After many requests, the graph is now available as a poster at adumb.store/. A portion of proceeds from each sale will be donated to the Wikimedia Foundation.
This comment was made 22 hours after the main comment: The original comment was: I love how almost all dead-end articles you mentioned have no longer been dead-end within a day of this video being uploaded. The second reply to this main comment is: This video was posted 8 days ago (sent 11 hours after original comment for anybody who is curious in the future). My comment: The Great Sun approaches. It grows. It spreads. Faster day by day. One day it shall expand to the point that it has exhausted all of its energy. Then the three inner planets shall be consumed in the fireball and Enceladus shall have liquid water. After mars with its rings shall cry. For three of its friends have died. And the sleeping monster shall fizzle away. And the life on Europa shall freeze and die. You atoms shall be consumed in the fireball. Unless…
This comment was made 2 hours after @@gregoryturk1275 's reply. I just want the exact time of the original comment to be documented for no reason in particular
Dude you should seriously submit this graph as a series to a modern art museum!! I know it sounds strange, but it’s so unique, so visually interesting, and there are so many parts of it that reveal truths about society, politics, human behavior, etc. I know so many galleries that would just love to have this as a series!
I was thinking it would be a great poster or graphic for merch to support the creator. But I think you are more on point, that a modern art museum would be an amazing place to display the visual graph and also an interactive version with the concepts explained in the video.
This is so wild to me because I'm one of the people who's edited that article but before this video came out. Wikipedia shows you the views on articles you've edited, so I was incredibly confused as to why the Fanta Cake article was abruptly getting oodles more than Lancelot's. Turns out it was this video! Another fun thing is that back when you started this in October, those two references on the Fanta Cake page are my contribution to the article. Small world!
@@Ten_Thousand_Locusts Rude lol. I'm glad she edited the Fanta Cake page, because I immediately saw that there were so many words/articles that could be linked to. I wonder how many "dead-ends" aren't real dead-ends, but just articles where nobody bothered to add any links. Any article, no matter how small, has potential for a link.
@Ten_Thousand_Locusts I think this was supposed to be a hate comment or something but I'm gonna be real with you I laugh every time I think about it, that's funny as hell
I work in graph and graph database research and i have not seen such a beautiful, succinct and well presented graph ever. I think an average person would never fathom the amount of computer science that backs this video up. Huge congratulations to the creator.
I completely agree, but I actually watched the video with one main question, which I believe didn't get answered: How are the articles or regions positioned on the X and Y axis on the graph?
The colors, the tone of narration, the jazz. It makes it feel like an instructional/educational video from the late 90s to early 2000s. Something I would see in a slow school day. I love it.
I miss the 16mm educational films of my 1970s youth. The exciting sight of the projector as we took our seats in the morning, the film canister(s), the teacher threading the film in the projector, the film's 3-2-1 countdown, the narrator's thundering voice exiting the projector's speaker, the anticipation of the film's subject. My favorite film, which I'd like to find as it shaped my skepticism, was about the filming tricks that toy makers used to make their products seem more than they really were.
Once when playing the wikipedia game in history class, the target article was "the French Revolution." We all had to start on a random page in order to demonstrate that essentially everything in the world is influenced heavily by the French Revolution. Some lucky duck's random article was "France"💀💀💀💀
I agree that the world was greatly impacted by the French revolution, but it's a very bad way of showing it, given it work with *any* page... It's a know "paradox", there is (almost) always less that 7link between two things: it's almost 100% certain that you know someone that know someone that know someone that know someone that know Jessica Alba (or anyone else). It's the same for wikipedia. It's mathematically proven that you can find *any* page in less that 7clics. Ps: I commented before finishing the video, but it's a good demonstration of the 7links rule. You can clearly see on that bell graph that almost all the articles were linked after 7clics.
Community 27 (Figure skating) is truly special. Almost all major Figure skaters have similarly formatted wikipedia pages with quite detaile info about their skating carreers. This hints towards that they have been majorly edited or set up by a very small group of dedicated fans.
Wrestlers articles used to be similar… and then it was hijacked by an asshole mod on Wikipedia who wanted users to use their shitty wikia for their information. Wrestler articles used to have their movesets, their finishers, their entrance songs, etc.
Putting the graphs shown in the video aside, just wanted to say this is a masterpiece of youtube storytelling. You had endless information to talk about and put together an incredible concise and compelling presentation. Kudos!
@@herpederpe4320 not everyone is cut out to be a Wikipedia editor. the folks who're self-aware of this fact thus respect Wikipedia and-in a way-help it by not breaking anything
Just wow. Bravo. As a hobby programmer and data scientist I have a glimpse of what it took to do this, but know the real effort and magnitude far exceeds that idea. You hide the complexity (and I'd suspect quite a few brutal bugs to solve) incredible well in your simple yet entertaining walkthrough. Incredible!
You just took a topic that I would probably spend my life without ever giving a single thought to, and made a video that was an absolute joy to watch. If there's a TH-cam Hall of Fame, this one belongs in it.
This is even more impressive. He talks about something I've seen done to dead by lots of other people (see his cheeky reference to mildly interesting reddit) and it's still new and fresh to me. I almost don't watch this article but when I finally budge and I don't regret it.
@@ameltedTSP is a completely different problem; to find pages 10 degrees of separation or greater, you could just use a breadth-first search, similar to what he showed in the video when analyzing degrees of separation. This is possible in polynomial time (I believe O(n^2) in the worst case, but feel free to correct me if it's wrong; it most certainly is polynomial, however).
@@ReliableExcavationDemolitionyeah and you probably take 20 clicks to bridge a 4th degree or so relation. Just because you didn’t personally find the shortest link doesn’t mean it’s the shortest link. 10 degrees or more would be absolutely brutal for a human
This sort of thing deserves to be an actual feature on Wikipedia, it's so well done. Would be super cool to play around with an interactive version of this, or have it regularly updated to take a timelapse of how it changes.
On the topic of degrees of separation and the longest paths. All we really need to do here is to add more links between articles and these numbers should decrease. People are already editing articles with the help of this video, and people are definitely interested in shortening these paths.
Absolutely fascinating video. The Fanta Cake bit at the end is a great example that for most orphans or dead ends, it's a matter of what could be considered bad article formatting/linking. I looked at the Wikipedia article for William Acton (senior) and someone has already destroyed the Acton group solitary-ness by adding links to the page for "Politics of England" off of the phrase "English politician" Great video! Fantastic work : )
@@opensocietyenjoyer not really, I find it beautiful as a work of art, but knowing that it’s actually a data graph with millions of nodes makes it SOOO much better!
One of my favorite wikipedia trivia bits is that, at least for a long time, by clicking the first non-disambunction or pronunciation link, you will eventually end up on philosophy. I think some of the natural sciences end up being recursive now but it used to all link to philosophy.
I remember doing this!!! I didn't hear about it from anywhere I just clicked the first article link (non pronunciation or disambiguation) and I always always ended up at philosophy where it recursed!!!
Part of the reason is that a couple of people found this fact, then checked it. The 1% that didn't end on philosophy where changed to end at philosophy.
I remember when I discovered this a killjoy had cut the link between knowledge and philosophy and broke the chain. Then they would revert any edits that added the link back
This is my favourite example of the power of being able to explain niche things disconnected from the general public’s interest well. You turned a seemingly useless thing: a graph of Wikipedia articles into an amazing, engaging, and thought provoking, inspiring video, highlighting each of the things that you’ve explored, with perfect transitions for dramatic effect and amazing animations and visuals. This is a mind-blowing video, keep it up!
