I'm Manny, I'm Noah.

This is Steven and this is no such thing.

The show where we settle our dumb arguments and yours by actually doing the research.

On today's episode, are the weather apps actually getting worse?

And how random is Spotify's shuffle feature?

Really?

I no, there's no no such thing, no such thing, no such thing.

Such those such.

So as the weather seems to be getting more and more unpredictable.

With flash flooding overtaking our subways in New York City, you.

Can see water spouting out of the walls, people walking across flooded platforms, and that rather dirty water.

To the tragic, deadly floods in central Texas, we.

Confirm that at least one hundred and thirty five people have been killed.

This is in addition to the more than one hundred people who remain missing across several counties.

This week, we are turning our attention to our weather apps and then nottedly.

Many people in our lives have been complaining that the weather apps are just not accurate anymore.

So we're going to get to the bottom of that and answer all your other questions about weather apps, including this one from my friend.

Alek Okay, So I want to know why we as society can't come together and use feels like temperatures instead of actual temperatures as the default, because right now, when they tell me it's going to be sixty five degrees, I then have to go and do other homework on like how wendy is it going to be that day?

And is it going to be really humid?

And it seems like they have come up with one number that can give me the answer that I'm really looking for here, which is do I need a jacket that day?

That's a great question, and it reminds me of when you guys used to make fun of me for saying that it feels quote unquote good outside like it feels good, and I maintain that many people say that maybe not from you know, Connecticut or Long Island, but I just wanted to get that off my chest.

Yeah.

I love the idea of like a Manny weather app that.

It just says feels good.

That's all I need to know.

But just to be clear that the feels like it's a part of the app, I actually don't see it that often.

It's a part is on the weather?

Which app is.

This all the apps kind of have it now, so there's usually app okay.

And they don't really have much.

So like, let's pull it up now.

It's usually you know, the air temp which is up top, and then feels like it's usually below that or very deeper.

So like this, this is the default Apple one.

You really got to scroll here.

So it's eighty seven right now from Brooklyn, and then we scroll all the way down, feels like ninety one.

Okay, that is helpful.

It's helpful, but I guess I do also want to know the real temperature.

That's my guess is that y'all do you want to know the real temp?

I don't care about the real temperature.

It doesn't matter.

This was especially true remember when we had those really really windy days.

Oh yeah, it's it's yeah, it's in the hot weather.

It's less of a thing I feel for us anymore.

I mean, well, no, still thing, because.

It can be in the cold.

I feel like that the wind can change it so dramatic.

They'll be like it's forty degrees out.

I'm like, oh, this is kind of warm for winter.

And then it was like the wind is making it feel like it's fifteen exactly.

But it's like I don't hear what the still cold day?

Okay, I can live in this versus like I'm about to die in.

Free because why are we checking the weather?

You're checking them whether to see what I gotta wear.

Yeah, it's true for me, it's.

We're not you know, like that's what most people checking other while they're checking the letter.

I might wear something different in sixty five degree weather than you might.

Yeah, I mean you still need to know what to wear for you.

That's why I want to know the real temperature.

Yeah, the real temperature is the real temperature, no matter who well look.

At no, no, but no.

But if it's all right it's sixty degrees but feels like ninety, yes, you you want to dress for ninety yes, yeah, but.

It feels like it's actually better for you because you want to know what it's going to feel like, so you should know what to wear.

The actual temperature doesn't know what.

It actually make a difference if it's actually if it's actually sixty.

Yeah, because I want to know if we've breaken up broken any records today.

Okay, okay, that's different, that's what you're wearing, breaking records is.

Over here.

I have a map the historical data.

I want to make sure that you can assume we.

Are breaking a record every day.

So I talked to meteorologists at the Weather Company, okay, which you may know for the Weather Channel brand.

So my name is doctor Loriana Gaudett.

I am a staff applied meteorological scientist at the Weather Company.

I've worked primarily within our forecast sciences team, so I'm not operationally forecasting the weather, but i am working on the science behind the scenes per se to improve our forecasts.

All right, So one of the first questions that we got on weather is this idea of real feel.

Right, you get our actual temperature and then we get the real field tamp.

So can we start by just having you break down what is the real field tamp?

How is that calculated?

So the real field temperature is more so known as the feels like temperature, and what it's composed of is a combination of the heat index and the windshiel temperature.

