E13: The Everything Store Reviewed, Waymo Surges Ahead

November 14, 2017 by Adam Walker

Random Talkers is the show where two data scientists (aka myself and Matt McKim) deliver piping hot takes on tech, sports, and more. New clips uploaded bi-weekly. Available in podcast form via iTunes or SoundCloud. Help Google's algorithms view us in a favorable light by subscribing on YouTube!

Book Review: "The Everything Store" By Brad Stone

Stone’s book tells the story of Jeff Bezos and Amazon, taking us from humble beginnings as a Seattle bookseller through its ascent into the e-commerce, media, and cloud computing colossus of today. Bezos doesn’t always come off well in the book (his wife actually gave it a one star review on Amazon) but we found “The Everything Store” an interesting read even four years after publication.

Waymo's Level 4 Self-Driving Cars Have Arrived

We discuss Waymo's driverless minivan program in Chandler, Arizona, then breakdown why self-driving cars have been getting into some fender benders.

E12: Superintelligence And The Future Of AI, ESPN Ditches Barstool

November 07, 2017 by Adam Walker

Nick Bostrom's Superintelligence Reviewed

We review Nick Bostrom's Superintelligence, a book about what happens when artificial intelligence becomes smarter than humans. How dangerous is superintelligent AI, and will it arrive soon?

ESPN Ditches Barstool

With ESPN cancelling Barstool Van Talk (aka Pardon My Take's TV show) after only one episode, we discuss why ESPN made the decision, along with what problems the network still faces.

Bill Simmons Doesn't Talk To Women Much

October 23, 2017 by Adam Walker

Quick post this week to analyze who has appeared on The Bill Simmons Podcast, a perennial top five sports pod. Data is from Simmons's Ringer era only - the ESPN/Grantland days are left to a more diligent researcher!

First, here are all the guests to appear at least 3 times over the pod's 275+ episode run. "House of Carbs" host Joe House leads the pack, reflecting his large role in the early post-Grantland shows, as well as a continued presence for discussions of the NBA, D.C. sports, and anything food-related. Proud gambling degenerate Cousin Sal is close behind, then there's a big gap before NFL commentator Mike Lombardi, resident Yankee fan (and Republican!) JackO, plus many other familiar Simmons collaborators.

While it's no surprise to anyone who follows Simmons that buddies like House and JackO are frequently heard from, it is a little surprising how few women have appeared...

Out of 433 guest spots across the 275 episodes tracked, only 18 (4.16%) were made by women. Nine women have appeared versus 135 men. The complete list is as follows: Mallory Rubin has been on six times, Juliet Litman four, Sarah Tiana twice, then Sally Jenkins, Abby Wambach, Charlize Theron, Diana Taurasi, Katie Baker, and Katie Nolan once. Mike Lombardi has made more appearances than all of them combined.

So why does Simmons have so few women on the pod? One reason might be that his show is a sports podcast featuring athlete interviews...and the majority of famous athletes in America are men. However, loyal listeners know that Simmons doesn't actually talk to that many current players or coaches (regular Kevin Durant appearances aside).

Sexism obviously comes to mind. It can't be disproved, but I think it is unlikely since Simmons has always done a nice job hiring female writers, both for Grantland and The Ringer. Is Simmons for whatever reason just uncomfortable interviewing women? I suppose it's possible.

Still, to understand who Bill Simmons talks to, it might be instructive to understand what he talks about. Looking at the words most frequently appearing together in episode descriptions (excluding guest names and stopwords), it appears Simmons is pretty locked onto the NFL and NBA:

This makes sense from both the perspective of Simmons's own personal interests (Celtics! Patriots!) and as a strategy to maximize podcast downloads - the NFL is America's most popular sport, while the NBA appeals to a younger crowd perhaps more likely to be into pods.

