Digg In the Future

jamie writes "A new site called Digg In The Future - created by 17-year-old high-school student Raj Vir as a research project - says that its algorithm can predict with 63-percent accuracy what shared links are going to make it to the front page of the Digg website. (Does it allow for brigades?)"

Would be useless for ./

[Pseudonym Authority]

You don't need any software to know that this story will be posted twice.

Re:

[Drakkenmensch]

That's barely more accurate than flipping heads or tails. Plus, a trained hand can force the result it wants from a heads or tail flip every single time, too! I drove several of friends nuts after challenging them that they couldn't call a coin flip no matter how many times they tried, and forcing the wrong side to pop up twelve times in a row.

Re:

[Drakkenmensch]

i dont believe you could do it 12 times in a row. on slashdot everyone pretends to be some uber-elite-skilled nerd. You sir, need to get laid.

Of course I need to get laid. Why do you think I had the time to practice a skill like that, other than not having someone to shag?

Re:

[egcagrac0]

Around the shop, we used to gamble for sodas, with coin flips being the deciding "random" element.

I didn't buy a lot of sodas for others. I did drink a lot of free sodas.

Coin flips are certainly controllable by a skilled throw. I expect this is why at sporting events, someone other than the coin-tosser attempts to predict the result while the coin is in the air.

In a self-fulfilling prophecy...

[oiron]

Did it predict that the article on it would be on top of Digg?

Re:In a self-fulfilling prophecy...

[MarkRose]

It looks at data after the fact. That is, who shared it, who diggs it, the frequency it is shared, etc. Thus it couldn't predict where itself would end up, at least not until after it was posted, which would be too late.

It's not like a Slashdot comment that says it will be moderated +5, Informative.

Re:

[Anonymous Coward]

I see what you did there.

Re:

[Ihmhi]

Quick, someone mod him down so we can fuck around with the timeline!

Re:

[pinkushun]

Ah, clever. I see what you did there.

Re:

[Joe Snipe]

You mean like this? [slashdot.org]

Designing such a site is easy...

[slime73]

Just redirect it to reddit.com

Re:

[MichaelSmith]

Yeah their comment system looks usable. The comment system on digg OTH looks like they decided to have a comment system but set out to make it as ugly and useless as they possibly could, and more than likely put a lot of effort into it too.

Re:

[account_deleted]

Comment removed based on user account deletion

I'm better than algorithm...

[ZeRu]

I can predict that this story will be one of those reaching digg front page (maybe it already did, but digg is too broken atm for me to go there and check it out)

Phooey!

[oldhack]

I don't need no stinking algorithm to tell me that 95.1189% of "stories" posted on slashdot are crock of shit.

Accuracy?

[thegarbz]

Could also say a 37% inaccuracy. The numbers don't differ that much. They seem closer to guessing than to certainty.

Re:

[Buggz]

Could also say a 37% inaccuracy. The numbers don't differ that much. They seem closer to guessing than to certainty.

If we're talking about large amounts of data (I imagine we are) it's actually significant, law of great numbers and such. Considering a random guess leaves you at 50%, it proves his algorithm actually does something.

Re:

[maxume]

It's actually a lot better than that, there isn't a 50% chance that each article on the internet will end up on Digg (well, at least not the front page).

Re:

[Buggz]

A random guess would be 50% correct because the only two options are yes or no.

Consider a sample of 100 articles, of which only 10 hit the front page. Guessing randomly I will have predicted a correct "yes" on about 5 of them, and a wrong "no" on the other 5. Of the 90 discarded articles I will have predicted a wrong "yes" on 45 items and a correct "no" on the other 45. Thus I have 45+5=50 correct predictions and 45+5=50 wrong predictions out of 100.

If there are 3 options a random guess is just over 33

Easy. Just use these keywords.

[MindlessAutomata]

This thing probably just scours the internet for these keywords in stories:

Obama
Apple, Mac, iPad, iPod, Macbook
Marijuana
Obama
Police
Linux
Sex
Westboro
Funny picture, cat picture
Obama

it's dead jim

[crossmr]

I assume it'll be a lot less stories. The new digg revamp is absolutely terrible. It's impenetrable. Most of the useful features before are gone, it's half broken, with comments not loading, or links to your own comments not working. It shows you basically none of the information it showed you before, the new main feed is completely out to lunch. Apparently the "most recent" story on digg was submitted 2 days ago, and I know I saw it on the front page yesterday since they busted it. So digg is telling me since the upgrade, no one has made any story popular.