The heat index itself is a combination of the temperature and the relative humidity and how those two interchangeably play together to affect how we feel when we're outside.

And so if it's really humid outside, less of that sweat is able to evaporate off of our bodies, and we are less effectively cooling ourselves down, which makes it feel much warmer.

As an example, if it's ninety degrees fahrenheit outside with a relative humidity of seventy percent, the heat index or the feels like temperature in those conditions is around one hundred and five degrees fahrenheit, which is very warm and it can become dangerous with heat stress impacts.

And then on the other side of the temperature spectrum, the feels like kind of points to the wind chill temperature.

So this time we're considering the effect of wind speed on temperature, and the reason that that's important is because the wind is effectively transporting the heat away from our body, which is cooling down our body or our skin temperature, and then over time core body temperature, which can also come dangerous in its own way through frostbite or through hypothermia effects.

So another example here is that if you have a temperature of thirty degrees fahrenheit and a wind speed of fifteen miles an hour, your wind chill or feels like temperature would be nineteen degrees fahrenheit, give or take.

Interesting.

So basically like seventy degrees with no humidity feels a lot different than seventy degrees with humidity, and some you know, X amount of miles per hour.

Wind exactly to take away I got from talking to her, it was like, the feels like takes into account how the temperature affects our body, and it's not just all right, here's the air temp.

Yeah, it's not so abstract.

I guess it's like kind of an algorithm basically.

Yeah, and yeah, it's about how you know your sweat's going to be a vap rated or not or whatever.

That's it's more thorough than I thought.

Yeah, I kind of thought.

A scientist and a lab coat walked outside, I was.

Like, okay, sixty five, Yeah, we wan'll make them feel good today.

But something is interesting about feels like, which we talked about a little bit already, is that someone like manny, if it's seventy tract Yeah, if it's seventy seven degrees outside, man's complaining as hot?

Yeah, well, not that it's hot, it's too It's like, does it's warmer than I prefer?

It feels bad?

It feels bad, bad, whereas with us that's seventy seven temperature bad, not bad.

So well, I.

Don't know, Okay, don't put that on that.

Okay, what do you what's what seventy seven.

Of you that could get pretty hot?

Okay?

Seventy three is comfy, okay, And I want a little bit of clouds.

And speaking of shade versus sun, when they take the temperature outside, did you know that they're taking into shade?

Whoa, oh, why is that?

I did not know.

Well, we're gonna find out.

So temperature and heat index are recorded in the shade, and so if oh, yeah, what I know, Yes, it's it's super surprising.

But the reason is we don't want the thermometer that's measuring the temperature to heat up and not be able to actually tell us what the air temperature is.

We want that not the temperature with the thermometer.

So when you were outside doing whatever activity in the sun, it's going to feel warmer than the reported air temperature and even the heat index.

So that's important to keep in mind.

And then the other component is that scientists when they were deriving these heat indices and the wind chill, they had to make certain approximations and assumptions.

So some of those involve like how tall are you, how much do you weigh, how much do you sweat, what kind of clothes are you're wearing, so on and so forth, and that obviously can't account for each of our unique body compositions and circumstances, but it's our best general approximation for what you might experience when you walk out the door.

The shade thing really blew my mind because I'm every time I go outside and I'm in the shade, I'm like, oh, this is cooler than what's the app because I'm in the shade.

But that's the best case scenario.

You're only going to get hotter if it's ninety degrees.

That means if you're in the sun, it's hotter than that.

Yeah.

Yeah, if you're walking around, you're gonna be sucked there sometimes.

Yeah.

There's so many variables, which is why I just try to distill them all into good or bad.

Yeah.

I know.

Now I'm kind of like many systems well because it's like, well, this makes me want them to have shade, Yeah, because I need to know.

You open the weather app.

It's just like an Excel sph Yeah.

Ever dial in and then you dial in your outfit yeah, and your skin tone calculator.

I mean.

Imagine you have an or a ring.

Yeah, that's that's a lot of data.

Yeah.

And then you didn't tell weather you take a selfie it skins your clothes.

You don't even need to take a selfie, yeah, right, your phone.

Yeah, and and you just exists and Peter Tiel buys all of this data.

Yeah, exactly, and then he but he tells you it's going to be only seventy two today, or it's only going to feel like seventy.

