S2 E6: Byte: Data side-hustle

0:00Hello and welcome to Datum.

0:02This is season two episode six and today is a bite and we're talking about the data side hustle.

0:07Rabbi, how are you?

0:09I'm not too bad.

0:09I'm fresh off watching Game of Thrones, so I am medium.

0:15That is my reaction.

0:17Medium, okay.

0:18That is my pop culture reaction for today.

0:21It's not been a uh a great season, I hear.

0:23I mean it's it's basically been like this the soundtrack and the cinematography and the if you're a film buff and you quite enjoy like ah look the

0:30the sort of angles and all that stuff.

0:32It's been great, like the soundtrack, the score, the sort of effects and like the where the camera's been and the some of the shots have been incredible.

0:39Um

0:40But yeah, the storyline just seems rushed.

0:42I think everyone's just I think the start of the season I've been some people was like honestly I just want it to finish and be like, ah, there we go, done.

0:48And the and there was a Starbucks cup and a water bottle.

0:51Of course.

0:52Apparently.

0:52Yeah.

0:53Two two two uh very unenvironmentally friendly references in the Middle Ages or wherever Game of Thrones is set.

1:01Yeah, right, exactly.

1:03Good stuff, good stuff.

1:04So yeah, no, we're back we're back on the grind after uh what I'd say is uh sort of a short break.

1:09Um we are hot off the

1:11hot off the heels.

1:11I don't think we're hot off the heels, but yeah, we're following up from the last episode on behavioral design, which we sort of quietly released along with our new branding and our new style.

1:20Yeah, exactly.

1:21I think the switch over to the new brand and then we just didn't really push this episode so hard.

1:26So th those of you that did listen, thank you.

1:28Um it was we didn't get much feedback on it.

1:30I think that that was in in lieu of the the quieter release of this one.

1:35But I think the content we covered in that was good fun.

1:37Like we had a few like um tangent rants and things like that, but it was good overall.

1:43Overall, I think we cut covered a lot of ground on that in that episode.

1:45I think that's definitely one we definitely mentioned to come back to in about six months' time and sort of see what's changed, especially with our opinions.

1:52Exactly.

1:52I think we did it as a bit as well because I think it needed that sort of precursor um episode to sort of take you through uh the the technicalities of it.

2:01And then when we come back to it as a byte, it'll be a very different topic

2:04where we'll be talking about the thing.

2:06So think of it as part one.

2:08It's definitely not something that we're done with.

2:10We'll come back to it.

2:11Completely, completely.

2:12Today we're talking about the data side hustle.

2:15What's a data side hustle, Raven?

2:17I think so this is this is your term, right?

2:19This is your term for the um the hobbyism and the things we do on the side to develop either our own skills or just a sort of a pet passion project to follow along with, right?

2:28So um

2:29For for you, I think that's quantified self, right?

2:31Like that's what you're doing.

2:32It is indeed it is indeed, yeah.

2:34I've had a very quiet um few actually I'm gonna say years with that.

2:38I haven't I haven't put much out into sort of the open open sphere component.

2:42to I think three or five uh three or four years ago when I was as pretty hard on it um in in the public sense.

2:48Um and and yours is football, right?

2:50Correct, yes, football analytics is sort was always my route into into the world of data and tableau and w who I am today in a weird way, right?

2:57Because it's um

2:58that this addiction I had to football manager or championship manager of the game, um, which then translates into is like a moment of, oh god, I can do this for a living.

3:06And you there is

3:07a faction of the sporting industry that can benefit from data.

3:11So that that's definitely my my passion project.

3:13And it's something I don't do as much as I used to.

3:15I used to

3:15blog semi actively on uh on a personal sidebog of mine.

3:19Um a lot of my Twitter followers are from those days.

3:21Um and

3:23It's sort of t tied down a bit, but then I I still produce some visualizations um, in my opinion, a bit better than I used to.

3:30Um to see when you scroll down to my the bottom of my Tableau Public Profile.

3:34Well yeah, no, no, it's good to see sort of your progression uh as you go on.

3:38We'll actually come on to this specifically later on.

3:41Um I have to say uh you mentioned something about Football Manager.

3:44I don't know many people who play that game who aren't

3:46Addicted.

3:49For reference, football managers, it's like a football simulation or for the Americans a soccer simulation game where you don't actually play uh the game like in FIFA

3:58But uh you just simulate the progression through a season, right?

4:01Yeah, yeah.

4:02It it's it's just it's just got more and more layers to it as you go on.

4:05Like you can play it as a really simple player, like you just assign the best players.

4:09set them up in a formation, then let them play and win all the games.

4:11But then you've got like you set your own s challenges like I want to build build a dynasty or build a team that has the average age of twenty-five and you basically you're doing all the facet things with within a team.

4:22You're doing the

4:23sort of m financial management, the training, uh you're then winning cup well, it's it's just good fun.

4:29You you're getting sacked and fired and all this stuff.

4:31I think even

4:32Uh they've even got Brexit in the game, right?

4:34So Brexit day comes in and then you can't sign foreign players.

4:37It sounds like a hide side hustle in itself.

4:40Completely, no, completely.

4:41There's so many people that have um

4:43Like they've got it on their CVs, like taking forest green rovers from the National League to the Champions League in seven years, for example.

4:50But that's a lot of time investment.

4:52Okay, okay, and and it's funny just coming back to this sort of idea of a side hustle.

4:56I think we both we we both have these side hustles, um A because I think we're passionate

5:00about them.

5:01But also I think it's because we have a sort of an inherent interest in this area.

5:05So um you you've sort of talked a bit about champions um championship manager.

