Alter Everything Podcast

A podcast about data science and analytics culture.
Alter Everything Podcast Survey

Be entered to win a prize while sharing your thoughts about our podcast.

Give your two cents here
It's the most wonderful time of the year - Santalytics 2020 is here! This year, Santa's workshop needs the help of the Alteryx Community to help get back on track, so head over to the Group Hub for all the info to get started!
Alteryx Community Team
Alteryx Community Team

In the third episode of our “Data [in the] Sandbox” mini-series, Susan tells Maddie all about psychics and predicting the future! Erm... well... actually she talks about predictive analytics, and how you can use patterns in data to make educated guesses on what is likely to happen in the future. Tune in to hear how!

 


Panelists

 

Maddie Johannsen - @MaddieJ, LinkedIn, Twitter

Susan Sivek - @SusanCS, LinkedIn, Twitter

 


Topics

 


Transcript

Spoiler

MADDIE: 00:00

Hey everyone. It's me again, Maddie Johannsen, and welcome back to Data in the Sandbox. This is our third episode of the miniseries, and I'm so excited because today my friend Susan will tell me all about predicting the future. Well, I guess actually it's kind of complicated, but luckily, Susan knows what she's talking about. So just keep listening and you'll see what I mean. Hey Susan, I want to tell you about something. I was asking my mom about her job the other day, and she said she makes predictions about stuff at work. I asked her what that means and she said it means she can see the future.

SUSAN: 00:42

Oh. Well, are you sure that's what she said?

MADDIE: 00:46

Well, yeah, it's almost like, I don't know, she's a psychic or something.

SUSAN: 00:50

Well, how do you think she's making predictions?

MADDIE: 00:54

I guess I don't really know. Fortune cookies? A magic eight ball? Oh, or maybe she has one of those sparkly crystal balls on her desk. That would be so cool. But yeah, I think it has something to do with data like we've been talking about.

SUSAN: 01:10

That is probably more likely than the crystal ball though that would be really cool. But it's funny there are kind of some similarities between what people who claim to be psychics do and what your mom is doing when she makes predictions with data at work.

MADDIE: 01:26

Wait. Seriously? Okay. Now, that sounds way more interesting.

SUSAN: 01:30

Yeah. So, tell me what you think of when you imagine a psychic? What does that scene look like?

MADDIE: 01:38

Oh, okay. So I picture it like a fancy tent at a circus with lots of velvet and gold and a fancy rug on the floor. And when you go in, there's this little table with a psychic sitting behind it on the other side, and maybe one of those crystal balls on the table and a chair for me to sit in while I get my fortune told. And maybe it's a little dark and there's a candle burning.

SUSAN: 02:01

Wow. Yes, that's what I picture too. So, okay. You walk into this little tent with the psychic and what happens next?

MADDIE: 02:09

So I sit down and the psychic looks into the crystal ball and somehow, they know things about my future and what will happen. And I can ask questions for the psychic to answer, like will I find love or be happy in the future?

SUSAN: 02:23

Oh, that's so romantic. So sweet [laughter]. So do you think the psychics' answers will be right? Will their predictions actually come true?

MADDIE: 02:33

They sure seem like they could. I guess I don't know yet.

SUSAN: 02:37

Well, that's a good answer. But yeah, it's amazing how good some people who say they're physic can be at saying predictions that feel perfectly reasonable, like they might really come true. Do you know how they do that?

MADDIE: 02:53

The crystal ball, remember?

SUSAN: 02:55

Oh, well, no, not actually [laughter]. Actually, what they're doing in a different kind of way is they're gathering data for different variables. They're making a hypothesis or it's just kind of a good guess and then they're coming up with a prediction.

MADDIE: 03:11

What? That again?

SUSAN: 03:12

Yeah. Variables and data all over again. So, yeah, because I hate to tell you this, but most of the people who claim to be psychic are actually just really good at collecting information quickly about the person they're talking to. They know that there are certain little pieces of data about someone that can make that psychic really sound like they're on the right track.

MADDIE: 03:35

So what's the data they're collecting?

SUSAN: 03:38

Well, it can be a lot of different things. They'll definitely make guesses, just to themselves, not out loud, about how old you are, where you're from, your background. They might look at your clothes or your hairstyle or your jewelry, like maybe a wedding ring or something. And in a way, you could say those are all variables for each person. Their age, whether they're married or not, whether they're rich or poor or somewhere in the middle and so on.

MADDIE: 04:04

Oh, yeah, variables. I remember talking about those. Those were the pieces of information that can vary or change, like whether I had a snack or not before running or how many shark attacks there are.

SUSAN: 04:14

Exactly. Yeah. So the psychic, they're just this really observant person and they're basically just quickly assembling a dataset. Remember, that's like this collection of all the information. This sort of mental dataset about each person who walks into the tent.

MADDIE: 04:31

Wow. They have to do that so fast.

SUSAN: 04:33

Yeah. Because then what they're going to do is use all of that data to make a prediction. They're using all the information they know about someone who fits your description to start making good guesses about you. And when they say those guesses, you agree or not, or maybe you just make a face if they get something wrong, and then they quickly adjust whatever it is that they're saying. So all of that information they're collecting from observing you and interacting with you, all that information goes into what we might call a model of what you would like to hear about things that might happen in your future. [music]

MADDIE: 05:10

Wait. A model? Like people who show off clothes or hairstyles?

SUSAN: 05:16

No. It's a different kind of model. So when people like your mom build models with data that's in the form of numbers, it's basically like building a math problem. And when she solves that math problem, the answer, just like when you're answering a problem on your math homework, that answer is a prediction.

MADDIE: 05:36

But what is she predicting? People aren't asking my mom about their love life or something?

