Daniel Parris

Kendell Kelton (00:02):

I'm Kendell Kelton, and I'm your host today on The Rough Draft. Featuring honest conversations with folks from across the creative industry, The Rough Draft explores the creative process, tools and resources used by some of the best in the business. From journalists to content creators and business leaders, we shed light on what it looks like to break into the industry, make mistakes, collaborate with others, and also the essential tools that help us along the way. This week I'm super excited to be talking with Daniel Parris, the Brain behind the popular newsletter, Stat Significant. Daniel's got a unique gig. He's a data scientist who blends numbers with pop culture. So if you've ever wondered about the statistical side of why you can't stop watching true crime documentaries or what makes Star Wars fans so dedicated, Daniel's your guy. So I want to kick things off by having you tell us a little bit about how you stumbled into this, what it means to be a data scientist in that sense to you, even though that might not have been what you intended to do with your career at first

Daniel Parris (01:15):

Start at the beginning. So when I was eight years old, I watched an American Film Institute special on the hundred greatest movies, and I printed out a list of the movies and I put it on my wall and I about to watch every single one. And I wasn't able to really get super far on the list until I was a little bit older. That list has things like a clock or Orange and Midnight Cowboy and some really upsetting movies,

Kendell Kelton (01:43):

Maybe some things that aren't as appropriate as that least that my parents weren't

Daniel Parris (01:46):

Psyched about. But I just went through the list. I think to this day I've probably watched 90 or so of them, and that just kicked off a love of film. I made a lot of movies when I was a kid. I would go to film camp, I would go after school to all these film programs. That was sort of my identity. I went to school and I majored in film and I actually worked in the entertainment industry for a little while, so I worked for the Conan O'Brien show and I worked for Steve. It sounds fun. And I worked, it sounds a lot cooler than it's, and then I also worked for Steve K's production company, which also sounds a lot cooler than it's, and I just really did not like the entertainment industry. I sort of saw what the next five to 10 years of my life were going to look like, and I realized that even as an intern at the lowest strong of entertainment, I was basically fighting with the other interns too, yet Coffee for People.

(02:41):

And so I just saw where my life was trending towards, and I hadn't aggressively bad intern manager when I worked at the Conan show. And I remember calling my dad walking off the Warner Brothers lot and being like, I'm sorry I can't do this. I know that this was the dream that I settled on when I was eight years old, but I think I have to choose another path. So at the time, my now wife was moving to San Francisco, and so I said, cool, me too. Why not? And I was told that if you move to San Francisco, there are a lot of tech jobs there. So I got a job working a travel startup doing sales, and I was actually quite bad at sales. And so the company decided to repurpose me as a data analyst. So I actually started going to night school to learn sql, Python, data science, machine learning, and I found I really liked it. So I just kept funding my own education and going to night school and I would buy textbooks and just learn about statistics and do all of these online courses. I ended up joining a company called DoorDash in 2016.

Kendell Kelton (03:44):

Never heard of them. Dunno about

Daniel Parris (03:46):

Never. Yeah,

Kendell Kelton (03:47):

I'm

Daniel Parris (03:48):

Kidding. At the time they were this crummy little startup, but I was one of the first data science hires and I stayed at the company full-time for six years doing a bunch of different data science projects. And by the time I left the company was around 10,000 employees. So just a crazy ride. It was a great environment to learn just because I basically lived through 30 different types of company. I saw it when it was a little startup. I saw it when it was growing really fast. I saw it when the pandemic happened and it was growing even faster. And I got to work on a bunch of cool different things. I left in 2022 because it was definitely different than the company I'd first joined. And since then I've been doing data science consulting and then I realized that I kind of just missed movies and entertainment and I had felt like I'd been away from that stuff for too long.

(04:40):

So I just sat down one day and I was like, well, what if I combined the stuff I loved when I was 10 years old with this vocational skillset that I've cultivated over the last few years and effectively what was born with Stats significant. And it's been a blast. A lot of people write about movies on the internet, and I wanted to figure out is there a way that I can come at this that's unique, both unique to my skillset and then also allows me to talk about these things and pay reverence to them, but in a way that's maybe different than how others are talking about these things on the internet.

