7 ms·

This part is true: "In practice, this data-driven approach never worked and campaigns that relied on it, like Hillary Clinton’s presidential bid in 2016, have

by hliyan 8y ago

This part is true: "In practice, this data-driven approach never worked and campaigns that relied on it, like Hillary Clinton’s presidential bid in 2016, have usually ended in failure." I think that a statistically significant set of detailed case studies is superior to a massive mountain of data where each data point is just an approximation of what it is supposed to measure (e.g. answers to a handful of questions, or some metric). Perhaps anecdotes, in large enough numbers, are data.

flatline 8y ago
> Perhaps anecdotes, in large enough numbers, are data. This was actually the original gist of that famous quote! http://blog.revolutionanalytics.com/2011/04/the-plural-of-anecdote-is-data-after-all.html http://blog.revolutionanalytics.com/2011/04/the-plural-of-an...
digi_owl 8y ago
The Clinton campaign had to so many sins that i suspect being data driven was the least of its problems.
- noobermin 8y ago
  In many ways being too data driven whilst disregarding basic political wisdom (like actually campaigning in swing states) was more of a downfall.
  - specialist 8y ago
    No. HRC's campaign knew MI, MN, PA would be close. They just didn't factor in the wide spread, targeted voter disenfranchisement. For instance, ~200k voters in WI were blocked, concentrated in urban areas like Milwaukee.
- specialist 8y ago
  HRC's greatest sins, why she lost, in rank order, are the electoral college, the 2m disenfranchised voters, not being a racist, not being given $3b in earned media (free advertising), and not being Putin's toady. PS- I'm not forgetting Comey's sabotage. I just don't know where the 2.5% swing it caused fits into this list. Probably #4.
_emacsomancer_ 8y ago
> "... this data-driven approach never worked and campaigns that relied on it, like Hillary Clinton’s presidential bid in 2016, have usually ended in failure." It's also, of course, a question of having the 'right' data, and therein (also, of course), lies the rub.
- Fnoord 8y ago
  My thought exactly. Consider the Cambridge Analytics debacle.
- ropeadopepope 8y ago
  Even if you have the right data, I fail to see how getting and interpreting data of any kind is somehow going to magically make people vote for you. The most it can do is give you a good picture of who will be willing to vote for you, but even that's pretty dubious when you consider the research that's been done on the inaccuracy of voter polling.
  - _emacsomancer_ 8y ago
    I think for the Clinton campaign it was used to make decisions about where to spend campaigning time. Arguably miscalculations/bad data on this point contributed to Clinton's loss.
garmaine 8y ago
Actually I think that quoted part is bogus. Hillary won a very contested and bitterly fought primary despite NOT being the favorite (among Democrats) by a large margin or having a clearly articulated platform. Big data seems to have worked very well there! She just found herself up against an even more data driven campaign.
- specialist 8y ago
  No one in the history of USA Presidential campaigns has ever had a larger, more detailed platform than HRC. That's kinda what she's known for. The more correct criticism, which she admits to, is her inability to be less wonkish and more arm wavey (visionary). The correct criticism, which no one will admit to, is the futility of trying to explain oneself thru a medium purpose built to sabotage rational discourse.
  - iainmerrick 8y ago
    Additionally, she was the clear favourite right from the start. It’s not true that it was a more tightly contested primary than usual, although Sanders gave her a surprisingly good run for sure. Clinton vs Sanders was much more one-sided than Obama vs Clinton, say.
    - specialist 8y ago
      re Obama vs Clinton and how it relates to Patrick Cockburn's article... One can be right about everything and yet still ultimately be wrong. Because not everything can be measured. Because it's easy to miss important details, angles, insights. Because rock, paper, scissors. For the 2008 primary, Clinton and Obama were playing different games, popularity vs delegate counts, so used different strategies. Obama novel strategy won. And FWIW, HRC's 2016 campaign knew the electoral college would be close, were doing pretty well, but then Comey threw the election to the Russians. Anticipating all the butthurt downvoters: Comey has all but admitted as much.
      - dang 8y ago
obelix_ 8y ago
All you have to do is ask Google - did inequality increase under 8 years of Obama? Look at the results that pop up. You will see articles filled with charts and data that tell two totally different stories. Every other link will dutifully fall on one side or the other. How is the average American going to figure out what the truth is?
- kadenshep 8y ago
  >How is the average American going to figure out what the truth is? Basic education that was provided for free in middle and high school?
  - obelix_ 8y ago
    Do you think that's working? Once you take this position, and from the outcomes we see all around us, its very easy to reach a conclusion that the edu system is failing as a large part of the population can't parse what they are being fed. And that includes highly educated people btw. Cause they land up on either side on any issue you pick. But that's just wrong. People constantly expect the education system to solve the problem (of info overload - so many facts that not all facts are equal as the article does a good job explaining). The education system was never designed to solve info overload. And social media/news media/Google are designed to overload everyone. We need a bridging of the two systems. Google/social media/news media has to understand they are constantly serving second graders tenth grade problems. And pretending there won't be any consewuences. The education system is atlest able to signal to a second grader he is in second grade and here is a path to tenth grade with feedback constantly being provided to ensure he stays on track if he wants to get to tenth grade. These signals don't exist on the internet of 2018 or are drowned out by 2000 other signals. Most people are busy, distracted, coping with their own life problems and with the minimum attention they have to spare are told to pick facts from a large buffet table. The tech world has reinvent what signals it's providing alongside those facts. Cause the highly simplistic current signals of upvotes/view counts/likes/retweets etc are not enough.
    - kadenshep 8y ago
roenxi 8y ago
America is large, multifaceted, diverse and complicated. The difference in the popular vote was large in political terms, but in absolute terms 48% of the voters voted for the winning candidate. To claim that the problem was "the data driven approach" seems to be very bold. There are a lot of aspects to a campaign, and flaws in any of them can be fatal. Ironically, we have very little data on the subject of what went wrong for Hillary; although given the election results it was probably a problem in specific regions. fivethirtyeight.com had Trump as the underdog with a ~30% chance of victory (so, we'd expect this sort of outcome every 4 elections). This suggests that data was painting an accurate picture to me, everyone was surprised but the election result wasn't unthinkable.
- incompatible 8y ago
  Speaking of facts, Wikipedia has it at 46.09% for Trump, 48.18% for Clinton, on a turnout of 55.7%. https://en.wikipedia.org/wiki/United_States_presidential_election,_2016 https://en.wikipedia.org/wiki/United_States_presidential_ele...
  - Izkata 8y ago
    The problem is that, because of how the election works, there's no way to know if that's relevant. States that reliably vote for one party could easily be discouraging a ton of people in the opposite camp from voting at all, since they know their vote won't matter. (In particular, I'm thinking of California, which I don't think anyone doubted would go Clinton. And yet it still managed to get around 30% votes for Trump. How much higher would that 30% be if there was uncertainty about which way the state could go?)
- baruchthescribe 8y ago
  FBIAnon had Trump at +20% with a 5% MOE in July. He was on the nail.
megaman22 8y ago
Maybe a data-driven campaign with a charismatic candidate would perform better. The sample size is so small, that it's impossible to make any real conclusions.
tzs 8y ago
What about Obama vs. Romney? I recall numerous articles after that election discussing the difference the Obama campaign's use of "big data" made. Googling "Obama vs Romney big data" or "Obama vs Romney data driven" shows I'm not misremembering.

