What does a TRUMP presidency look like?

[grumbler]

To be fair, while playing with the y-axis is often a deliberate attempt to be misleading, there are legitimate reasons to do so.  You don't want any dead space in the plots, you want to accentuate the plots on what is being displayed.  A lot of statistical software also automatically determines minimum and maximum axis values from the data points, precisely for that purpose.

To be fair, focussing on the range of the data isn't a problem if you present the big picture (zero to whatever) first.  That's not what VOX is doing.  Manipulating the Y-axis without indicating you are doing so is deliberately misleading.

[DGuller]

To be fair, while playing with the y-axis is often a deliberate attempt to be misleading, there are legitimate reasons to do so.  You don't want any dead space in the plots, you want to accentuate the plots on what is being displayed.  A lot of statistical software also automatically determines minimum and maximum axis values from the data points, precisely for that purpose.

To be fair, focussing on the range of the data isn't a problem if you present the big picture (zero to whatever) first.  That's not what VOX is doing.  Manipulating the Y-axis without indicating you are doing so is deliberately misleading.

What's the practical effect?  The y-axis is more or less the same between the two graphs.  The range of y-values is not so small that blowing it up is misleading on its own.  I do lots of plots almost every day, and the only time I play with the y-axis is if default values for some reason make the plot look awful, or if it makes the readers reach a misleading conclusion without an explanation/warning.  Most of the time neither happens, and the range of min(y) - margin to max(y) + margin does exactly what I need it to do.

None of these things happen with the graphs presented here either.  Given the lack of effect of "manipulation of y-axis", maybe a good default explanation to fall back on is that the y-axis was not manipulated, but rather the y-axis for both plots was what the plotting software chose, and the authors correctly concluded that it would not be misleading to leave it that way.

[crazy canuck]

To be fair, while playing with the y-axis is often a deliberate attempt to be misleading, there are legitimate reasons to do so.  You don't want any dead space in the plots, you want to accentuate the plots on what is being displayed.  A lot of statistical software also automatically determines minimum and maximum axis values from the data points, precisely for that purpose.

To be fair, focussing on the range of the data isn't a problem if you present the big picture (zero to whatever) first.  That's not what VOX is doing.  Manipulating the Y-axis without indicating you are doing so is deliberately misleading.

What's the practical effect?  The y-axis is more or less the same between the two graphs.  The range of y-values is not so small that blowing it up is misleading on its own.  I do lots of plots almost every day, and the only time I play with the y-axis is if default values for some reason make the plot look awful, or if it makes the readers reach a misleading conclusion without an explanation/warning.  Most of the time neither happens, and the range of min(y) - margin to max(y) + margin does exactly what I need it to do.

None of these things happen with the graphs presented here either.  Given the lack of effect of "manipulation of y-axis", maybe a good default explanation to fall back on is that the y-axis was not manipulated, but rather the y-axis for both plots was what the plotting software chose, and the authors correctly concluded that it would not be misleading to leave it that way.

Or we could just jump to the conclusion that someone was being intentionally dishonest and not worry about all the probable innocent explanations. 

[mongers]

To be fair, while playing with the y-axis is often a deliberate attempt to be misleading, there are legitimate reasons to do so.  You don't want any dead space in the plots, you want to accentuate the plots on what is being displayed.  A lot of statistical software also automatically determines minimum and maximum axis values from the data points, precisely for that purpose.

To be fair, focussing on the range of the data isn't a problem if you present the big picture (zero to whatever) first.  That's not what VOX is doing.  Manipulating the Y-axis without indicating you are doing so is deliberately misleading.

What's the practical effect?  The y-axis is more or less the same between the two graphs.  The range of y-values is not so small that blowing it up is misleading on its own.  I do lots of plots almost every day, and the only time I play with the y-axis is if default values for some reason make the plot look awful, or if it makes the readers reach a misleading conclusion without an explanation/warning.  Most of the time neither happens, and the range of min(y) - margin to max(y) + margin does exactly what I need it to do.

None of these things happen with the graphs presented here either.  Given the lack of effect of "manipulation of y-axis", maybe a good default explanation to fall back on is that the y-axis was not manipulated, but rather the y-axis for both plots was what the plotting software chose, and the authors correctly concluded that it would not be misleading to leave it that way.

Or we could just jump to the conclusion that someone was being intentionally dishonest and not worry about all the probable innocent explanations.

Well this could be a first for Languish, CC, Grumbler and Mongers in agreement. 

And to echo what Grumbler said, why over-egg the pudding, the two graphs drawn down to the zero would still be powerful, but also present a fuller picture.

[jimmy olsen]

Oh? Tell us more.

https://twitter.com/realDonaldTrump/status/1024656465158721536

Russian Collusion with the Trump Campaign, one of the most successful in history, is a TOTAL HOAX. The Democrats paid for the phony and discredited Dossier which was, along with Comey, McCabe, Strzok and his lover, the lovely Lisa Page, used to begin the Witch Hunt. Disgraceful!

[garbon]

Vox didn't manipulate the graph, it came that way from the original source.

[jimmy olsen]

Will the Donald's words come back to haunt him once again?

https://www.nbcnews.com/politics/donald-trump/trump-s-undermining-obamacare-violates-constitution-new-lawsuit-charges-n896626

[Legbiter]

The man has around 88% approval rating among Republicans. Only a nasty recession would dent that. 

[crazy canuck]

The man has around 88% approval rating among Republicans. Only a nasty recession would dent that. 

I wonder whether that is a function of people abandoning the party.

[Tamas]

The man has around 88% approval rating among Republicans. Only a nasty recession would dent that. 

I wonder whether that is a function of people abandoning the party.

I don't think so. I think if you make the (sub-)conscious decision to believe Trump's version of reality, then it is easy to see him as the not very refined but heroic fighter of the people sorting shit out all over the place, from North Korea through Russia to preventing dirty Mexicans crossing the border.
Not to mention making sure nothing remains of anything Obama has ever done, which was made out to be the cornerstone of preserving civilization by the GOP during the previous 8 years.

Why exactly would Republicans dislike him?

[garbon]

I can't find anything easily showing trends into this year but here are recent trends and then past decade from Gallup.

https://news.gallup.com/poll/223124/democratic-party-maintains-edge-party-affiliation.aspx

[garbon]

Though if you look at affiliation strictly speaking, no lean, they are seeing slightly less than 30% typically reporting that they are Republicans...not really impacted from when Trump because president to now.

https://news.gallup.com/poll/15370/party-affiliation.aspx

[Oexmelin]

 "Independents" who always vote Republicans are typically more numerous than "Independents" who always vote Democrats.

[garbon]

"Independents" who always vote Republicans are typically more numerous than "Independents" who always vote Democrats.

They wouldn't be part of the approval rating though. Or rather, they wouldn't be counted in the Republican portion of the approval rating.

[FunkMonk]

The gradual downward trend in Rs and leaning-Rs since the election is very interesting. Thanks G.