Canada and Hockey being one community/category is amazing. The fact that you know 100% for certain that the article for Tim Hortons is in that category is just the glue of perfection.
@@MrMickio1So famous I’ve literally never heard of them. I didn’t even know they were a genuine person until this thread! (Then again, that could just be a logical side effect of my complete lack of interest in sports in general, and the complete lack of TH outlets anywhere near where I live!)
Watching this for a class and when you mentioned dead end and orphaned articles it reminded me of pre- synaptic and post- synaptic neurons, since they’re neurons which have no input from other neurons or don’t output to other neurons respectively (I.e. sensory neurons and motor neurons (?)). It’s really interesting how universal networks are
I love how you went over so many different things related to the graph in this video. It really satisfied that curious urge you get when learning about something new
I wish social media, especially platforms other than TH-cam, had more fascinating stuff like this. I want to experience that childlike curiosity again!
Seriously, he totally downplays the significance of what he's accomplished with this graph. There are so many fascinating insights here, not just about Wikipedia, but English-speaking culture.
"A place where anyone has the power to free an article from an existence of solitude" is such a beautiful line I did not expect to hear in a wikipedia graph analysis video
As a former Wikipedia editor, this is really cool to see! I regularly make use of the SpecialPages Orphaned, Deadend, Unconnected or Redirect to try and improve the linked data structure. I would really love seeing Wikipedia take this project as a source for more linked improvements!
this project just does not use the right data- this optimal paths like shown in this video so often just not exist this 1 example when path ends with buffalo is acutally findable but the other one is not : flood does not link to this list of non-water floods this link dump data is just not all that great, it shows many links that are not visible in the page. basicly this video is wrong, the guy would need to have his own script that would scrap all those links that are shown in the page, those links in wiki dump data are often very random.
@@koks49045 The flood article does link to the list of non-water floods. In the first line, in fact. The text of the links is not always the name of the article they point to. If you want to find a link in a page you can search the URL in the HTML source.
@@koks49045the 120iq understander has logged on we just having fun bro u might be 140 for all i know, 120 is like a step below the average phd and 140 is like a bright student at a top program (a step and a half above the average phd)
I think a way to find the absolute longest path would be to start at the "list of highways numbered 825" article and start mapping pathways backwards from there. Whatever you end up with, you can add the links from 825->999 to that
1. How many Wikipedia editors are now looking to make sure all pages are linked. Eliminating orphans and dead ends. 2. Wikipedia should add this somehow to give a visualization of its vast knowledge.
long-time wikipedia editor here. There have existed entire projects which have tried to eliminate orphan and dead end articles. At least for orphan articles I believe there are hidden categories that flag them. I think the OP could've made use of Wikiprojects in order to link related articles together instead of just using outbound links, though I guess that if his analysis is based on the wikipedia game that it makes sense why he wouldn't. Wikiprojects already give you something similar to his idea of "communities" of articles
@@drbuckley1 Won't ever be fixed. People try, and I salute them for it. Wikipedia is a guidebook not an answer book. Wikipedia usually gives enough of an overview, correct or incorrect, to start looking up information elsewhere. Trying to make it an answer book is impossible.
I originally thought this video wouldn't be too interesting, but I clicked on it out of curiosity (and like the saying goes, you had my curiousity but now you have my attention). The amount of detail, effort and production value put into this video astounded me and I was hooked. I also appreaciate the informative description. Thank you for this wonderful video. The only thing missing is the raw data and code.
I checked them out. Seems they are no longer an orphan comunity, as many links were introduced in the article, like"Englis Politician" and other nonsense. Or maybe I misunderstood the situation completely...
@@zo2oI haven't checked, but to clarify, links included in the articles don't exclude them from orphanage, it's the fact that no other articles link to them.
The issue not explored is that there are Wikipedia editors who have an intense interest in one topic or narrow groups of topics. That shapes the style and linking for many groups of articles which become mainly the work of a single author. At 3:50 it's not at all surprising that the principal authors of articles on Association Football are completely different from the authors of Gridiron Football; each set of authors will know comparatively little about the other topic and will therefore be far less likely to cross-link them. If you want to look deeper into how editors shape articles, you will want to study the various Wikiprojects -- groups of editors who work together to improve a specific topic.
14:56 in solving a rubiks cube there is a similar concept called "gods number" which is the number it takes to go from any scrambled position to solved. It was calculated a few years ago and came out to be 20. That's why I was really suprised when you said the biggest link you found was 166 which is crazy to me, but makes sense. Especially since Wikipedia isn't as easy mathematically explained as a rubiks cube. Awesome Video!
I can’t believe I watched a 20 minutes video about Wikipedia graphs to be finally be surprised with Fantakuchen as one of the most special articles. I just had Fantakuchen on Easter this year and it was one of my favorite birthday cakes all childhood long (next to Donauwelle, wave of the river Donau). Applause!
Lmao typical german, he wants to make clear that he's from germany. Writes fantakuchen despite being able to write everything else in english, then donauwelle and then the pretentious applause. All germans are the same, I have no idea why the chauvinism.
I have a specific memory of hearing about Fantakuchen when I was in like second grade, thinking "huh?!?" and then it never coming up in my life until now. Fits the video perfectly
Dude. 200 views in an hour on this is *criminal*, this should be blowing up. I wouldn't be surprised if Ye Algorithme picks this up and it's in the 100,000s sooner or later.
Your editing and sound editing in this highly commendable by the way. Extraordinarily smooth and intentionally timed without being too obnoxious in anyway.
In the segment of “longest path” you should’ve more clearly said the “longest shortest path” (or: shortest path with the most nodes.) since obviously longer paths can exist if you just deviate away from the goal
That seems like a given because if that weren’t the case almost any article could loop forever between 2 or 3 things. If you consider the path as not being able to repeat pages, that would likely be several million articles long.
yea, but also to actually find the longest shortest past and not just guess that it was the 166 one shouldn t he had just picked a random page, do a bfs, go to the the page which was the farthest away and do a bfs again? And do it for all conex components
I'm actually kinda interested in the path since "list of highways numbered 1000" and "list of highways numbered 999." point at each other. If I'm not making a mistake this implies the same length from the start to either "list of highways numbered 998" and "list of highways numbered 1000." otherwise the path would be shorter or going to the next article would be a longer shortest path.
This has be the quickest growing channel I’ve come across in the last year with only 6 or so videos. It just goes to show that if you have solid production with an intriguing subject that relates to many different people whilst hitting all the marks with title and thumbnail you too can make it. This channel is a perfect template. If you start a new channel and have several initial videos flop with only 500 views there is a good chance you’re gonna struggle to ever breakout of the algo blackhole. You must hit it off from the first video and by hit off I mean at minimum 20k views
Regarding about Fanta Cake page, the page has just updated again several times since the beginning of April, now the page is even expanded with more information in it (Soda Cake section added), "Fantakuchen" redirect page removed, and is neither Dead End nor Orphan page anymore, much like a normal article now!
You deserve an award for this work and this video! It is really interesting how some of the "communities" are structured. BTW: If you write up your findings in a Wikipedia page, linking to all the orphans in your graph, you would drastically reduce the number of orphans in Wikipedia.
