Transcript Synced · select any line to jump ▾ 0:00 Thea Ngo: Go for a run and your watch gives you 20 stats back. Play a full game of basketball and you get nothing. 0:06 Cordelia King: People were saying that they were better than they were, and sometimes clubs are getting like catfished. We give them back all the data, all the stats, and all the highlights for every player on a court. 0:15 Thea Ngo: Cordelia's first company put local footy players in front of clubs, and the clubs kept asking to recruit on stats that didn't exist. So they built the company that made the stats a reality. They tried it on footy first, an oval the size of a farm 18 players on either side, the model just couldn't see. 0:35 Cordelia King: They liked the idea, but the market just wasn't big enough. And then Sam was like, why don't we try doing this for basketball? Like, the court is the same size every time. It's 5 players, you can see everyone, the stats are really easy. 0:45 Thea Ngo: 2,000 games later made for parents and coaches because a kid's shot at a life-changing scholarship still came down to if a scout was at the gym that night. Kai sold his car, Sam dropped out of uni, Cordelia moved back home. 1:01 Cordelia King: Mom wasn't happy. 1:02 Thea Ngo: And her only regret was not doing everything much sooner. 1:06 Cordelia King: My name is Cordelia King. I'm a co-founder of Superstat, and this is Founders in Motion. 1:18 Thea Ngo: Before Superstat, you and your co-founders built TrainStop, the recruiting app The Herald Sun called Tinder for local sports. It was in hundreds of clubs and also where the spark for Superstat came from. So maybe take me back to your sports tech journey all the way at the beginning, at the beginning of TrainStop. What made 3 friends decide to build for local sports? 1:39 Cordelia King: Yeah, good question. I think it started with our co-founder, Kai. The platform TrainStop was connecting local sporting clubs with local players, and that recruitment process was very, very manual. It was people calling up people they knew who used to play, people asking their friends if they could come and play for this club. Players were getting paid to play, so there was definitely money in the system, and these clubs had money to be able to put towards a recruiting budget, but there was just no platform for them to be able to do that. And so Kai and Sam actually created the first app to connect players and clubs. And then I came on a little bit later and we had some press yet again calling us Tinder for local sport, which we were excited about the press. 2:21 Cordelia King: And then when we heard the title, we were like, oh, is that what we want to be called? And it was because players didn't have enough data on their profiles to be recruited accurately. So people were saying that they were better than they were, and sometimes clubs are getting like catfished, so to speak. 2:37 Thea Ngo: So very much like Tinder. 2:39 Cordelia King: Yeah, exactly. So we asked these players to add data into their profiles. And the players came back to us and said, like, we don't get, we don't get data, we don't get stats at a local level of sport. So we can't do that. And we were like, okay, that sucks. How can we get data for local sport? And that became the next problem to solve. And a lot of local sport is filmed now. So that footage existed, and the data could be attainable. It was just how to get it in a way that was scalable. And that's when Sam built out our first computer vision model. 3:13 Thea Ngo: Cool. Exciting. Well, you've literally answered like 3 of my questions in like one. So obviously the media training has like kicked in. 3:21 Cordelia King: It's kicked in. 3:22 Thea Ngo: But I'm super curious. So maybe like with TrainStop, so you built the app, you built the first version. How did you get those first few clubs to start using to get like a mass number of people on the platform? 3:36 Cordelia King: Yeah, good question. We grew up playing sports. We had a lot of connections to local clubs and we knew them very well and the people that ran them. And so that first app that was built was built pretty much alongside those clubs. And it was like, okay, what would you look for in a player? What would you need to know? Things like age, height, weight, where you live, how far you're willing to drive for a game, that sort of thing. So once we had those details on there, we were able to build it. But building a marketplace is like the hardest product to build when you have to build up both sides. So you kind of got to really have a good hold of one to be able to bring the other in. 4:10 Cordelia King: So we had a good hold of these clubs. And thankfully, those first few clubs were able to come on the platform for us so that we could say to players, hey, there's clubs on here wanting to recruit you. That brought on the players. Once we had players, you go back to the club, more clubs, hey, we've got players on here looking for a new club. And then you kind of build both sides at once. 4:30 Thea Ngo: That's cool. Yeah. So At the peak of it, how many players or clubs did you guys have? 4:35 Cordelia King: I think we had about 10,000 players and probably 500 clubs recruiting. Yeah, mainly in local footy and local cricket. 