5:09Is it championship manager?

5:09Have I got a football manager now.

5:11Football manager now okay

5:12Great.

5:12You've talked about football manager and uh for me quantified self um is you know really started when we we had things like fitbits

5:21I I I kind of when Fitbits came out, I envisioned this world where you'd walk around with a sort of health device on your wrist and I was always fascinated about the data that you

5:29could sort of get from these devices.

5:30And it's taken a while to get to where we are today with Apple Watchers that can, you know, do ECG readings off your wrist.

5:36Um but um what's been fascinating is that I discovered this world of people who are doing way more interesting analysis than just

5:43step counting, using data, some of it probably mostly manually collected, but it's been a really interesting kind of uh subject because it's involved a lot of technical skills.

5:53and soft skills.

5:54So that's that's sort of what draw me drew me into quantified stuff.

5:57I was I was getting familiar with data.

5:59I found this hobby that was data oriented.

6:02Uh and at the same time I was seeing sort of different soft sides to data.

6:06I was seeing the

6:07kind of um you know psychological effects of monitoring yourself and improving the quality of your life.

6:12That's that's really interesting because that in in terms of being drawn into football analytics it's it's it was always moneyball as the concept, but then the second you started figuring out

6:20um and looking at the data that actually was collected from football matches by companies like Opta and all these other data providers, um you realise that the the granularity was so important and then you sort of start thinking about like, hang on, can I spot players that

6:34people aren't knowing about do I do what can I start looking at team styles?

6:39Can I s see patterns that you might not be able to see because you're watching one match and um

6:43There's always the anecdotes about um scouts going to watch players.

6:47Right.

6:47You're going to watch a single player and during a 90-minute match you'll basically focus on that one player and what they're doing.

6:53Whereas data then allows you to watch a hundred, a thousand players at one time and just say set KPIs to saying if they're hitting this consistently, then you know we we look up and listen and start looking at the video behind that.

7:04Exactly, exactly.

7:05And but there's something to be said about uh sort of the aggregate nature of that, right?

7:09So you you find people who are on paper technically good, but actually when

7:13When you go see them, there's a softer side to the finesse of the way they do it that as a scout I think gives you an edge over sort of the big the big data approach, right?

7:21It's not just that they're making those tackles.

7:23and uh you know scoring those goals, it's the way they're doing it that maybe leads you to believe that they can do it at a higher standard in in a more sort of competitive play.

7:31So so you mentioned your sort of starting with Fitbit um and and going from there.

7:37Like I I remember we had a conversation recently where you were like, ah man's stepcun, that's so bushly, that's old school stuff.

7:42No one looks at stepcun anymore.

7:43Have you seen that development like how you understand and look at wearables or Yeah, so it's interesting because wearables themselves are trying to do more as well.

7:51So there's this uh big push at the moment to do with uh

7:55uh well being and healthy healthy mind and yeah mindfulness.

7:59Exactly.

7:59And the only thing that these devices can really do is remind you to do them and sort of nudge you to kind of do them more often.

8:06Um but in in in in real terms uh the technology hasn't really developed uh to a really sort of advanced state.

8:13I think if you take uh an Apple Watcher of Fitbit

8:16At their heart, they're just doing what they did, you know, three years ago.

8:19They're doing it at better, uh doing it a lot better at more increasing levels of granularity, if that makes sense.

8:25sense.

8:25And the other thing is accuracy as well.

8:27So whereas a step counter three years ago was a guesstimate, uh today's step counter a little bit more nuanced.

8:34They can detect things like swimming, running, cycling and so on and so forth.

8:37So they can more accurately

8:39detect activities that you're up to.

8:41But um the classic example I always give is that let's say you took your Apple Watch off and you you left it in a drawer for a day

8:47And then the next day you put it back on.

8:49Uh, it would probably nudge you and say, Hey, yesterday was really unproductive day.

8:53Yeah.

8:53Do you want to do stuff today?

8:54It has no awareness of the fact that you weren't wearing it.

8:57Um, the prompts they give you kind of punish you.

8:59Like the classic one is

9:00You know, ten o'clock you get a notification from the Apple Watch saying, stand up.

9:04Uh or go for a short, brisk walk.

9:08You know, this this kind of like, hey, you're not doing what I've been programmed to tell you to do.

9:12uh do you want to do it right now and there's no there's no context to say hey you know what I'm sitting on the sofa um maybe now isn't the right time but the next time I get up maybe that's when you should notify me uh and tell me to go on a brisk walk or do something because you know that that's

9:30I I always find it funny when you sat in a room with Apple Watch users and it's like

9:3310 to the next hour and everyone sort of looks at their watch.

9:37Yeah.

9:38Exactly.

9:39Exactly.

9:39It's like everyone's part of the same sort of machine and

9:42Yeah.

9:43Um there's not really much sort of um customization, but uh uh we sort of digress.

9:48I think going back to sort of you know the value I found from it is that

9:52kind of tackling these soft questions and understanding how the data doesn't necessarily tell the full picture.

9:58Now that's something I've really found valuable in my professional life.

10:01life and it's something that I learned purely through QS.

10:04I learnt it quite simply because what I did is um uh one of my sort of uh things I do all the time is track my music

10:11So since two thousand and nine I'm coming up to a decade actually.

10:13Last FM?

10:14Exactly, exactly.

10:15In July I'll have a decades worth of music listening history

10:19available to me through a service called LastFM.

10:22And about three years ago I did a very basic viz.

10:25I joined up all the dots.

10:26This was back before Tableau was really kind of, you know

10:31And I used all tricks to bring the day together.

