Stands to reason. When a bunch of quants run the same variables through the same ML engines and optimization algorithms, they all get the same results.

I haven't voluntarily listened to commercial broadcast radio in well over a decade, but music playlists, outside of independent stations, were already pretty repetitive pre-Moneyball.

As the analytics revolution in music grew, radio playlists became more repetitive, and by some measures, the most popular songs became more similar to one another

I fully agree with what the author said about what this approach has done to movies---why bother to try to make a good original movie, when it's now far more important to make a financially successful one and the most likely way to do that is to make a series of sequels

Wonder why Doug's submission of this article went through, and mine didn't (put in Sun. evening).

Wonder why Doug's submission of this article went through, and mine didn't (put in Sun. evening).

Because Doug ran his through a series of complex algorithms first and his was determined to be the more efficient submission, so you missed out.

I'm trying to liven up my music collection a bit, especially in the non-jazz area, so I finally broke down and signed up for free Spotify which immediately offered me 2 free months of premium (with no commitment) so I'm giving it a go.

I'm not sure if I confound algorithms or if those algorithms are just not as smart and detailed as they are portrayed. For example, near as I can tell, just due to being an old, white male, facebook assumes I'm a rightwing nut job, fan of the Confederacy or at least want to follow WW2 websites cuz ... I don't know, that was the last time men were men or something. As an experiment, I spent several days actually going through fb suggestions saying "don't ever want to see this again" and all I achieved was replacing a daily pic of Debbie Harry ca 1977 (no objections) with a daily pic of Muddy Waters which I do not consider an upgrade. It was just this endless "you must like the 'War between the States' (always a giveaway). No? You must like WW2. No? Well you must be pissed off that the US left Afghanistan or that women now sometimes do stuff men do. No? Well you must love 60s TV ... 50s TV ... old movies ... classic rock. Ohhhh, we get it now, you're an old hippie, you must be fascinated with Janis Joplin or something. Educational in its way but tiresome.

Anyway back to Spotify. I started with a playlist of 29 non-jazz and 11 jazz songs. Don't do this. It generated a playlist for me that was over 90% jazz. Now it did a pretty good job of connect the dots but all dots I'd already connected (or intentionally not) and why whould a 75% non-jazz list generate a 90% jazz list? Statistically I can see how this might happen -- us jazzies probably cluster close together so plug in enough jazz and the top scores on the similarity algotihm will nearly all be jazz.

I did manage somehow to get rid of those matches and do it again and, for whatever reason, this time it came up with a list that was only about 2/3 jazz. That gave me enough non-jazz to get some idea of whether it would do a good job of finding me unfamiliar non-jazz that I liked. The results weren't encouraging. Mostly the logic seemed to be "if you liked this Patti Smith song, you'll probably like this other Patti Smith song." Fair enough, that's true but it's not helpful. When it did manage to break out of my circle even a little bit, it got stuck on a single artist -- apparently I should really, really like Mary Gauthier. I can see why it would think so (and maybe I do) but I'd rather it offered up 5 different people rather than 5 songs from one artist in a playlist about 150 songs long.

So I did a second playlist with 38 non-jazz and 2 jazz. It still probably turns up more than the 5% jazz I gave it but the balance is fine. But it continued that behavior of "here are more songs from this artist you like" and "here's the one other artist we think you want 5 songs from." In fact, it was probably worse in that regard. The oddest example is the Spanish pop star Rosalia of whom I am actually a big fan. She's got 3 albums and. heaps of singles/collaborations (most of which I'm not a fan of). I gave it a track off of her latest Motomami and all that did was give me several other tracks off Motomami. It couldn't even make the leap from Rosalia to Rosalia, I don't know how long it would take the algorithm to get from Rosalia to Georgia Anne Muldrow.

Then I made the mistake of telling it I "loved" a few of the things it spun up. I was hoping that meant "yes, you are moving in the right direction" but all that did was more of that behavior. One band was Heatmiser. I'd never heard of Heatmiser before (more in 2 sentences). The song was catchy so I ticked it. That led to another 5 or so Heatmiser songs. Turns out Heatmiser is an early Elliott Smith band. It couldn't make the leap from Heatmiser to Elliott Smith even though I'm pretty sure the only reason anybody particularly remembers Heatmiser is Elliott Smith (though there were other quasi-famous indie types in there). Maybe I need to "love" more stuff which I would gladly do if (a) it was any good at finding stuff that earned my notice and (b) after doing so, it didn't just give me more and more of that band -- I can do that on my own once I know they're worth checking out.