4:43 Thea Ngo: That's amazing. And then, okay, if you have to have an estimate in terms of like who was like a fake profile slash catfishing versus who was like authentic and normal, like how would you split that? 4:55 Cordelia King: Good question. We had a few like things looking through players to see what was real or not, like if there was duplicate images, If there were— you could do a Google search, a Google image search of, like, profile pictures to see if it had occurred elsewhere, and then flag it if it has anomalies in terms of, like, age, height, weight. People that were getting— players that were getting, like, a lot of messages who were in demand, you'd kind of quickly review their profile. We had a few instances where ex-AFL players would jump on the platform to find a new club after they'd been playing professionally, and it was hard to verify if they were— if it was actually the player or someone else. 5:30 Cordelia King: So they put a phone number on their profile, so we'd give them a ring and then— And crazily, it was them. So that really did wonders for our PR as well, going to the clubs and saying, hey, there's an ex-AFL player on the platform. Of course, the club's going to sign up and try to recruit them. So yeah. 5:47 Thea Ngo: Just one last question here on like the actual club side of things. Is there any way for them to sort through these players really quickly? 5:53 Cordelia King: Yeah, you can filter. So you can search by name, you can search by— there's like a map. So a lot of these clubs are like very regional. So they want to try and find someone in their area rather than someone from like inner city Melbourne. So you can search on a map where across Australia nationally. So that's really helpful. You can search by position, age, height, weight, whether they're looking to travel, whether they're looking to relocate, what type of club they might be looking for, what type of salary they might be looking for per game. So it's really easy to kind of filter down to find that perfect player, and then you can just message them and go from there. 6:26 Thea Ngo: Cool. That's super interesting. And then do you take sort of any commission on like connecting people or? 6:32 Cordelia King: No. So the clubs will pay a subscription every month to be able to message all the players in our database. And players can make an account for free. 6:40 Thea Ngo: Beautiful. Can you imagine if Tinder had a toggle where you can toggle on weight? That'd be crazy. 6:47 Cordelia King: Yeah, it's important in sport, but maybe not so in dating. 6:52 Thea Ngo: Like you mentioned, sort of like when you start to try to add— when you tried to ask the players to add their own data into TrainStop, they didn't really have any. So when you think about fixing that data problem, like what were some of the things you explored first before the computer vision model? 7:09 Cordelia King: We looked at what data was being like collected by whoever was like scoring the game. 7:12 Thea Ngo: Yeah. 7:13 Cordelia King: But it was mainly only goals and often just how many games you played. So if you're a defender, like it's very hard to have any actual tangible data on your game if you're not kicking goals every game. Same for a cricketer, it would be how many runs that you make on average, how many wickets you're taking on average if you're a bowler, and that's about it, which can vary game to game. And sometimes if you're playing in different leagues, that data is not going to be like in one central location to put into your profile. So it's very difficult. And often these recruiters, like, they just want to see you play. They just want to see something that they can kind of get an idea of what kind of athlete you are. 7:50 Cordelia King: So the vision is actually really important as well. And like highlight packages, people are always asking for highlight packages from a recruiting perspective. So we tried to find those data points first, and they just didn't really exist in a way that was usable or watchable or easily collectible for players. And luckily, at the point when we started building the vision model, like, AI had just gotten good enough where it can start identifying things off images. So that's how we started building it. And GPU and compute had come down in cost to something that was actually reasonable and affordable if we were to build a computer vision model. So that's what kind of set it off. 8:29 Thea Ngo: So, okay, so SuperStat is what you built to fill that gap. So, okay, a parent or coach is filming their kid's game on a phone this Saturday. In 2 sentences, what Do you give them back? 8:40 Cordelia King: We give them back all the data, all the stats, and all the highlights for every player on a court in an easy one-click upload process. And you're able to use that for development and for recruitment, recruitment, so you can get better as a player. You can be seen by those pathways in a way that's really easily, easily readable and understandable. I think that was probably 3 sentences. 