions going up is only one of them and the fact no one will post a recent graph, because that change occurred more than 3 weeks ago and I'd like to know if it is an ongoing trend or simply a one off, only compounds the fact that there is no data to support said claims.

In this data that @HenrySalayne has here covers several days.  So even smoothing out the smaller fluctuations we can see a clear before and after.  We see a randomly fluctuating "noise" and a strong signal. 

The signals that we use to detect particles like the Higgs Boson look like this.  

*

Trust me I know.  While I was not personally involved in finding the Higgs boson I was taught physics by the same people who, working at Fermilab here in IL, helped build the detectors that did.  The graphs presented, including the last one  by Henry are smoking gun, PROOF.  We know it as well as we know the Earth is round! 

 

*This plot is something called a histogram.  It has equally spaced energy bins on the X  axis and the number of events on the Y axis.   Like a bar graph with all the bars pressed together.  When we see a lot of events in a given energy bin that means there is a particle with a mass equivalent to that.  The plots that show interactions with a twitter feed VS time indicate that something occurred to make that tweet more popular.  

• A single viral tweet would be a sharp peak and drop off fast like the red outlined peak on the chart above.   That would be natural. 
• The noise of the background would be the random fluctuations at the bottom of the chart. 
• Seeing a sustained increase above the baseline as seen in Henry's charts and the ones I found on Twitter is an indication of some... external force... being applied.  The conditions of the experiment being changed.  Like the boss wanting to be sure he gets more engagement than the President of the United States as the original story claimed.  The data bears it out.  

Edited March 3, 2023 by Uttamattamakin
Making cool Lincoln Smaller. No reasonable person should have a problem with Lincoln in this day and age.

[wanderingfool2]

9 hours ago, HenrySalayne said:

You countered "he is going way faster than before" with "but this other person also moved a distance". You can't just take the impressions on one post from one account and compare it to the total impressions over time for another account. It's incompatible data.

"total impressions over time" is a foolhardy approach when the data clearly has 12 hour gaps in it's sampling rate.  It does also invalidate claims, given that the spikes occurred on a similar size account and on approximately the same days.  You try claiming what he is providing is proof, but it's literally a counter example of me showing that there is a similar sized account which had a 6x increase on roughly the same day.

 

The Reimann hypothesis is not trivial to show as being true (to the point that it's still unsolved), yet it only requires a single number to show that it's invalid.  The general claim is that because of the superbowl event and tampering of the algo there was the spike in numbers for, yet I just showed that there was a similar sized spike on a similar sized account.  I also showed how the data that was shown cannot be viewed as being accurate in regards to hourly time-frames (because if you haven't clued in Uta is arguing a horizontal slope somehow means he's true, which is 100% incorrect when there are only 2 points involved during that period).

 

9 hours ago, HenrySalayne said:

No, it was not. It was written after the event but before API data was published by a researcher showing the changes.

That is just backwards reasoning, I could watch a Twitter account right now and observe a spike, write an article saying that the person bribe twitter, and then pull API data that shows the observed "event" happened. (Then pretend that the fact the person was plastered all over the news in a scandal doesn't explain the spike, and try claiming that a post analysis of it doesn't have meaning to the event).

 

9 hours ago, HenrySalayne said:

We have somebody hand-picking* a few posts to compare the data on the one side and Tom Graham simply pulling data from the API and put it into a graph on the other side. And you are telling me that Tom Graham is the one with the bad data?

 

*) of course not at all biased

I already mentioned before, that's what happens when you cherry pick data to fit your narrative.

 

You are being foolhardy if you believe that Tom Graham hasn't presented data that misconstrues the events.  Again, you can tell it's garbage data by looking at the 12 hour period that Tom Graham is trying to claim is 8 hours.  Strange how you magically seem to never address that point of contention that the guy claiming to have pulled it hourly seems to have 12 hour gaps in his data.

 

9 hours ago, HenrySalayne said:

And maybe put on some glasses because you could easily miss the intricate details, it's really subtle.

Uta literally is claiming the presence of horizontal slopes is the event in question.  Again, I have said multiple times that there is justification behind a spike.  Including the fact that somewhere along around that time Elon deleted his block list, modifying the fanout service, etc.  All those tweets regarding that were *gasps* written before the event.