I wasn't expecting it to get from, say, Graham Parker to Peter Blegvad (and I'm not sure I want it to) within the space of 150 songs but Parker to Elvis Costello or Joe Jackson ought to be move one ... instead it was Parker to Parker to Parker. Smith, Kristin Hersh, Horsegirl I would expect to quickly suggest Sleater-Kinney or Bikini Kill or Juliana Hatfield or Pretenders or something in that vague neighborhood. I stuck in one David Bowie song (Heroes, not obscure) -- a bit worried that would consign me to classic rock hell -- and it led ... absolutely nowhere, not even to more Bowie. WTF? Richard Thompson's Beeswing got me to 52 VBL which got me to a live version of Beeswing which got me to a live version of 52 VBL which got me to Teddy Thompson (his son) covering Leonard Cohen. The next time I tried a somewhat more obscure RT song which got me to Beeswing, 52 VBL and that same Teddy Leonard Cohen cover.

OK, fair enough -- maybe the algorithm has me pegged so accurately that it gets stuck in the same rut I'm stuck in.

All that said, it generated a perfectly pleasant set of songs that didn't have me reaching for the "god no" button too often. There were no "that was awesome" moments but it would be fine for car radio, housecleaning purposes. Maybe if I cobble together a 500-song list it will behave better.

Wonder why Doug's submission of this article went through, and mine didn't (put in Sun. evening).

This is actually the second time I submitted it. The first time I didn't add the info about "Subscription maybe required but you get a number of free articles per month:". That seemed to do the trick. So, that's the complex algorithm wrangling!

-- a bit worried that would consign me to classic rock hell -

My experience with these algorithms is that, seemingly no matter what I do, it always starts to feed the Beatles to me.

Back to baseball ... I'm not sure that (brief, knowingly simplified) statement really captures the influence of "analytics" per se. Velocity has always been fetishized, it's not like Gene Mauch didn't want guys throwing 100 MPH. Spin rates were something we couldn't really measure until recently but the level of "snap" and grip were always recognized as important. The revolution there isn't in the analytics, it's in figuring out ways to boost velocity and spin rates. Analytics has had influence on the mix of pitches and some on usage patterns (probably more than I'm willing to give it).

Somewhat similarly on the hitting side -- once you've got the data, it's easy to find "HRs usually have a EV of at least X and a launch angle between Y and Z." There wasn't anything particularly insightful there -- surely you've got this by week 2 of physics 101. The revolution has been in training batters to do it.

It's a bit like the underpants gnoems:

Step 1: Discover/quantify some important relationships via analytics
Setp 2: ????
Step 3: Win

It's seemed to me that network sitcoms went anti-analytic over the last decade. The one remaining dominant sitcom, Big Bang Theory, was basically a clone of the dominant sitcom, Friends, but the networks resolutely refused to try to remake it for yet another generation. They instead fixated on recreating The Office, which I loved but was never a ratings hit. NBC, the sitcom factory, kept re-doing The Office until they finally abandoned sitcoms altogether and have recently been talking about reducing the number of hours they program. Looking at the Nielsens, it seems that at least on the networks The Office clones have a lot of difficulty in building an audience through word of mouth as Cheers or Seinfeld did, but the networks don't seem to have ever noticed that.

The broadcast networks mainly followed the Marlins model -- you can make a reality series for essentially nothing so even if almost nobody watches it, you make money.

I think I read that Abbott Elementary (a Parks & Rec clone of a sort) was the only network show to win any Emmys at the last awards -- even that I suspect was half pity and half that the streaming services haven't hit on a winning sitcom formula either.

Before Moneyball we had Steve Jobs, Johnny Cash, and Bob Hope. Now we have no Jobs, no Cash, and no Hope.

Big Bang Theory, was basically a clone of the dominant sitcom, Friends

That's like saying Bryce Harper is a clone of Ty Cobb because they are/were both good at baseball.