9:04 Thea Ngo: Sorry. No, no, that's beautiful. So one of the things that was really interesting about your journey is obviously kind Kai has a background in AFL and you guys started out in AFL, but then you've sort of pivoted away from AFL. So could you tell me a little bit about like what happened the first time you ran the model on like a real AFL game? 9:23 Cordelia King: Yeah, for sure. We started building the model for AFL because we had that immediate distribution to all the clubs that were recruiting with us. So it was a no-brainer and we knew that they were feeling that pain point of not having the data. AFL is very hard to build a computer vision model for because the oval is huge. So when you're trying to film a game, you're constantly zooming in and out, and it's hard to get an idea of where you're filming on the oval. There's 18 players on each team, so again, it's a lot of people to identify and constantly track across the ground when you can't see them all the time. The stats are very difficult to identify. 9:58 Cordelia King: Like, there's a lot of really niche stats like clearances. Hitouts are a bit easier, goals are a bit easier, handballs might look a bit different every time, tackles are a bit different every time. So it was a more stats that we had to annotate for the model to be able to collect. So that was difficult. And a lot of occlusion, which means it's like people running in front of each other. So it's hard to like see where everyone is. So we tried to build this model for AFL and we had all these problems. And we worked for a couple months. And Sam and I both played basketball growing up. And Sam actually, the first ever thing he built was a shot tracker for basketball when he was 14. 10:34 Cordelia King: And we were kind of like Atlas, like pushing this rock up a hill, trying to figure out how we could do AFL. And then Sam was like, why don't we try doing this for basketball? Like, the court is the same size every time. It's 5 players. You can see everyone. The stats are really easy. And we were like, okay, like, let's try. And the model which we'd worked so hard building for AFL worked really well on basketball because we'd fine-tuned it so well for occlusion, for the lines of the ground. And we kind of just took it and ran from there. 11:03 Thea Ngo: Cool. And I feel like AFL has a lot of randomly niche rules. 11:08 Cordelia King: Yeah, it's really hard. You can like go to a basketball game and not know anything about basketball and still like know what's happening and enjoy it. And AFL is kind of crazy. Like if you're a tourist watching AFL, I don't, I don't know how you'd understand it or know what's going on. It's really hard. 11:21 Thea Ngo: Yeah, my first AFL game, I really do not understand at all. I just know that like when everyone's standing up to cheer, you're supposed to stand up and cheer. And I'm like, great, amazing. But that's super cool. So now it's basketball, but you still have sort of the problems, maybe like one really shaky phone filmed from the top of the bleachers. How does the model process footage that's like not super clean like that? 11:46 Cordelia King: Well, the accuracy of our model depends on how good the footage is. The things that it needs to be accurate is it needs to see what number you have on your jersey to be able to identify who you are on the court. It needs to see the lines of the court to figure out where you're standing on it. It needs to see the ring and it needs to see the ball. So as long as it can see all of those things, it's pretty accurate. It's just a little bit more accurate if it's from a higher vantage point because you've got a better view of the court and where everyone's running. 12:17 Thea Ngo: So you mentioned like the number on the jersey. What if a player like turns around? 12:22 Cordelia King: Oh, we backtrack them. So it'd be like a human watching it. It's like if you— if a few seconds previously you can see that they're 17, but then they turn around to shoot the ball, you're kind of tracking them through that footage to know that they're number 17. So it uses tracking. 12:35 Thea Ngo: Cool. And is it accurate enough to say, like, if say a player like runs in the front of them and like kind of covers them a little bit, does it still track the player? 12:46 Cordelia King: Yeah. So that's occlusion. So that's the problem we had in AFL. And it was really hard because like in a tackle in AFL, it can be like 20 people and it's like, okay, where is everyone in this frame compared to this frame? But basketball, it's just running a lot of the time is running in front of each other back and forth. So you're just tracking them through the frames before and after to figure out where they are in the middle. 13:06 Thea Ngo: Cool. That's super interesting. So as you develop this like computer vision model, what do you think is the hardest problem that you've had to solve so far that not a lot of people estimate? 13:16 Cordelia King: Yeah, the really hard part is not like the computer vision model is obviously very difficult and Sam is a genius for building what he's built. But the hard part is running it over like hours of footage. Like you're not running it on 10 seconds or a minute of footage. So you see a lot of like LinkedIn posts of people building these sport models with like YOLOv8, SAM 2, and they're like, look how well the analysis has been done, the demos, and it's 10 seconds of footage. And it's like, okay, that's great. But it's, that's pretty easy relative to processing like an hour and a half of footage, especially where there's a lot of dead time. So like timeouts, fouls, halftime, quarter time, it's hard to teach the computer vision model like, hey, the game's not in play. 13:59 Cordelia King: Like, don't analyze. any of that bit because no one's on the court or some people are on the court. So it's the length of footage that we have to process in cloud GPU can get very difficult. Just big videos are really hard to like upload, stream back, process. But technology is getting better and we're getting better. So hopefully it won't be a problem for much sooner. 14:21 Thea Ngo: One thing that I thought was pretty interesting is obviously computer vision is quite a heavy model. So do you guys develop your own local model? In order to do that? 14:29 Cordelia King: I like to describe it as like Frankenstein. So there's a lot of really good open source papers out there that we've learned from, and we use those models, but we have to add in a lot of things and fine-tune a lot of it for the sport that we're catering for and for the vision that we're catering for. So it's a lot of different things piled on top of each other, which is why it's so heavy to process and why we have to use cloud GPU. But eventually, as the tech gets better, we'll be able to process that on people's phones. Like you go for a run with Strava and it tracks it on your phone, it's not tracking it in the cloud. cloud. And that's what we want to do, which brings compute cost to zero if you're running it on your phone, and we can give away stats for free to every player. 15:07 Thea Ngo: Yeah, that's amazing. An on-prem computer vision model. I love that. So one thing I thought was pretty interesting is that you also keep a very live thread going with your actual users, coaches, and parents. Do you ever feel the pull of users asking for things you know will move the needle? And how do you kind of like hold that line? 15:28 Cordelia King: Yeah, 100%. I mean, I was just on my phone like 10 minutes ago replying to some people. But yeah, we'll get like a lot of messages every day, which is awesome, about things that people want to see, um, things that people want to add into the platform, but also things that people don't use, bugs that people are having, which are equally as important if not more important. So I think it's about knowing who your customer is. In our group chat, we have parents, we have coaches, we have players. So it's understanding where that message is coming from, who it's for. And a lot of the time it's understanding why they're asking for it. So someone the other day asked for the ability to export like the timeline of the stats. 16:07 Cordelia King: And I was like, oh, just out of interest, like why, where are you exporting it to? Like, what are you doing? They're like, oh, I export it into my video editing software because I want to cut up all these specific moments. And I was like, oh, you can do that in SuperStat. You can cut up the moments here. And they were like, oh, right. I didn't know that. So I think it's like digging a little bit deeper to understand and why they're asking for what they're asking for. So you can build for that rather than build what they're asking for just straight off the bat. 16:32 Thea Ngo: Now, I really like that point because I feel like sometimes people want— I think people try to problem solve themselves and then give you like the solution. But maybe that's actually not the right solution that like applies for that problem the best way possible. And like maybe that problem is actually experienced by a lot of other people. So like just fixing the problem is better than like doing whatever whatever the solution that they propose. So one of the things that I thought was pretty interesting too is like you've kind of had this like journey of like pivoting around the sports tech space. So if you pivoted off something that was already gaining like real traction, so like 10,000 users on the platform, because of structural concerns and basically started from scratch. 17:11 Thea Ngo: But how did it feel like internally when you guys made the decision to like kind of like step away and focus on something for the time being? 17:19 Cordelia King: Yeah. Are you talking about stepping away from Train Stop into Superstat? Train Stop actually still runs. So we've got people, we've brought in people to keep it going while we were working on Superstat, which is cool. Shout out to my brother. 17:33 Thea Ngo: Like people, but yeah, it's my brother. 17:35 Cordelia King: He's very good at it actually. He plays footy semi-professionally, so he really knows that world and he's very good at talking to the players in the clubs, which is great. But it was hard. It was our baby, that, that first product. It's like moving overseas as well. Now it's even harder to step even further away from it. But I think when you have such an opportunity in front of you, like for us, Superstat, and the funding that we've gotten and the audience that it has, like to resist that kind of pull is a risk. So I think you've got to take risks. Yeah. 18:06 Thea Ngo: So like you mentioned, you're moving to Texas. 18:09 Cordelia King: Yes. 18:09 Thea Ngo: Which is incredible. I've actually never been to Texas. 18:12 Cordelia King: I only just went last month for the first time. It's very cool. I'm excited. I loved it. Austin is so cool. It's got— it's very Melbourny. It's got a big river through the middle of the city. You can swim in it, the river, which is amazing. It's very warm, massive sporting city. So like the Longhorns, the university there, huge footy team— football team, sorry, not footy. And it borders on Mexico. So the Mexican food there is really good. 18:38 Thea Ngo: Oh, I can imagine. So you're moving to Texas. What are you hoping to get from the US comparative to here in Melbourne? 18:47 Cordelia King: Yeah, I mean, our initial product offering in basketball, like you look at the size of basketball in Australia, which is still very big, but then you look at the US and the opportunity is huge. I mean, you've got to go to where your customers are, you go to where the pull is, and that pull is very, very, very strongly from the US. Like over half of our customers on our platform are from the US now, despite not having done much marketing or reaching out in the US. So clearly that's where we need to go and continue to build. 19:15 Thea Ngo: Yeah. So is Texas sort of a basketball capital for the US? 19:21 Cordelia King: It's one of them in terms of junior sport. Texas is just a very strong sports stronghold in the US. There's a lot of schools, there's a lot of colleges, there's a lot of passion for sport over there and getting into those pathways into college and certain schools. So we're going to go take advantage of that. 19:38 Thea Ngo: Oh my God, so fun. 19:40 Cordelia King: Cool. 19:41 Thea Ngo: So, okay. So you guys are 3 co-founders. You and Kai have known each other since primary school, and then Sam joined the 2 of you at 16 to build your first company. Okay, so you picked a co-founder out of a primary school friendship. What did you see in Kai back then that turned out to actually matter a lot now? 20:02 Cordelia King: Yeah, this is funny. He was very, very sporty in primary school. He was always playing markers up at recess and lunch. He was very, very competitive as well and always wanted to win and be the best. He was very ambitious, and I think that still rings true today. His goals are very high for himself. So I think partnering yourself up with someone like that, you're always going to strive for the top possible outcome that you can. And that's probably what I saw back then that still rings true today. 20:30 Thea Ngo: Cool. And then what about Sam? How did, how did the trio, like, met? 20:36 Cordelia King: Yeah, Sam— well, when Kyle had the initial idea for Train Stop through footy, he was coaching a bunch of boys at the time and was telling them about the idea. idea and what he wanted to build. Um, and these boys who I think were Year 9 or Year 10 at the time was like, oh, there's a kid in our year level at school who's like a, like a gun at coding, you should work with him. And Sam was 16, and it just took off from there. 20:58 Thea Ngo: Oh my God, the rest is history. Beautiful. So, okay, so 2 companies, 10 years, same 3 people. What do you think is the hardest thing the 3 of you have ever disagreed on? 21:09 Cordelia King: I think the pivot from AFL to basketball was a pretty big one. And it was something that was very, very easy to decide in the end. But having that immediate distribution in AFL was hard to not utilize. So we did talk about that for a while. I don't think we've ever disagreed, so to speak. I think there's been a lot of like trying to figure out what the best path is and not knowing and just really discussing it and like mapping out every possible opportunity and picking the one that seems to have the best rate of success. 21:40 Thea Ngo: A lot of whiteboard sessions. 21:41 Cordelia King: A lot of whiteboard sessions, yes. 21:42 Thea Ngo: That's cool. Um, so who is more pro basketball and who is like not so pro basketball? 21:48 Cordelia King: I'd say Sammy was definitely pro basketball. Um, he plays, he plays every week, he plays pickup, he's very good. And I think Kai, just working in footy, he used to be an analyst for St. Kilda as well, and knowing that footy landscape inside out, saw that pain point which still exists very, very strongly in footy, not having that data at the local level. And it's something we still want to solve. I'm just waiting for the tech to catch up a bit and then we'll be able to solve it. 22:12 Thea Ngo: Yeah. And I can also imagine, like, sort of as you refine the model for the basketball, you can generalise it to something that's a little bit bigger, that's a bit more difficult, like footy or even American football as well. 22:27 Cordelia King: Yeah, that's our next one that we really want to get into, American football. 22:30 Thea Ngo: That one is also pretty hard in the sense that, like, I feel like there's 20 people tackling the same guy, right? 22:34 Cordelia King: Yeah, a lot of occlusion in that. But the vent, there's always like It's such a fancy word. 22:40 Thea Ngo: Occlusion. 22:40 Cordelia King: Occlusion. People running in front of each other. There's always big bleachers, which is like big stands in high school football where you can get a good filming vantage point from. So you can actually see everyone on the ground. And there's a lot of good things like everyone's got a number, the lines on the ground, you're able to calculate distance really easily. There's a lot of things like that that will help us along the way, but we're keen to crack into that one next. 23:04 Thea Ngo: Yeah. And I can also imagine it's one of the biggest recruiting— 23:08 Cordelia King: Huge. 23:08 Thea Ngo: Sport in the US in general. 23:10 Cordelia King: Yeah. I mean, like college NFL players will make $2 million a year. Like it's professional sport and the recruiting is still so manual and the data is still like nonexistent almost at that level. So huge pain point. 23:23 Thea Ngo: Yeah. 23:23 Cordelia King: Yeah. 23:23 Thea Ngo: And I can imagine that's like probably a little bit easier than maybe like golf or swimming, which has no lines or grids or anything to measure from. 23:32 Cordelia King: Yeah. It's hard to like film that, like golf, every hole walking around. 23:37 Thea Ngo: Where is the ball? 23:37 Cordelia King: Yeah. 23:40 Thea Ngo: So one thing that I thought was pretty exciting when I looked into your journey as well is that, like, you mentioned that none of the 3 of you took a salary for a very long time. And Kai actually sold his car, Sam left university, and you moved back home. So each of you gave something up. So giving up your stable income, when did you feel sure enough about this to take a bet on it? 24:06 Cordelia King: It was hard. Like, I had a full-time job in graphic design previously to this. price, which is hard to get if anyone's a graphic designer. It's a competitive market. And it was in fashion, which I loved, which was very cool. But I think just that natural pull, it was like, hey, this can be something cool. This is— it's got potential. We've seen the growth. Like, we all knew and trusted each other from building before. And it was like, look, if not now, then when? Like, fuck it, let's do it. Let's try it. Like, what have we got to lose? We can always go back to those jobs. jobs, you know, after if we try this and it doesn't work. So I think the safety net was like, you can always go back, like nothing's stopping you. 24:44 Cordelia King: Although my mum wasn't happy moving back home. 24:47 Thea Ngo: Is she happier now? 24:48 Cordelia King: Yeah, she's happier now. She's happier now. She was a bit worried. She's, she's quite risk-averse like myself. So it was hard. But I think we all kind of see the big opportunity now. 24:57 Thea Ngo: And she's just looking out for you. 24:58 Cordelia King: Yeah, she is. She is. 25:01 Thea Ngo: So you mentioned moving back home. What did that actually feel like? 25:07 Cordelia King: Weird. It was kind of fun. I love my family, so I was quite excited to go back home. I don't know if they were excited, but, but now I'm off to Texas, so they'll miss me. 25:16 Thea Ngo: But no, I always think that idea of moving back home is sort of like a weird feeling in the sense that like, oh, like, I guess they're happy to have me, but also like it's, it feels like a regression back to my old self. 25:27 Cordelia King: It does feel weird. 25:28 Thea Ngo: Yeah. But I guess now you're gone, so. 25:30 Cordelia King: Yeah, I think you just have to see it as like a temporary sacrifice. Like a lot of being a founder is like a temporary sacrifice. for greater good, for greater good. It's like pain now, like it pays off later. So I think it's just like trusting in the outcome and using that sacrifice to work hard as well. 25:46 Thea Ngo: Okay, so speaking of regret, what do you think you regret the most throughout this entire journey? 25:52 Cordelia King: Probably not, not doing it earlier. Um, so I'm 25, which is like— Sam calls me Unc. It's like old age. Don't, don't. 26:02 Thea Ngo: Um, everyone has their own timelines, you know. 26:04 Cordelia King: That's true. But everyone always says, like, if you've got an idea, just like, just do it, try it sooner than you think you're ready. And I think that's, that's so true. I think, as I said, naturally I'm a very risk-averse person, so I will like overthink something and think a lot about it before I maybe try it, when it would most likely be quicker to just try it and then see what happens. So I would definitely say not starting earlier. 26:27 Thea Ngo: So how early would you have started? 26:29 Cordelia King: It's a good question. Well, to be honest, I kind of like the idea came very naturally, so it was— it would be hard to bring that idea like further back in time. So I'm not sure. I think I would have liked to pivot faster from AFL to basketball, but when we had that immediate network of all those footy clubs, it was— it's hard to say that in hindsight. Um, but so far, no, no huge regrets. Like, we kind of wiggled our way along and found the right path, and trusting our gut and instincts and trusting the data. So So I think following on the journey that we've had so far, we should be okay. Yeah. 27:03 Thea Ngo: I mean, also like hindsight bias, you know? So, but I think it's good to like look back at a decision and be like, oh, like I wish I made it sooner, rather than look back at a decision and it's like, oh, I wish I like didn't really do that. So at least that's the good thing on this bias. But yeah, I appreciate you coming on. Like I've learned so much about sports tech. And I feel like I need to go back to my basketball roots these days. 27:31 Cordelia King: And film a game, upload it. 27:33 Thea Ngo: Yeah, I really do. Okay, thank you. 27:36 Cordelia King: Thanks, Leah. 27:37 Thea Ngo: That's a wrap. If you like this episode, please hit the like and subscribe button. It helps us bring on more awesome guests, level up production, and bring on new series you'll want to watch. And if you want to hear more early-stage builder stories, check out our other episodes. Okay, see you next time.
▶
▶
▶