Two when we look at the weather app.

Besides many lying and saying he wants to look at historical data, we're looking at it to see what we should wear to go outside.

Do I need to bring an umbrella?

Why don't we have this as the default?

If I open up the app, I want to see feels like and then if I want to search for the actual air temp, I could find that lower down.

Why are we doing the reverse?

We take a certain approach within our Weather Channel app, and we've supply both temperatures, so not just the temperature, but also the fields like temperature, so that you can have both data points and use those to make decisions about how you want to plan your day or go about your day.

But yeah, it's a really good question because it gets down to the core of like the human experience of course, but there's other impacts and implications of the temperature that also don't relate to humans.

So some of those might be like if a farmer, for example, is trying to figure out when to plant their crop or protect their crop, they are looking at the actual temperature.

So that's what our weather forecast models are actually outputting, is the real air temperature, and that's what we are measuring and looking at when we are monitoring the climate, for example, and it allows us to compare the observations of temperature across the country, across the world, and over time.

Whereas, like with everything that we talked about with the feels like temperature, there's a lot of approximations that are being used, and so some weather services given the same environmental setup might come up with different values.

That makes it really difficult for that inter comparison component.

But if you yourself, as the user of a weather app find that the feels like temperature gives you everything that you need to know you can set that as your own to fault temperature.

But we just don't want to make that assumption for all of our users.

So they're just very considerate of me at this point.

The thing that impacts most of us would be it feels like temperature that should be to default.

Make everybody else look for the other stuff.

Please contact your company.

I guess do you feel I understand all everything that she said.

When I'm looking at my app, I'm not looking for climate.

You know, over the last ten years, I don't have any crops.

I guess I can do it within my own but I feel like as a country we need to we neither do real feel I'm not saying get rid of the real number.

I'm just saying put that lower.

See why I can't?

Why not just side by side?

Oh slash yeah, pretty much?

Okay, big one, I mean, pick whichever one you want, big one then basically an arrow or slash, Yeah, basically, and you just know that you just know, okay, this is you know, you know the real thing.

And here's the thing.

It wouldn't take up that much space.

They're digits and like the numbers are hopefully not Yeah, yeah, hopefully four digits.

You're look at that total.

I'm fine with the normal one.

I'm fine with the regular Yeah.

My main concern with these things is doing an umbrella or not, which I forget to look for most of the time.

Anymore, Speaking of rain, Manny, do you want to explain this Twitter argument that people have been having.

When it's raining or when it might looks like it might rain.

You go on the app and says, let's say, for example, thirty five percent.

Doesn't say what the thirty five percent is.

So some people think that it's thirty five percent chance it's going to rain where you're standing.

Other people think it's in a rain on thirty five percent of the area around you, yes, and a third group of people think that's effectively the same thing.

Well, well what do you what do you guys think?

Before I did this interview, Yeah, my thought was it's a thirty five percent chance it's going to rain.

That's not that's my thought too, because the area doesn't make because it's like, well, what's what area?

A lot of people, a lot of people online, these are not experts, but they were like, no, that is what they're telling you.

Well, what did you think I was thought, thirty five percent chance is going to rain.

Chance, Yeah, where you are?

I agree, That's what I'm looking at, and that's how I feel like my lived experience is.

You know, it's like, well, that's where it gets tricky.

It's it's never right to me.

But yeah, I always assumed it's basically there is a percentage of chance that it will rain.

Not whatever area I'm taking in, it's going to cover sixty percent of this.

Yeah, yeah, that.

Doesn't really that would never cross our behind me that much.

So this is a common area of confusion for users, and even in our own field of meteorology, it can be really hard to pinpoint what we are actually communicating with probability of precipitation or pop In my experience and knowledge base, I think both of those interpretations are true in their own right, which of course makes it even more confusing.

But I can I can share from the Weather Channel app perspective, the way that we're communicating it is really in your surrounding area.

What is the likelihood of precipitation or of rain impacting you within whatever time frame you're interested in, So, whether it's today, the next hour, fifteen minutes, and less so the amount of area that is going to be impacted.

A couple misconceptions that also come up to are about the intensity of precipitation from pop.

So sometimes people will see like, oh, there's a ninety percent chance of rain, it's going to be a downpour.

We are indicating the likelihood of rain, of measurable rain, any rain, and so generally it has nothing to do with the intensity of that precipitation, which is important to keep in mind.