I've commented this on another comment as well, but what about a page "List of Wikipedia Orphan Articles"! Same can be done for "List of Wikipedia Dead End Articles" lol
@@lajawi. Both of those pages already exist! The reason orphan articles don't become automatically un-orphaned from being in that list is because the list is not in the "main" Wikipedia space, but a special section of Wikipedia that has editing guides and the likes, and so it doesn't really count as a link
ahí la tiene Maradona,⚽ lo marcan dos👥, pisa la pelota Maradona🚶♂, arranca por la derecha↖ el genio del fútbol mundial🏃♂, deja el tendal y va a tocar para Burruchaga... ¡Siempre Maradona!🎖 ¡Genio!😯 ¡Genio!😯 ¡Genio!😯 Ta-ta-ta-ta-ta-ta-ta-ta...🗣 ⚽🥅Gooooool...😱 Gooooool...😱 ¡Quiero llorar!😢 ¡Dios Santo, viva el fútbol!🗣🔥🔥🔥🔥🔥🔥🔥 ¡Golaaazooo!🔥🔥 ¡Diegoooool!🔥🔥 ¡Maradona! Es para llorar😭, perdónenme... Maradona, en recorrida memorable, en la jugada de todos los tiempos... Barrilete cósmico🌌... ¿De qué planeta viniste para dejar en el camino a tanto inglés🪐👽, para que el país sea un puño apretado gritando por Argentina🔵⚪🔵? 💯Argentina 2 - Inglaterra 0. Diegol, Diegol, Diego Armando Maradona... Gracias💫, Dios, por el fútbol⚽, por Maradona🥇, por estas lágrimas💦, por este Argentina🏆🏅 2-Inglaterra 🥈💩0” Pelota para Xavi⚽, asistencia de Xavi👀 en esta pelota para Messi👀 Messi🏃♂, Messi⚽, Messi🏃♂, Messi⚽, Messi🏃♂, Messi⚽ y inmenso Messi🏃♂, Messi⚽ Encara Messi🏃♂, Encara Messi🏃♂, Encara Messi🏃♂, Encara Messi🏃♂, Encara Messi🏃♂, Encara Messi🏃♂, Encara Messi🏃♂, Encara Messi🏃♂, Encara Messi🏃♂, Encara Messi🏃♂, ⚽🥅Gol, Gol Gol Gol, Gol, Gol, Gol, Gol, Gol, Gol,⚽🥅 Gol, Gol, Gol, Gol, Gol, Gol
The articles on the Altons made me think of constelations and galaxies. Like how most stars are bound together by gravity in galaxies and clusters, but then you have intergalatic or rogue stars, that are just not bound to any galaxy. I just find neat how we can find similar patterns in so different parts of reality.
The idea of an intergalactic rouge star is kinda terryfing. Like, how did it even get there?? Why did it just get lost. Is it just incredibly ancient and has just always been there since the beginning of the universe? Or did some ungodly cataclysmic event rip it out of it's galaxy? How do you even rip a star out of it's own galaxy?? It's easy to imagine rouge planets. You hear about them all the time. Galaxies are relatively dense so it's easy to imagine how a passing star could rip a planet out of it's home system. But even then, rouge planets still exist within their own galaxies. What ungodly apocalyptic catastrophe has to occur for a star to end up in intergalactic space???
@@a2izzard It's still crazy that their sent flying off all the way into intergalactic space. Galaxies are humongous, you'd think that along the way the star would eventually get attracted by the gravitational pull of the galaxy itself. Then again, for as big as they are, galaxies are still mostly empty. It's like how neutrinos can seemingly phase through matter. They are just so tiny that things that appear solid to us just aren't to them. The gaps between atoms are like the gaps between planets at that scale. I guess stars are just so tiny when compared to an entire galaxy that they can just pass through without interacting with anything.
I'm chronically online goon and spend avg. of 5 hours a day on TH-cam for like 8 years, but to find a the video so well crafted and with topic that interesting.. man good job
The video and work behind the graph itself is commendable, but the visualizations throughout the video is just as commendable. A lot of quality work, good job!
Wow! This is probably the most beautiful presentation of connected data I've ever seen. Would be awesome if you were able to make this into an interactive website 👀
Finally a video about a very interesting topic that is not clickbait and really well presented. First one in quite a while after I accidentally flooded my recommended with meme clips...
You mean the one where you need to search for links to get you to a goal article? You can cheat by editing the original article and adding the link to the goal article
@@badgermcbadger1968 you don't have to play a game you find boring though xD you can just tell your friends no, it's not like it's a tournament with prize money on the line or anything, all you'll get by cheating is break the trust of your friends. Idk that just seems like a really strange and low stake situation to cheat in.
the problem with this game is most articles are some random weird stuff like list of something or a village, if there was a fillterted version where only articles that are like concepts or some important stuff like country existed, then it would make sense, but also like each article should have atmost 10 hyperlinks that link to stuff that is rly related
this has so much untapped potential. if no company was looking into this before they certainly are now, most likely from commercial companies and software companies
It would be genuinely incredible if you made this interactable so you could filter out different communities, zoom in, look around, click on dots to go to a page, and whatever else you can think of. I would love to explore this graph
I think this might be one of my favorite vids now. From start to finish, the writing and pacing captivated me. The visuals were stunningly animated, the music was bopping and the whole time I wanted to know more. I hope your channel becomes a staple in the "computer science experiment" side of youtube
This was one of the most interesting videos I've ever seen. Absolutely fascinating! This could easily have been a master's thesis. And the production and presentation were stellar!
3:58 I'm glad you brought up "sports" when you did. As an avid (maybe even "rabid") sports fan, my first thought upon hearing "44 communities" was "that seams very low!" When you consider Wiki covers EVERYTHING, my initial thought was "we'll, then 1 of the 44 must be a broad and all encompassing 'sports' community. "
44 does seem pretty low overall, but it gets slightly easier to understand the more he breaks it down. I'd still like to get my hands on this data and look at it myself. But the reasons why "sports" isn't a singular community is because the algorithm he used isn't connecting things by human defined concepts. It's connecting articles based on which other articles they link to. You'll be in a tough spot if you were to try to find a connection between, let's say Darts and Baseball. So those two get their own "communities." Each sport is pretty insular when you really look at them. Even between the "big 3" in the US: baseball, basketball, and football. As huge as these three are, there's still very little crossover between them. Very few players have played in more than one of these in their professional careers and so links between the three will be very few and far between on the Wiki. And if the three most popular sports on the English speaking Wiki aren't very well connected then you can forget it for any of the other sports as well.
You can "tune" algorithms like this to be stricter or more lenient. I don't think 44 is some objective figure that someone doing a similar analysis would come to as well - presumably making the requirements stricter leads to more communities, and more lenient leads to fewer communities, similar to a fuzzy select in an image editor.
Thank you so much for making this video. I've wondered about exactly these questions for years, especially how "connected" wikipedia is as a graph, and suddenly your video popped up in my feed and answered basically all of them. I really hope you'll make a second part and extract some more info from all the data you've gathered, I guess there are definitely some nice graph-theoretical features left to explore here! (for instance how tightly linked the wiki graph is, are there bridges whose removal would disconnect the (main component of the) graph? Does this only happen at dead ends, etc.)
Amazing work on all this stuff! Also the music is somehow weirdly perfect for this video like you're some random dude that I got stuck with on an elevator and the elevator music is still playing in the background while we just get into this long conversation about Wikipedia.
I'd love to see a video about the technical side of this project. The crawler you used, the database it's stored in and the format in which it is stored, the queries and algorithms used to find the paths in each of your headings. Super interesting project, thanks!
![[MV] A leap of faith...ความเชื่อใจครั้งสุดท้าย (From "Petrichor the series")](http://i.ytimg.com/vi/U1piZH2CNXA/mqdefault.jpg)
![The moment we stopped understanding AI [AlexNet]](/img/n.gif)
After many requests, the graph is now available as a poster at adumb.store/. A portion of proceeds from each sale will be donated to the Wikimedia Foundation.
ty baby
peak
Buying it rn :D
We meant release the code to make it :P
Are you willing to share the data for academic use?