Kendell Kelton (05:14):

Your topics are so unique. They're really fascinating

Daniel Parris (05:18):

And I feel like my favorite part of the articles is I usually lay out a question, then I provide a few different graphics and visualizations, things that I've found by digging into data. And then there's usually a section that I call final thoughts where I feel like I effectively shoehorned a more editorialized opinion.

Kendell Kelton (05:36):

One of my personal favorite ones that you wrote, it was from a few months ago, the when do we stop finding new music? I'd love to hear a little bit more of how do you stumble upon some of these topics and then how do you get going in terms of gathering the data that you need?

Daniel Parris (05:55):

This is usually heavily driven by something that has happened to me in my day-to-Day life. So as an example, when do we stop finding a new music article was because Spotify launched this new DJ feature, it's an AI dj, and I realized that it was just spoonfeeding me music from when I was 13. And to be fair, when I listened to Spotify, I only listened to music I liked when I was 13, but the idea that this person, the DJ would reference, here's the year of the songs that I'm playing for you. And I kept realizing like, oh, this is within a very tight band of years. And after the third, my Chemical Romance song or the fourth Fallout voice song, I started realizing that it crystallized as something that felt more than a trend. And I had thought about this topic a few times before, and so I guess I put it in this long running list I have of ideas.

(06:56):

And at this point it's got about a hundred ideas. Some of them are absolute garbage. I'd say 90% are never going to be made into an article. And then some of them I am slowly collecting different data points and data sets on and trying to figure out is there maybe something topical that's about to happen that could be a good tie-in, or is there an in to this article? A good question to lead with. And I guess going back to the when do we stop finding new music? I found a lot of really good data. So this has been studied by different researchers. This has been studied by Spotify employees. This has been studied by the New York Times, and so I was able to gather all of that data.

Kendell Kelton (07:42):

Let's chat through a bit more about your writing process because it takes a lot of discipline from what I've read and would like to help our audience get a little bit of a peek behind the curtain of what you do on the daily because it is a labor of love.

Daniel Parris (07:59):

Yeah, it's a labor of love. Sometimes I joke to myself that I've created the world's worst business, like 10 to 20 hours for each article. Right now it's a free substack, so it's 10 to 20 hours that just goes out into the abyss and there's no compensation, which is fine because I really love doing it.

Kendell Kelton (08:20):

Writing these in depth essays takes a lot of discipline. So I've heard you have a pretty specific writing routine. Can you kind of share how that works for you and what started that path?

Daniel Parris (08:35):

Yeah, so I actually write at least the first draft of my articles either while on walks around my neighborhood or while on the treadmill at the gym. And so I'll use one of two tools. If I'm walking around my neighborhood, I'll bring a voice recorder and I won't bring my phone. And so I'll just dictate what I think the article should be. If I do it at the gym, I basically am writing directly into a note taking tool called Notion. And basically I say as soon as I get on the treadmill only, I'm obviously going to be doing my workout, but for 30 minutes I'm only in notion I'm not going anywhere else on the internet and I'm just writing as quickly as possible. When I do that, I can usually write in 30 minutes on the treadmill. I usually can write about half of an article. And so between two days, 30 minutes on the treadmill, I can usually knock out the first draft and then I'll ultimately go through, take whatever garbage I wrote for first draft and edit it into a second draft. But it is a really good way to just kick off the creative process and just go through that writer's block.

Kendell Kelton (09:43):

So shifting gears a bit, I'd like to focus on how you turn raw data and to these captivating essays. So you're one person show, we've talked about that on the last episode. So how do you keep all of your data and ideas organized? I believe you're a fan of Notion, right?

Daniel Parris (10:04):

Yes, I am a big fan of Notion. I have basically, I organize it into sort of like a Kanban board. So I'll have things that are in the pipeline, I'll have things that I basically have moved into a column that is effectively like contemplation, so trying to flesh out and develop those ideas. And then I'll have a column that's actively writing or researching, and then within those I will keep notes. So within Notion, within these tiles, you can click into them and they effectively become a Word doc. I will typically compile either research or ideas. This morning I was thinking through an article on pop punk. I just got so inspired that I wrote, I think, and I was at the gym of course, so I wrote the intro and the final thoughts. I don't have the data yet, but I was just like, I will figure out a way to write about pop punk. So that's what I use for organization, for Notetaking, and then also for at least my first drafts in writing.