10:33I could probably use tablet prep for that now.

10:35And what was interesting when I hope so.

10:38I hope

10:39I don't think you connect to the API through that.

10:41I I can't connect to the API, but I could use it to sort of connect uh the datasets together, which is predominantly what I'm doing.

10:48So I'm downloading lots of different data sets.

10:50different services in Altrex.

10:52And then once I have them, I have what's essentially like a a music metadata library.

10:57So each track has its own metadata.

10:58And I'm basically just

11:00Just joining that data onto my music listening trends over time and then visualizing it in Tableau.

11:10I think this just goes back to when I had my Fitbit.

11:14So my my my my main two motivators for buying a Fitbit was first of all I wanted to track my heart rate and step count and third of all I wanted to do sleep.

11:22Now

11:23When I had a Fitbit, I I I have an Apple Watch now, but when I had my Fitbit, it was like notoriously difficult to get your super granular second by second data out.

11:32Never gonna give a thing.

11:34No, exactly.

11:34And it's like

11:35I'm buying this wearable.

11:37I'm tracking my data.

11:38I want to know at the lowest level of granularity what I can get.

11:42But I can't even see like how many steps I did in a minute.

11:45I can't go down to granular and bring it back up.

11:47Um the only thing I can do is look at the the Fitbit dashboards.

11:51And I think this is where um going back to your earlier point about the data collection and the the the development of uh wearables, one of the big things I've seen personally is they're slowly getting better at visualizing this data.

12:02Right.

12:02I think um Apple do this quite well.

12:05I think we've I've seen a few people um who have criticized the use of rings, but I think the rings work well because you're not looking at that granular number, you're looking at whether you're hitting your target and

12:16Completion to goal is like, am I at that completion to goal?

12:19Am I at the 100%?

12:20Yeah.

12:20And you can see that with the ring, that's fine.

12:21But when you look at the actual um visas behind it,

12:26um you can actually find out whether or not um it's able to you know pick up those visualizations.

12:33So yeah if you look at the actual app itself, the the chart, the bar charts and whatever the distance to go on giving that overview of the last seven days.

12:40etc.

12:41They're actually pretty good, I think.

12:42Um and that's getting better as well alongside the the granularity of data.

12:46I think you've actually pulled some of the Apple data out using Ultrax, right?

12:50I have, I have.

12:51So this is this comes down to uh philosophy, right?

12:53So Fitbit uh they don't really have much.

12:55They don't have an ecosystem to lock you into.

12:57So that's one of the reasons they're not prepared to just easily give you the intraday data.

13:01Intraday data is a minute by minute blur of what's happened.

13:05They actually do collect it at that level, but what they do is they aggregate it up to the minute because if you think about it from a second point of view, the databases would be humongous.

13:13Humongous.

13:14Yeah, absolutely.

13:15So they they aggregate it to the minute and they store it away

13:18And in the apps you see kind of readouts based on the minute-by-minute blow, but in reality when you export it, I think the easiest form to get is a day-by-day blow.

13:26Now what's interesting about the Apple ecosystem.

13:29Is that because it uses a privacy-first model, I think it's actually one of the best systems because uh essentially what happens is your device is the database.

13:39And when I exported data from Apple Health, um

13:42The it it collects data at the most granular level that the device that's tracking that information collects it at.

13:49So if I think about my phone, my phone will count things like steps

13:54um as granular as every second.

13:56So it will say, okay, for this second, Tim did this many steps.

13:59Okay.

14:00I use a heart rate monitor that's capable of recording multiple readings per second.

14:05Again, the heart rate monitor will record to Apple Health at multiple

14:08readings per second.

14:09But something else might only do it per minute.

14:11So there's no sort of specified level of aggregation in Apple Health.

14:15Because of that, my Apple Health library

14:17is about three and a half gigabytes when I export uh ex explore export it and it comes out as a JSON readout.

14:24So um it's it's a good thing that this is synced to uh to iCloud because it means across devices you can carry kind of

14:30carry on.

14:31But it's an enormous wealth of information and most of the stuff on there is heart rate readings because my Apple Watch has been tracking that and then I also do a lot of sports and my chest track has been

14:39tracking that.

14:40So this is something that's getting better and better over time going back to my earlier point about accuracy and fidelity.

14:46Depending on the platform you choose, they have various reasons to give you as much granular information as

14:50as as they can uh and at least in Apple sense and also to an extent Google as well, they're actually pretty good at giving you access to the raw data because what they're providing is a platform that other apps can hook into.

15:02And in order to do that, you need to

15:03kind of be as open as you possibly can.

15:05Now the downside is that um you know this stuff is on your device.

15:09It's actually incredibly hard to get it out of your device unless you know what you're doing.

15:13And like any sort of

15:15Yeah, I've always criticized exports from services.

15:17They're never that useful because they come in formats that the the the the everyday person is.

15:21Just a runs format, yeah, yeah, has no clue what to do with like

15:24XML, JSON, and CSV, you might be out of push, be able to do that, but then you get ridiculously sized CSVs like you know a gig, a gig and a half that you can't pass in Excel or something.

15:35Nice.

15:35So this is a really nice segue.

15:36I think this is a nice segue back into football analytics.

15:39So one of the reasons I picked up on this as a side hustle is um so providers like Opta and uh more recently got Prozone and Statsbom, well Prozone is now Stats as a company.

15:49So a lot of these companies do collect all this data.

15:51So for example, if you look at Trackab, which is tracking data of a player, which is using uh multiple cameras on a stadium, and they're looking at the number on the back of a player's shirt and mapping it back to the database.

16:01That's taking data at every point.