Making a Wikipedia article about an orphan article creates a paradox where any orphan example given in the article automatically stops being an orphan
Make it happen
Lmao
I'm pretty sure there is already a list of orphan articles. But I don't think the guy counted lists or similar.
_Bertrand Russell breathing heavily in the distance_
we could call it the adoption service
I love how almost all dead-end articles you mentioned have no longer been dead-end just within a day of this video being uploaded.
This video was posted 8 days ago (sent 11 hours after original comment for anybody who's curious in the future)
This comment was made 22 hours after the main comment: The original comment was: I love how almost all dead-end articles you mentioned have no longer been dead-end within a day of this video being uploaded. The second reply to this main comment is: This video was posted 8 days ago (sent 11 hours after original comment for anybody who is curious in the future).
My comment: The Great Sun approaches. It grows. It spreads. Faster day by day. One day it shall expand to the point that it has exhausted all of its energy. Then the three inner planets shall be consumed in the fireball and Enceladus shall have liquid water. After mars with its rings shall cry. For three of its friends have died. And the sleeping monster shall fizzle away. And the life on Europa shall freeze and die. You atoms shall be consumed in the fireball. Unless…
This comment was made 2 hours after @@gregoryturk1275 's reply. I just want the exact time of the original comment to be documented for no reason in particular
what?
@@micheal5117 yes
Dude you should seriously submit this graph as a series to a modern art museum!! I know it sounds strange, but it’s so unique, so visually interesting, and there are so many parts of it that reveal truths about society, politics, human behavior, etc. I know so many galleries that would just love to have this as a series!
honestly, this is something genuinely worth of the name *modern* art
@@DarknessDShadow yeah it looks like paint spilled all over the place so its 100% worth the name of modern art tbh
I was thinking it would be a great poster or graphic for merch to support the creator. But I think you are more on point, that a modern art museum would be an amazing place to display the visual graph and also an interactive version with the concepts explained in the video.
Imagine having this with a UI would be interesting. Letting you cycle through the categories or showing all of an articles specific connections.
@@electralumen165this is so important!
This is so wild to me because I'm one of the people who's edited that article but before this video came out. Wikipedia shows you the views on articles you've edited, so I was incredibly confused as to why the Fanta Cake article was abruptly getting oodles more than Lancelot's. Turns out it was this video! Another fun thing is that back when you started this in October, those two references on the Fanta Cake page are my contribution to the article. Small world!
Yeah, you look like someone who would edit an article about Fanta Cake. Checks out.
@@Ten_Thousand_Locusts Rude lol. I'm glad she edited the Fanta Cake page, because I immediately saw that there were so many words/articles that could be linked to. I wonder how many "dead-ends" aren't real dead-ends, but just articles where nobody bothered to add any links. Any article, no matter how small, has potential for a link.
@@uamsnof rude how?
@Ten_Thousand_Locusts I think this was supposed to be a hate comment or something but I'm gonna be real with you I laugh every time I think about it, that's funny as hell
@@charlottemacmillan4845 it honestly wasn't, just a stupid little comment I thought up after seeing that you had a pfp of what I assume is yourself.
I work in graph and graph database research and i have not seen such a beautiful, succinct and well presented graph ever. I think an average person would never fathom the amount of computer science that backs this video up. Huge congratulations to the creator.
Agree!
Many maps are useless until you have the key.
lol same, im wondering how many days it took to run the visualization. Not to mention editing any errors notes 😅
I completely agree, but I actually watched the video with one main question, which I believe didn't get answered: How are the articles or regions positioned on the X and Y axis on the graph?
@@H2-HQ There are no X and Y axis on a graph, you can arrange nodes in any way you like
Explaining overly complex charts over smooth jazz is my favorite TH-cam genre.
Royalty free music accompanied unhinged rants are a close second
I swear the light jazz helped me understand it better lol
i didnt even notice it lol
@aaronnekrin5150 i wonder if school played smooth jazz over a complicated class
I think you'll like the vaporwave music genre. Have fun in that youtube rabbit hole
Petition to run the code to make this graph yearly to see how it changes.
I second this
Now that's something that will sorta reflect the human knowledge base and how we evolved to an even tighter better connected world.
I'm in
Signed
Should run continously to see how it changes continously.
The colors, the tone of narration, the jazz. It makes it feel like an instructional/educational video from the late 90s to early 2000s. Something I would see in a slow school day. I love it.
More like an instructional video made in the 70s and endlessly regurgitated well into the 90s/2000s as they kept rolling out things on tape.
With a splash of computer science skill 999. Like PHD level analysicssoring been happened, wha!!?
Got that vibe too) Pretty sure that was the author's intent
This has Jon Bois all over it
I miss the 16mm educational films of my 1970s youth. The exciting sight of the projector as we took our seats in the morning, the film canister(s), the teacher threading the film in the projector, the film's 3-2-1 countdown, the narrator's thundering voice exiting the projector's speaker, the anticipation of the film's subject.
My favorite film, which I'd like to find as it shaped my skepticism, was about the filming tricks that toy makers used to make their products seem more than they really were.
Another thing I’d like to point out is how 97% of all Wikipedia articles will end up in philosophy if you kept clicking on the first hyperlink
Try also Wiktionary, and click on hyperonyms
fun!
this was surprisingly true. I tried doing this from the Fanta Cake article and ended up on 'Existence'
when i tried it it just kept going in a loop at science
Edit:
its not only at science but also at other things
that's probably because 97% of articles start with the name of a language as the first hyperlink
Once when playing the wikipedia game in history class, the target article was "the French Revolution." We all had to start on a random page in order to demonstrate that essentially everything in the world is influenced heavily by the French Revolution. Some lucky duck's random article was "France"💀💀💀💀
Oh that just ain't fair 😂😂
@@DAMfoxygrampa Well, some people just get luckier than others. A lesson from the french revolution.
@@echoplots8058 pfffffffffffffft
who would think that france was impacted by the french revolution
I agree that the world was greatly impacted by the French revolution, but it's a very bad way of showing it, given it work with *any* page...
It's a know "paradox", there is (almost) always less that 7link between two things: it's almost 100% certain that you know someone that know someone that know someone that know someone that know Jessica Alba (or anyone else).
It's the same for wikipedia. It's mathematically proven that you can find *any* page in less that 7clics.
Ps: I commented before finishing the video, but it's a good demonstration of the 7links rule.
You can clearly see on that bell graph that almost all the articles were linked after 7clics.
Community 27 (Figure skating) is truly special. Almost all major Figure skaters have similarly formatted wikipedia pages with quite detaile info about their skating carreers.
This hints towards that they have been majorly edited or set up by a very small group of dedicated fans.
Didicated.
Dictated.
Wrestlers articles used to be similar… and then it was hijacked by an asshole mod on Wikipedia who wanted users to use their shitty wikia for their information.
Wrestler articles used to have their movesets, their finishers, their entrance songs, etc.
The same for alpine ski racers at the World Cup level. I don't know about the other snow sports sanctioned by FIS.
@@vcom741 Isn't that more appropriate for a specific Wiki than Wikipedia? The latter just tells you what wrestling is.
Putting the graphs shown in the video aside, just wanted to say this is a masterpiece of youtube storytelling. You had endless information to talk about and put together an incredible concise and compelling presentation. Kudos!
Too bad I didn't see a "Tetris" community in the top 28. It's part of the "video games" community I guess. Boring...
this looked like insane and I mean insane amounts of effort to make lol just to get 2.6 mill views
The fact that the "Fanta Cake" was noticeably edited during the making of the video is hilarious
I did not know that I had a single thought about Fanta Cake . . .