Kendell Kelton (11:04):

Got it. And then when you, you had mentioned this a bit in the last episode because it takes you about 15 to 20 hours to actually write and compile everything, but you will sometimes use a voice recorder when you're walking. What voice recorder do you use? Why do you use that one? I think people, if they are looking for a way to remove themselves from their phone, it'd be good to link to.

Daniel Parris (11:34):

Yeah, I really can't tell you the exact name. It is a very small Sony voice recorder. I recently read Cal Norton's Deep Work, and I realized that whenever I brought my phone along, it was a distraction. And ultimately I found that my mindset was fragmented and I wasn't able to come up with ideas as easily. So I was just like, what if I brought out a voice recorder and whenever an idea came up, I would just dictate it. It's really nice to walk around my neighborhood and feel like I'm both outside in nature and effectively writing an article. It's kind of like a weird thing to balance, but it's actually quite a nice experience.

Kendell Kelton (12:15):

I mean, it's whatever gives you creative joy, it's whatever sparks that willingness to sit down and write. So I'm here for it. So for those of us who aren't data scientists like myself, finding the right data sets can feel daunting. I do a lot of writing, a lot of different types of writing, but when I'm trying to find things to source or to back up my own claims or ideas, there are a plethora of AI tools available today to help synthesize information, and it can be overwhelming for people who don't quite understand intimately data. So how do you source data for your essays? What are your go-to tools?

Daniel Parris (13:07):

Yeah, so my number one Go-to is a site called cagle, which is a site for people who are interested in upskilling their data science and machine learning skills. They just have this giant repository of data sets and their fair use data sets as well. And so for me, I've been able to grab, there's a data set of 7 million records of movie grosses and budgets and actors and actresses, and it's this extremely robust data set that I've been able to use in a bunch of different analyses. There's stuff on tv, there's stuff related to music and Spotify. So that's usually my first step, and that's where I get about 80 to 90% of my dataset. And then other than that, I subscribe to a listerv called Datas Plural, where each week I get sent five unique data sets. And so I found some really unique data sets. And oftentimes, as an example, there's a data set compiled by a bunch of economists that's basically every notable, notable person, historical figure who's lived since, I forget, it was like 3,500 ad not AD bc. And it's a super helpful data set because I wrote an article on how has celebrity changed over time. And so it's really whenever I can get something that's truly unique, that's always a treat. And so those are really actually my main go-tos. And

Kendell Kelton (14:35):

Then you also feature when you're trying to bring these to life some pretty cool visualization. So graphs sometimes you throw in some names and gifs, but I think you rely on one particular tool maybe for those. Do you mind sharing

Daniel Parris (14:52):

That? Yeah, yeah, of course.

Kendell Kelton (14:53):

Yeah.

Daniel Parris (14:54):

So I do my data analysis, so that's transforming the data, aggregating the data, joining together data sets. I do that in Jupyter Notebook, which is kind of the go-to data science manipulation tool. And then after I've basically transform the data to the point where it's ready to be graphed, I'll then put that in a tool called Flourish. So Flourish is a free graphing tool that I think was purchased by Canva a few years ago. So it also integrates with Canva. So you can very easily take something on a flourish, put it into Canva, and then actually continue to do design or design around the graphic.

Kendell Kelton (15:35):

Are there any other tools off the top of your mind that you want to share that you also just have an affection for?

Daniel Parris (15:41):

Yeah, I write in Grammarly, which has been extremely helpful. My biggest mental block to starting a newsletter was I don't want to look stupid on the internet. And I think my biggest fear was having just a stupid, it's kind of like those anxiety dreams you have about showing up to school and not knowing that there's a test. Before I published my first stat significant, I had these anxiety dreams that I was going to have a major typo, and everyone on the internet was going to be like, look at this idiot. And so Grammarly is at least a good way to de-risk, make sure that you don't make some pretty basic day one mistakes. And to be honest, I still have made big mistakes and it's been fine. I just edited and it all turns out great.