16:04um uh zero two five of a second.

16:07So it's like every quarter of a second it's taking.

16:10So you've got four data points per second.

16:12And it's funny when you when you and that comes in a format which is dot dat and an XML and then you get opted data, the F twenty-four or um which is the event level data, they've got a new one which is event plus a few other metrics.

16:25Um that that comes in XML format.

16:27So if you think about the people that work in analysis at clubs, until more recently, it's generally been

16:34Uh sports scientists, right?

16:36So people that have done physiology, they've done sports science, but they've never really been people that use computers.

16:41Like yes, they've done coding because as part of sports science, they do some coding on matches, similar to how Opti do it, where they tag events

16:47and clip videos together to say, okay, here's where we had a good shot, a f a fast attack, here are the key events in the game.

16:54So when you're playing back a 90 minute match to s to your teammates or the players and the or the staff

17:00they would just quickly see what's going on.

17:02Now, when I started working um with the public data, it sort of again it was aggregated because you know Optera was selling this as a business.

17:10Like they it's it's a couple of thousand multiple thousands of pounds uh to buy the the access to the data either by API or exports, whatever you want, uh they're able to service that

17:20And one of the most interesting things we found was like, well, these people who work in clubs, yes, they have the idea, they have the the grounding in the game, they understand the technical aspects, but they're not really data people.

17:31Right.

17:31Um I've been going to the OptoPro Forum now, I think this is my third or fourth one this year.

17:36And it's it's fascinating to see like the amount of statistical analysis that goes on.

17:40people using R and Python and they're combining all these events together and finding patterns and telling stories with with the data sets.

17:47But what they're fundamentally no one ever talks about is the back end.

17:50Yeah.

17:50How are you passing the data?

17:52How are you bringing all these things together?

17:53What is the the function?

17:54What's the best way to show this back to a coach or a player to get them to get that insight?

17:58And this is the sort of the ground that I find myself in, like increasingly thinking like

18:04Building that platform, the clubs that I've worked with or I know of, there's only a few doing it in the right way where they've got a database, they've got a a data uh processing platform that they're doing in-house.

18:15not a second secondary company.

18:17And then they're also doing some bespoke things on the front end.

18:20That's that the the club's doing that few and far between.

18:23But I think that that is changing given that you know um that the game is changing

18:27Right.

18:27Exactly, exactly, exactly.

18:29And just going back now to a high level, it's funny because um our side hustles have actually given us this sort of unique ability to take

18:37uh this data and it's it's allowed us to sort of dig into this discussion.

18:42So from your perspective, I think that's one of the things you bring to the Opta forum when you

18:46you when you go there because you actually critique the back end systems, you know, because a lot of s a lot of people buy the data but don't necessarily you know they have the data and they're in Excel that

18:55You know, they're kind of working it hard and working it the hard way, sorry.

18:58And um in contrast, you know, by doing what you've done, you've actually been able to go and critique the

19:03this this this process.

19:05Exactly.

19:05And and and that's exactly my my my overarching point.

19:08It's like this the the sort of scrambling way it was just sort of an interest and now I can sort of I I feel I can speak a bit more intelligently, not just about the data and how to pass it and the the issues that there are is in it.

19:18But also saying, well, it what happens if you go beyond counting stats?

19:21One before you go beyond saying we took seven shots in this match.

19:24Like, when did you take them?

19:25What was the game stake?

19:26Were you two goals up or one goal down?

19:28Like how does

19:29How is the team impacted by those things?

19:32And then with the tracking data, it's fascinating because this is such granular data.

19:36And if you think about football as a sport or soccer as a sport, it's a game of spaces

19:40And when you start being able to quantify those spaces and say and detect patterns and spot things that players do or teams do or you know and then tag that to events, that's when you start getting really insightful stuff.

19:52And that's what's exciting for me at least.

19:54When you start bringing those things together and getting that insight in, you can start having those really interesting conversations with coaches on a level that they understand and they can see.

20:02Um beyond just like when people generally get out tracking data, all they do is plot the X and the Y's and watch the games but with scatter plots basically.

20:11Um which which defeats the object in some way.

20:15Right, right, exactly.

20:17And that I think um that's one of the things I think side hustles do is they because you're more passionate about the subject, let's take um uh football analytics.

20:25You're naturally football about naturally football about

20:28about analytics.

20:29You're naturally passionate about uh uh football analytics here.

20:33And so um that passion actually drives you to a deeper level of understanding within that area.

20:39Maybe more so than let's say uh, you know, some of the sort of the drier stuff we do on our day-to-day work, you know, you know, whether with with clients or or with our own businesses, right?

20:49Because I think your own business can kind of suffer from this um well several issues.

20:53Uh first of all, access.

20:54You don't get as much access to data internally sometimes because

20:57because of uh governance and bureaucracy.

20:59And so when you have ideas, it's really hard to run with them because you have to make sure you've got everything in place, right?

21:04Yeah.

21:04Uh you just want to prove the value of something.

21:07By experimenting and innovating, but that you do you don't necessarily get that in business.

21:11Now the other thing is that it it naturally develops your skills to be able to communicate the challenges you're having.

21:17I think one of the first things I learned when I was working with quantified self-data

21:20was that bringing these things together wasn't a simple thing.

21:23Like I had to really understand that if I was going to make my music data and my moves location data work well together, I had to understand at a very granular level, uh, you know

21:34what I was doing in each minute.

21:36So I built out this minute level data set and then I was able to say okay in this minute I was generally here and this is what I was listening to.

21:43And then I was able to sort of map those two onto each other

21:45and and then go from there.