BUTT . . .
NOW That You Mention It . . .
blah . . . Blah . . . B L A H !
And obviously they were both wrong and I should edit it 😂
Fanta cake the choice of the nazis
So have the Acton family articles.
The observation of the phenomenon changes the results…
@@ACOUSTITRON-mp6tc Same with Veritasium's 37 video
I really hope the Wikipedia groups start talking about this, this is really cool to see
I hope that this causes people to add links to orphans and dead ends.
It will probably be included in our internal newspaper the signpost.
Maybe you should lift a finger instead of asking others to do it
@@herpederpe4320 no offense, but if I was a Wikipedia editor I would be part of the problem
@@herpederpe4320 not everyone is cut out to be a Wikipedia editor. the folks who're self-aware of this fact thus respect Wikipedia and-in a way-help it by not breaking anything
I like how someone fixed the Fanta Cake article but didn't bother to replace the sad sopping excuse of a fanta cake picture lmao
Nobody else can bear to make one
I like how by mentioning this you got them to fix it lol
@@FineTuxedo I just checked to see lol. The wiki community is generally pretty swift when it comes to resolving issues that they are made aware of.
@@popcorn8153and now it has almost a dozen references too.
@@FineTuxedo yeah, they "fixed" it with an ai generated image...
Just wow. Bravo. As a hobby programmer and data scientist I have a glimpse of what it took to do this, but know the real effort and magnitude far exceeds that idea. You hide the complexity (and I'd suspect quite a few brutal bugs to solve) incredible well in your simple yet entertaining walkthrough. Incredible!
Dude came up with one of the most significant and important studies of Wikipedia ever conceived for a game. Amazing.
Honestly, I studied network graphs as part of my PhD, this analysis was better and more interesting that 99% of them for aure.
For real! We love visionary data nerds!
A study of about 1/10th of wikipedia
@@sprgeorge333 honestly i did too and this is NOT more interesting or intelligent than any of the papers i read.
@@sprgeorge333 this is more interesting than my bachelors thesis on github collaboration networks xd
You just took a topic that I would probably spend my life without ever giving a single thought to, and made a video that was an absolute joy to watch. If there's a TH-cam Hall of Fame, this one belongs in it.
This is even more impressive. He talks about something I've seen done to dead by lots of other people (see his cheeky reference to mildly interesting reddit) and it's still new and fresh to me. I almost don't watch this article but when I finally budge and I don't regret it.
I want a CURSED wikipedia race as a prank. You host, you select at "random" but all of them are 10th degree separation OR HIGHER.
good idea!
But most Wikipedia races already take 20+ clicks so having a minimum of 10 clicks won’t really change much
i have a feeling that trying to calculate that would quickly turn into the traveling salesman problem
@@ameltedTSP is a completely different problem; to find pages 10 degrees of separation or greater, you could just use a breadth-first search, similar to what he showed in the video when analyzing degrees of separation. This is possible in polynomial time (I believe O(n^2) in the worst case, but feel free to correct me if it's wrong; it most certainly is polynomial, however).
@@ReliableExcavationDemolitionyeah and you probably take 20 clicks to bridge a 4th degree or so relation. Just because you didn’t personally find the shortest link doesn’t mean it’s the shortest link. 10 degrees or more would be absolutely brutal for a human
This would honestly make such a cool website. You could zoom in, click on a node, and see all the information regarding it addressed in this video.
This is honestly one of the most interesting videos I've seen on youtube in the past 6 years.
What was the most interesting video you watched 6 years ago?
@@sa88 That is a great question
why exactly 6
we are wondering bro
what was the last one
The algorithm is sleeping on this one
update: The algorithm was sleeping on this one
It's ok...I'm here now.
no the video is for nerds
i thought it already had like 7m views
Wake it up then lol
It brought me here.
This sort of thing deserves to be an actual feature on Wikipedia, it's so well done. Would be super cool to play around with an interactive version of this, or have it regularly updated to take a timelapse of how it changes.
We'll be doing well for them to get regular charts back working first.
The computing involved with an interactive, LIVE version of this would be...non-trivial to maintain.
There already is something very similar already.
I could see a tool like this being particularly useful for cleanup. Those "highways numbered 9xx" articles could probably be consolidated.
@@AL-lh2ht which is?
On the topic of degrees of separation and the longest paths. All we really need to do here is to add more links between articles and these numbers should decrease. People are already editing articles with the help of this video, and people are definitely interested in shortening these paths.
Absolutely fascinating video. The Fanta Cake bit at the end is a great example that for most orphans or dead ends, it's a matter of what could be considered bad article formatting/linking.
I looked at the Wikipedia article for William Acton (senior) and someone has already destroyed the Acton group solitary-ness by adding links to the page for "Politics of England" off of the phrase "English politician"
Great video! Fantastic work : )
My first thought too - this video is 5 days old, no way those Acton family articles are still their own group. Sure enough. ;)
While those Acton links seem a bit forced to me,
They did also link them to bailiff and Newcastle-upon-Tyne,
Which they always should've been.
real
Solitude?
Fanta cakestic
“A complete waste of time”, “Mildly interesting”, hell no, I was thinking that graph looks freaking BEAUTIFUL!
" This is your brain . . . "
< sizzle . . . >
" And this is your brain on wikipedia . . . "
B---)
but only after understanding what it is
@@solarnaut My brain isn’t famous enough to be on Wikipedia 😋
@@opensocietyenjoyer not really, I find it beautiful as a work of art, but knowing that it’s actually a data graph with millions of nodes makes it SOOO much better!
It's the color palette and the way it is elliptically shaped, as if observing something through a gravitational lens
One of my favorite wikipedia trivia bits is that, at least for a long time, by clicking the first non-disambunction or pronunciation link, you will eventually end up on philosophy. I think some of the natural sciences end up being recursive now but it used to all link to philosophy.
Trying it now, my first try unfortunately got stuck in Telecommunications Network Node
I remember doing this!!! I didn't hear about it from anywhere I just clicked the first article link (non pronunciation or disambiguation) and I always always ended up at philosophy where it recursed!!!
Part of the reason is that a couple of people found this fact, then checked it. The 1% that didn't end on philosophy where changed to end at philosophy.
Holy shit it actually worked. It still works to this day
I remember when I discovered this a killjoy had cut the link between knowledge and philosophy and broke the chain. Then they would revert any edits that added the link back
this needs to be a interactible website
This is my favourite example of the power of being able to explain niche things disconnected from the general public’s interest well. You turned a seemingly useless thing: a graph of Wikipedia articles into an amazing, engaging, and thought provoking, inspiring video, highlighting each of the things that you’ve explored, with perfect transitions for dramatic effect and amazing animations and visuals. This is a mind-blowing video, keep it up!
Sound like chatgpt output 😂
@@euli_mo Well, it’s not.
ChatGPT ahh comment
@@theoverreactor8731it's not, there is a ponctuation error. Why would he use ChatGPT anyways?
@@euli_mo so anyone who can put together three sentences that aren't basic af sounds like chatgpt, ok then
This graph is really beautiful, you should make it into a poster
It's gorgeous, i hope he uploaded it to wikimedia commons
I would genuinely print it out and put it up on my wall
Great visualisation! Would like to put it on my wall as food for thought poster.
THEY FORGOT PHYSICS MATH BIOLOGY CHEMISTRY AND ALL OTHER SUBJECTS THAT ARE NORMALLY VERY THOUGHT BASED
"What's on your wall?" "Oh nothing, just Wikipedia"
Canada and Hockey being one community/category is amazing. The fact that you know 100% for certain that the article for Tim Hortons is in that category is just the glue of perfection.