And then the other component two is you know, sometimes people think, Okay, there's a fifty percent chance of pop in New York City, so only half of the day will be wet, And again you could have that fifty percent chance, there might only be a couple hours that are impacted, and then you have a beautiful remainder of the day.

I think that it really lays on us meteorologists to make sure that we're communicating what that likely hood and probability means, especially if it does vary, you know, from a news broadcaster or a meteorologist who's on TV trying to communicate the likelihood of rain, to a push alert from your weather app telling you that there's rain coming within the next thirty minutes.

That is really funny because the idea that the percent chance is the total coverage is such a like, actually, I'm smart, and this is what they're really saying, fools in conversation.

I've heard it so many times and it's just wrong.

And then I mean it's kind of fascinating, like people are just taking this and twisting it in any direction where it's like, okay, ten percent, that's ten percent of the day.

Yeah, so yeah, okay, so whatever ten percent of twenty four hours is.

It's like, yeah, thinking of all the different ways to map it out, it's kind of that's actually like a SAT question or something.

Yeah, I heard the rain is only coming from ten percent of the clouds.

Yeah, exactly.

Yeah, but there.

I thought it was interesting that they're kind of they could both be right depending on your weather.

You need explanation.

I mean, I think that you know, the apps that we're looking at or telling us, I.

Would hope so frankly, I think that you know, maybe they could do a better job than of getting the word out there hopefully.

You know, we're going to do our part today with our you know, hundreds of thousands of listeners.

Some would say millions, So yeah, you could argue that millions.

Maybe all right, we are going to take a quick break and when we get back, we are going to find out are the weather apps actually getting worse?

And it's Trump to blame.

So in last week's mail bag, we were debating who gets the right of way between a school bus with the stop sign out, an ambulance with the siren on, and a funeral procession, and from North Carolina called in with this insight.

I am a state wide law enforcement officer here in North Carolina.

If I am running lights and sirens, I am allowed.

To go through a soft sign.

People legally have to yield to me.

But if I run a soft sign and hit someone, it is still my fault, and I, as a driver, would.

Not only be responsible to my agency, but I could also be personally legally financially responsible funeral processions.

You guys are correct, that is a courtesy.

It's not a legal thing, and but it is it drive me.

Grace to get down here in the South, they love them at funeral profession.

So we're going to try to include more listener responses to our episodes like this in the future, So you can email us at Manny Noladevin at gmail dot com or leave us a voicemail at the number in our show notes and hey, you may includes you in a future episode.

All right, let's get back to the show.

We do have one final weather question from Mia.

I have a question, and this is just bugging me so much, especially this summer, our weather apps getting less accurate because I feel like they are like these days, I just don't feel like I can trust the weather app to tell me exactly what's going to happen, specifically if it's going to rain, like over this the course of the summer, It's been a rainy summer in New York, and now I feel like I'm becoming this kind of like medieval farmer where I'm sort of gauging what's happening in the sky, and I feel like I am becoming a much better predictor of what's actually going to happen than what my phone is telling me.

I'm just starting to be like, what's big weather even there?

So if you could answer for me, is it getting less accurate?

What's the reason should I trust this app anymore?

That would be great.

Thank you.

I put this question to our meteorologists from the weather.

She was sweam bullets at this one.

Is the weather more unpredictable or is this on in our heads?

And you know, has nothing really changed.

So that's a super intriguing question.

No, as a short answer, the aps are not getting worse at predicting rain.

I think there's a there's a few different things at play here, one of which you pointed at.

It's psychological, so we're more likely to remember when the weather apps get it wrong, versus that when we are seeing a prediction of, oh, it's going to rain in ten minutes, let me make sure I grab my rain jacket before I go out, and it ends up being correct, and so you're prepared.

It becomes a non event and you forget about it.

But let's say that if someone is outside it starts pouring, there's absolutely no warning.

Of course, you're going to be annoyed that you had no notice that that was going to happen.

And there's a few reasons that those specific circumstances can crop up.

This example of stepping outside and being met with an unexpected downpour of rain can largely happen because even if the radar is able to pick up on a storm let's say it just started raining within the past few minutes.

The weather radar just got that data.