21:46And it's got you.

21:47It's a really, really important thing to start to understand.

21:50And having that passion allows you to ask deeper and deeper.

21:53Deeper and deeper questions to the point where there y you realize that actually there is no end to what you can ask.

21:59There is no end to how you can interrogate something.

22:02And then when you apply that back to your everyday, you suddenly realize there's so much more you can do with your everyday day

22:07Agreed, agreed.

22:08And I think you mentioned this earlier when you were looking at your quantified self-tech through Alterix.

22:12Yeah.

22:13It sort of exposed you to the download tool and things like this.

22:17And I think that definitely harks back to something I've done.

22:19So when I started using um

22:21Ulterics, I learned so much about things like the detour tool, like control parameters, like having the basically the painful things when you're trying to combine and pivot and pass XML from a really ugly thing.

22:34And try and create this repeatable process which is a macro that does it what it works.

22:38I think one of the issues this was like one of my train projects.

22:41So I've a

22:42or hourish commute.

22:43Um I'd spend an hour in the morning just like tinkering with this sort of thing and getting getting it as far as possible.

22:48And basically ended up with this working prototype until I got to a match where it was nil-nil

22:54So basically I I I I had one of my passers saying if there's a goal, pass it out.

22:58But then if it was a nil-nil, there were no goals and there were no assists.

23:01So basically it was like, well I can't do I can't pass this match.

23:04So I had to create a de Exactly.

23:06So I had to do a detour tool to be like

23:07Is there a goal?

23:08Is there an assist?

23:09And then if not, do a detour to go like this way or this way.

23:12And it's like so many of those small nuances about working with data, I was able to work out through a p p passion.

23:20Like if I was working with, I don't know

23:22um website visitor data and trying to figure out how to pass hours.

23:26I'll be I would have been asleep and I would have moved on really quickly.

23:28Like I would have basically got it working and left it.

23:31I wouldn't have tried to make this really perfect, efficient macro that does it automatically.

23:36Um I wouldn't have thought about it in such a way where I'm like, I really want to get this done because of X, Y, and Z.

23:41Um and then finally, um, I think your your point on wanting to do better because it's a pet passion and trying to drive

23:48the other people around you further?

23:50That's definitely something I've done.

23:51Like I I remember I think I mentioned my um crusade on s long for long sc and wise scatterplot

23:56Right.

23:57That would never have happened if I wasn't like getting angry at people posting wide scatterplots on Twitter.

24:03Right.

24:03Like I would never have written that and I wouldn't have thought I need to educate these people about why it's wrong because three years ago I probably would have done the same.

24:10But now I'm a bit more educated and a bit more um formalizing data biz practice and the reasons behind doing things, I feel I'm in a place to educate that back to people.

24:18And that's the thing and

24:21It's funny because that in itself is is is another skill that you pick out of it.

24:25So um you you learn actually you kind of become humbled by the experience.

24:29Right.

24:30You correct you go you go in, I think, you go in with this sense of confidence when you're passionate about something.

24:34I know I know everything about football.

24:36I know everything about quantified stuff.

24:38I've got a fit, you know

24:39Yeah.

24:40And as you go into it, you slowly start to realize, oh wait, I can't just export this data.

24:44And you go, uh huh.

24:46Then you you someone tells you, oh there's an API, and you go, okay, API, I wasn't expecting

24:50to have to do this this way and you start working with it and it humbles you it kind of shows you it kind of teaches you to give these challenges the right level of respect and that in itself is portrayed by the way you then communicate it to people

25:03You treat the topic with respect and therefore you you you're more encompassing about bringing people into your sort of your your skill, your your fold and kind of making them understand the challenges that you go through.

25:13Completely.

25:14And I think that's that's definitely fueled my my recent like um change in tone has been about a lot about like actually it does matter a lot about who the player is.

25:23like what they're doing and all this stuff.

25:25Like you can't just purely base something on, you know, statistics or numbers.

25:29You know, you can't just say, um, one of my favorite players, um, Cole Skews.

25:34is the most pivotal player in the team.

25:36Like you can see that.

25:37You can see that without him and the team in Ipswich, he's a lot the team are a lot worse off.

25:41But also you've got to take into account his fitness data, which we don't have access to do.

25:45You don't see him in training.

25:46You don't see his attitude.

25:47You don't see how he changes and the things he brings to the team.

25:50Like all of these different things you can't really quantify.

25:53It's a lot harder to quantify, and that's what a coach brings to the table.

25:56Again, going back to our day-to-day

25:58We go into companies and we try and solve their problems with limited to no knowledge of the field.

26:04We don't have that depth of knowledge.

26:06at a company and I think those pet passions, these data side hustles, really help you to realise like you're not an expert in those things and you know

26:15It's you're really what you're doing with these things is continuing your learning based on something you're passionate about and seeing what else you can pick up along the way.

26:23Exactly, exactly.

26:24And I I think that's that's one thing.

26:26Um I remember an experience where I I I presented to the London uh Quantified Stuff London Group.

26:31The first time I presented, uh I really didn't know what to expect.

26:34Um so I just I just

26:36you know, did a classic talk about um music and I talked a lot about altrix.

26:42I I kind of felt like I was selling Tableau and Ultics at the time 'cause I was

26:45was uh a year into my journey so I was trying to kind of showcase that a little bit.

26:49Um and I you know the first few questions I got like the first question just just knocked me off the floor and someone just asked me like

26:55Like, so why are you doing this?

26:57And yeah, and I at the time I was like in my head, I was like, because it's cool, okay?

27:02Like Yeah, right.

27:03Exactly.