Completely irrelevant. What matters is China
@@Ps5prolite OK grandpa, go take your meds.
I mean, yea. Tim horton is a famous hockey player so its pretty much impossible for it not to link to hockey.
@@MrMickio1 Someone tells a joke. This guy: "That is factually accurate."
@@MrMickio1So famous I’ve literally never heard of them. I didn’t even know they were a genuine person until this thread!
(Then again, that could just be a logical side effect of my complete lack of interest in sports in general, and the complete lack of TH outlets anywhere near where I live!)
Watching this for a class and when you mentioned dead end and orphaned articles it reminded me of pre- synaptic and post- synaptic neurons, since they’re neurons which have no input from other neurons or don’t output to other neurons respectively (I.e. sensory neurons and motor neurons (?)). It’s really interesting how universal networks are
I love how you went over so many different things related to the graph in this video. It really satisfied that curious urge you get when learning about something new
your feelings are irrational
@@Fire_Axus everyone's feelings are irrational, it came free with being a human
I agree, I really wanted to see it graphed and I got to see it!
I wish social media, especially platforms other than TH-cam, had more fascinating stuff like this. I want to experience that childlike curiosity again!
The amount of complex work this guy has presented here with understandable tone suggests that this can as well be a PhD thesis topic.
Seriously, he totally downplays the significance of what he's accomplished with this graph. There are so many fascinating insights here, not just about Wikipedia, but English-speaking culture.
Why would you say that? I wanted to understand your perspective 🤔
as a student currently in alogrithms and graph theory this is insane. wonderful project and video man
I was thinking that creating a graph database of this data might make for an interesting project.
"A place where anyone has the power to free an article from an existence of solitude" is such a beautiful line I did not expect to hear in a wikipedia graph analysis video
As a former Wikipedia editor, this is really cool to see! I regularly make use of the SpecialPages Orphaned, Deadend, Unconnected or Redirect to try and improve the linked data structure.
I would really love seeing Wikipedia take this project as a source for more linked improvements!
Thank you for your service
Wikipedia is one of humanities greatest creations
also editing my language version of wikipedia and I'm just curious - was there a reason why you stopped?
this project just does not use the right data-
this optimal paths like shown in this video so often just not exist this 1 example when path ends with buffalo is acutally findable but the other one is not :
flood does not link to this list of non-water floods
this link dump data is just not all that great, it shows many links that are not visible in the page.
basicly this video is wrong, the guy would need to have his own script that would scrap all those links that are shown in the page, those links in wiki dump data are often very random.
@@koks49045 The flood article does link to the list of non-water floods. In the first line, in fact.
The text of the links is not always the name of the article they point to. If you want to find a link in a page you can search the URL in the HTML source.
@@koks49045the 120iq understander has logged on
we just having fun bro u might be 140 for all i know, 120 is like a step below the average phd and 140 is like a bright student at a top program (a step and a half above the average phd)
I think a way to find the absolute longest path would be to start at the "list of highways numbered 825" article and start mapping pathways backwards from there. Whatever you end up with, you can add the links from 825->999 to that
Longest path is an NP-hard problem, it would take an absurd amount of time on a graph like this.
I went backwards from 825 and found 530 and it goes down from there
The path he found also starts with a small chain too, stepping through the articles, "Athletics at the [year] Arab games".
like a highway
just start at number 1 and end up at 999
1. How many Wikipedia editors are now looking to make sure all pages are linked. Eliminating orphans and dead ends.
2. Wikipedia should add this somehow to give a visualization of its vast knowledge.
long-time wikipedia editor here. There have existed entire projects which have tried to eliminate orphan and dead end articles. At least for orphan articles I believe there are hidden categories that flag them. I think the OP could've made use of Wikiprojects in order to link related articles together instead of just using outbound links, though I guess that if his analysis is based on the wikipedia game that it makes sense why he wouldn't. Wikiprojects already give you something similar to his idea of "communities" of articles
@@redcoat4348 A more pressing problem is correction of false information.
Why eliminate orphans and dead ends? The entire internet is not just wikipedia links.
@@drbuckley1 Won't ever be fixed. People try, and I salute them for it. Wikipedia is a guidebook not an answer book. Wikipedia usually gives enough of an overview, correct or incorrect, to start looking up information elsewhere. Trying to make it an answer book is impossible.
Before I even finished the video, I went to check on the Actons and found one of the pages edited an hour ago, adding more links. RIP community 42
0:25 wikiverse
Cool name!
17:20 omg I just realized "fanta cake" was why I clicked on this thumbnail and for 17 minutes I was just lost in the sauce
Update: the fanta cake article is no longer a dead end, and now has 15 links.
Is googlewhacking an orphan an indictable offense?
you know you've made a good video when every second of it can become a wallpaper or a T-shirt
Fantakuken wallpaper
good luck with 1:36
*DISGUISED DEAD END ORPHAN* 18:39
@@davisjian8250id wear that
10:41 wallpaper
I originally thought this video wouldn't be too interesting, but I clicked on it out of curiosity (and like the saying goes, you had my curiousity but now you have my attention). The amount of detail, effort and production value put into this video astounded me and I was hooked. I also appreaciate the informative description. Thank you for this wonderful video. The only thing missing is the raw data and code.
He also has BDE
Just that family making up community 42 made me smile. Seeing a robust algorithm in action is awesome.
I checked them out. Seems they are no longer an orphan comunity, as many links were introduced in the article, like"Englis Politician" and other nonsense. Or maybe I misunderstood the situation completely...
@@zo2oit’s been a month since the video. It’s been updated
@@zo2oI haven't checked, but to clarify, links included in the articles don't exclude them from orphanage, it's the fact that no other articles link to them.
666 LIKES
The issue not explored is that there are Wikipedia editors who have an intense interest in one topic or narrow groups of topics. That shapes the style and linking for many groups of articles which become mainly the work of a single author. At 3:50 it's not at all surprising that the principal authors of articles on Association Football are completely different from the authors of Gridiron Football; each set of authors will know comparatively little about the other topic and will therefore be far less likely to cross-link them.
If you want to look deeper into how editors shape articles, you will want to study the various Wikiprojects -- groups of editors who work together to improve a specific topic.
Good point!
There's something beautiful about how telling us of this information causes alot of it to be improved on.
14:56 in solving a rubiks cube there is a similar concept called "gods number" which is the number it takes to go from any scrambled position to solved. It was calculated a few years ago and came out to be 20. That's why I was really suprised when you said the biggest link you found was 166 which is crazy to me, but makes sense. Especially since Wikipedia isn't as easy mathematically explained as a rubiks cube. Awesome Video!
I can’t believe I watched a 20 minutes video about Wikipedia graphs to be finally be surprised with Fantakuchen as one of the most special articles. I just had Fantakuchen on Easter this year and it was one of my favorite birthday cakes all childhood long (next to Donauwelle, wave of the river Donau). Applause!
Lmao typical german, he wants to make clear that he's from germany. Writes fantakuchen despite being able to write everything else in english, then donauwelle and then the pretentious applause. All germans are the same, I have no idea why the chauvinism.
Donauwelle and Fantakuchen are just a whole new level
But don't forget the good old blondes Blech
I have a specific memory of hearing about Fantakuchen when I was in like second grade, thinking "huh?!?" and then it never coming up in my life until now. Fits the video perfectly
omg i buy donauwella at aldi all the time i love it. also stollen omfgg
Happy birthday! I am glad you enjoyed your kuchen!
Dude. 200 views in an hour on this is *criminal*, this should be blowing up. I wouldn't be surprised if Ye Algorithme picks this up and it's in the 100,000s sooner or later.