I know the music culture business better than others -- still no expert. Anyway, the internet/streaming vastly improved access, distribution and acquisition of recorded music ... as long as you know what you're looking for. It destroyed revenue, the exposure methods (advertising, getting into rotation, get on TV) and the gatekeeping. The gatekeeping was at best a mixed bag, dominated by the major labels, but it went overnight from 10 gatekeepers to 100 million influencers/curators/bloggers/blowhards.

One spot where I'm ignorant is the whole viral thing -- pretty sure the only internet viruses I ever caught were the blue/gold dress and "some hear rutabaga, others hear kumquat" thing and that was probably only after reading it in the Guardian or something. I don't know how some teen youtube/tiktokkers break out while the other billion don't. But that's your word of mouth model these days.

So I see Dylan Moran (Black Books) has a new sitcom where each episode is only 15 minutes. I assume we'll see more of this sort of thing -- why not, if there's maybe one good 3-minute skit on SNL, why do we need the other 40 minutes? (to sell stuff of course)

Another thing that's changed in the last 20 years or so ... or not so much changed as the long-term trend went exponential -- is the end of almost anything that can legitimately be considered "the main stream." Taylor Swift is the last one in music (and she keeps changing up her style to do it), Marvel and a few others in movies. There too the internet has been great for diversity -- you want Ethiopian jazz? you want Tibetan cuisine delivered? But who can make money in that setup? Sure, taxis were a ripoff and you can get around cheaper now ... with 100,000 unregulated taxi drivers most of whom unknowingly are spending more on depreciation, fuel and maintenance (or not??) than they make.

On the one hand, all this access is awesome. On the other hand, it's info overload and confusing. And it sure doesn't look sustainable to me. Art will survive of course because artists are addicted to creating stuff.

The Office clones have a lot of difficulty in building an audience through word of mouth as Cheers or Seinfeld did, but the networks don't seem to have ever noticed that.

Seinfeld wouldn't have a prayer of being shown on a broadcast network today.

On the one hand, all this access is awesome. On the other hand, it's info overload and confusing. And it sure doesn't look sustainable to me. Art will survive of course because artists are addicted to creating stuff.

One of the problems with the whole system, art, finding content of any kind on the internet in fact, is that there is a lot of garbage in the way of finding actual good stuff.

My experience from examining "who becomes an influencer" is that usually they had funding to start with.

So, all the 21st Century has been at some level is the increasing "reality-TVization" of everything. Influencers aren't usually organic (some are, most aren't it seems). They're created. And, they're cheaper than paying an actual celeb to do an ad for you. So, its just fueling higher profits while preventing some of the celebs from making an actual jump to being genuinely rich. I'd rather go back to Halle Berry being the one selling makeup rather than Jeffree Star personally (I know this example through my daughter).

As for algorithms on facebook and me? I thinks I like Stalin and Hitler at the same time so I'm genuinely concerned for what it thinks is going on with me.

The one remaining dominant sitcom, Big Bang Theory, was basically a clone of the dominant sitcom, Friends, but the networks resolutely refused to try to remake it for yet another generation.

wait, what?

I'm sorry, but I guess I have to be the old man telling you kids to get off my lawn! I couldn't for the life of me understand the Moneyball phenomenon. Ooh, get Scott Hatteberg because he gets on base! Wow, on base % is more important than batting average? Do tell, it's news to me, teams should be paying attention to this! Where the hell had these teams been for the last 20 years? Nothing Moneyball was preaching about wasn't material already covered by Bill James and Pete Palmer two full decades earlier. Why was this considered new? Suddenly front offices were interested in analytics? After ignoring them for decades when they were readily available? Hell, I understood this when I was a kid playing Strat-O-Matic and would see cards like Jim Wynn's. Guy doesn't have a great batting average, but his card is full of WALKS and HOMERUNs, this player doesn't make many outs!

I don't think moneyball had anything to do with the way movies and music are nowadays with the exception of Moneyball the movie.

Re 18. The book wasn't breaking ground inside baseball. It was informing readers of what was going on inside baseball in the 80s through early 00s. Amd of course it had a healthy dose of Michael Lewis and his form of narrative license.