27:04Like, like I know what I've been listening to for the last eight years.

27:07Like, that's cool.

27:08And

27:08Uh in my head I could hear myself saying, no, that's not that's not the right response to him.

27:12Come on.

27:13Like why are we doing this?

27:14And and and and and what I realized is okay, so why do I think it's cool?

27:18So I drilled into that question in my head and I just started talking out loud here.

27:22I was like, okay.

27:22So why don't I think this is cool.

27:24Um then I explained like listen you you you have this idea of listening to music and you the the the really simple thing I always say you ask someone what do you listen to and they come back with two or three arts

27:35artists, okay?

27:36In reality, you listen to like a plethora of artists.

27:38You're not the sum total of two artists, two or three artists.

27:42And the exact opposite, someone goes, Oh, I'm really eclectic, right?

27:45And more often than not, when someone says that and I'm able to look at their data on something

27:49that last of them, they're actually not that eclectic.

27:52There's like a hidden secret between what they listen to.

27:54And I'm only able to see this through the attributes of the individual tracks that they listen to.

27:58to to actually arrive and say, no, no, you don't you're not eclectic at all.

28:02You happen to listen to all these songs.

28:03But here's the thing they have in common.

28:05They're all acoustic, they'll have the

28:07same level of energy and all these attributes are measured by other companies.

28:10But I'm actually able to sort of draw the common the commonalities between what they think is eclectic and what is actually the same.

28:16So what you've described as eclectic is actually one thing.

28:19It's not eclectic at all.

28:20And so just just having that discussion out in public was was the cool thing.

28:24And I'd realized, yes, okay, that is actually why I do it.

28:27You know, understanding uh underneath the skin what's going on

28:31Yeah, yep, yep, yep, yep.

28:33And I think the quantified selves is this really interesting like intersection of you sort of sit there and be like, yeah, but why am I doing this?

28:40Like it's funny, like everyone's like, yeah, but why are you collecting all this fitness data?

28:43About yourself, yeah.

28:44Exactly.

28:44And it's like, well

28:45In some ways it's actually it's changing my behaviors because you start looking at these things and you're like, Oh, I did less than last week or you start doing these like you compete with yourself and others.

28:54Um but also it's it's intr I think the what one of the key things you mentioned uh

28:59That's really interesting is like the patterns between people uh and the insights you can find with that.

29:05And that that that then goes back to like

29:08If you're able to do this within an organization, like you're saying, right, we're not going to look at X1Z, we're actually gonna we actually only care about this week what you're doing outside of work.

29:17Right.

29:22Your workplace and show me something interesting.

29:25Okay.

29:25Exactly.

29:26And it's almost almost like this five to ten percent time, right?

29:28Or five percent time, whatever you might call it, where you're able to give back to someone and because inherently you're giving someone time back

29:35to develop something which is relevant and you're saying use these tools because you're using the tools that you'd use every day anyway um it would just end up being like

29:45a benefit in kind.

29:46No, maybe not today or tomorrow, but in a a month's time you're like, oh, I I worked on that because I had to change this data set from this ugly thing to something else.

29:55And then I visualize it like this.

29:56Maybe I can use a similar visual to portray this finance data.

29:59When we're trying to compare seven different products.

30:02Right, right, right.

30:04And and then that's and that's that's one of the sort of um unique aspects about it is that

30:11I always call quantified self as digital narcissists because that's totally what you guys are.

30:24Yeah.

30:24But but it it it's funny because like why wouldn't you take an interest in who you are and what makes you you, right?

30:31And understand that because at a high level you have no you you

30:35You you think you know what's going on around you.

30:37Like you think you know what you did yesterday in incredible amounts of detail.

30:40But actually you're very transit uh humans are very transit beings.

30:43We focus about the now and the future.

30:45We we we although we like to think we reflect on the

30:47the past.

30:48We don't actually keep that much uh in the past apart from painful experiences.

30:52That's that's what that's what we that's what we carry forward.

30:54And so the good stuff just kind of blows on by.

30:57And so just having this understanding of how you work and the transition, I think is really interesting

31:02And I think in a business context, this is absolutely the same skill that you need to have, but you're applying it not to yourself but to your own business.

31:09And so it's the same sort of inquisitive nature that you'd

31:13If you apply it to your business and you create a culture where people can do the same thing, then people start asking really fascinating questions about the way you do things rather than just assuming that

31:22That's the way they're done.

31:23The questions will always get better, right?

31:31no matter how difficult or not so difficult it might be, um you just your questions inherently just get better and better, right?

31:38And I think that that that holds true with quantify yourself, even holds true with football analytics.

31:41The more you dig into this stuff, the more you find out

31:44I think one of the big big like topics these days is all about automation, right?

31:49Um and automation is a funny one because with quantified self you always talk about active and passive, right?

31:54Right, right.

31:54And I think for the last six months I've been using an app called Reporter

31:58And in the last like I'd say two or three weeks, so report is this app done by a guy called um created by uh Nicholas Feltron

32:05um who had these incredible, incredible reports um about himself, basically.

32:10He did a yearly report about his year based on data and infographics.

32:14And one of the tools he used to do this was a app called Reporter.

32:18Now I started off using this quite enthusiastically.

32:20Oh yeah, this is gonna be great.

32:22I know what you're about to say.

32:23Yeah.

32:23I'm on to now and then I'm just like whenever I get a ping, so it's it picks um six random times throughout a day, I'm just like, oh

32:29Fatigue, yeah.

32:30I'm just like, do I need to go through I'm now I'm now just like sporadically and I'm just gonna basically think about this in years' time and say, right, let's look at the frequency of reports and also

32:40Let's also look at the fact that the thing that I'm I'm now trying to get out of it is when am I reporting?