Honestly I can see this getting millions lmao.
It could get way more but I personally think he really missed the mark with the thumbnail and title choice.
As a member of the thousands club I don't doubt the video could get 100,000 views given enough time
@@BombsanTheCommenter ye, its a very unique and interesting video
@@pepsalt It's also culturally significant and important. We few can say, I remember when that had only 3000 views!
Your editing and sound editing in this highly commendable by the way. Extraordinarily smooth and intentionally timed without being too obnoxious in anyway.
I would love an interactive version of this
In the segment of “longest path” you should’ve more clearly said the “longest shortest path” (or: shortest path with the most nodes.) since obviously longer paths can exist if you just deviate away from the goal
That seems like a given because if that weren’t the case almost any article could loop forever between 2 or 3 things. If you consider the path as not being able to repeat pages, that would likely be several million articles long.
@@somedude4832 That'd also be the most nightmare-inducing Hamiltonian path problem possible
yea, but also to actually find the longest shortest past and not just guess that it was the 166 one shouldn t he had just picked a random page, do a bfs, go to the the page which was the farthest away and do a bfs again? And do it for all conex components
I'm actually kinda interested in the path since "list of highways numbered 1000" and "list of highways numbered 999." point at each other. If I'm not making a mistake this implies the same length from the start to either "list of highways numbered 998" and "list of highways numbered 1000." otherwise the path would be shorter or going to the next article would be a longer shortest path.
Took a complex graph theory class in college and loved creating / analyzing these node graphs. Your videos are awesome!
one of the most well made youtube videos on data visualisation i have ever seen. good job
there's a website that does this for any link
This has be the quickest growing channel I’ve come across in the last year with only 6 or so videos. It just goes to show that if you have solid production with an intriguing subject that relates to many different people whilst hitting all the marks with title and thumbnail you too can make it.
This channel is a perfect template.
If you start a new channel and have several initial videos flop with only 500 views there is a good chance you’re gonna struggle to ever breakout of the algo blackhole. You must hit it off from the first video and by hit off I mean at minimum 20k views
I need an overtime with how this graph changes. Great work. This is actually extremely important and someone needed to make this. Seriously good job.
Regarding about Fanta Cake page, the page has just updated again several times since the beginning of April, now the page is even expanded with more information in it (Soda Cake section added), "Fantakuchen" redirect page removed, and is neither Dead End nor Orphan page anymore, much like a normal article now!
"f" for Fantakuchen😢
NOOOOOOOOOOOOO 😭😭😭😭
@luiskerscher5047 don't worry, I checked and fantakuchen still exists. Only the link to fantakuchen in the fanta cake page has been removed.
@@maksrambe3812 Thank you, thats one issue off of my list. Well appreciated!
we should call this process adoption💀
Great video. Superb yet subtle editing skills. The kind where you dont even notice how good it was. Also the jazz was a killer choice.
You should make an interactive version of this for us to play with!
You deserve an award for this work and this video! It is really interesting how some of the "communities" are structured.
BTW: If you write up your findings in a Wikipedia page, linking to all the orphans in your graph, you would drastically reduce the number of orphans in Wikipedia.
That's brilliant!
I've commented this on another comment as well, but what about a page "List of Wikipedia Orphan Articles"! Same can be done for "List of Wikipedia Dead End Articles" lol
list of every wikipedia page
@@lajawi. Both of those pages already exist! The reason orphan articles don't become automatically un-orphaned from being in that list is because the list is not in the "main" Wikipedia space, but a special section of Wikipedia that has editing guides and the likes, and so it doesn't really count as a link
I can not express how joyful I am that Rugby, on its own, managed to become an entire category.
i had the exact same reaction, i love rugby
Same with norwegian politics... on english wikipedia and I checked, it's very fleshed out.
same importance as category 42, 4 members of the former british parliament as an orphan group
ahí la tiene Maradona,⚽ lo marcan dos👥, pisa la pelota Maradona🚶♂, arranca por la derecha↖ el genio del fútbol mundial🏃♂, deja el tendal y va a tocar para Burruchaga... ¡Siempre Maradona!🎖 ¡Genio!😯 ¡Genio!😯 ¡Genio!😯 Ta-ta-ta-ta-ta-ta-ta-ta...🗣 ⚽🥅Gooooool...😱 Gooooool...😱 ¡Quiero llorar!😢 ¡Dios Santo, viva el fútbol!🗣🔥🔥🔥🔥🔥🔥🔥 ¡Golaaazooo!🔥🔥 ¡Diegoooool!🔥🔥 ¡Maradona! Es para llorar😭, perdónenme... Maradona, en recorrida memorable, en la jugada de todos los tiempos... Barrilete cósmico🌌... ¿De qué planeta viniste para dejar en el camino a tanto inglés🪐👽, para que el país sea un puño apretado gritando por Argentina🔵⚪🔵? 💯Argentina 2 - Inglaterra 0. Diegol, Diegol, Diego Armando Maradona... Gracias💫, Dios, por el fútbol⚽, por Maradona🥇, por estas lágrimas💦, por este Argentina🏆🏅 2-Inglaterra 🥈💩0”
Pelota para Xavi⚽, asistencia de Xavi👀
en esta pelota para Messi👀
Messi🏃♂, Messi⚽, Messi🏃♂, Messi⚽, Messi🏃♂,
Messi⚽ y inmenso Messi🏃♂, Messi⚽
Encara Messi🏃♂, Encara Messi🏃♂,
Encara Messi🏃♂, Encara Messi🏃♂, Encara Messi🏃♂,
Encara Messi🏃♂, Encara Messi🏃♂,
Encara Messi🏃♂, Encara Messi🏃♂, Encara Messi🏃♂,
⚽🥅Gol, Gol Gol Gol, Gol, Gol, Gol, Gol, Gol, Gol,⚽🥅
Gol, Gol, Gol, Gol, Gol, Gol
And what is so special about Norwegian politicians versus the politicians of 200 other countries? It seems quite bizarre!
The articles on the Altons made me think of constelations and galaxies. Like how most stars are bound together by gravity in galaxies and clusters, but then you have intergalatic or rogue stars, that are just not bound to any galaxy. I just find neat how we can find similar patterns in so different parts of reality.
Well, now the Actons are not an orphan group anymore.
You might like emergent phenomenon and universality of dynamical system
The idea of an intergalactic rouge star is kinda terryfing. Like, how did it even get there?? Why did it just get lost. Is it just incredibly ancient and has just always been there since the beginning of the universe? Or did some ungodly cataclysmic event rip it out of it's galaxy? How do you even rip a star out of it's own galaxy??
It's easy to imagine rouge planets. You hear about them all the time. Galaxies are relatively dense so it's easy to imagine how a passing star could rip a planet out of it's home system. But even then, rouge planets still exist within their own galaxies.
What ungodly apocalyptic catastrophe has to occur for a star to end up in intergalactic space???
@@qwertydavid8070It's actually pretty dull. They just get to close to a supermassive blackhole at a wrong angle, and they're sent flying off!
@@a2izzard It's still crazy that their sent flying off all the way into intergalactic space. Galaxies are humongous, you'd think that along the way the star would eventually get attracted by the gravitational pull of the galaxy itself.
Then again, for as big as they are, galaxies are still mostly empty. It's like how neutrinos can seemingly phase through matter. They are just so tiny that things that appear solid to us just aren't to them. The gaps between atoms are like the gaps between planets at that scale. I guess stars are just so tiny when compared to an entire galaxy that they can just pass through without interacting with anything.