Re:

[crossmr]

I would, but I can't figure out where to do that. I can't get through the damn interface.

Re:

[PopeRatzo]

Everyone always complains yet they keep going

I assume that by "everyone" you are referring to yourself and the mouse in your pocket?

Re:

[elrous0]

His name is Algernon!

Re:

[not_a_product_id]

"Everyone always complains yet they keep going and eventually get used to the site again"

Not this time. Before I've gone on after a redesign and thought it looked 'odd'. This time I just thought it looked shit

Re:

[EngivalX]

I actually thought I accidentally went to facebook when it first loaded. Every other time they redesigned the site, it was pretty easy to shrug it off. This redesign is horrible.

Re:

[elrous0]

Digg is the classic case of a site that started out kind of useful and interesting. But it quickly got way too much press and hype, then subsequently expanded itself into something that tries to be everything to everyone--and lost its focus and value in the process. The comments at Digg have always been pretty lame compared to more mature sites like slashdot (and here you thought you'd never see "mature" and "slashdot" in the same sentence). But it was a pretty good place for tech articles before they start

Ditto.

[antdude]

Since end of 2005, I was a big Digger until yesterday's v4 release to production. I even got a free T-shirt for getting 300 stories promoted/vote up onto its home page. They didn't even change anything from others' and my feedbacks. I asked support about being to filter categories (e.g., World News and Politics), but they said it doesn't exist. No upcoming, no following, can't show 100 comments per load, etc. Well, frak that.

Why can't they keep the old design for old school users like /. did? If /. didn't h

on digg because predicted to be there

[ludwigf]

First setup a site that dispays random digg entries with "digg it" links, name it "digg in the future" and get some publicity. As the stories displayed get more diggs, your site gets accuracy > 50% even if selecting random entries.

Re:

[halfaperson]

your site gets accuracy > 50%

I put on my robe and wizard hat.

So...

[cybrodroid]

it scans reddit?

Re:

[account_deleted]

Comment removed based on user account deletion

Re:

[Z8]

Basically, yes. From TFA:

Vir says that since many of the links that make it to the front page of the site have already been shared on other social networks, the Digg In The Future software looks at frequently shared URLs from Twitter and gives those added weight.

Is his program called Reddit?

[ZuchinniOne]

Because that seems to predict what Digg will do in the future too.

Re:

[matt_gaia]

Because that seems to predict what Digg will do in the future too.

with a much higher percentage than 63%, I'll bet...

Observing disturbs the observed

[pinkushun]

... factors taken into consideration are what I like to call “power submitters” ... and “power diggers”... The algorithm also relies on other factors, Vir says, including the time of day (since stories submitted in the early morning hours are unlikely to reach the front page) and whether the link comes from “preferred” sites that appeal to Digg users

If people adapt their submission procedure to increase their chance of it reaching front page, it will drop the algorithm's accuracy initially, as submitters that don't fit the user profile suddenly match the 'best submission time criteria'.

The curve will then level out, and climb again, eventually increasing the algorithm's accuracy beyond the initial point as more people conform to the best criteria.

Digg has no future

[ThatsNotPudding]

Not after their MeTooBook redesign. /checking out reddit

...and top-40 radio predicts iTunes top sellers

[Powercntrl]

Digg seems to be heavily slanted towards news that falls into the "odd/strange" category and whatever useless viral videos happen to be hottest at the moment. It's the top-40 station of news sites, and it doesn't surprise me much that what will be posted on Digg can be easily predicted by looking at other social networking sites.

The new Digg is terrible

[British]

The comments won't load. They also try too hard to make it like twitter(following newsfeeds, etc). It lost any unique charm it once had.

Re:

[denis-The-menace]

New Digg is annoying, confusing and the content is static.
The content is SO static that /. puts out more post than New Digg

Part of algorithm leaked

[Megane]

if (poster == "MrBabyMan") { return 100.0; }

And wow, I just checked their "front page" for the first time in a couple of months. This is not even the Digg I knew that I barely cared about any more.

A day late and a dollar short

[Lucas123]

Digg just completely revamped its site, basically killing Digg brigades and turning itself into a giant RSS feed.

Re:

[denis-The-menace]

Maybe the RSS feed work but the web interface is almost static.
New Digg, in old Digg lingo: "sucks donkey balls"

hii

[dudley28]

Hi, I just found your blog via google. Your viewpoint is truly relevant to my life currently, and I’m really happy I discovered your website. http://www.happyfacetv.com/ [happyfacetv.com]