Not All Facts Are Equal | Hacker News Reader

Not All Facts Are Equal

independent.co.uk

83 points_1sta57 comments

This confusion is rampant in the business world. Some people prefer to bamboozle and be bamboozled by case studies, some people prefer to bamboozle and be bamboozled by pie charts. Powerpoint is a common medium for both. However, despite their differences, they share their audience - people who can't do a good analysis and have no idea what one would look like. Inevitably, if this realization ever sinks in it will have to be worded in a way that doesn't make it sound like there's anything wrong with the people who don't know what they're doing. Now that journalists have started to phrase it in a way that makes it sound like the "facts," are at fault, I think it has a good chance of becoming accepted.

This is an excellent point. Although, I think as far as (most) journalism goes, it's less about "facts" in general than it is defending an authority/legitimacy that was previously accepted (generally) but is now being put into question.

It seems the author is pointing out how the tendencies that lead people to consider something as "fact" haven't changed despite the mechanism of communication improving in capability/efficiency.

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.

> 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-an...

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

In many ways being too data driven whilst disregarding basic political wisdom (like actually campaigning in swing states) was more of a downfall.

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.

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.

> "... 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.

My thought exactly. Consider the Cambridge Analytics debacle.

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.

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.

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.

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.

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.

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.

Would you please not take HN threads into partisan flamewars? That's the opposite of what we want here, regardless of how right you are.

I appreciate that you haven't been using HN primarily for this (other than today), but still, it damages this place, so please don't.

https://news.ycombinator.com/newsguidelines.html

What was partisan in the comment you are replying to? Looks like standard (and on-topic) political science reasoning not politicking to me.

Edit: did you mean to respond to specialist’s other comment in this thread?

That was in reply to multiple political flamewar posts, but the Comey bit here obviously qualifies. Also, "butthurt downvoters"? that's embarrassing.

Fair enough. The last line did dirty it a bit.

You're right. Thank you for keeping me in line.

I'm a repeat offender. My immediate solution is to not login, so less tempted to engage. If I feel the need to continue participating, I use a real-name account, which usually keeps me from trolling.

Back when I was moderator, we had offline mail readers. I miss two features. Time shifting and twit filters.

I really want to read the threads a week later, after the commentariat has a chance to weigh in. That's also how I consume my "news". Keeps things "cool" (vs "hot" a la Marshall McLuhan).

And I really miss having twit filters, which allowed you to fully block mail from marked accounts. I scanned the various 3rd party HN clients, but nothing like this jumped out. Not that I have enough to do, but maybe I can morph an existing client.

Again, thank you for your efforts.

The phrase I said was “clearly articulated platform.” I don’t think you and I are in disagreement.

She was the worst possible candidate of Democrats. Maybe even some time in future we'll learn that Russians helped her win that primary in order to give Trump easiest opponent to beat :)

I seriously doubt Putin, who famously despises HRC, would help her win anything.

But since you brought it up...

HRC helped Trump win the GOP nomination. Mostly because Rubio, the favorite, would have crushed HRC in the general. (And because why not monkey wrench the other side's primary?)

In a weird way, I think we're lucky Trump won and not Rubio. The backlash against Trump is akin to a tidal wave, a sea change. Whereas Rubio would have been less motivating than say GWB & Cheney, so the GOP would probably hold onto power for longer.

Doubly so because Rubio would have had the legitimacy of a clear victory, whereas no one to the left of Steven Bannon considers Trump legitimate.

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?

>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?

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.

>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.

I think the problem is deeper than any education system could reasonably be expected to fix.

>Basic education that was provided for free in middle and high school?

I can't tell if you are trolling so tell me which is true?

A > B

B > A

Where only one is true, and you have no more information I have provided here.

This is what you seemed to dismiss with your statement about basic education.

That's not what the studies do. Some show A' < B', while others do A" > B". Both can be true yet both mislead about how A and B compare.

My comment was in response to the parent comment, that a that "education" (which nearly everyone has) would cure the problem being discussed.

Uh, that 'education' typically involves more than evaluating simple inequalities.