I'm chronically online goon and spend avg. of 5 hours a day on TH-cam for like 8 years, but to find a the video so well crafted and with topic that interesting.. man good job
It's amazing how low key influential Jon Bois is on informational TH-cam videos. Great video!
I wasn't thinking it was a waste of time, I was thinking that it was beautiful and looked like a universe.
Why do you have 3 furaffinity accounts
what is wrong with you
Why are you a broken human being
>Springing a leak
>Sparkly skunk mascot~
This is some beyond level data scraping and analyzing. Well done adumb! Diving into your other videos now.
I wish TH-cam would suggest things like this to me all the time, this is literally so cool
The video and work behind the graph itself is commendable, but the visualizations throughout the video is just as commendable. A lot of quality work, good job!
Wow! This is probably the most beautiful presentation of connected data I've ever seen.
Would be awesome if you were able to make this into an interactive website 👀
That would be super awesome
If you're a fan of interactive websites, I recommend you try out Wikipedia and TH-cam and all the websites they are connected to.
9:58 according to all known laws of aviation, there is no way a bee should be able to fly.
Yeah, there are some really weird interactions that go on at small scale that we haven't observed anywhere near enough.
@@horsemumbler1 We now know exactly how bees fly. You can google it :)
but it does, because nobody told the bee :)
@@mst4309 We know exactly how a bee flies. For a long time we didn't, but it has been several decades since we found out.
Those american scientist headlines really ruined an entire generation of science communication
Its midnight and in 5 hours I have a German test and instead of sleeping I’m learning about Wikipedia lore
Finally a video about a very interesting topic that is not clickbait and really well presented. First one in quite a while after I accidentally flooded my recommended with meme clips...
This is unbelievably fascinating! You did a beautiful job investigating, explaining, and illustrating all of this. Well done!
This is possibly one of the best videos I've ever watched on TH-cam. Amazing work
'history of the entire world, i guess' was better
@@Kriae" one of the best"
A german seeing the thumbnail: Whats wrong with Fanta Cake? Had it with my last birthday.
I love the Internet!
14:25 The Acton family was immediately welcomed -back- into the wider Wikipedia community xD
We will leave no orphans behind!
@@Streetcleanergaming lmao
I thought he said the afton family and i was like fnaf?!
you should absolutely render HD images of the graph, it's beautiful, 4k at least if not super high res
The image is 19200 x 10800 btw. 4K is 3840 x 2160
@@007i1 awesome, is there a link?
@@chigginheadD No i just looked at the code
@@chigginheadD we need a linkkkk
It's like modern art, except it actually means something! lol
Absolutely amazing video. Can finally beat my friends at the wikipedia game for the first time
You mean the one where you need to search for links to get you to a goal article? You can cheat by editing the original article and adding the link to the goal article
@@badgermcbadger1968 why would you though?
@@Laezar1 because i find it boring and it's pretty funny the first time
@@badgermcbadger1968 you don't have to play a game you find boring though xD you can just tell your friends no, it's not like it's a tournament with prize money on the line or anything, all you'll get by cheating is break the trust of your friends.
Idk that just seems like a really strange and low stake situation to cheat in.
the problem with this game is most articles are some random weird stuff like list of something or a village,
if there was a fillterted version where only articles that are like concepts or some important stuff like country existed, then it would make sense,
but also like each article should have atmost 10 hyperlinks that link to stuff that is rly related
this has so much untapped potential. if no company was looking into this before they certainly are now, most likely from commercial companies and software companies
Community number 42 is about family, how poetic 🥰
It's also no longer in isolation,
Well, okay, as far as I could find it's still impossible to get in,
But you can get out of the community now.
Unexpected intersection of Hitchhikers guide to the galaxy and Fast & Furious 😄
@@johannesandersson9477Is that how many of those films there is now? 😂
"What do you get if you multiply six by nine?"
@@dielaughing73 69?
This’s insanely well made, some things are very insightful or hilarious, it’s a shame the videos doesn’t have many views:(
This is an incredible video. Despite being 19 minutes long it felt like it was ending before it even started, I was so interested.
It would be genuinely incredible if you made this interactable so you could filter out different communities, zoom in, look around, click on dots to go to a page, and whatever else you can think of. I would love to explore this graph
I think this might be one of my favorite vids now. From start to finish, the writing and pacing captivated me. The visuals were stunningly animated, the music was bopping and the whole time I wanted to know more. I hope your channel becomes a staple in the "computer science experiment" side of youtube
Literally everything in this video was so wholesome, for a number of reasons.
can I like... buy a poster of this?
You can now.
The presentation, editing and music make this such a good video, along with the of course fantastic data!
This was one of the most interesting videos I've ever seen. Absolutely fascinating! This could easily have been a master's thesis. And the production and presentation were stellar!
Currently on 700 views, but I already know this is going to blow up. One of the most beautiful videos on the internet.
Now it’s almost 2000, I think the same. This is going to explode soon
@@lucasklein448 5.600 and counting!
EDIT: +3.600 in 3h
12k four hours after 2k comment. It's coming!
43K views now
52k
3:58 I'm glad you brought up "sports" when you did. As an avid (maybe even "rabid") sports fan, my first thought upon hearing "44 communities" was "that seams very low!"
When you consider Wiki covers EVERYTHING, my initial thought was "we'll, then 1 of the 44 must be a broad and all encompassing 'sports' community. "
44 does seem pretty low overall, but it gets slightly easier to understand the more he breaks it down. I'd still like to get my hands on this data and look at it myself.
But the reasons why "sports" isn't a singular community is because the algorithm he used isn't connecting things by human defined concepts. It's connecting articles based on which other articles they link to. You'll be in a tough spot if you were to try to find a connection between, let's say Darts and Baseball. So those two get their own "communities." Each sport is pretty insular when you really look at them. Even between the "big 3" in the US: baseball, basketball, and football. As huge as these three are, there's still very little crossover between them. Very few players have played in more than one of these in their professional careers and so links between the three will be very few and far between on the Wiki. And if the three most popular sports on the English speaking Wiki aren't very well connected then you can forget it for any of the other sports as well.
You can "tune" algorithms like this to be stricter or more lenient. I don't think 44 is some objective figure that someone doing a similar analysis would come to as well - presumably making the requirements stricter leads to more communities, and more lenient leads to fewer communities, similar to a fuzzy select in an image editor.
probably the best video I‘ve seen this month
Thank you so much for making this video. I've wondered about exactly these questions for years, especially how "connected" wikipedia is as a graph, and suddenly your video popped up in my feed and answered basically all of them.
I really hope you'll make a second part and extract some more info from all the data you've gathered, I guess there are definitely some nice graph-theoretical features left to explore here! (for instance how tightly linked the wiki graph is, are there bridges whose removal would disconnect the (main component of the) graph? Does this only happen at dead ends, etc.)
The craziest part is that it is just the english wikipedia, there isn't other-languages-only articles
Amazing work on all this stuff! Also the music is somehow weirdly perfect for this video like you're some random dude that I got stuck with on an elevator and the elevator music is still playing in the background while we just get into this long conversation about Wikipedia.
It looks like a universe where each galaxy is its own Wikipedia page. Would make for a cool movie/show
18:15 imagine a story about a sort of Cyberpunk2077 rogue AI (this is one of the settings features) being stuck in the web because of this
We need a one hour deepdive!!!
would be cool if the data was on a website to browse
I always wanted someone to do it, but presenting it in such an entertaining and insightful way is way more i could have dreamed of. This is awesome
I'd love to see a video about the technical side of this project. The crawler you used, the database it's stored in and the format in which it is stored, the queries and algorithms used to find the paths in each of your headings. Super interesting project, thanks!