by hliyan 8y ago

flatline 8y ago

> Perhaps anecdotes, in large enough numbers, are data. This was actually the original gist of that famous quote! http://blog.revolutionanalytics.com/2011/04/the-plural-of-anecdote-is-data-after-all.html http://blog.revolutionanalytics.com/2011/04/the-plural-of-an...

digi_owl 8y ago

The Clinton campaign had to so many sins that i suspect being data driven was the least of its problems.

noobermin 8y ago
In many ways being too data driven whilst disregarding basic political wisdom (like actually campaigning in swing states) was more of a downfall.
- specialist 8y ago
  No. HRC's campaign knew MI, MN, PA would be close. They just didn't factor in the wide spread, targeted voter disenfranchisement. For instance, ~200k voters in WI were blocked, concentrated in urban areas like Milwaukee.
specialist 8y ago
HRC's greatest sins, why she lost, in rank order, are the electoral college, the 2m disenfranchised voters, not being a racist, not being given $3b in earned media (free advertising), and not being Putin's toady. PS- I'm not forgetting Comey's sabotage. I just don't know where the 2.5% swing it caused fits into this list. Probably #4.

_emacsomancer_ 8y ago

> "... this data-driven approach never worked and campaigns that relied on it, like Hillary Clinton’s presidential bid in 2016, have usually ended in failure." It's also, of course, a question of having the 'right' data, and therein (also, of course), lies the rub.

