200 Countries, 200 Years, 4 Minutes | Hacker News Reader

200 Countries, 200 Years, 4 Minutes

blog.sokanu.com

91 pointssthomps21 comments

The original video link (http://www.youtube.com/watch?v=jbkSRLYSojo) was submitted 5 days ago (http://news.ycombinator.com/item?id=1954315). Then a slightly different url (http://www.youtube.com/watch?feature=player_embedded&v=j...) was submitted 10 hours ago (http://news.ycombinator.com/item?id=1973202)

It seems quite unfair to me that the person who submitted this content first, and did so without linking to their own blog, gets none of the recognition and karma. While 100% of sthomps submissions (http://news.ycombinator.com/submitted?id=sthomps) are links to his own blog, where he mirrors or links to the original content.

I just said it (http://news.ycombinator.com/item?id=1974170) but I'll say it again: I think this is a flaw in HN's ranking scheme. If the original link isn't upvoted quickly, it never gets on the front page. After that, the rewards (recognition, traffic, and karma) go to people who mirror the original content or simply link to it from their own websites.

How would you improve the ranking system?

I don't know of a complete fix, but a start would be to rank based on upvotes in the past hour (or 2 hours, or whatever works best) instead of since submission. This would allow links submitted days ago to end up on the front page if they were recently upvoted a lot. The new algorithm could be put on an alternate front page (similar to http://news.ycombinator.com/classic) until it was tweaked to get the desired result.

Since detection of duplicates and content mirrors is currently done best by humans, I think another aspect of the solution is cultural. If we want to stop this sort of thing, people need to point out mirroring/linking and flag submissions. If we do it enough, the incentives will change to discourage this behavior.

Sounds like a good solution that would also decrease the asymmetry between first submission, new submission and up-voting. (There's one asymmetry that we'd like to keep: Who gets the karma, but the asymmetry of those actions on ranking aren't as interesting.)

I would final duplicate content and automatically merge likely submissions together. Spread the karma between the first 3 submitters in a 50/35/15 manner, so that the latter two still have incentive to spread the link outside of HN.

That sounds easily abusable to me. People would just submit a duplicate article when they see an article gaining upvotes.

If there were a "dup" button that let you actually mark what's a duplicate of what, you can try all sorts of things to shift karma and points around.

EDIT: I fleshed this idea out a bit and submitted it as an Ask HN: http://news.ycombinator.com/item?id=1975950

I totally agree with you. In terms of not getting the original content not getting the recognition it deserves, this is exactly right. We write a lot of original content, that unless it gets many votes very quickly, does very little on HN. And often someone may link to it, or re-blog later on and get credit.

When it comes to videos, especially TED talks, etc... we try and put up really high quality stuff that our readers will appreciate. Is it fair that we can get traffic off of other people's work sometimes? Probably not. TED talks, and this video from today are a tremendous amount of work and we should probably not always link to our blog.

I will say, however, that on HN, I have more than just this one account. My other "dummy" accounts are used to submit others content etc... Although this account can start doing that as well. Anyways, thanks for the comment, it is very true.

He gave a similar presentation at TED. It's a lot longer though.Check it out. -> Hans Rosling shows the best stats you've ever seen: http://www.ted.com/talks/hans_rosling_shows_the_best_stats_y...

More stuff from Rosling: -> http://www.ted.com/search?q=hans+rosling

In the documentary Swedish national television made about him, he says this regarding that TED video:

"If I count the number of minutes I've been teaching, this video teaches more than anything I've done in my life. All the articles, books, courses... None of it can compare with this one video. That's a humbling experience."

The documentary is now available on youtube, with English subtitles: http://www.youtube.com/watch?v=y_7howQzatw

I knew I saw something like this before. Thanks for reminding me.

His visualization reminds me A LOT about gapminder (http://www.gapminder.org/world/), if anyone wants to play with this kind of data with this kind of visualization.

Shouldn't be surprising since Hans Rosling (the guy in the video) co-founded Gapminder.

You can also build pretty much the exact same visualization in Google's public data tool: http://tinyurl.com/2az8dm2

This data, and tons more, has been available through Google's public data explorer (http://www.google.com/publicdata/home)

Albeit, without an enthusiastic British (Edit: Swedish!) guy's explanations...

Swedish (http://en.wikipedia.org/wiki/Hans_Rosling)

It's interesting that around the 70's the Asian countries (mostly China) spiked in life expectancy but still are at the same income level. There's a year where China's life expectancy just drop dramatically. I wish he had explained why.

Just found out that it's the Great Chinese Famine (http://en.wikipedia.org/wiki/Great_Chinese_Famine)

Nice use of sound effects too!!

Income is not linear. So the disparity in income today is less evident. Thus helping paint the picture of a perfect future.

That and the selection of medians for thing like income and life expectancy makes that a very poor data viz choice. But a very good motivational talk. Really Ted material...

Income is not supposed to be linear. It makes it easier to see the relationship between log(income) and life expectancy visually.

Reply on news.ycombinator.com