SUSAN: 05:42

No, probably not. Well, here's one example. [music] So maybe your mom needs to predict how many people will visit one of her company's stores this year. She's going to gather a lot of data for a bunch of different variables and she'll design a way of putting the data together. That's like the model, using math. It might look something like an equation if you've maybe seen one of those in school. The model then uses math to predict the number of people who will go to the store, or more likely, she'll say something like, "I predict the number of store visitors will be somewhere between, oh, 10,000 and 12,000."

MADDIE: 06:20

But the psychic isn't using math, right?

SUSAN: 06:23

No, probably not. That would be pretty complicated to do right there on the spot. But instead, the psychic kind of has a mental model for what you might be like as a person and the kinds of things that you might like to hear. For example, with a young person going in to get their fortune told, they might talk about how, "You'll find love when you go off to college." So since that young person is probably not going to college right now and maybe not even dating yet, that seems like a very reasonable prediction, all based just on the variable of your age.

MADDIE: 06:56

Oh, okay. That's pretty neat. I guess they're not actually seeing the future, but just making a good guess based on what they observed about me when I walked into the tent?

SUSAN: 07:05

Precisely. Yeah. And that's kind of what your mom is doing with her models and predictions at work too. She's making the best predictions about the future that are possible with the information, all that data, that's already been collected, but she is using math to analyze those numbers and try to see what will happen in the future.

MADDIE: 07:24

So, no crystal ball, just math? All right. But I want to go back to something you said earlier. You said my mom would probably predict something like the number of store visitors will probably be between 10,000 and 12,000. That is kind of vague. It doesn't sound like my mom is a very good prediction maker.

SUSAN: 07:43

Oh [laughter]. No, no, no [laughter]. Your mom is doing great. That that's exactly the right approach. The reality is that even when we have a lot of information and some very cool math, we can't and really shouldn't try to make super precise predictions about the future, especially if things are kind of uncertain, or they might change a lot between now and that time in the future. So if your mom told your boss, "I predict there will be 11,000 store visitors exactly next year," but then there were actually only 10,000, what do you think your mom's boss would say?

MADDIE: 08:18

Her boss would be like, "But you told me there would be 11,000! Where are my other 1,000 people?"

SUSAN: 08:23

Yeah. Exactly. Right. So it's good to plan for a range or a variety of possibilities that could happen based on the data and all our fancy math. That makes it more likely that our plan can be flexible and can work with a bunch of different potential situations. Models aren't perfect.

MADDIE: 08:41

That makes sense, I guess, since, I guess, psychics aren't really real.

SUSAN: 08:45

Yeah. Yeah. If they were real, we could just hire them to do all the predictions. But instead, we use cool software and code and math to figure out things as best as we can. And we also kind of do what the psychic in the tent is doing. We can constantly make our models and our predictions better by asking more questions, gathering new data, and adjusting as we go.

MADDIE: 09:09

So what about some of the other things we talked about? Could we make a model to try and predict how fast I could run around the track at school on a certain day?

SUSAN: 09:17

We could try. What kinds of information or data do you think we would need? Which variables do you think would be important?

MADDIE: 09:24

Well, I know I run faster on days when it's not too hot. So I guess the temperature outside would be good. And then maybe whether I had a snack or not. And I guess maybe whether I ran the day before or not because I might be tired if I ran the day before.

SUSAN: 09:40

Sure. Yeah. Yeah. Those all sound like really good things to include. So what we could do is look at how fast you ran around the track on previous days, and we could make a note about the temperature, your snack or no snack situation, and whether you ran the day before.

MADDIE: 09:56

So does that mean we have four different variables? Four different things that could change?

SUSAN: 10:00

Yeah. Exactly. Nice. So yeah four variables to include in our model or our mathy kind of way of putting all that data together. Our prediction is going to be the time that it takes for you to run around the track. Now, if we had a lot of data on your past runs, we could actually maybe do a pretty good job of predicting how fast you would run on a future day if we could find out what the temperature might be and if you could plan for a snack or not and if you would know whether you would run the day before or not.

MADDIE: 10:32

Wow. So I don't need a psychic, I just need math. Awesome. Now, I'll figure out exactly how to win at the next school track meet [laughter].

SUSAN: 10:41

Indeed. Yeah. So, I mean, math is much better than a psychic. You just need to use these new powers of prediction for good and not evil. And remember, don't expect the predictions to be exact. But yeah, now you're just like a lot of athletes who actually use a lot of data to try to do better in their sports.

MADDIE: 10:59

Cool. Yeah. I'm really getting into this data thing. And maybe I can use data to win something else too. Have I told you about the phone game I'm playing?

SUSAN: 11:07

No, I don't think you did. [music]

MADDIE: 11:08

Well, it's super annoying but also so much fun and maybe you can help me win it.

SUSAN: 11:13

Sure.

MADDIE: 11:14

There's all this driving around that we're doing and I have to go collect coins all over town, but it's hard to know where I should go.

SUSAN: 11:21

Well, I have some ideas about that. [music] Let's talk about that on our next episode and maybe we can build a treasure map.

MADDIE: 11:33

Thanks for listening to Data in the Sandbox. This miniseries was written by Susan Currie Sivek and our theme music is by Andy Atley. If you know a K through 12 educator or student or are one yourself, we're excited to offer a new learning and certification program designed for kids and young adults. To sign up or learn more, visit alteryx.com/forgood. That's alteryx.com/F-O-R-G-O-O-D. Catch you next time. [music]

This episode of Alter Everything was produced by Maddie Johannsen (@MaddieJ) and @TreyW.
Special thanks to @SusanCS for writing this episode, @andyuttley for the theme music track, and @jeho for our album artwork.

Comments
12 - Quasar

Awesome! Love this!

Alteryx Community Team
Alteryx Community Team

@the_jake_tool awesome! Glad to hear it!