 

A spike means absolutely nothing in isolation, because again if the guy made the same claim back in Dec the graph would look exactly the same

[Uttamattamakin]

12 minutes ago, wanderingfool2 said:

Uta literally is claiming the presence of horizontal slopes is the event in question.  Again, I have said multiple times that there is justification behind a spike.  Including the fact that somewhere along around that time Elon deleted his block list, modifying the fanout service, etc.  All those tweets regarding that were *gasps* written before the event.

 

A spike means absolutely nothing in isolation, because again if the guy made the same claim back in Dec the graph would look exactly the same

The two graphs have different time bins.  One day by day the other bins multiple days.  The huge peak above the background shows that something was done which is not what was done before.  No other account shows this effect. 

I even gave you an easy enough to see example of how this kind of analysis is used. It is generally applicable.  Here is another one. 

 

[wanderingfool2]

25 minutes ago, Uttamattamakin said:

The two graphs have different time bins.  One day by day the other bins multiple days.  The huge peak above the background shows that something was done which is not what was done before.  No other account shows this effect. 

You literally argued the horizontal line meant something earlier as well.  You keep claiming you are an expert and you know the math and can prove things, yet you seem to lack the knowledge of basic level logic.

 

You can't even get it through you head that a horizontal line doesn't make it proof (when the line was due to bad sampling intervals).  You try making up some long winded explanation that explains a horizontal line but are too blind to notice that it's because they only used 2 data points during that period.

I bolded the claim you are making, that you have yet to substantiate.  Show the API calls of similar accounts.

 

Again, just because there is a spike doesn't mean it's artificial or that what you claim is correct...because as was stated multiple times there are situations which were mentioned in public prior that could have similar types of effects.  [Fan out service bogged down, block list, I think there was an implied reset, actual bug where replies were treated as full tweets, etc]

[HenrySalayne]

8 hours ago, mr moose said:

and the fact no one will post a recent graph, because that change occurred more than 3 weeks ago and I'd like to know if it is an ongoing trend or simply a one off, only compounds the fact that there is no data to support said claims.

Are you trolling or are you simply uninformed?

 

3 hours ago, wanderingfool2 said:

"total impressions over time" is a foolhardy approach when the data clearly has 12 hour gaps in it's sampling rate.

3 hours ago, wanderingfool2 said:

It does also invalidate claims, given that the spikes occurred on a similar size account and on approximately the same days. 

3 hours ago, wanderingfool2 said:

You try claiming what he is providing is proof, but it's literally a counter example of me showing that there is a similar sized account which had a 6x increase on roughly the same day.

This is honestly hilarious.

3 hours ago, wanderingfool2 said:

You are being foolhardy if you believe that Tom Graham hasn't presented data that misconstrues the events.  Again, you can tell it's garbage data by looking at the 12 hour period that Tom Graham is trying to claim is 8 hours.  Strange how you magically seem to never address that point of contention that the guy claiming to have pulled it hourly seems to have 12 hour gaps in his data.

Since Tom Graham is an actual researcher and not just a rando on the internet (like you if this wasn't clear), I think his credibility is a lot higher than yours. If you think his data is flawed, maybe give him a call. I bet he can help you understand what happened in what you call "gap".

And you completely ignore the second graph with daily impressions. Why exactly?

3 hours ago, wanderingfool2 said:

I already mentioned before, that's what happens when you cherry pick data to fit your narrative.

Since you already mention this, is there something you want to tell us about Obama's account? Your prime counter example?

Because it seems to be fresh fuel for the "something was done to benefit Musk" theory.

[wanderingfool2]

24 minutes ago, HenrySalayne said:

Since Tom Graham is an actual researcher and not just a rando on the internet (like you if this wasn't clear), I think his credibility is a lot higher than yours. If you think his data is flawed, maybe give him a call. I bet he can help you understand what happened in what you call "gap".

And you completely ignore the second graph with daily impressions. Why exactly?

I didn't ignore it. I am not denying there is a spike, so stop with that asinine argument trying to shoehorn that I'm saying there wasn't a spike.  I didn't ignore it, if you bothered even looking I've been saying there are other reasons why spikes can happen.

 

Get your head out of the sand, I don't care if it's a world class researcher.  The fact is the graph he posted where he clearly marks 2 points that he claims are 8 hours apart while on the bottom of the graph it indicates it's over 12 hours apart IS PROOF that Tom either has terrible data, biased opinion added, or fabricated information.  Uta literally argued that the horizontal line was essentially the smoking gun, yet failed to understand that the graph had 2 points of data.