So for example, that data needs to be ingested into our back end systems or any other weather app systems to process that data, figure out where that storm is using proprietary software and algorithms, predict where that storm is headed and if it will intensify or again dissipate, get that forecast to our technology system downstream, and then those then deliver that information to your app.

So that takes a bit of wall clock time for all of that process to happen.

So unfortunately, there are those situations where rain can impact you within that window of time that we're trying to get that information to you as fast as possible, but perhaps missed that window of opportunity.

I've never thought about a situation where the weather app was wrong, because it just was like trying to catch up to the real time m hm.

You would think.

You're supposed to be predicting.

It's like, don't blame me, I'm not the weather company.

Hello.

And now that I remember what Mia was complaining about, I think her concern is that the apps are more frequently incorrect, not that they're ever incorrect, but that it's happening more often.

Yes, yeah, she thinks it's getting the apps are getting worse.

So the Weather Company said, no, that's you know, the apps aren't getting worse, but they work at the Weather Company.

So I wanted to call up someone who can maybe speak a bit more freely about whether or not these apps are getting worse and do we need to worry, especially with these Doge cuts, about them being worse in the future.

So I called the Mary Glacken.

I'm the former deputy undersecretary for NOAH for operations there and I had worked for Noah for just about thirty five years.

I also am on an AI board and things like that, so today these days I'm mainly in an advisory capacity.

She also worked at the Weather Company back in the day.

All right, great.

Job finding, Mary.

I always have stuff to say about my former employees.

It should be good to take.

So, you know, the big thing that we were concerned about and seeing the headlines is that with the Doge cuts, that there's just less people working, which is why the apps are getting worse.

So I asked Mary, like, where do we currently stand.

So let me talk about it in terms of personnel first, because one of the major things that happened was the exodus.

You know, there was the Fork in the.

Road, another unprecedented action from the Trump administration, offering more than two million federal workers a buyout.

The workers receiving an email with the subject line fork in the Road, giving them a choice resign and be paid through September or risk being fired.

So the National Weather Service and all of Noah are down significantly in terms of people, something approaching as much as twenty percent in some locations.

So the National Weather Service did lose a lot of senior people.

In terms of funding, they had actually been funded by Congress last year for a full year at a reasonable level, but the Trump administration has delayed spending of those funds, so they've really been slow rolling contracts to go out for various things.

I don't have any figures here, but it's safe to assume there's a fair amount of money still on the table.

But the direction that we saw this week was a spend plan for the rest of this fiscal year is clawing back that money.

It's saying you're not allowed to spend that anymore.

So that's that's kind of where we are.

We've seen some news about them hiring.

So earlier this year, President Trump's Doze Office cut more than five hundred Weather employees.

Well now WS says they're going to hire four hundred and fifty meteorologists, hydrologists, and radar technicians.

They're all entry level positions, So you know you're not going to hire back the expertise that's been lost that one of the senior officials that Noah career official had quoted, that was twenty seven thousand years of expertise that walked out the door between February in April.

It's just going to be a lot of people, a lot of college who don't have the expertise, and now they also don't have the leadership to mentor those people.

They just went from an NFL team to like JV high school team.

So I wanted Mary to answer the question because I couldn't take the Weather Company's word for it.

Are the apps getting worse?

The apps haven't degraded at all.

I think the worries that we have with the Trump administration is if they move forward with their plans.

Literally one of the plans is to eliminate all of Noah's research arm So and that's all our future.

You know, how do we do a better job in forecasting that whole section would be eliminated?

Which is crazy, but that's the plan.

I do want to kind of remind folks that Congress has a say in this whole thing.

Both the House and Senate that have given their plan on how Noah would be funded.

The Senate does a really pretty good job and keeping Noah funded.

The House has some significant cuts in it, but nothing as draconian as the administration is asking for.

I think the problem is that Congress won't be able to agree on the overall spending, which will really basically allow the Trump administration to continue to do what it wants to do.

I was curious if we've kind of hit the point of no return, you know, see if President Biden runs again and.

Wins, yeah, you know which, we're in eight can we get back to where we were?

One of the things I'd like to make clear with you know, all of my experience with Noah, both in Noah and then working with Noah outside is I would not be the one arguing for status quo.

You know, I'm not here saying oh, it was a perfect world and now this has happened.

I think it will be possible to get us back if we can keep people in the field and interested in our problems.