32:46Um and also why exactly when am I sending these reports myself?

32:51Um and and then comparing that to the passive data collection that I have, right?

32:55Like steps, my heart rate, um, all like I don't know, Twitter history, when am I tweeting, all of this stuff.

33:01Um and it could combining all that together I think is will be more interesting, but I think activity passive is a really interesting conversation as well.

33:07Yeah.

33:08Especially with quantified self.

33:09And again in business.

33:11I I I'll reiterate this again.

33:12It's you know the um the the the thing I the reason I always uh harp on about passive uh collection.

33:18So active and passive what Ravi's talking about

33:20about is essentially this philosophy in uh quantified self which uh I think uh I don't know I don't know if I came up with it, but I think I d I read it from somewhere.

33:28And the idea is that uh active is when you're physically collecting the data yourself.

33:32So you're getting a ping on your phone, you're filling out a form and then you

33:34you're storing the data somewhere.

33:36Passive is where you're not having to do that.

33:38Something is doing it for you in the background.

33:40So for me when I listen to music, I don't have to log that somewhere.

33:43There's a service called LastFM which is doing

33:45doing it for me.

33:47Exactly.

33:48I've set up another passive way of tracking when I'm at home or not.

33:51And the way I do that is I have a Wi-Fi router that's hooked into if

33:55this than that.

33:56And so every time my phone or my watch or my laptop connects to the network, it logs it.

34:00And so I have a pretty accurate reference for when I'm at home and when I'm not at home

34:04just based on when my devices connect to the network because I'm never at home without one of those three devices, right?

34:10So that's a very good active and passive way rather than me saying

34:15uh oh hello foursquare okay swarm okay let me check in or you can Amazon dash button to be like I'm home let me press this button let me press this button right exactly it's

34:25Like the thing about this Wi Fi collection is that yes it might overreport when I'm at home because you know I might leave my phone at home and go out to the shops very quickly

34:33But my watch will disconnect and connect.

34:35So I've got I've got such a good uh level of accuracy there in aggregate compared to me manually maybe forgetting one or two times

34:44And so the quality of the analysis I then get from that is much higher.

34:47Same with location tracking.

34:48I don't do that myself.

34:49I have an app that does it.

34:50And it doesn't have to tell me exactly where I am, but it just needs to capture the general trend.

34:55And when you have that quality of data, it's much, much richer

34:58Now the thing where you know passive can't work is for feelings and emotions because going back to where you you're talking about when you track, when you do some active forms of tracking, what t it tends to work better for um

35:12when you're trying to measure sort of your mental health or well being or things that are you know things that trigger and therefore you think let me monitor this, right?

35:20And over time when you become fatigued, what you end up doing is only recording them the more negative aspects.

35:25So when if you're sort of tracking your mood and you have to log it manually, what ends up happening is you only ever log it when you're unhappy because you want to make sure you've got that as a reference point.

35:33But you don't like stop in the middle of I don't know, roller coaster ride to say, let me log this while I'm going up and down this, you know.

35:40This tru uh this truff right.

35:41This feels great, but you're not pulling out your fern and and logging it uh straight away.

35:45In fact you just you're just enjoying yourself.

35:48So uh uh active passing tends to suffer from this sort of very negative skew where

35:53you capture more about the bad stuff than the good stuff because the good stuff well why why would you monitor so so I'm actually having a quick look at my reporter and um it it's honestly like uh so when when am I reporting I'm reporting when I'm watching TV, traveling texting or chatting

36:06Yeah, exactly.

36:07Mostly when I'm at home.

36:08Um what else am I like yeah.

36:12And my my s my cause there's also like you can turn it off when you sleep when you wake up, so you've got questions to ask when you wake up.

36:17about your quality of sleep and also when you go to sleep about your day in general and your feelings towards it.

36:22And again, that's so sporadic.

36:24I have so few data points on that because it's like before you go to bed.

36:26Am I gonna press that button?

36:28Probably not because I'm doing other stuff.

36:30Um like I'm busy or I'm just like flat out and I've just gone to sleep.

36:34Exactly.

36:34Um

36:36But yeah, I I think your your point on um when you do it, I think that that's really interesting because um it's definitely something people use in mental health like

36:44to to track behaviors and how you feel.

36:46Um just so you know psychiatrists and psychologists have that information to go back to saying like, okay, cool.

36:52So

36:52When I'm around Tim, I'm always feeling really depressed, right?

36:56For example.

36:57Nice.

36:58And that that that that that sort of feeling and tracking just to help people be more aware of their emotions.

37:02Um, it does help.

37:03It does help.

37:04Yeah, exactly.

37:05And so if I take this back to sort of a business context, whereas this has really helped me is that, you know, whenever I'm talking to clients about the way they collect their data

37:12and how they bring it in.

37:13Yeah.

37:14I'm always leaning towards the more passive systems because I inherently know that for the things that they're trying to track, apart from the qualitative stuff

37:23Like this is actually the more valuable side of things.

37:25So look if you can just spend a little bit more time getting out of your form culture and into your sort of um, you know, uh sort of automatic data collection.

37:34and just have have systems that collect data on behalf of people.

37:37So A, they're more likely to input when they need to, because they know that that quality of data isn't the

37:42to be there inherently in a good way.

37:44And you're also taking a cognitive load off the user for having to remember to do a whole bunch of other stuff.

37:48So I find it easy to track my mood because I don't have to track all these other things.

37:53I have I think about 35 different things that I track passively.