But it doesn't provide anymore data to determine the answer, therefore education alone is not a valid solution.

Specific information is required to solve the problem, not general education.

The information is there. Nothing you've said in your responses have been remotely realistic.

I will take the bait, answer me this, is this equation true or false?

C > D

I propose to you that you do not have enough specific information to answer this question, even if you had a PHD in mathematics.

But no one is talking about that kind of problem. We have the information. You're using an extremely contrived example just to make a point. One might even call it a straw man. I just prefer the term dishonest based on current online discourse.

This was already pointed out by another extremely patient commenter.

>We have the information...

No, actually you don't. _That_ is the problem.

Here's the original comment I responded to:

---

>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?

---

This person's comment is that a simple black and white "basic education" solves the problem of "how is the average American going to figure out the truth?"

I disagreed with my initial comment stating that, education is simply not enough.

I have a good friend with a PhD, and he's believed some incorrect things before, and I showed him more data, and he agreed he was wrong.

So, no, just "having an education" doesn't automagically allow anyone to determine the "truth" of a matter.

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.

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_ele...

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?)

FBIAnon had Trump at +20% with a 5% MOE in July. He was on the nail.

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.

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.

I think the headline is great, the article not so much.

I learned about the inequality of facts while processing accident claims. Accident claims are a little like a court case where a single judge determines a verdict. That verdict determines whether or not a check is cut to cover the claim.

In any claim, you will see strong facts, weak facts, and even red flags that can indicate it is fraudulent. With reading it, you need to make one of four decisions:

1. Pay the claim.

2. Deny the claim.

3. Request more information.

4. Refer to the fraud department for review.

Any given claim can have indicators to pay and contradictory indicators to deny. The decision to pay or deny hinges not only on how many such indicators there are, but on the strength of each of them. Kind of like DNA evidence will outweigh hearsay in a court case.

Some facts are stronger than others. This is true even if you take out the element of politics.

I feel like literature can often have this problem too; the information density has become less thoughtful since there is so much data to quote and people spend less time just sitting and thinking. It is easy to quote data; it is much harder and more time consuming to connect the data yourself and come up with something meaningful to guide your life.

"Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?" - TS Eliot

The author has a long track record of being one of the most careful and factual journalists in the English-speaking world. His reporting on the wars in the Middle East has been consistently some of the most informative, and unlike most of his colleagues he has put his body on the line by being physically present there. So an article like the OP is more worth listening to than it might seem.

I always read carefully when I see this authors byline. If he did not quite often tell uncomfortable truths, he would probably be much more well known. As it is, he does not make the mainstream media very often.

FYI, the author, Patrick Cockburn, has a last name pronounced “co-burn”.

Obligatory relevant xkcd:

https://xkcd.com/882/

And if you think this only applies to statistics, you need to think about the problem being described here a bit more.

I think the title of this piece is right (if a bit banal and obvious), but the op-ed is rambling and not at all convincing.

The facts that support one's narrative are the most important.

Mainstream media 101

Bret Weinstein talked about his direct experience with that in a recent appearance in Vancouver. How left-leaning and right-leaning publications reported his story differently and to varying degrees of accuracy. He suggests it's exactly because of the underlying narrative of each side.

https://youtu.be/bz0oxIZ3xIg?t=311

I agree that all mainstream media, no matter if it's on the right, the left, center, government owned or not -- it's all full of bias to push the narrative each organization cares about.

Maybe 5% of the news reported is worth paying attention to, the rest may as well be garbage.

They put the narrative people want to read about. Have you seen what's on the clickbaity ads ? That's what people click on ! Most people want garbage.

Very true. I internally weep for humanity every time I use a browser without ad block.

And often lacks nuance or historical context. I can't blame people for wanting to avoid complexity, that's why headlines exist, but there will always be information left out. Even from more honest reporting.

I understand why but it still saddens me.

Maybe so, but please don't post unsubstantive comments here. Especially not generic ones on ideological topics.

https://news.ycombinator.com/newsguidelines.html

Reply on news.ycombinator.com