You know, you have the whole advent of AI, and AI is impacting our value chain all the way from taking an observation to helping somebody make a decision on what to do that day.

So when I look at AI, it's pretty clear to me and many is the private sector is ahead of where Noah is with AI, and that's no great chock.

You know, the private sector can pay somebody a half a million a year because they're a genius in AI.

You know, if you go to nowhere, you're going to be making ninety thousand.

So that says to me that we need to find a way to work with the private sector that meets public good.

You know, how do we ensure that everybody gets warning that they need and how do we ensure people get the basic information.

To do that?

So I do think we need to think more creatively about public private partnerships.

You know, right now, there's not a lot of mechanisms for the private sector to work with the government sector.

And I think if we you know, if we started rebuilding tomorrow I think that's what we would be thinking about, Well, what really makes sense here to do in the future.

Yeah, I'm like, I'm like your run of the mill kind of lefty, right, who has some skepticism about what the private sector can do, like for the good of the people, quote unquote.

But it's interesting how this situation is so die or that, like we do just kind of have to look at it.

All.

Right, So we've been, you know, trying to make sense of the world through weather.

We're going to take a little break.

I'm going to leave you guys, and then we're going to try to make sense of the world through Spotify's shuffle feature.

This is Devin, Welcome back to no such thing.

A few weeks ago, actually, we got an email from this listener named Austin.

He wrote, after listening to your latest mail Bag episode, it got me thinking about an issue I've run into over the years.

Whenever I hit shuffle play for a music playlist on Spotify, the shuffle feature never seems to be truly random.

I want to know how does shuffle work for music apps and is it truly random.

I'm Heather mccoldon and I'm a writer, artist, and sometimes a tarot reader.

All right, So I reached out because we got a question from a listener who wanted to know if we could figure out how Spotify's shuffle feature works.

And I came across your article in the Financial Times cool in which you dive into this.

Would you mind, I guess kind of setting the scene for me on what happened on December fifteenth, let's say, twenty twenty three.

Yeah, when all these random acts started coming together.

Yeah, basically, you know, it's like one of those things.

It's like holiday time, Brooklyn, like the aras chilled.

Everyone's excited and happy for like the end of the year beginning of the new year.

I was visiting some friends at a dinner party in Williamsburg and you know, too much wine, too much festivities, like everyone leaves on masks to go home, and yeah, I was feeling like kind of nostalgic, like I had a partner in that neighborhood, like, and you know, it's just brought back memories and you know, my friends were like, oh, like we opened the door from the building, there's a cab right there.

And I was like great, Like the universe is working out for me.

So yeah, like I get in and I'm just like, oh, like went into my liked songs on Spotify and just hit shuffle and the song Maps by the yayaya As comes on, which is, you know, like this epic contemporary love song and basically like the taxi just like decides to, you know, go a certain way and we literally like pass my ex's building as like this song is about like, you know, saying like you know, they don't love you like I.

Love you.

Then, and I just had this moment of like, what the fuck is happening, not just like in my life in the universe, but like what is this app even?

So, yeah, that led me to investigate those Spotify shuffle works in a very randabout way.

But you know, like I think that's like what's kind of cool about existing now like in this era where it's like, how are these technologies and their glitches affecting our emotions in our personal lives?

And you know, is it random?

Is it all random?

Et cetera.

Can you talk me through I guess starting with the first iteration of like what Spotify shuffle used to do.

So basically there's like a very simple, elegant, mathematical description of randomness.

That's called the fisher Yates algorithm, and essentially this is like from Spotify's insight option.

Until somewhere between twenty twelve and twenty fourteen, it was using fisher Yates from my understanding from what's available fisher Yates basically, it's an algorithm that will take an.

Array of values.

Let's say, like you have numbers one through ten, it removes five, and five becomes the first track of the shuffle list, and then it just keeps kind of you know, randomly selecting the other values to like create a new list, Sasha, So every number in that list has an equal chance of appearing in the first position or the consequential one.

It's great because technically that's completely random.

Right.

So Spotify engineers were like, this is awesome.

You can do this in between like two or three lines of code, Like it was so simple.

They're like, yeah, we sold this, and like kind of instantly the user complaints came in of like my shuffle is broken.

How is it possible that you could screw this up?

It's so simple?

And the complaints were at the time because I remember seeing them on social right would be like, how's you know Spotify Shuffle Ranmouth.