 

24 minutes ago, HenrySalayne said:

Since you already mention this, is there something you want to tell us about Obama's account? Your prime counter example?

Because it seems to be fresh fuel for the "something was done to benefit Musk" theory.

Gee now lets see, basic logic.  Oh this account spiked so there must be manipulation; counter example, show a similar account that had a similar spike over that course of time.

 

 

Here's a hint for you.  If the article was posted back in December, your so called "proof" that you keep pointing to would have seen a similar spike.

[HenrySalayne]

5 minutes ago, wanderingfool2 said:

Gee now lets see, basic logic.  Oh this account spiked so there must be manipulation; counter example, show a similar account that had a similar spike over that course of time.

Maybe, I don't know. But the spike is a lie. There is no spike.

[wanderingfool2]

40 minutes ago, HenrySalayne said:

Maybe, I don't know. But the spike is a lie. There is no spike.

Okay, let's look at Feb.

Oh yea, no spike at all.  Lets just ignore that.  Lets ignore the fact that I already showed a screenshot of the Feb 13, vs Feb 14.  Lets ignore that all his tweets in Feb averaged less than 4 mill.  Again, time periods aren't necessarily a good tool to show something when you are cherry picking.  Again, look at December time for Musk you have a large spike that is larger than the current one.  Under your illogical reasoning if the article was posted then you would be "proven correct" because there was a spike.

 

Again, all the reporter need to see was a spike, and then decides to write about it using "sources".  Here's the hint as well, you keep claiming you did it by deduction, but ultimately proper deduction is essentially the form if p then q.    You still are trying to prove p by showing q which is still faulty logic.

[Uttamattamakin]

4 hours ago, wanderingfool2 said:

You literally argued the horizontal line meant something earlier as well.  You keep claiming you are an expert and you know the math and can prove things, yet you seem to lack the knowledge of basic level logic.

Yes they both mean something.  

 

One shows a constant addition of interactions to Elons account.  

The other shows the result of that that forcing.  

[wanderingfool2]

15 minutes ago, Uttamattamakin said:

One shows a constant addition of interactions to Elons account.  

It doesn't show that.  It shows they used 2 data points during that period.  It shows they are claiming 8 hours while it's 12 hours, and claiming it's an 1 hour interval when it clearly isn't.

 

So try using your "expertise" to explain away the fact that your data source is showing a 8 hour period as 12 hours.  The fact is you can't.

[BabaGanuche]

15 hours ago, HenrySalayne said:

And maybe put on some glasses because you could easily miss the intricate details, it's really subtle.

Let's look at this like a scientific study. Your hypothesis is there is a user that is getting special treatment in the algorithm. Like any good scientist you start assuming the null hypothesis, nothing is happening, and have provide evidence that something is happening.

 

You start by looking a one user and notice a their impressions are higher than you expected. You would then need to compare that user to other users to see if they were getting an unfair boost in impressions.

 

Example, say there was a bug the showed ever users tweet to twice the normal number of users. How can you tell that scenario apart from your assertion, that one particular user is just getting a boost? And that is just one possible explanation for the increase in impressions. There are many more such as a particularly interesting or conversal tweet. You need more data to be able to say the your hypothesis is real reason over any other explanation.

 

In conclusion that graph proves nothing for this user getting a special boost. You would have to compare against other users to even attempt to show that that user was getting a unfair number of impressions.

[HenrySalayne]

1 hour ago, wanderingfool2 said:

Okay, let's look at Feb.

Oh yea, no spike at all.  Lets just ignore that.  Lets ignore the fact that I already showed a screenshot of the Feb 13, vs Feb 14.  Lets ignore that all his tweets in Feb averaged less than 4 mill.  Again, time periods aren't necessarily a good tool to show something when you are cherry picking.  Again, look at December time for Musk you have a large spike that is larger than the current one.  Under your illogical reasoning if the article was posted then you would be "proven correct" because there was a spike.

 

Again, all the reporter need to see was a spike, and then decides to write about it using "sources".  Here's the hint as well, you keep claiming you did it by deduction, but ultimately proper deduction is essentially the form if p then q.    You still are trying to prove p by showing q which is still faulty logic.

Source? And what exact data does your graph show?

That you still miss the point by a mile...

[SansVarnic]

-= Topic Locked =-

I believe this has run its course.

The course not being a straight line but an oddly shaped segment with a beginning and an end.

 

Spoiler

“The keenest sorrow is to recognize ourselves as the sole cause of all our adversities.” ― Sophocles