Kendell Kelton (16:27):

So you wrote an essay, why do people Hate Nickelback so much? And people absorbed it like I did, and you kicked it off with how does one explain Nickelback and how does one explain the band's remarkable Second life as a meme? So I want to ask you to walk us through why Nickelback, what were you doing on some sunny day in California walking around and all of a sudden thinking about Nickelback?

Daniel Parris (16:52):

Yeah, I was talking with my friends and Nickelback came up and think it's one of those sort of picture perfect moments that you see in a movie where someone says the thing and then in the next scene they're doing the thing that they said where I think I said something like, it'd be cool if you could quantify how terrible. And then I went home and was like, can I quantify how terrible? And for the record, I actually am very lukewarm on Nickelback. I like some of their songs. I think some of the complaints are valid. I don't have a super strong belief that they're undeserving of the attention that they get. So basically it came out of a conversation with friends and I was like, okay, if I can find two or three valid reasons why people dislike this band, then yeah, I'll definitely write this article.

Kendell Kelton (17:45):

Well, you laid out four hypotheses in your essay about this, so I'd like to kind of break those down a bit. So starting with the idea that they're overplayed, what data did you pull to back what they really are in the public eye?

Daniel Parris (18:01):

Yeah, so I think the thing about Nickel Beck, they're sort of chief complaint against them is that all of their songs sound the same. But what's interesting is if you had a band that was maybe not as heavily played and all their songs sound the same, people wouldn't be angry about it. I think what's angry is that they're a fairly heavily exposed band in the sense that they actually have a lot of hit songs and they're one of the bestselling bands of all time. At the same time, I wanted to figure out, can I also then quantify whether their songs are fairly similar? So ultimately, I basically graft both record sales against a bunch of different Spotify variables, like an amalgamation of Spotify variables. And so I basically said, okay, let's look at the variance of Nickelback song composition versus the amount of sales that they have. And ultimately what I found is they're actually pretty unique in the fact that their songs are quite low variance when you look at all of these different aspects of song composition. And for some, given that low variability, they're also extremely high selling. So they were one of the main two or three outliers that is low variability, high record sales. And so what you get is people are listening to a lot of songs that actually quantifiably do sound pretty similar.

Kendell Kelton (19:25):

Well, and it was interesting, I guess I didn't, if you would've asked me a while ago, if I would've associated Nickelback with having country genre influence, I would've been like, no. And then I read your article and I was like, oh, that actually kind of makes sense. And they were falling into that category alongside a lot of country artists.

Daniel Parris (19:49):

Yes. I think the other artist that was most similar to them was Toby Keith, who's an American country star. That was something that I hadn't previously thought of, and I actually found other data subsequent to that view that really framed it as they are kind of reminiscent of a country band. I think they often are called a post grunge alternative country mashup. So there are actually a lot of different micro genres put together in this one band.

Kendell Kelton (20:19):

Well, and then the other interesting thing, and I think a lot of people associate this with the country genre, is that sometimes they get put in this box of, it's just the same song on repeat over and over and over again. And I think you found that also, and within one of your hypotheses for this, there was an interesting quote of somebody who was saying They have 89 songs, but essentially it's the same three songs over and over and over again.

Daniel Parris (20:49):

Yeah, I think I quoted a music journalist who many people on the internet just took their shot at Nickelback,

Kendell Kelton (21:01):

But you also mentioned that sociopolitical factors might play a role also, Annette, which was really interesting.

Daniel Parris (21:07):

I didn't go into it being like, ah, I see the world and and left. But I was like, okay, you probably have music critics who are mostly left-leaning. And so I was like, why? One of the cuts that I looked at earlier was basically sales versus critical acclaim. So I looked at meta critic reviews for Nickelback, and then I also looked at sales, and they were also an outlier in the fact that their albums received quite low scores from music critics, but they obviously have a lot of sales. And so my question was like, okay, who's listening to these albums? And it wasn't in a disdainful way. It was like, I imagine that there's probably a different slant for the people who are writing music criticism.