37:57And those are 35 things I don't have to manually track.

37:59And that makes it so much easier to track one simple thing, like say my diet when I'm eating or something, because there's still no good way of tracking food automatically.

38:07Um I think the the best way I found is um MyFitnessPal uh barcode scanning.

38:12So it has a really good barcode database.

38:14So you can scan something you're eating and it will grab the calorie data off its own database and

38:18input it to health for you, which is great.

38:21But I still have to do the manual task of opening my camera apps to try to scan in the barcode.

38:26And I'm hoping this is one of those features that Apple Sherlock can then just build it into the camera app so that you can just take a picture of a barcode and it'll tell you all the nutritional information about the food without you having to open another app

38:38Genuinely, genuinely, that would be a life-changing uh feature for so many people, honestly.

38:43Um but anyway, that's I digress.

38:45Um these principles carry through to the business and it's

38:50it's so it's so valuable when people do that because it also brings a different um a different what is it um what's the word ethos um uh

39:02Uh uh no subject areas, different subject areas, people who think differently, who are creative in a different way, end up interacting with your data in the

39:11These awesome ways and they borrow concepts.

39:13I borrow concepts from quantified stuff all the time.

39:15And they've got nothing to do with most of the client stuff I I I work with, but the chart types, the ways of manipulating data, dealing with sort of

39:23b obscure things and like you say, coming across edge cases and realizing that when you build s systems you have to test test these edge cases thoroughly, right?

39:32You can't you can't just build something, A it looks great and then as soon as you have a nil nil game rather realizes how

39:38Yeah, I mean a perfect a perfect example was my World Cup dashboard, right?

39:41That was a lot of manual data collection.

39:42But like w w once we got to the second round and

39:46My um my thing where players come on and off, the the the the team's sort of structure, I didn't factor into, you know, extra time.

39:53So it's like taking left

39:54two of the numbers like oh yeah yeah 90th minute that's all they're gonna be and then suddenly you've got a player coming off in the 12th minute it's like hang on a second there definitely wasn't interesting but it was because it was 120th minute oh man just rookie errors all around yeah yeah

40:09Yeah.

40:09But yeah, it's funny it's funny how that that um that all all kind of plays out.

40:14And I think it's a really valuable thing.

40:16I guess I guess the now um the so what is that

40:19Um we've touched a bit already, but actually enabling people to have these um things and hackathons is such a good way of doing that, right?

40:26Yeah, agreed.

40:27Like running a hackathon internally, even if it's within your own data.

40:29So we've I've sort of seen this in two different places.

40:32So you can either have a hackathon and internal data where you're saying you're saying, right, okay, the incentive is you get to present this back to the C suite or senior management and show your ideas and what you can do.

40:42Or also seeing this as like, here's a theme, here's some data

40:45Tell us something interesting, use these tools.

40:47Right.

40:47And that that both work in their own sort of way.

40:50Um the first one probably takes a bit more like coaching on the tools.

40:54Um, the second one is a bit more fun uh and perhaps gets a bit more engagement.

40:58But at the end of the day, it's about giving employees time and sort of saying this is this is interesting.

41:04This matters to us.

41:05We really do want to have our employees work on this sort of thing.

41:09Exactly, exactly.

41:10And no and also I think um in terms of uh how else do you support these uh people to do these things, um like it's it sort of goes without saying we have access to incredible

41:21software that does all this stuff and part of that is being able to actually you know take my uh copy of tableau or altrix and be able to do what professionals do with the data

41:32But to do it with with this hobby.

41:34And it's it's one it's really important thing that to to recognize that you know people like Tableau, they actually enable you to do that through things like Tableau Public.

41:41So you know take advantage of free software

41:44software and training courses and let people go on that one-day training course that might not sound like it's got anything relevant to do with your everyday business uh work but actually the skills they'll learn they'll pick up or give them a passion in data

41:57And also data literacy, like you're you're fundamentally talking about building people's literacy levels, everything we're talking about, whether it's reading or writing as we talked about before

42:05Right.

42:06And so uh just investing in people, literacy alone in this sense really, really helps give back to your business as well.

42:14Exactly.

42:16Cool.

42:17That's a good episode.

42:19Yeah, it's been uh it's been a good episode.

42:20I think we we we we b we we indulged a bit, we came a bit off track, we went into our passion.

42:25But it's all worth the while, hopefully, just to try and highlight how valuable we think it is to our own development also, how it can be to others.

42:34We'll be back in a couple of weeks with another

42:36another episode.

42:37Please keep the feedback coming in.

42:40As a reminder, we've changed our name and branding.

42:43So we're at Dayton Podcast.

42:45com and we're on Dayton Pod on Twitter.

42:47So much, much easier to find

42:49No what not what nots anymore.

42:54I keep digging that because I'm just literally I'm so glad every time I have to say this now.

42:58I'm just like, ah, it's so much better.

43:03Agreed.

43:04Agreed.

43:04It's so much easier.

43:05It rolls off the tongue a lot more.

43:06Yeah, exactly.

43:08And and as ever, we'd rarely ask for the support in this way, but if you can uh give us a rating on whatever podcast that form

43:15uh you listen to that would help because apparently that's becoming the currency for uh the podcast uh walls that are going to be coming up in the future.

43:22So platforms are gonna start uh

43:25kind of siphoning off podcasts to lots of different places.

43:27So yeah, if you can rate us that would really help us as well.

43:30So we we we know what we're doing and we can improve and make sure we're getting to the right places.

43:35Cool.

43:35Fantastic.

43:36Cheers everyone.

43:37Have you got a good one?

43:37No worries.

43:38Take it easy.