The metrics angle is worth considering further. It's well-known that The Ringer monetizes primarily from podcast ad revenue, and given this dependence one can't help but wonder if Simmons has found that episodes with female headliners (e.g. Abby Wambach, Diana Taurasi, or Charlize Theron) perform worse than the rest? This would be disappointing, but could at least provide an economic rationale for how Simmons chooses interview subjects.

Taking this possibility into account, I would guess there are a couple of reasons for Simmons's overwhelmingly male guest selection. First, Simmons just likes talking to his old college buddies - and these people are men. Second, by bringing on guests he already knows pretty well, Simmons reduces prep time for podcasts (allowing more episodes) while also generating stable download numbers due to listener familiarity. People really like hearing Simmons and Cousin Sal guess the lines each week.

However, for someone who has been fine featuring obscure male interviews (Jason Stein anyone?), it would be a refreshing change for Bill Simmons to broaden his podcast's guestlist.

Analysis code and data is available on Github.

Slightly off-topic random thoughts on The Sports Guy: I read Simmons's columns religiously from 2005 to about 2015 or so. He's undoubtedly one of the most influential sportswriters of the last two decades. Furthermore, the 30 for 30 series was a phenomenal accomplishment, and despite his endless Celtics homerism Simmons stands as a legitimate NBA historian. While everyone has one now, podcasts were not always an obvious choice - yet Simmons recognized the format's potential as a compelling alternative to sports radio. Simmons's defining ESPN project, Grantland, was awesome, both in terms of content and the careers it helped launch, and more recently The Ringer has begun to carve its own niche too. In my mind his shtick has unfortunately grown stale - Simmons has clearly been left behind by some of the more analytical NBA commentary that's arrived, along with certain podcasting upstarts - but he has undoubtedly had an amazing career.

E11: “The Upstarts” Book Review, The USMNT’s World Cup Rejection

October 22, 2017 by Adam Walker

The Upstarts Review

This week we review “The Upstarts,” Brad Stone’s lively new book on the founding of AirBnB and Uber - perhaps Silicon Valley’s two most divisive startups. Although the story of Uber’s rise has been well-documented, the book still managed to hold our attention throughout.

Summer 2018 Is Now Free

We also react to that 2-1 defeat. No stranger to sporting let-downs, Adam helpfully reminds everyone of the time England failed to reach Euro 2008, then we analyze why the United States may never develop much world class soccer talent.

Creation Stories: Analyzing YouTube Data With R

October 11, 2017 by Adam Walker

Earlier this year, YouTube introduced Creators On The Rise, a showcase for up and coming stars on the platform. For those of us with somewhat modest YouTube followings, these creators are VERY intriguing. What secrets enable their lavish subscriber growth? Is it the cadence of posts? The eye-catching thumbnails? Perhaps it's the video titles perfectly written to elicit a click...

Or maybe channel popularity is a function of plain old survivorship bias. Life is unfair!

Anyway, to analyze these new media success stories I turned to tuber, a handy R package that facilitates access to YouTube's API - just make sure to register via Google's developer console for the necessary ID and key. With a little wrangling I was able to compile a dataset of 92 YouTube channels previously featured as "Creators On The Rise," including comprehensive metadata (title, description, tags, duration, views) for 9,410 videos. All of this plus my code is available on Github.

Screen Shot 2017-10-10 at 3.19.04 PM.png — "With more than 1,000 creators crossing the 1,000 subscriber threshold every single day, new talent is constantly emerging."

**YouTube introductory blog post**

Now, with data in hand, time to tackle some burning questions:

1. How Long Till I Become A YouTube Superstar?

The 92 creators represent an interesting mix of fast success versus long-term grind. Channels like Deestroying and Hailey Reese have amassed subscribers with abandon since their debut, while 007craft and Amazing Grays have taken a more leisurely approach:

Screen Shot 2017-10-10 at 7.49.22 PM.png

YouTube's selection criteria for a "Creator On The Rise" likely means this chart represents an overly optimistic view of the time it takes to become a YouTube superstar. In addition, it's worth noting that some creators may have had an existing following prior to creating their channel - Sophina The Diva, for example, has garnered 64K subscribers in a little over a month, but had the advantage of already being rather famous. Nevertheless, building a YouTube empire does not have to be a decades-long affair.