Kendell Kelton (21:50):

Well, they're obviously doing well to some degree, so who is listening to them?

Daniel Parris (21:55):

And I actually found it not on purpose. What happened is I was looking at Google Trends and I typed in Nickelback, and I noticed that when the United States search map came up that Nickelback was mostly showing up in states that you associate as right-leaning. What I did is I took 2020 presidential election results, and then I looked at the correlation between how often the band is searched in that state and then what the voter percentage was for the GOP presidential candidate. And I found that there was actually quite a strong correlation. And so in the article, you can see there's a nice regression line. And so what I found is that the band is more likely to be listened to in right leaning states. And so that's not to say that that is a reason that they're bad, but I think what that answered for me was, okay, there's some divide.

Kendell Kelton (22:49):

It segues nicely into the idea of them just being a victim of bad publicity. They had a few really unfortunate series of events that happened.

Daniel Parris (22:59):

This is the one that blew my mind. This is the one where when I first heard about this, I was like, this can't be right. So basically what happened is that there's this random standup show on Comedy Central where a bunch of comedians sat around and talked about current events, and this one comedian made just an offhand quip where he was like, oh, I hate Nickelbacks music. If I had a gun, I'd chew Nickelback. And everyone's like, ha, ha ha. And then they clipped that, and then they made it a cornerstone of the advertising campaign for that show. And apparently that ad ran for quite some time, like six months to a year. And so I basically did McKinsey esque sizing of how many times was that joke repeated? And I forget the actual amount that I was able to size, but it was like, oh yeah, I think it was billions of times. But then I found this quote from Nickel Beck's front man, his name's Chad Kroger, where he basically points to that as the genesis of Nickelback hatred.

Kendell Kelton (24:04):

And so many people were watching TV at that point. So everyone saw the ad.

Daniel Parris (24:08):

And this is also Comedy Central in their heyday, I think you had South Park Chappelle show. And then the other thing that I found where they were were the beneficiary of Bad Luck is kind of how I put it, was that they were slated to perform a halftime show for the NFL Thanksgiving game in Detroit. And the city of Detroit wrote a petition saying, this is the city of Motown. Why is Nickelback playing our halftime show? And the petition got nearly 60,000 signatures. And so what I ended up finding is that if you look at Google searches for Nickelback hate, just the term Nickelback hate, which gets a decent amount of Google traffic, it didn't come into being until this very specific time. So basically just from a pure search perspective, people searching and seeking out the Nickelback meme didn't seem to be as much of a thing before this NFL halftime show.

(25:06):

And then maybe just to add insult to injury, you can watch a video of the halftime show and people are actively booing them as they perform. An interesting aspect of all of this is that I think now the band cite the Nickelback meme as a reason for their longevity. The fact that they are still relevant is now a function of the Nickelback meme. You have bands like Three Doors Down, and I'm honestly going to Exhaust, yeah, I'm going to exhaust my understanding of post-run bands, scrunch bands. But if you had a thought experiments like Why is Three Doors Down Not Relevant? And Nickelback is they're very comparable bands probably have the same number of hits, and it's because Nickelback has had this second life as an internet meme, and that's also galvanized their fans. And so it's actually been this phenomenon that, I mean, it's probably not great to be Nickelback on the receiving end of all that hatred, but there has been some benefit whole to the meme and the relevancy.

Kendell Kelton (26:02):

Would you take a minute to tell people where they can read your essays and then we'll link to it on the show notes?

Daniel Parris (26:10):

Yeah, so you can read the essays that stat significant.com, all one word, obviously. And then, yeah, it's the stat significant newsletter on Substack. I post it once a week on Wednesdays. And if you like pop culture and you like thinking of pop culture in a slightly different way, subscribe.

Kendell Kelton (26:29):

Well, that's it for today's episode of The Rough Draft. To learn more about our guests and to find links and resources related to the conversation, check out rev.com/podcast. If you enjoyed today's conversation, be sure to rate and subscribe in order to stay up to date with the latest episodes. Thank you for listening, and we look forward to seeing you again on the next episode of the Rough Draft.