Nothing Moneyball was preaching about wasn't material already covered by Bill James and Pete Palmer two full decades earlier. Why was this considered new?

Because those guys were shouting into the wind, what was new was someone in baseball was finally listening.

Anyway, the internet/streaming vastly improved access, distribution and acquisition of recorded music ... as long as you know what you're looking for.

Oh, I know exactly what I am looking for alright. I just kind of doubt that these mindless algorithms do (c.f. the Spotify example above). MusicalLy it has gotten to the point that I fully expect Sturgeon's Law to relentlessly apply to pretty much any music I try to sample. As a result I am not sure if the stuff I like is hidden behind like 6 magic doors that I have to find and unlock, or if it doesn't exist at all anymore.

The relentless move away this millenium from dynamics, tension, thematic development, and power in favor of endless musical stasis (be it hip-hop or modern alternative) doesn't help. I'm just worn out from hearing stuff that remains in the same narrow groove and never moves beyond it for the entire cut.

Wonder why Doug's submission of this article went through, and mine didn't (put in Sun. evening).

It's a matter of timing of when someone checks the queue. I don't check it that often, and I wasn't involved with this one.

When a thread is approved for posting the default is to put it on the Newsstand page based on submission date/time. If we have the same article submitted 5 minutes ago and 2 days ago and we approve the one submitted first, it will show up further back in the Newsstand page, making it more likely it will be missed. There is no set protocol for dealing with that, but my approach (again, wasn't involved this time) is either to approve the most recently submitted one or to override the default date/time to something closer to now, so it shows up near the top of the list.

the only reason anybody particularly remembers Heatmiser is Elliott Smith (though there were other quasi-famous indie types in there)

I see what you did there

https://en.wikipedia.org/wiki/Quasi

I haven't voluntarily listened to commercial broadcast radio in well over a decade, but music playlists, outside of independent stations, were already pretty repetitive pre-Moneyball.

I think the issue here is not that the playlists all have the same songs in them, but that the songs themselves are all the same. A video making the rounds of my musician friends a few years ago was a spliced-together version of 6 different country hits -- four that were #1's and two others climbing the charts at the time -- showing that they were all simply different lyrics written over the same chords, same beat, same key, same tempo, same everything.

Walt, thanks for post number 6. I don't have Spotify, but I wondered about it because it seems everyone else does. Your post is a great argument for never getting it.

Walt, thanks for post number 6. I don't have Spotify, but I wondered about it because it seems everyone else does. Your post is a great argument for never getting it.

I use Pandora if I'm using anything, and I will reiterate the part Walt said at the end: it does a really good job of providing a series of songs you won't hate. And, like, if I'm using it for background music that's exactly what I want.

Spotify is nice because I can actually find the obscure music I am interested in. If I want to listen to all of Handel's oratorios, in chronological order, I can go there. I don't use it for recommendations.

The only online recommendation algorithm that is semi-accurate as to what I like appears to be YouTube, which has directed me to content that I didn't already know about that I find interesting, at least some of the time.

A video making the rounds of my musician friends a few years ago was a spliced-together version of 6 different country hits -- four that were #1's and two others climbing the charts at the time -- showing that they were all simply different lyrics written over the same chords, same beat, same key, same tempo, same everything.

This is especially notable in country music, at least radio country, the so-called "Bro Country" movement.

Atlanta used to have a classic country station (The Eagle). That station went anti-woke really early ("The Station Where You'll Never Have to Push #1 to Speak English!") and is now off the air, but they did play great music.

For music discovery, I can't recommend Bandcamp enough. Forget having robots give you recommendations. Buy something you like on Bandcamp. Then you'll be able to see everyone else who also bought it, as we as everything else that they've bought. So you can go through their libraries (streaming stuff for free) to find other stuff you might want to buy. It's more work than just pressing play on Spotify, but browsing is part of the fun anyway.

Thing is music has always been the same stuff over and over. Always. The Star Spangled Banner ain't a unique piece of work nwver heard before. People have always taken tunes or arrangements amd basically copied them to death. Before nowadays you'd have someone write a song and then dozens if not hundreds of artists would cover it.

i think some of what the author posits is true and inevitable. but - this thesis runs counter to another thing which strikes me as true and inevitable, the movement toward recognizing and catering to the long tail of culture.

people talk about changes in tv and convergences around certain trendy (and often cheap to produce) things, like true crime and reality programming. but we've also absolutely nothing like the monoculture of my youth, much less what was around when my parents were young in a three network (plus pbs!) and no home video world. many more people can make tv or movies compared to the before times and i've access to many more different types of programming than ever before. i have to look for them, i may have to depend on curation (where there are many good ways to do this, though what those are require a little effort/searching on the user's part and are often custom to the thing that person is interested in), but they are there. this is more true with tv - where we've been in a "golden age" in the eyes of many - than with movies; it's harder to get mid-sized budget films made than in a very long time and big budget movies have gotten homogenized. but the small film scene is still relatively rich.

team sports are a little different, they're in a closed continuum. i think there are two responses to what thompson is describing that can help. one - leagues can tweak rules to incentivize entertaining play. we've talked plenty about this on this site, things like pace of play, incentivizing contact, and so on. the other is if your leagues are in an enviroment where there are possible rewards to heterodoxy / being ahead of the curve. i think basketball still has some of this, as teams haven't really optimized how to defend the three yet (shout out to sam vecenie for crystalizing a thing that's been on my mind there). baseball is trickier, given how richer teams invest much more heavily in info and tech than poorer ones, which makes it harder to catch up or benefit from new or weird strategies. rule changes (see point 1) that alter the environment should create short run opportunities here.

---

i think there's a lot of overlap with my above ramble to your posts, walt. anyhoo, i'm not a spotify fan (apart from the big catalog). it's been a few years since i've found a music curation process that has worked for me - i now do things like block off time in december to go through, like, fluxblog's year in music mix and make notes of what i do and don't like. that said, my music tastes seem to be calcifying anyway and i'm drifting into the old man trap of "band x is just group y mixed with artist z and with more modern production".

---

edit: oh yeah, ziggy, bandcamp is pretty good. good call.

YMMV, but I've yet to hear a song about Moneyball.

My family insists on spotify, so we use that. I listen mostly to albums, and make sure I have the repeat album button clicked, so spotify can't offer me their own suggestions.

Everything devolves to playing the Grateful Dead or Rush for me. which I don't mind.

I may be a dinosaur, but I prefer a good radio station, if I'm not listening to albums I already own. I get exposed to new things. No experience with Spotify, but I generally find algorithms suck at giving me stuff I want. When i used to buy books on Amazon, I pretty much never bought one of their recommendations that wasn't "books by the same author".

Pandora seems to do a much better job of recommending things I like...Put in Kris Kristofferson, and I might get a lost Leon Russell jam next. Listen to Elvis Costello and it will play some great, early Billy Bragg next song. Bobby "Blue" Bland, well up next is Bobby Womack...seems to work better for me.

Radio is pretty much iHeartRadio playing the same 30 to 40 songs over and over with a couple new songs getting added every couple of months and a couple of old songs getting pulled.

Satellite radio is just as bad except you don't have commercials so they burn through their list really fast. It's really noticable on long car trips.


I used to use Pandora and would pick an artist i liked and they did a decent job adding new artists and songs into the mix. Of course the issue is if you are a really active thumbs up or down person pretty quickly Pandora will be reduced to playing the same couple of dozen songs over and over.

There were no "that was awesome" moments but it would be fine for car radio, housecleaning purposes. Maybe if I cobble together a 500-song list it will behave better.

Walt, you may want to check out Matthew Perpetua's playlists at Spotify---he has dozens (hundreds?) broken down into straightforward genres but others categorized by mood, or attitude, and it covers everything "pop"---very loosely defined, including jazz, folk, metal, emo, hiphop, etc.---from the 1960s to right now.

Every now and then I dip into one of the playlists---even of a subgenre I'm very familiar with---and there's always something to discover. To keep current, I'll flip on his "2022" (or whatever year it happens to be) playlist, where he has a few hundred releases from the current year across many genres, and while I'm indifferent to much of it, I almost always find something interesting that I can pursue further on my own.

I much enjoyed the news clip I heard about people making money in music by catering to pre-schoolers. Apparently pre-schoolers love to make Alexa say potty words. Musicians noticed this in the Alexa data and started recording "Poopy Poop Face" songs for Alexa to stream.

https://www.npr.org/2022/09/09/1122122287/when-kids-yell-alexa-play-poop-youll-hear-these-songs

I like Spotify, and I don't really care about the quality of their algorithms because I barely use them.

Do I understand the article correctly to say that reliance on hard data led to all music sounding alike?

If so, I disagree strongly.

I won't spend 1000 words defending my disagreement (yet), as maybe I have misunderstood that conclusion.

“As the analytics revolution in music grew, radio playlists became more repetitive, and by some measures, the most popular songs became more similar to one another.”

I fully agree with what the author said about what this approach has done to movies---why bother to try to make a good original movie, when it's now far more important to make a financially successful one and the most likely way to do that is to make a series of sequels

This is not from analytics.

#41 ... I'm the one who took it in the music direction (although the article probably does too). The "sameness" argument has just been applied to radio. I don't think anybody is arguing that music (broadly) is homogenized ... just the new stuff you can easily find although even there you are talking about relatively fine sub-genres.

Oh, I know exactly what I am looking for alright. I just kind of doubt that these mindless algorithms do (c.f. the Spotify example above). MusicalLy it has gotten to the point that I fully expect Sturgeon's Law to relentlessly apply to pretty much any music I try to sample. As a result I am not sure if the stuff I like is hidden behind like 6 magic doors that I have to find and unlock, or if it doesn't exist at all anymore.

Your first sentence and your last bit seem contradictory. Because believe me, whatever you are looking for is available on the internet somewhere. And you (generally) don't need any magic doors, you just need a name and a web search. It's finding out the name of the folks making the music you like that is the challenge. You can either troll through youtube, Spotify, Pandora (I'd completely forgotten about them), whoever and trust their algorithms or you can paw throuh bandcamp profiles or you can bounce around youtube for hours and hours on your own and hope you stumble across something.

But if there's, say, a spec of 60s-70s Philly soul that hasn't been reissued I'd be surprised. Near as I can tell, Fela's entire catalog -- not his entire US-released catalog, the entire thing -- is available for download. Interested in funk from behind the Iron Curtain? You want a wide range of stuff way to obscure for me to know? There are even more labels out there than craft brewers not to mention all the folks just uploading their own stuff cuz why not.

Obviously a lot of folks just want music that was recorded a long time ago. Nothing wrong with that, that's most of my listening but as a "serious" music guy I've got lots and lots of that stuff and presumably you have lots of stuff. Or of course some folks just want to be in a crappy 1967 club watching the Who -- who doesn't but alas only a time machine can truly deliver that one but I assume there's some grainy live Who footage on the internet somewhere.

But there's no good way to find the stuff unless you know pretty precisely what you're looking for. Curators/gatekeepers used to be very centralized (not good for diversity); then it became more decentralied and indie which (for me) worked better but already posed challenges; now that decentralized curation has become anybody with a ISP can post their music on the net and anybody else with a ISP can call themselves a curator.

So for example, the rec (thanks) for Matthew Perpetua's playlists. How in the world would I know to ever look at this guy's playlists? Until the rec (which is still made largely in ignorance of my tastes), why would I delve into this guy's playlists and not some other randomly chosen person's playlists. Playing connect the dots through bandcamp buyers is fine -- but that's what the f'ing algorithms are supposed to be doing. (One suspects the algorithm companies are paid by "big" labels to channel you in certain directions.) Like I said -- this is word of mouth now. It's not 10 people saying "have you seen this show called Seinfeld yet?" it's your bestie saying "there's this Danish political TV series" or some dude on a baseball website or (the closest to Seinfeld) something trending on TikTok or it's you getting obsessed for a few weeks and doing the work.

Alas, my bandcamp collection doesn't overlap that much with other folks -- I've even got something in there apparently owned by nobody else on bandcamp (people Regina Hexaphone are a very fine band! "The Beautiful World" is a very good album.) If it wasn't for me letting my inner depressed teen girl out, I don't know that I'd have anything co-owned by more than 1000 people (Phoebe Bridgers in a landslide). Most of those I found in end-of-year lists from obvious sources (Pitchfork, Guardian, etc.)

Finding out about "cool new stuff" has always been a problem it's just interesting that the nature of the problem has shifted from far too few sources of information to far too many. In baseball too -- we've gone from James, Palmer, etc being locked out of the inner circle even though their stuff was good to anybody with a ISP and maybe some R scripts being a "baseball analytics" expert overloading us with info -- sure he's got a 4.45 ERA but he's 77th percentile in spin rate and 84th percentile in vertical movement but just 18th percentile in horizontal movement so his projected ERA from my model is 3.65, even better if he works with a catcher who's good at framing pitches in the lower left quadrant (has anybody looked at that yet? have I created a cottage industry?)

just a footnote - matthew perpetua is the same fluxblog guy i mentioned in #32; he's quite useful and covers a broad range of stuff.

also, phoebe bridgers is tremendous. heard her once described as making "music for sad dads". i'm not sad? but that is right on.

Now it did a pretty good job of connect the dots but all dots I'd already connected (or intentionally not) and why whould a 75% non-jazz list generate a 90% jazz list?

My experience with the Amazon recommendation engine was that it would take the single most unusual thing you had reviewed favorably and suggest a million variations of it.

You like Bulgakov's Master and Margarita? Then you'll really like another edition of Master and Margarita!

You can see how the math would work, right up to the point where they forget that the things you've already bought are substitutes for the things they're recommending.

Baseball is not a leading indicator of anything in American culture, but a laggard. Baseball integrated well after the army did, after colleges, and the best companies. The drawbacks of excessive analytics and the winner take all society are real, but baseball and moneyball were late to that game. The sequels and lack of original music, novels, etc. is more properly a critique of Reaganism, Japanese just-in-timeinnovations and the abandonment of most measures of success that aren’t pure profit maximization. Moneyball is a symptom, not a cause.

YMMV, but I've yet to hear a song about Moneyball.

Talkin' Softball? After all, Mr. Burns did pull Strawberry to have Homer pinch hit due to the platoon advantage, rather than going the simplistic "Well, Darryl Strawberry is a phenomenal batter" route.

Also, I don't know if you remember the state of the game in 1992, but players with tragic illnesses were seriously undervalued at the time. And along those lines, Burns had not just the obvious Mike Sciosia, but also Ken Griffey, Jr. with his grotesquely swollen jaw! And it's at the very least arguable that Roger Clemens, in his seemingly-permanently-hypnotized-to-think-he's-a-chicken state, could be counted here too.

Baseball integrated well after the army did

Truman's order letting black and white troops serve in the same units came in July 48. Prior to that it was more like the co-existence of the NeL and MLB with Black units separate from white and with white leadership. Once Jim Crow hit, Blacks were mostly driven out and not recruited.

By World War I, “separate but equal” and the rise of Jim Crow laws had created a new wave of racially-charged discrimination. The Navy is a striking example of this. Despite representing a quarter of the Navy during the Civil War and consistently filling 20-30% of the Navy’s manpower in the latter half of the 19th century, African American’s opportunities in the Navy were abruptly curtailed in the early 1900s. During World War I, Black Sailors only represented 1.2% of the Navy, and these men were only allowed in the galley or the coal room.

Despite the proven valor of Black troops, Black Soldiers represented only 1.5% of the Army in June 1940, and roughly the same percentage of the Navy. The Marine Corps and Air Corps, on the other hand, were off limits completely.

Black Army units were sidelined during WW1 and still through early WW2. Even after Truman's order, the Navy/Marines basically ignored it entirely and the Army didn't make any serious efforts until the Korean War. In contrast, by 1959, African-American players had won 9 NL MVPs and 8 NL RoYs (plus Cepeda). The racist morons in the AL mostly took the Navy/Marines approach.

Obviously when Black units were allowed to fight, they fought on the same battlefields as the white units and, in that sense, fought "side by side" but to say the army was "integrated" before baseball is employing a misleadingly technical definition of "integrated."

A brief history

Talkin' Softball? After all, Mr. Burns did pull Strawberry to have Homer pinch hit due to the platoon advantage, rather than going the simplistic "Well, Darryl Strawberry is a phenomenal batter" route.

Yes, it's what smart managers do to win ballgames.