They're playing four Taylor Swift songs.

In a row one hundred percent.

But you can see, like based on how I described how it works, Ye, randomness does create clusters.

So it creates like, you know, clusters of songs from one album or clusters of songs by the same artist, Like each of those songs has the exact same chance of like emerging, right, there's nothing preventing that.

So what the engineers finally determined was users were falling for what's called the Monte Carlo fallacy.

So this is basically a human perception of randomness that is not correct, and it's named after this very specific event that occurred August eighteenth, nineteen thirteen, at the famous Monte Carlo casino.

We're essentially at the roulette table.

The ball kept falling on black.

It starts to create this weird chaos and like sensation in the casino because people are thinking like this isn't possible, right, like you know, two or three times sure, then you get into like five to ten times, you know, like whirl, like this is unusual.

So yeah, something's up, and people kind of come down with this fever of just expecting like on the next turn of the wheel, red, Red is going to happen.

Red is going to happen because Black has happened all these times of beforehand, and that night all fell on Black twenty six times in a row, which is just like it's like an astronomical this happened, and yeah, it just this whole situation pointed out that like it's also called the gambler's fallacy.

It's where you think previous events somehow affect the current event unfolding.

So that's why people get angry at like the clusters of you know, Addison Ray or whatever.

Yeah, it's like, okay, you're already played in Addison Ray song.

Therefore their Monte Carlo Felas is like, yeah, it has to be something different now because we've done Black so many times in a row.

We played the Addison and Ray song, so therefore an X song cannot possibly be diet pepsi, you.

Know, right, and then like what the fuck it's diet pepsi?

Like are you kidding me?

Right now?

The engineers were like, oh, okay, people are having like this weird issue, like they don't realize what randomness actually is.

So then they created essentially, like it's not like a hybrid algorithm, but they created a different one that mimics or attempts to mimic, the human perception of randomness.

So it does this way.

It takes your playlist and it's like, oh, like, I don't know, You've got a lot of like Justin Bieber in here, So it's spreading Justin Bieber and then with being Justin Bieber, it's spreading out the albums.

Yeah yeah, yeah, yeah.

So it's called cluster breaking.

Yeah, you're not going to get to Justin Bieber songs within a ten track shuffle from Swag.

We're going to make sure if you get to Justin Bieber songs, you're from completely different albums.

One hundred percent.

Yes.

So that was their solution.

And keep in mind, like what I've just described with the cluster breaks, like that's from twenty fourteen, so like we there's been nothing really that they put out since then that describes what's going on.

People are now in this weird position where they're like they want, like the Fisheryates algorithm to come back as like an additional like like if shuffle could have yet exactly, and they're like yeah, like I know it's bad, but like it would just be preferable to like whatever this weird thing is.

Our human minds are so small that, you know, we're only able to comprehend like a very like tiny wedge.

Of how the world functions.

Right, So it could be that like we're seeing all these things as random, but really it's like, oh, if we could actually take a step back and like have more of a bird's eye view, potentially these things could be connected.

And they just operate at a pace that's so slow that like over the core of one human lifetime it appears random, but over the course of like five hundred human lifetimes, it's actually like one slow process that makes perfect sense, but we're just not able to comprehend it, right, So you know it just randomness is so interesting because in a sense, it's undefinable.

Yeah, right, So that's why it's so interesting when you have something that's essentially like a math problem where it's like, come on, like I have like I don't know, thirty five hundred songs.

It can't like how how complicated could it possibly be?

And it turns out it's really complicated.

HEATHERN.

Mccalldon is the author of the Observable Universe.

We're going to put a link to it in our show.

Notess Hews Hews her howss.

No Such Thing is a production of Kaleidoscope Content.

Our executive producers are Kate Osborne in Mangesh Hadi Cardur.

The show was created by Many Fidel, Noah Friedman and Me Devin Joseph.

Our theme in credit song or by Manny, mixing by Steve Bone.

Additional music for this episode by certain Self.

Our guests this week we're doctor Loriana Gaudette from the Weather Company, Mary Glacken and Heather mccauden.

Thanks to Alec Mia and Austin for the questions.

If you have something you want us to get at the bottom of you can email us at Manny Noah Devin at gmail dot com.

Sign up for our newsletter No Such Thing that shows We'll

See you guys, those by those such things