Fnoord 8y ago
My thought exactly. Consider the Cambridge Analytics debacle.
ropeadopepope 8y ago
Even if you have the right data, I fail to see how getting and interpreting data of any kind is somehow going to magically make people vote for you. The most it can do is give you a good picture of who will be willing to vote for you, but even that's pretty dubious when you consider the research that's been done on the inaccuracy of voter polling.
- _emacsomancer_ 8y ago
  I think for the Clinton campaign it was used to make decisions about where to spend campaigning time. Arguably miscalculations/bad data on this point contributed to Clinton's loss.

garmaine 8y ago

Actually I think that quoted part is bogus. Hillary won a very contested and bitterly fought primary despite NOT being the favorite (among Democrats) by a large margin or having a clearly articulated platform. Big data seems to have worked very well there! She just found herself up against an even more data driven campaign.

specialist 8y ago
No one in the history of USA Presidential campaigns has ever had a larger, more detailed platform than HRC. That's kinda what she's known for. The more correct criticism, which she admits to, is her inability to be less wonkish and more arm wavey (visionary). The correct criticism, which no one will admit to, is the futility of trying to explain oneself thru a medium purpose built to sabotage rational discourse.
- iainmerrick 8y ago
  Additionally, she was the clear favourite right from the start. It’s not true that it was a more tightly contested primary than usual, although Sanders gave her a surprisingly good run for sure. Clinton vs Sanders was much more one-sided than Obama vs Clinton, say.
  - specialist 8y ago
    re Obama vs Clinton and how it relates to Patrick Cockburn's article... One can be right about everything and yet still ultimately be wrong. Because not everything can be measured. Because it's easy to miss important details, angles, insights. Because rock, paper, scissors. For the 2008 primary, Clinton and Obama were playing different games, popularity vs delegate counts, so used different strategies. Obama novel strategy won. And FWIW, HRC's 2016 campaign knew the electoral college would be close, were doing pretty well, but then Comey threw the election to the Russians. Anticipating all the butthurt downvoters: Comey has all but admitted as much.
    - dang 8y ago

obelix_ 8y ago

All you have to do is ask Google - did inequality increase under 8 years of Obama? Look at the results that pop up. You will see articles filled with charts and data that tell two totally different stories. Every other link will dutifully fall on one side or the other. How is the average American going to figure out what the truth is?

kadenshep 8y ago
>How is the average American going to figure out what the truth is? Basic education that was provided for free in middle and high school?
- obelix_ 8y ago
  Do you think that's working? Once you take this position, and from the outcomes we see all around us, its very easy to reach a conclusion that the edu system is failing as a large part of the population can't parse what they are being fed. And that includes highly educated people btw. Cause they land up on either side on any issue you pick. But that's just wrong. People constantly expect the education system to solve the problem (of info overload - so many facts that not all facts are equal as the article does a good job explaining). The education system was never designed to solve info overload. And social media/news media/Google are designed to overload everyone. We need a bridging of the two systems. Google/social media/news media has to understand they are constantly serving second graders tenth grade problems. And pretending there won't be any consewuences. The education system is atlest able to signal to a second grader he is in second grade and here is a path to tenth grade with feedback constantly being provided to ensure he stays on track if he wants to get to tenth grade. These signals don't exist on the internet of 2018 or are drowned out by 2000 other signals. Most people are busy, distracted, coping with their own life problems and with the minimum attention they have to spare are told to pick facts from a large buffet table. The tech world has reinvent what signals it's providing alongside those facts. Cause the highly simplistic current signals of upvotes/view counts/likes/retweets etc are not enough.
  - kadenshep 8y ago

roenxi 8y ago

America is large, multifaceted, diverse and complicated. The difference in the popular vote was large in political terms, but in absolute terms 48% of the voters voted for the winning candidate. To claim that the problem was "the data driven approach" seems to be very bold. There are a lot of aspects to a campaign, and flaws in any of them can be fatal. Ironically, we have very little data on the subject of what went wrong for Hillary; although given the election results it was probably a problem in specific regions. fivethirtyeight.com had Trump as the underdog with a ~30% chance of victory (so, we'd expect this sort of outcome every 4 elections). This suggests that data was painting an accurate picture to me, everyone was surprised but the election result wasn't unthinkable.

incompatible 8y ago
Speaking of facts, Wikipedia has it at 46.09% for Trump, 48.18% for Clinton, on a turnout of 55.7%. https://en.wikipedia.org/wiki/United_States_presidential_election,_2016 https://en.wikipedia.org/wiki/United_States_presidential_ele...
- Izkata 8y ago
  The problem is that, because of how the election works, there's no way to know if that's relevant. States that reliably vote for one party could easily be discouraging a ton of people in the opposite camp from voting at all, since they know their vote won't matter. (In particular, I'm thinking of California, which I don't think anyone doubted would go Clinton. And yet it still managed to get around 30% votes for Trump. How much higher would that 30% be if there was uncertainty about which way the state could go?)
baruchthescribe 8y ago
FBIAnon had Trump at +20% with a 5% MOE in July. He was on the nail.

megaman22 8y ago

Maybe a data-driven campaign with a charismatic candidate would perform better. The sample size is so small, that it's impossible to make any real conclusions.

tzs 8y ago

What about Obama vs. Romney? I recall numerous articles after that election discussing the difference the Obama campaign's use of "big data" made. Googling "Obama vs Romney big data" or "Obama vs Romney data driven" shows I'm not misremembering.