2. Will I Need A Viral Hit?

It doesn't hurt. Let's return to the previously mentioned 007craft, a solo creator channel with around 39K subscribers. Mr. 007craft (I'm not sure anyone knows his real name) attracted attention earlier this year as the guy who was living in a storage locker. His video on the experience has collected over 3.5M views, representing an outsized portion of his channel's 11.8M total.

Of course, creating a viral video is tough. One could toil for years without a fleeting whiff of fame. Only 55 videos in the (very much non-representative!) data have broken 1 million views, representing about 0.6% of the 9,410 total clips. Still, to have a shot you probably want to keep your clip under 15-20 minutes:

Screen Shot 2017-10-10 at 4.21.32 PM.png

"FEEDING THE DEVIL | Spiders and Centipede" can be found here for those interested. I thought insect-averse readers would appreciate me omitting the thumbnail, so check out the second place video instead:

In any case, have all 92 Creators On The Rise had a true viral hit? Subsetting the data to only the highest viewed video from each channel, we find that these top clips have a median view count of 553K. This is certainly a lofty bar for non music videos, but at least your channel doesn't have to garner a Charlie bit my finger level of fame to gain subs. Overall: viral hits help, but aren't essential.

3. What Should I Make My Videos About?

All the channels analyzed used tags to help people discover their content. So tag liberally! For the graph below, I parsed tags from each individual video and computed pairwise correlations between all tags appearing in at least 10 channels and 100 videos. Groups of tags with high positive correlations indicate common topics covered by many creators:

Screen Shot 2017-10-10 at 3.03.45 PM.png

Most topics are fairly down-to-earth! There are fitness channels, family-oriented content, make-up tutorials and the like. And no politics! Perhaps YouTube is consciously refraining from naming political channels to the "On The Rise" section? Or maybe political takes aren't that compelling after all...we can only hope.

4. How Much Content Does A Top Content Creator Create?

Now that you have some ideas on potential channel topics, you might be wondering if it's possible to build a large following without quitting your day job? Maybe! Most creators are posting steadily but certainly not daily. The data below shows publication frequency over the past 52 weeks for all the channels created prior to the start of the year:

Screen Shot 2017-10-10 at 2.28.59 PM.png

Side-note: CatPusic might be my favorite of all the channels in the data. Consider adding a cat to your channel for an instant popularity boost.

In terms of timing, Friday is the most popular day of the week for publishing new videos:

Screen Shot 2017-10-10 at 2.37.12 PM.png

5. The Rich Life...What's That?

The attentive reader will have noticed an outlier in the second to last chart. The Rich Life lead the pack by far in terms of posting frequency, cranking out a ridiculous 6.5 videos weekly. For the uninitiated, The Rich Life follows a "homeschooling family of 7 that loves to share the good, the bad and maybe the occasional ugly." Of course, the idea of family as #brand is nothing new, but adding 200K subs in a little over a year is impressive regardless.

A cursory glance at The Rich Life's recent clips reveals a channel that clearly has its act together:

Screen Shot 2017-10-10 at 12.43.39 AM.png

The ALL CAPS titles alone entice a click, plus The Rich Life's thumbnails are attention-grabbing and effective. Also, the sheer variety of hijinks encountered is incredible! Note that if you have recently been kicked in the head by a horse, suffered a break-in, evaded a tornado, AND dealt with a police run-in...then you may have what it takes to be a top YouTube creator.

Final Thoughts

I hope this article has served as a somewhat tongue-in-cheek introduction to analyzing YouTube data via tuber. There's much, much more that could be done with what's available, from running sentiment analysis on video titles and descriptions to building models for subscriber growth or a video's likelihood of going viral. Once again, all project code can be found here. Thanks for reading!

Further Adventures In R: