Carv Digital Ski Coach - “We want to give skiers fast, objective feedback.”

There was a skier king of the leaderboard, highest IQ, end of last season.

Yeah, so it's an instructor called Tyler who was at the top of the leaderboard last time.

With some of the changes, maybe filters and algorithm tweaks, that sort of stuff, can you say who actually was— had the highest SkiIQ?

Based off our current analysis, so we're still going through the process of finalizing everything from last season, Ted now goes to the top of the leaderboard with, with the new version of SkiIQ, which is obviously great. We want the best skiers with the highest skill to be at the top of the leaderboard. And the real goal for SkiIQ and the metrics this season was to make it reward more dynamic carving instead of it being kind of getting to a higher jangle, but, you know, more of a park-and-ridey turn.

I must say that that news was like music to my ears because I just— but it's also a testament to you guys. You look at where you're at at the moment and you go, hang on, Ted Ligety, we know how good he is, ski with him, but then he's not sitting at the top of the leaderboard with SkiIQ, someone else is. What's going on there?

Tweak it.

So, so then, like you said, so it's not that you want to reward that, it's that you want your system to be truly honest. Hey everyone, this is Tom from Big Picture Skiing. We've got a great podcast for you today. I'm in London, Carv HQ today, with founder of Kārv, Jamie, lead data scientist Charlie, and head of product and marketing Alex. And we've got a really interesting episode where diving into what exactly this tech product is, uh, we're looking at what changes, what improvements are happening. What Jamie's going to give us the story of how it all began, why it all started, find a little bit more about the science, uh, behind things, some of the new metrics that Kārv are bringing out this, uh, winter, and then how Alex is sort of shaping the experience you get as a user when you ski with Carv. So great episode coming up, tune in. Let's, uh, jump straight into it. Carv has recorded 344 million turns. Take us back to the beginning, like how did it start?

Well, so, um, yeah, so my background is physics and, um, and then finance as well. And so I, and I also learned to ski very late. So imagine I was about 18 years old, like, desperately trying to ski, trying to catch up with my friends, and basically thinking, how can I hack this sport to actually be better at it? And I was very lucky to have some great coaches— well, instructors— right at the beginning for, you know, about a week or so. But then I was really just trying to teach myself, and I was watching videos of how people were skiing, and I was trying to get tips from friends. But The whole time I found it like very kind of like stop-start and very subjective, 'cause I wouldn't know if I could trust a friend or, you know, if I was copying someone right, if I was trying to copy them. And so I really kind of thought, you know, given I was spending all my time, so I was doing a PhD in financial economics and I was looking at spending all my time looking at these time series of data from finance and thinking like processing that data into something useful. And so I was thinking to myself, I know it's possible to put sensors on, you know, anything. And I know it's possible to see how, you know, I actually ski.

Ski.

And then the missing element is if you could take that data and, you know, process it to give some meaningful insights, then I could actually improve my skiing technique. And so I got really excited by this idea of, imagine if I didn't have to, you know, wait to see a video of myself. Imagine if I could get that feedback to improve myself in the moment, then I could create this like really powerful experience where I'd just be getting better, you know, every run or even every turn. And so I got really excited by this idea. Um, and yeah, very naive at the time. But this was kind of like, you know, sort of 2014, and I was kind of finishing my PhD. And I basically put an advert out around Imperial while I was doing my PhD for anyone that could essentially code, because I was very good at like making algorithms. And, you know, my whole PhD was trying to process data into something useful. But I didn't know how to actually make an app really. Um, I'd made my own app to, to measure my own data, but it wasn't enough to actually kind of create a fun, powerful experience. So that's when I met Pruth, uh, my co-founder, and he basically took all the really kind of basic concepts I had and actually turned that into an app. And we started to sort of put sensors in ski boots and then pitch it to investors who very, very thankfully invested in us in 2015. Uh, and then we went to China for about 3 3 years in total, but 3 months initially. Made the initial prototypes, got some things working. We then started skiing with the US, um, so we then started skiing with the UK team, first the UK team, um, at one of their ski camps, and actually with David Riding, and then got a lot of data about how to, about what great skiing is. And then from that, we kind of built, built up and up, had a Kickstarter campaign and grew the team. So yeah, that's, that's the brief introduction.

Okay, now like, like part of it, if people don't— this is the first time they're being introduced to Carv, this is kind of what it looks like physically. Do you want to explain the physical part and then the app part? How, like, what is Carv? Yeah, no problem.

So, so Carv is a digital ski coach and it consists of a physical product, which is this hardware we see in front of us that fits into your ski boots. So it retrofits into your ski boots. You take out the liner and then you put this layer in the bottom of your ski boots where it measures the pressure and the motion as you ski. All of that data is then sent to your phone, and then as you're skiing, it's collecting all that data to give you two kind of experiences. The first one is when you finish a ski run, you can get a SkiIQ, which is essentially our measure of how good you are at skiing. Um, average person is 100, um, and if you're very, very good at skiing, you can get to 150, 160. I, I don't know what your SkiIQ is.

I think, I think the best I've got is like some maybe 150. 156, 157 or something.

Yeah, some real top ones there. So, um, and so that's— and then you also have a tip. So that's our kind of analysis of what you can do to improve. And then from that, you can drill down into all the different metrics that, you know, Charlie's been working so hard to, to basically characterize skiing technique. So that's one format. And the other format is that real-time coaching. So basically, as you're skiing, you can actually have Carv talk to you about one specific part of your skiing to create this very powerful, uh, real-time feedback experience.

So that's the, that's the product.

Yeah, I mean, I think I love your idea, the first idea of an uninterrupted learning process that is kind of just based on, on this information you've got. And then so you're sort of— it allows you to go through the self— a bit of self-discovery process, right? You're trying to put the power in the hands of, of individual, individual skiers. Have there been some What, what would you say? I'm going to ask the other guys their, their biggest hurdles they've come across. What's been the biggest hurdle so far to get to this point? Because I'm sure with a tech device there's a lot.

Oh, the same.

This, this could be a podcast. Um, I think one of the biggest ones was really, really getting a physical product into market is, is really tough. Um, you know, very, very low funds, doing things in China. Um, doing— running a production line. Yeah, right, right at the beginning was definitely a—

definitely, because you're holding like that. Our mic today is part of the insole in here. So what it— like, in there is— you're holding like, like that part. It's got to deal with people putting a lot of pressure on it.

And yeah, it's got to be waterproof, and it's got to cope with people taking in and out of boots. So yeah, essentially what we're looking at is this is a pressure sensor with a motion sensor actually put into it here as well on the, on the chips. With these chips here. And you've also got a kind of metal board on the back to make sure it's strong enough. And so what we've got here is about 3 layers, so about 5 layers actually, of different materials that build up to measure the pressure and the motion. So the pressure is essentially, there's essentially a layer of foam that squeezes when you push on it. Then there's an array of sensors that sit underneath it. And so that means we have 32 sensors that actually kind of show us how your weight is moving across. So there's a lot of complexity in that.

Yes. Yeah. Okay, so then Charlie, like, you're involved, and I mean, his role is super important, right? Your 344 million turns, can you give us a bit of an insight into like what a data scientist does and some of the things you— Yeah, yeah, yeah.

I mean, so basically what we're trying to do is convert the raw data we're collecting into something useful for people to use and then build that into their own skiing. So yeah, the big question is, can we measure skiing and is it going to correlate with different ability levels? So if you're an expert skier, as you are, and I'm an advanced skier, we need you to be getting higher scores than me. So that, that's basically the kind of, uh, biggest, uh, problem that we try to solve initially.

And so in what Maybe you're probably the best person to ask, like, what are some of the metrics? So all this data coming in, there's probably numbers. What do they get turned into that the skiers who use Carv and they look on their phone, they see? Like, run us through the things that measures.

So we basically sample from the hardware 20 times a second, and that is some measurement of, uh, of the sensor. So we measure in your, your motion and your pressure. So with your motion, we're looking at, you know, 20 times a second, where is your ski pointing. And then not only where your ski is pointing, but how that progresses through the turn. And then when we translate that into a metric, what we have to do is turn something that is measuring throughout the turn into one single number that describes something important about your skiing. So, uh, one of our best metrics is called edge similarity, and basically looks at how similar your edging is between your inside ski and your outside ski throughout the turn. So when we visualize the raw data on, uh, on a graph, you can see this super clearly for an advanced skier versus an intermediate skier. The advanced skier will be doing almost the exact same thing with their inside ski as their outside ski. Their outside ski might be a little bit higher because that's the, the ski that you're mainly loading against, Um, but when you look at an intermediate skier, it's fascinating. So someone who's doing a kind of wedge parallel turn, starting off with a wedge because they're not as confident to go straight into a parallel turn, they actually, in their raw data, you can see it remain flat throughout the first bit of the turn, then they catch up later. So visually we look at that and we're like, that's completely different to, to your skiing because you're doing the exact same thing pretty much with your inside ski and your outside ski. Um, so yeah, our job as data scientists is we can see this is different, but how do we convert this into a number that agrees with what we can see? So yeah, day to day, that's basically what we're trying to do and make sure that we've got the systems in place to evaluate that we're doing it well.

Um, yeah, so yeah, okay, cool. Because I know I actually got to try it in New Zealand, um, I think maybe just after the Kickstarter campaign. Had, had, had taken off. And I do— I remember being so fascinated because you had the mode where I could see, like, I'd go through all the metrics, like, I could look at how much edge angle I produced in a turn and even, like, pinpoint exactly what turn that, that happened on, and then where my pressure was fore and aft at specific parts of the turn. Like, it was amazing, the level of detail you could pick up with, with this thing. And me being fascinated with the body and with skiing, I was like, I love this objective— like, subjective, you know, non-biased data. It's telling me what is happening. And I had a breakthrough because I thought I was in the right position, and I looked at it, and these sensors are telling me, no, you're not. You're on your heels more than you think you are on the front of the foot. And I knew what to do with my, my boots, changed it. Next day, took out Carv and then went saw the, the change in pressure. And anyway, it's fantastic. I think you guys have developed something that gives, gives people, um, power. Alex, like, you're involved in, in marketing. Like, how have you— your job is to like take this great stuff, this awesome idea, the science. How, like, what have you found interesting bringing it to the, to the skiing market?

Yeah, it's a super interesting question because, look, I think when, when most people first think about Carv, they think this thing can't possibly work, right? It's probably a gimmick, or like, is it really going to give me advice? That's probably the first reaction people have. And then especially because if you're a good skier, it's kind of hard to get good advice. You know, if you're, if you're a good skier, you, you often might even think with an instructor, am I going to get the right advice from this person that's going to help me? And it's a bit of a gamble. And so actually, a lot of great skiers will look very hard at which instructor they're going you can— they're actually quite, um, you know, choosy about that. I think the first thing is, like, we have actually the inside of our insert here, and I think when people see some of the graphics or whatever we put out in our marketing, it kind of looks like showy marketing. But it's kind of cool actually seeing inside it, and, and maybe it's kind of hard to see on the camera, but you can see just how much data it's possible to collect. And we can even see, because we've got all of these sensors We have the ability to see the pressure roll from like your 5th metatarsal to your 1st as you enter the turn. So we can actually look at things in skiing that an instructor might need to freeze frame a video. Well, we can see that in the data profile. And then the cool thing is we get to work with people like yourself, and we try and balance this across a lot of different instructors so we're not becoming too biased in a certain way of skiing. Um, and we get to try and construct something which people can use to get leverage on their own skiing. Because once you give some data, it kind of gives them something to play with which maybe they didn't have. And you'll find, you know, especially people who have quite a sort of self-starter, like, hacking mentality, they'll go out and they'll just hack away at a metric and it'll get better and better and better, and they'll suddenly start having an unlock. And I think the challenge for us is trying to persuade people that that's going to happen. Yeah, we see it all the time.

Do you think there is— have you found there's definitely a type of personality that that really is drawn to Carv currently and maybe a type that's not?

Yeah, definitely. I mean, we're, I'd say we're in the early adopter phase of Carv's life. So I think we certainly have probably the more tech-enabled person. But really for us, if you're, if you think you can improve your skiing and you want to find a new way to get to the next level, Carv is perfect. And we do find that kind of person really levels up.

Yes, I would agree that, that— I mean, some of the testimonials you've had of people— I remember going into a ski shop in Sun Peaks and one of the, the staff there, the like instructor friend, he's like, oh, that's, you know, I can't remember his name. He's like, he's, he's only been skiing a couple of years and he hasn't had any lessons. He just bought Carv and it's amazing to see.

Yeah, yeah.

And it's— and he's, his SkiIQ, boom, boom, boom, up. And you know what, it makes me think of, like, there's a question I want to bring up here, but like, here's a guy with kind of a blank canvas, comes to a sport, he doesn't have any biases from other people, he just wants to learn it, have fun. And then he's got this device that tracks everything, gives him feedback, he utilizes it, goes practice, you know, gets in this sort of pattern, in this habit of doing it, and and, you know, a season of skiing and he's like really improved. So you guys are in like this unique position to really influence and change people's skiing. And we were discussing a little bit before this morning around, you know, like you've mentioned, you've worked with some associations, different instructors, and used like the Balance Edging rotary pressure kind of the skill set to measure things based a lot of the metrics off that. However, there are some, some newer metrics coming out. So maybe my question would be, tell us about the newer metrics, and then do you want to maybe change how people learn and analyze skiing?

Yeah, I mean, I can kick us off. I'll probably hand over to Charlie in a bit. Um, I think, you know, when I think about what we're trying to do at Carv, we basically have some principles at the heart of our coaching that we're always trying to bring in. And whether we're trying to coach something very specific and technical in certain areas, there's some things that we really care about. So the first thing is we really try and build this product so it doesn't get in the way of your skiing. I think, you know, technology, you know, when we're skiing, it's like very valuable time. We're on the mountains with our friends. It's expensive, but it's just an incredible place to be, and we're privileged to be able to be there. So with Carv, we try and not get in the way. So we're always thinking about how we adapt our coaching and how we interface with the user to give enough to help an improvement, but not enough to basically break the skiing day. I think that's kind of an interesting thing for us.

Yeah.

And the other thing, which I think you're driving at here is, is around giving fast, objective, instant feedback. And for us, we kind of have the luxury to change what that looks like over time. I think, you know, ski associations, they have to teach through so many layers of instructors, just the communication problem is really, really hard if you try and drive consistency through a big Ski Association, it's a real challenge. We can change what we look at every single year if we find a better way to look at something. So I'll hand over to Charlie in a second, but we look a lot at edging. And a good example this year, we were working with Ted Ligety, a 5-time world champ, 2-time Olympic gold medalist, and he was doing something in the way he was edging his skis that we could see in the data that we hadn't really looked at before. And without giving too much of the story away, I'll pass over to Charlie, who did a lot of this work. And Charlie, you can explain about our new metrics.

Yeah, thanks. Um, yeah, exactly as, as Alex says, objective real-time feedback is so powerful because the way to get a higher score is to do this aspect of your skiing in a better way. And if you can stop at each run or each turn, depending on which mode you're using with Carv, you can hear this number and, you know, you can try different things to try and improve the number. But to get the number higher, you need to be doing the skill in a better way. So, uh, the metric Alex was talking about, that is, yeah, pretty much inspired by, uh, looking at Ted's data, talking to him about the way he skis. Um, you know, we've been so privileged to be able to work with him. Uh, the metric we're calling, um, progressive edge build. And one of the things Ted does is, you know, he gets on his edges super early and he's building the edge angle, but he's not, uh, just kind of doing everything early and then letting things go from there. He's continuing to build his edge angle throughout the turn. So we, we can see this in our graphs of the real-time— of the, the raw data, as I mentioned earlier. We can see him continue to build, whereas a less advanced skier who gets to a higher angle but then is just kind of holding it there, or even coming off it— park and ride turn, longer radius— Ted's able to continue to build that edge angle whipping his, uh, his ski around super quickly, tight turn radius, because the higher your edge angle, uh, the more the effect of the reverse camber of the ski, uh, whips the, the turnaround. So, you know, seeing, seeing Ted skiing, trying to keep up with him as, as he's skiing in front of you, you can see him doing these incredible turns, and it's, yeah, an absolute privilege to be able to then look at 20 times a second what is that actually doing to his edge angle. Um, so yeah, that's one of the new metrics that we've, uh, built through working with him.

Yes, um, it's pretty neat that you can— like, it won't— it'll be new for this year. People will be able to take a run, look at their progressive edging score, and see what— see where it's at.

Exactly. And then the next run, you know, you might try something else. Maybe you would, uh, you know, really committing to the turn, inclining a lot, trying to get lots of edge angle but you're, you're, um, you know, overcommitting, and then actually you've got an inside ski pressure, you're, you're kind of, uh, not balanced against your outside ski, so your turn suffers. Um, you know, the next time you do a run, you can try, oh, maybe I'll try and slow it down. Um, and you know, when you, when you get that, that click, that feeling of the, of the great turn, the number agrees with it. So you don't have to be skiing with someone like you who really knows their stuff. You can just look at the numbers, if the numbers are, um, you know, uh, that they can provide that objective feedback, which, yeah, it's just awesome for, uh, for improvement.

Yeah, I can't wait to— yeah, yeah. So there's something also that we, that we see with, with CARV is, um, and, and I think you do this extremely well, Tom, but if you give someone more of a structure to think about their sport, you know, it's not like people don't want to improve, it's just it's so hard to work out what you might want to do next. And we actually survey this a lot with people who come into our website and we ask them a few questions. And actually knowing what to do next is one of the biggest challenges in any sport. And I was listening to a surf coach talking about this and he coaches the best guys in the world like Kelly Slater and people. And he was saying, if you can just give someone one thing that's going to help them change something, you make them intentional about it and they reflect on how it's going in that process, they will improve. Even if it's not like the perfect thing, you know, there might be one perfect thing for someone but actually just getting someone to know what other things I could change and let me just be intentional about that and focus on something I can, you know, do I feel something different? You know, I'll try something else. How does it feel again? I think with Carv, you know, we have this opportunity with these metrics we're creating to create like a new way to think about skiing. And that in itself is actually an unlock. So Progressive Edge, for example, is something people thought, okay, I'm either carving or I'm not. And they don't think about the dynamic nature of the car. And now we can think— people might think, oh, I'm scoring well, but I'm not actually that progressive. What is that? And that will build a lot of improvement just in and of itself, showing people all the mechanics of a really good ski turn.

Because I must say, what— something I find, uh, quite commonplace, especially when you look on the internet and maybe you ask someone what they think their problem is that they're skiing, they might give quite a could be terms like a static answer. Oh, I don't angulate enough, or I need to get more forward, or something like that. There's not— yeah, it's, it's definitely not progressive edging. I think it's going to be— you guys, I know from like my— the APSI, we have this term and it's in there, but you don't really learn about it or really teach it much until higher levels. I think it's fantastic. I remember we did the Zoom session, you said like should a beginner be taught about this, or do we just like tell experts? I was like, no, everybody should. It's like everybody should have the chance to look under the hood and know what good skiing is. I, I don't believe you should really hold stuff back. Yeah, I think to someone's skill level, when they go do it, they'll feel they won't be able to do it. But if you're an athletic person, you find out about that, you could cut off your, your first intention of not stuffing around, getting to being a really good skier faster.

Yeah, and you know, we've actually been quite cautious about— we wouldn't actively recommend Carv to absolute beginners, for example. There's so much that an instructor needs to do to acquaint you to this whole sport. But we do find total beginners pick up the product and do pretty well with it, and they actually get really good, and they sometimes get really good really fast. And I think, you know, we're quite cautious because we wouldn't go out and recommend it. We've designed it for intermediate to advanced skiers, even experts. And just to be clear on that, if you're confident parallel skiing, you're at the right level for this product, right the way up to top-level instructors. But we do find total beginners, actually, because they're getting that feedback, they can go probably faster than we would have thought possible, like just using that mechanism.

Yep. Jamie, within— with the app, what area of the Carve app do you wish more users engaged with?

Um, so we have, um, this mode that we made with Ted, and we do have a lot of people use it, but we think, uh, if even more use it, they get so much more value from it because it's, uh, it's one of these real-time coaching modes. And essentially what, what happens is, um, you get an intro from Ted, tells you about how it's going to improve your carving. We have 40 different levels, and then as you go through the system, it's— it initially focuses on one part of your skiing, and then it moves it to more and more complex parts of your skiing to really get you carving, get you on edge. And the interesting thing is, when we look at the data, we can see that when people use that modes, they really improve. So I think, I think there's so much more potential, you know, to— so we obviously designed the app with this in mind, but when we think about how to design the app, we think, okay, how can we make sure that we get as many people into this because it's having such incredible effects on their skiing?

Yeah.

Um, so maybe Charlie's got one as well, I don't know.

Yeah, what do you think? Because I mean, I guess the most popular mode function is Free Ski mode. Yeah, yeah, yeah. Is, is just turn it on, ski what, do what you're doing, and then see if it has a suggestion. Yeah. And then the, the Ski with Olympian one's a little more of a structured type sort of progression, correct?

Exactly. Yes, it's exactly what I was saying at the beginning. So we have these two kind of ways to interface with the product. First is that free ski mode, so you're, you're just setting it, and every, every time you finish a segment, so when you stop skiing, you hear SkiIQ and then when you get to the end of the run and the lift, you hear a tip about what you can do to improve next run. So it's a really powerful, kind of like fun way to initially sort of start thinking about your technique and, and start making technique into a game. And then once you really want to level up on a specific part of your technique, so say you're getting a tip, what the, what the system will do is it'll actually recommend you a certain part of the system. So to, to have a real-time, so you can, you know, maybe it's like focusing on your edge similarity, and so you can really start to dial in and basically improve at that thing with the real-time coaching. So yes, it's really about kind of minimizing the time between you doing something and, and having that correction to create this like very, very powerful feedback loop. So the idea is that at the end of a run you could actually have, you know, maybe sort of 30 or 40 interactions with, with the system and sort of slight nudges to improve your technique. So it's, it's a lot faster, it's a lot quicker way of, of learning.

Yeah.

So yeah, most of the time people will ski on that free ski because it, you know, it's fun with friends and, um, you don't have to kind of think too much about where you're going to ski and things. But once they, you know, once you really get into it and you want to kind of really level up. Yeah, yeah, yeah. Once you get through the easy things, then yeah.

And that's— I think that's just human behavior. People are on the lazy side in general. So you heard it from like the founder here, he's like, use the app. People, you know, like you buy a ski and instructors talk about like, use the whole ski, use the tip, use the tail. You got this, this app, this technology, you utilize it? Because I think a lot of people don't get right into it. And, and you, you guys are pushing more in that direction, aren't you? This kind of active coaching mode.

Exactly.

Not just— yeah, people can look at stuff, but then, then it is really getting involved in like sending you down a pathway.

Yeah, it's kind of like halfway between freeski and one of these like very active coaching modes because it naturally guides you through the whole coaching system. So it's a, it's a really kind of easy kind of step from freeski into, into a more kind of real-time experience.

So I got— I heard Alex said you're, you're actually doing a ski improvement camp. Yeah, coming up soon.

Yeah, yeah, yeah, yeah.

You excited for that?

Yeah, really excited.

Yeah, awesome.

Yeah, a lot of fun.

What are you— what are you gonna go in there, uh, I'm sure you're not gonna— you're gonna be able to switch off from like thinking about translation to car, but like personally, like what are you trying to— what do you hope to come away from that camp improving in your own skiing?

Well, so if we, if we get some— it's quite early season, but if we get—

This episode is brought to you by Carv, the digital ski coach. Now, one of the features I've had the most fun with all time is simply the fact that Carv tracks every single turn you make. I treat it a bit like Strava, so at the end of the day it tells me exactly how many turns I've done, and those numbers keep adding up across the season. So last season in Australia, I had a few days where I was doing nearly 2,500 turns, and my legs were tired, but it was really cool to try and and have that number there motivating me. And I'm quietly working my way towards my 1 millionth turn with Carv. So the other metrics I've been diving into lately, and these are new for this season, are very unique. And these are turn corridor and turn radius. And I don't see many skiers using them, but they tell you a lot about the shape and flow of your skiing. So the turn corridor shows you how far you're traveling across the hill, or the time almost between apex to apex, and how long you're taking to transition. While the turn radius measures the actual arc you create once you start the turn. So these can be different, you can have like a very wide corridor, say 18 meters, but a lower turn radius, or the other way around. So when you compare the two, it opens up a whole world of experimentation, and it would be a bit like a race coach will set different types of race courses to influence a racer or skier to change the timing and mix it up. So I encourage you to look into this if you've got Carv and play with it and see if you can manipulate it, because it's only going to improve your versatility. So I keep my Carv on my boots every single day because you never know quite know when like a change in technique or a new ski resort or skis will show up in the numbers, and it's great at the end of the day being able to check that, or even after the fact. And I'm really looking forward to the day I open the app and I see turn number 1 million So if you want to explore your skiing in a completely different way, check out Carv, the digital ski coach. And thank you, team, for sponsoring this episode.

Some off-piste, I'd love to do some, some off-piste skiing because that's definitely something I want to improve. Um, on-piste, it's really, yeah, it's really getting, getting those high edges and really, really carving. Um, you know, looking like Ted. Yeah, yeah, yeah. My SkiIQ is 153, and so yeah, it's definitely Um, it's good, but I think, you know, I'd love to try and get it to like the very top of the leaderboard. So yeah, I'll be asking Riley for that.

Yeah, okay, okay, cool. Um, it's interesting, uh, for me anyway, that comment about the off-piste we were kind of discussing before. I think some people like— that's, that's definitely super fun, like the, the most. I, I love it. I think off-piste skiing is, is the best. But I think sometimes people don't realize how much you will improve that area by just stepping back, going back a level, doing a camp, kind of like, or working on some skills back to like that work in Olympian mode instead of kind of just having a bash at it kind of thing. So it's just interesting. Um, yeah, like that's where people want to go, but the pathway there, I know that makes it easier is, is work on these like fundamentals that might be a bit, you know, not in this— that you think not in the exact direction of going towards your goal.

So yeah, I think, isn't, you know, I mean, obviously you're the expert, but I think it's a lot to do with when you're, when you're on piste, it's, it's quite forgiving. So you can make these mistakes and you can get away with them, but then you can't make them off piste. If you actually try and focus on not making the mistakes on piste, then you can learn the muscle memory.

Yes.

You know, so you're, you know, so you're not just like unbalanced and you're tilting your edges together. So when you do get off-piste, you're actually, actually skiing naturally.

Yes.

Yeah, absolutely.

And, uh, the, the—

Charlie, there's— back to the metrics thing, I got to try out a new metric this season and we were chatting a little bit. Turn, turn g-force.

Turn g-force.

Turn g-force. Yeah, yeah, bringing the G's in. Yeah, that's my favorite metric metric so far.

I think mine too.

Yeah. Um, can you explain what it is? And then you said it was actually— it's not super simple, like you had to put some, yeah, some maths into it to figure out. Yeah, can you talk how you did that? And then, yeah, explain the metrics.

So I mean, turn G-force is something we've been after for a few years. It, you know, I think even back when Jamie, uh, first got going, I'm sure it was something you guys were looking at as well, but Um, yeah, we, we try to, to quantify it. So, uh, what we're trying to measure basically is the acceleration in the turn, uh, that is inside. So it's the centripetal force that you experience as you turn across the mountain. It's the force that you balance against to actually do a turn in the first place. So it's, it's super important for skiing. Um, and yeah, we wanted to try and work out a way to measure it. One of our sensors in the insole is an accelerometer. So this measures acceleration, which is, you know, mass times acceleration is force. So if you can measure the acceleration, then you can work out the force. Um, so yeah, basically we always wanted to do it, but if I have a look at the, the insole, uh, yeah, so the sensor basically measures in 3D space. But when you ski, your sensor doesn't just stay flat, it moves with your ski because we're in the boot and the boot moves with the ski as well. So we can measure in a certain axis, but we then— when you turn, that axis changes. So what we, uh, the breakthrough that we were able to do— so one of the team, Sam, she, she did this, made this discovery— was to use the angle of the ski to then isolate the axis that is the horizontal force the acceleration in, in the turn. So by, uh, yeah, basically every step of the, of the turn, we're adjusting for the right edge angle. Then we found suddenly, uh, this metric, the accuracy of, of it and its correlation with ability had increased a whole load.

So, so you don't actually even need the pressure sensors for this one, even though like when people do it, the sense is you're getting thrown to the— like, the best feeling in a ski turn is that G-force feeling.

Yeah, yeah, we don't need the pressure sensors, it's the accelerometer, which is part of the, the motion sensor.

Yeah, amazing.

So yeah, and, and like you said, it's so fun to ski with. I, I tried it out over the summer as well, and, you know, it kind of felt to me like it was correlating how much fun I was having.

Like, that's what I— yeah, yeah, that's what we both agree on.

Yeah, exactly. It's just like you lay over, you do a, um, you know, your best turn you feel the G-force and balancing against it. And, um, yeah, just, it was super addictive to try and push it up, push it up. Um, and yeah, I saw Ted's scores, you know, he can get over 4 G's, which is just crazy.

What did you get?

I could just about get to like 2.5, maybe 2.8. Not— but yeah, I want to try and get to 3, but Ted obviously gets incredible scores there. Yeah, when you think about, you know, what skiing is, you're mainly balancing against your outside ski. And, um, you know, it's kind of like bench pressing 4 times your body weight on one ski when you can, um, you know, withstand that force. Uh, it's super impressive. So yeah, it's a really cool metric. It correlates with how well you're skiing. It's really fun to try and kind of push it up. Um, so yeah, we're really excited about this metric.

Yeah, I am too, yeah.

Yeah, I think like a rocket ship taking off is around 3 to 4 Gs, so you know, when you hear that Ted's getting 4 Gs, he's like going through a rocket launch every turn, it's pretty cool. I think fighter pilots at max are pulling 6 Gs, and then they start to like have problems with blacking out and stuff, so you know, if people are playing with the GeForce monitors, you should keep these numbers in mind, if you're getting this before, it's pretty darn impressive.

But you know how I was mentioning you guys have the unique opportunity to to maybe not just teach people with the skill system. You've, you've now got something like the G-Force sort of thing. And then, you know, if that's what— because the kind of the cool thing is, as we said, like, that correlates with fun factor, which is totally what— why people keep coming back to skiing, like, make it more fun. So if people have kind of a measurement almost of that and see where they're at, then they can go back, work on some of these other finer elements, go back, maybe play with the G-Force things. Hopefully you know, feel a change, see a change in the score. I think it's really cool, like it's very unique in that way.

Yeah, I, I kind of see our history in terms of some big steps of massive like unlocks that have taken so much time and energy. So just making this product work took years, like years in China trying to make it so it doesn't break, so it's sensitive in the right places. Even how we route the cables, we looked pressure maps of, you know, thousands of people's feet in boots from skiing to work out exactly how to route the cables to minimize pressure on the foot. Um, once we had all that, we could develop great data science because now we have accurate data, uh, and then we can develop a metric system that allows people to improve. I think now, and the stuff that we're really excited about that we've done a lot of work on this summer, is building context. So the system up till now has been brilliant on a well-groomed slope where there's quite stable conditions and you can ski and improve and you can kind of work, but the system also kind of doesn't have to adapt to loads of changing terrain. Obviously, we're all skiers, like our whole team are skiers, and we know that, you know, sometimes your best skiing is when the terrain isn't quite so simple. And that's when we also feel like we're doing our best skiing. It's a sport of versatility, skiing. So yeah, we spent a lot of time working on, okay, well, when we have fun as skiers, how can Carv have more of a role in that environment? So we developed really, really smart ways that Charlie and his team have worked on all summer to determine what the terrain is underfoot, what the steepness is, so we can start to adjust for that and make this apply much more to all the variety of contexts that you ski in. And for us, that's where the fun is coming in as well, because as skiers, like, that's kind of the fun part, isn't it? It's as much about terrain as it is about putting in that like one perfect turn.

And that's probably a good like topic to go into just for the people that are listening that, that own Carv and have, and have used it. You know, that's probably one thing they've maybe complained about is, oh, I can only get good scores when it's great snow, this certain pitch. So you're, you're basically saying you guys have now worked on ways to detect that put in almost like a filter or an algorithm to, to kind of show that there's difficulty there so you don't get deducted points because it's rougher snow. It sort of filters it out. Yeah, because I remember when I had— when I sent in some, uh, when I was beta testing some stuff, I had to say what the snow type was like. Was it rough? Was it groomed? That sort of stuff. But yeah, you're also now getting pitch. Yeah, yeah, which is crazy. I mean, that's going to be a That's a big one, right, to get all the angles on all the slopes.

Yeah, I mean, it's pretty amazing, like, what this can add to our dataset. You know, as a coach, for example, when you're coaching, you take that into account when you see a video of someone or you're working on snow. You don't, like, ignore the terrain.

Yeah.

And just coach them based on, you know, the turn in 3D space. I think for us, Carv was a really amazing kind of carving arc measuring device. and if you can hold that quality of turn as you move into harder and harder terrain, you're going to get rewarded for it. And we have these really amazing graphs. Charlie's able to look at our whole population, and we take the really brilliant skiers and the average skiers and the less average, like maybe lower intermediate skiers, and we look at how your SkiIQ deteriorates as the terrain gets harder, because now we can see if the terrain's harder. And it's pretty interesting. It's like almost like your versatility score. So someone like Ted or yourself, you probably find really good consistency. You're still getting a great, like, turn shape in as the terrain gets harder. Maybe someone like me, maybe a bit less polished, I start to see that fall off, like, quite severely. And we actually have this really cool graph where we can see, like, normal skiers, it, like, drops down a lot. Uh, instructors, it drops down.

And this is terrain and snow conditions.

As the terrain's getting harder, the IQ it's dropping more.

Yeah, yeah, yeah.

And then you see someone like Ted and it's pretty much flat, and it's like, I think I remember you showing me that graph.

Yeah, yeah, yeah, yeah.

And so he should be rewarded for that. And so we wanted to make sure that when you are doing something more like awesome, it feels great to you, it's kind of challenging you, and you're skiing well, well, carbs gonna really reward that now. So it should be really fun in different conditions.

Yeah, absolutely. And the Any other new changes coming this winter?

We've always got changes. That's kind of the nice thing is like we've got a team who work all summer and try and pull the product forward. So I think we've got more data that comes in every year. So we've got new metrics coming out. I think those will be a game changer, especially for advanced level skiers who are looking to get the edge in how they carve. The terrain changes are massive and we're also working on a new real-time coaching mode. So With TED mode that Jamie spoke about, we call it TED mode, but Train with an Olympian if you're in America.

Yeah, yeah.

I think with Train with an Olympian, we had this amazing, like, jump forward working with Ted where he wanted to make a very progressive mode that built skill upon skill and it adapted as you skied. And what we're now looking to do is to take how that's working but adapt to the terrain around you. So if you're skiing on a harder pitch, the level is going to adjust to the terrain. So for us, I think if there's one word for the season, it would be context, because we tried to take account of context in almost every feature that we've built. So whether it's how the metrics or the SkiIQ is working, or whether it's how the active coaching, the actual real-time training is working. Carv is trying to adapt now to the terrain around you so it can give you more contextual and more relevant advice. So yeah, so quite a lot being worked on.

Yeah, because you, you have a leaderboard like within Carv, uh, monthly and seasonally.

It can be daily, weekly, and all season, and it does reset.

And resort. Yeah, yeah.

And you can be the king, which is kind of like— I was the number 2 skier in Cortina for quite a long time until about 5 people knocked me down. So I was pretty happy with that for like most of the season. I screenshotted it, don't worry. Um, so yeah, I think, uh, that resort level view is really great because a lot of us ski— like, we know our local resort and we even know some of the people there it's nice to have that, that leaderboard a bit more local.

Yeah. And so you— can I mention that? So that who was king of the leaderboard? Well, there was, there was a skier king of the leaderboard, highest IQ, end of last season.

Yeah.

With some of the changes, maybe filters and algorithm tweaks, that's the stuff. Can you say who actually was— had the highest SkiIQ?

So who had the highest SkiIQ last season? Yeah. After you've done your So it's an instructor called Tyler who was at the top of the leaderboard last time. Um, and then, yeah, we've gone away, we've created 3 new metrics, we've tuned 2 other metrics, um, all kind of, you know, looking at Ted's data in particular and talking to him and yourself and many other instructors. Um, and the real goal for, for SkiIQ and the metrics this season was to make, make it reward more dynamic carving instead of it being kind of getting to a higher jangle, but you know, more of a park-and-ridey turn. So we've worked really hard to, to try and reward this type of skiing. And based off our current analysis, so we're still going through the process of finalising everything, but of all the skiing from last season, Ted now goes to the top of the leaderboard with, with the new version of SkiIQ, which is obviously great. We want the best skiers with the highest skill to be at the top of the leaderboard. But he'll have to set a run as good this year to get to the top of the leaderboard. As Alex says, it resets, and no one's skied with the new number that we're going to release. No one else has skied with it. So once you can hear the new SkiIQ every time you stop, people will be able to kind of optimize for that in particular. And yeah, we'll see who's there at the end of the season.

Yeah, that was— I must say that that, that news was like music to my ears because I just— but it's also a testament to you guys. You, you look at where you're at at the moment and you go, hang on, Ted Ligety, we know how good he is, ski with him, but then he's not sitting at the top of the leaderboard with SkiIQ. Someone else is. What's going on there?

Yeah, tweak it.

So, so then, like you said, so it's not that you want to reward that, it's that you want your system to be truly honest Exactly.

Yeah, so it's helped us tune the actual metrics. So one of our metrics looks at, um, how you move forwards at the start of the turn. And by looking at Ted's data, we could actually see that he was moving forwards so early it was before our window was, was looking for it. So, uh, we were able to extend the window. And then, you know, the— when we develop metrics, we want to make sure that they're based in kind of coaching, uh, literature, that, that, you know, it's what good skiing is, is all about, that we can measure them to be accurate, and then also that they're understandable for, for skiers. And in this case, you know, we knew the, the concept to move forwards at the start of the turn to pressure the, this, the tip at the start, we knew that was the right thing to do from, uh, from a skiing technique perspective. So we're like, why is Ted not getting the highest score here? Or not the highest, but a high score. And then, yeah, looking at the data, you can see he's, he's doing it earlier than we were looking for. So So that's one example of how we've been able to, um, make the system better by, you know, we need the best skiers to be using the system so that we can then make it even better. If you didn't have him there, then, you know, we're not going to be, be looking for that. Yeah.

And constantly improving. And yeah, what is— Jamie, what does— what if you had the, the looking glass into the future, what, what would you kind of love Carv to be, say, in 10 years' time? What, what could it do? What would it look at?

Yeah, I think we're seeing some really interesting, uh, you know, developments in technology to think. I think we've already talked a lot about it in terms of, um, you know, really working across the whole mountain is like something really big for us. So yeah, you know, thinking about when we started Carv, uh, you know, it's called Carv and you carve one piece obviously, and that's obviously a very big part of our skiing. But you know, when a lot of people think about skiing, they also think a lot about the backcountry and, and really just properly getting coaching across all, you know, across the whole mountain is, is our long-term vision. Um, I think in terms of creating like a very powerful coaching experience, there's some really interesting stuff happening at the moment with the large language models. Obviously you've seen ChatGPT, and into— and the thing about a coach and the thing about when you're skiing is you, you want the interaction to be verbal. You, you don't want to have to look at your phone. And so there's some really powerful things that we, you know, obviously in, in the pipeline where we can take this huge data set that we have about the skier, we have a huge data set of all the knowledge of, you know, instruction, we have a huge data set of the ski resort, we know what the weather's like, you know, there's so much data coming in that, that long term, as you, as you kind of factor all of those things in, I, I just think the, the potential, and, and you use the, uh, a large language model or ChatGPT as a, as the interface with great audio recognition and great audio pronunciation with the, with, um, with the models we're seeing, you can just create such a powerful experience. So I think, you know, it's really kind of the combination of all these technologies that are kind of all coming together, and it's really exciting because we're basically at the perfect point, and we've already got the data because we've already got this huge data set of people who've been skiing. Um, and I always look at our data and it's, you know, it's like we're kind of scratching the surface. We have terabytes and terabytes of data on our servers. We're constantly thinking about how we improve things. But the exciting thing is we've kept all that really detailed data. So as we go back into the systems and rewrite our models and get better and better, we can start incorporating those.

When you first started, was AI, these kind of systems, ChatGPT, did you even know? Had you heard about that?

Yeah, so, so my PhD, one of the, one of the parts of it was machine learning. Um, so one of the things I thought we should do, and, you know, and, and we're already doing it, it's really exciting, is really thinking about the feedback loop. So, you know, if you think about what a coach does, first they have to be really effective at understanding what your problem is, and then they have to be really effective at actually giving you the right piece of feedback. And so the interesting thing about what we do is obviously we've talked a lot at the moment about about, you know, thinking about, um, thinking about like what measuring what good is, you know, thinking about with TED and thinking about what, what good is in that context. But the other thing, um, and you can use machine learning to optimize that, and we do a lot of thinking about like how can we optimize the whole system so that we make sure that TED is given the best score and the best people are getting the best scores and the weightings are appropriate.

Yeah.

And then the other side, which we're getting better and better at, which I think is really exciting and is kind of borrowed from education technology, so these big kind of edtech companies like Duolingo, is essentially optimizing the feedback to be the most effective piece of feedback. Because not only are we collecting loads of data about people skiing, we're also giving them all these interventions. So we're saying, you know, lean forward. We're saying, like, next run, focus on tilting your edges at the same time. All of those interventions and the actual language we use is all measured, and we actually see the impact on the customers. So, over time we're creating this very, very powerful database, not just of how people ski, but also how people react to feedback.

Yeah.

And so it's a very, very interesting place to be. I don't think there's any company that's ever done this on a wide scale, is to have a huge data set of what you've actually told people and you actually see how they react to it.

Yeah.

Because you think like, yeah, there's, there's not really anyone trying to like teach technique, uh, across any sports. So that was really like the philosophy of Carv was like, okay, we can, this is, we can get to this really interesting place and collect really interesting data and then hopefully make a very powerful product with it. So yeah, it's very, very exciting to see that kind of coming to fruition. Obviously there's lots more, lots more work to do.

Yeah, yeah.

Um, and yeah, and, and the cool thing is, is that everyone, everyone learns differently, right? And as Jamie says, we've got this, uh, diverse data set with lots of different people. So not only will we find like, oh, this is an effective bit of, bit of feedback, but linking that to the, the type of person potentially, You know, that's another key step that I think will give us a big unlock. Um, but yeah, there's, there's definitely a long way to go. As Alex was saying with the context, like, you know, that's super important. You know, if you're getting a low score but you've, you've done like one of your best runs, but it was just super steep, it was bumpy, uh, you know, we need to work out that that is, um, the, the reason that you had a lower score and either adjust for it or compensate in, in some way. So yeah, we need to we need to make sure that we can link improvement, we can understand what improvement is, which we can, and make sure that's all over the mountain. And then we can, uh, yeah, hopefully find like this optimal learning pathway for, for people.

Yeah, back to like that, the turn g-force one, it's just very comforting to have like a metric that, that directly correlates with a felt sense Now you're adding something to your ski setup. This, uh, insole is going inside the ski boot. And I know you guys, like, the big question probably first of all, so many people, is, is it going to fit? Is it going to ruin, like, my, my boot fit? Can, can you speak to that?

Yeah, definitely. I mean, we have this question a lot, and it's because, look, we get it, the boot is a really precious environment. If you've, if you feel like you've got your boots dialed, well, it's probably taken you a while to get there, and you're really, um, going to be protective about that. I think the first thing to say, you can't sit here, maybe I just tear this sort of low-key lovely stand we've made. It's actually pretty thin. I think most people don't really see the thickness, um, but it's, it's generally around 3mm. And what we've done when we've built it is we've routed some of the circuitry and put, we put anything that we had to put, we put it in places we know that for 99.9% of you there's no issues because we have pressure data from the boot from literally thousands, tens of thousands of people. Um, I'm wearing almost every single boot that's available on the market, but the most important thing is it goes under the liner, so it sits on the plastic boot board. So you take the liner out, you put it in under the liner on that plastic boot board, and it's very neutral. It's not adding volume in different— it's like, it's not like a footbed where it's adding volume on the sides or anything. It's just very, very thin piece of material that sits under your liner. And for a lot of people, even, I dare say, it might even improve boot fit because it creates a little bit of lift. It stops a little bit of that float that might be commonly happening, especially with leisure skiers. And you might actually find you get more control, weirdly, by adding a small bit of volume to the boot. But even if you do have very dialed race boots, you can opt for a slightly thinner insole. In very extreme cases, you can actually shave down a boot board or go to a thinner boot board. And some people, boot fitters, are happy to do this in many cases. But we have almost no issues with people returning it for boot fit. But there's a lot of anxiety up front. And even Ted skied with Carv without any boot mods. For example, if Ted Ligety, you know, 5-time world champ, is able to put it in race-fit boots, I think most of us will be okay. So I think, yeah, you know, we worked really hard on this issue and we— and, you know, in the case that you do have an issue and it just doesn't quite work with your particular setup and you can't get it to work, then we do offer a boot fit guarantee, and you can just send them back to us and we'll issue you a refund. So, you know, we get that this is a really precious issue, and, you know, we've done everything in our power to make sure that it's— if anything, you wouldn't even notice they were there for most skiers.

Yes. Yeah, I would agree with that too, because in— I think unless you have a really high volume foot, and those people would know that, and they've got pressure on the top most people are actually, like you said, are going to benefit. It's going to give a more snug fit inside. And if it's— if it is too snug, you can reduce volume in other places. The boot board can be ground, and there are ways around it. And yeah, I didn't realize, but yeah, basically no one really returns it because of, of that. And no, yeah, so a lot of you—

no, honestly, no. It's kind of like one of those things you're so worried about until you try it, and then I think for almost for most of us skiing in off-the-shelf boots that, you know, it fits with custom molded boots, it works with custom insoles because it goes under the liner. There's actually no problems.

Yep, excellent. So in like kind of wrapping up here, Jamie, if I say I'm the skier that went, this is pretty cool, this sounds interesting, and they go and they get a set of Carv units, they download the app, and they head to the snow they turn on Carv and they do a run, what essentially are they, are they seeing? What's their experience when they then stop and look at the, look at the app?

Yeah, exactly. Um, so, so first thing is, as you finish the run and you're on the, on the chairlift, you're going to get a Ski IQ. So that's our definition of how good that run was. So average is 100, and then it obviously goes all the way up to 150, 160, maybe even 170. Um, and so that's a kind of like first measure of like how well you're skiing. But that's like the very top of the kind of pyramid of, of details. So under that, we separate the skiing technique into 4 different skills. So you have balance and edging, rotary, and pressure. And then we go one step deeper to look at individual metrics that con— individual skiing characteristics that correspond to that skill. So for instance, edging, we've got edge angle, we've got edge similarity. Um, edge similarity is actually a really good example of, of a metric which is really powerful. So say on that ski run you're actually doing a bit of an A-frame, then so, you know, you might have got like 125 SkiIQ, you're putting on that outside ski and you're not bringing them nicely together. And so Carv will pick that up and it'll say your SkiIQ is 125, and then it will go one step deeper into to actually identify your— the problem is your edge similarity. So if you're doing an A-frame, you're probably getting an edge similarity of about, of about 30 or 40 maybe. And what we're going to do is we're going to give you some tips. So you'll get a recommendation. The first one will be just an audio one. So you'll hear, okay, next run, focus on rolling your skis together so that you can have that nice, nice fluid motion and not one ski and then the other, which, which creates the A-frame. And then what you can also do is in the app, you can obviously track your edge similarity every, every run, but you can also get videos on how to actually improve your edge similarity. So the idea is that you have that simple SkiIQ view, then you have a tip, then you have— that tip will be around a metric, so you can actually kind of drill down into what that metric is. You can watch videos on that metric, and then as you start focusing on that metric, you can actually improve it. So good edge similarity, if I'm, if I'm on a good run and I'm like pumped up, I can probably get like 70 or 80. Um, I've seen, uh, I've seen Ted get basically 100, I think. So 100, so his edges are basically moving like perfectly together all the way through the time. Yeah.

Which is hard, like, because it is based, right? It's not just, oh, when one snapshot moment. It is measuring it, like you said, however many seconds or times a second.

Yeah, it's, it's very hard to do.

Yeah.

And it's incredible indicator of skill.

Yeah, it's—

you can't fake that. You've got to be good to get that.

Yeah.

Awesome.

Well, guys, thank you so much for your time. We covered a lot of stuff. I think there's some really exciting things coming coming for, for new users and current users this season. Some great new metrics, changes to the SkiIQ system. The leaderboard, I think, is going to be— you know, I'm going to be watching that closely and, and yeah, see where, see where a couple of runs I put down go this season. Yeah, like, thank you so much, and you know, good luck with, with the future.

Oh, thank you. Yeah, thanks for coming in.

Yeah, it's been a great day.

This episode is brought to you by Carv, the digital ski coach. Now, one of the features I've had the most fun with all time is is simply the fact that Carv tracks every single turn you make. I treat it a bit like Strava, so at the end of the day it tells me exactly how many turns I've done, and those numbers keep adding up across the season. So last season in Australia, I had a few days where I was doing nearly 2,500 turns, and my legs were tired, but it was really cool to try and have that number there motivating me. And I'm quietly working my way towards my 1 millionth turn with Carv.

Love.

So the other metrics I've been diving into lately, and these are new for this season, are very unique, and these are turn corridor and turn radius. And I don't see many skiers using them, but they tell you a lot about the shape and flow of your skiing. So the turn corridor shows you how far you're traveling across the hill, or the time almost between apex to apex, and how long you're taking to transition. While the turn radius measures the actual arc you create once you start the turn So these can be different. You can have like a very wide corridor, say 18 meters, but a lower turn radius, or the other way around. So when you compare the two, it opens up a whole world of experimentation, and it would be a bit like a race coach will set different types of race courses to influence a racer or skier to change the timing and mix it up. So I encourage you to look into this if you've got Carv and play with it and see if you can manipulate it, because it's only going to improve your versatility. So I keep my Carv on my boots every single day because you never know quite know when like a change in technique or a new ski resort or skis will show up in the numbers. And it's great at the end of the day being able to check that, or even after the fact. And I'm really looking forward to the day I open the app and I see turn number 1 million. So if you want to explore your skiing in a completely different way, check out Carv, the digital ski coach. And thank you team for sponsoring this episode.