Waterhouse VC host a discussion about the world of professional betting with Tom Dry, who runs one of the world's leading tennis betting syndicates.
The discussion covers:
How elite betting syndicates find and maintain a profitable edge
The systems and processes involved in daily betting operations
The technologies creating new opportunities in the betting landscape
This webinar was recorded on 26 March 2025.
Watch the full webinar here:
Timestamps
0:00 Introduction
1:19 About Waterhouse VC
2:45 How Tom Dry became interested in betting
3:48 Working for StarLizard
7:55 Starting a betting syndicate
12:54 Going all-in on tennis
16:14 Putting models into practice
18:06 Approach to data
21:25 Example of unique insights used to win
25:32 Placing bets
28:00 Bankroll management
32:07 Maintaining an edge
35:00 Recruitment
38:40 Impact of AI
40:45 Diversifying to other sports
44:45 Book recommendations
46:36 In-play betting
47:50 Market resistance
50:40 Tournament focus
Transcript
Michael Donohue: Welcome to today's webinar on the business of professional betting. My name is Michael Donohue, head of distribution at Waterhouse VC. We've got something pretty special lined up today, I think. A rare look inside professional betting with someone who's built a successful operation from the ground up. We'll go into a few things, like the work that goes into finding an edge, managing risk, building a sustainable business, and scaling up in what is a pretty competitive space. Joining me today are, first, Tom Waterhouse, founder and chief investment officer at Waterhouse VC. Many of you joining today will know Tom's background, but just very briefly, Tom ran a bookmaking operation at the track in the 2000s, built a fast growing online betting business, sold that to William Hill, served as CEO of William Hill Australia, and for the past six years has been at the helm of Waterhouse VC, investing in technology companies within the gambling ecosystem. Welcome, Tom. And also with us today is Tom Dry. He's a special guest today. He runs a professional betting syndicate focused on tennis. Tom cut his teeth working for one of the biggest betting operations in the world before venturing out on his own, and we'll hear more about his journey very shortly. Welcome, Tom Dry.
About Waterhouse VC
Michael: A quick word about Waterhouse VC before we begin. We manage an investment fund focused on what we call the technology layer of the global gambling industry. So, the businesses supplying software, data and services to the gambling operators. It's a segment of the industry with a lot of growth, and it plays to our strengths in being able to source and analyse these opportunities. The fund has delivered returns of over 90 per cent annually before fees for almost six years now, so if you'd like to learn more about that fund, please just visit our website, or you can drop me an email after the webinar. And of course, a quick disclaimer before we get stuck in. Today's session is intended for wholesale investors only. Everything we discuss today should not be considered investment advice, or betting advice for that matter. And I should also mention that Waterhouse VC has a financial interest in Tom Dry's business. If you have any questions during our discussion, please just look at that question tab in the bottom right hand corner of your screen and pop any questions in there during the presentation. I'll either weave them into our conversation during the presentation, or we'll leave them to the end, and we'll have a dedicated time for Q&A.
How did Tom Dry first get interested in betting?
Michael: I'd love to understand a bit more about where you were in your journey, and how you first got interested in betting.
Tom Dry: Thanks, Mike. Yeah, I can't actually pinpoint the first moment when it became more serious for me. I mean, the very first bet I ever struck was probably when I was about nine years old, at Goodwood racecourse, and my parents had probably given me ten or twenty pounds or so to use at the track. We were all there, it was like a family day out, and I remember backing a horse called Ringmoor Down at 8 to 1, which came in. I was nine years old, and I went and collected my money, and it was the best feeling ever. I can't remember really when I stepped it up. After university, I started doing a little bit of betting on football, and at this point I was really struggling to find a job in the real world. And just really by complete chance, and I'd never considered a career in betting, I didn't think any existed beyond working for a bookmaker, my dad found an advert for a London company and betting syndicate called Starlizard. I did a little bit of reading online, there's not much to be found, and I just couldn't believe it existed. At this point, all the other avenues I was pursuing, the regular stuff people would do, like finance, just went out the window. I decided that I wanted to work on edges in betting, and I was really lucky to get a job at Starlizard, and that really kicked it all off.
What was it like working at Starlizard?
Michael: I don't know how much you can say about that job, but was it a great environment to learn the craft?
Tom Dry: Yeah. So, I was there between 2017 and 2019. My job, roughly speaking, was to adjust the model for the particular team-specific and player-specific stuff that affected that particular match. The model is a broad, general thing which estimates team strengths on average, and it's blind to all the random things that can happen in football, such as motivation, team news, all of these things. So my job, and this actually later became very important in my career, was to adjust the model and make sure that the final numbers that Starlizard would use for their betting were very accurate. That's definitely something I've taken with me massively to this day, in not idolising models, and understanding that they're most powerful when they've been sense-checked rigorously, and you've included every single factor and adjustment you can think of. It's that combined process that actually makes money, not just the model.
And sorry, I actually deviated from your question, but the environment was absolutely unbelievable. I think the most unique thing about it, from my perspective, being about 22 when I started, was just how receptive everybody was to new ideas, even if they were terrible and they'd been tried and failed, or they were just non-starters. There was no concept of hierarchy. People were not just trying to get ahead of one another by doing odd things, or sending pretty emails, things like this that you might see in normal corporate jobs. All anyone wanted was just the best for the syndicate, because we were cut into that. And so if someone had a good idea, it doesn't matter if you're 22 or you're Tony himself, that gets heard. And look, most of my ideas for probably the first year were just terrible, but I got a chance to say them out loud, and that mattered a lot. I really valued that experience immensely.
Why did you decide to go out on your own?
Michael: And then how did you make that decision? You're in this pretty good job, working for a pretty good place. How did you make that decision to then go out on your own and bet on your own?
Tom Dry: It was a really tough decision, mainly because, and I don't know how to say it, I worked constantly when I was at Starlizard. I didn't really follow the concept of contracted hours or anything like that, and it's kind of based around the football calendar anyway, so not many people did. But I just loved the job, so I was there all the time. And I had some really good friends there, and guys I'd learned a lot from, who were very kind and fair to me as I was coming up. So just from the emotional side, it was really tough to make that decision. I guess what swung it in the end is that the Starlizard syndicates were having their bets, and I was fortunate enough to be a member of that syndicate and have exposure to those bets, but my own betting, which I was doing on smaller sports like snooker and golf at this time, was ramping up to the point where it was becoming more meaningful for me than the bets at work. And I was making money on these bets. Not inordinate amounts of money, but I was profitable, and I loved that process of finding edges and executing bets. The pendulum swung, I guess is what I'm saying, towards my own betting. And to be honest, I probably made that decision three or four months before I plucked up the courage to tell the guys at Starlizard that that's what I wanted to do. That's how much that job meant to me. And not just my friends, but the Starlizard establishment was also very, very helpful to me in my early days. They stayed in touch, and they were rooting for me and supported me. So I have nothing but positive things to say about that experience.
How Waterhouse VC met Tom Dry
Michael: This might be a good time, Tom Waterhouse, for you to just talk about how you first met Tom Dry.
Tom Waterhouse: Yep. Look, so we look at probably 30 businesses a week, and generally we're looking at businesses that are tech suppliers to the gambling space. But quite a few times we have seen businesses that are looking at finding an edge through data, and think that they can sell that data to bookmakers. Some of them have a small edge, some of them have just had a period of positive variance, and some of them have been con artists. But very rarely, almost never, do we come across anyone that's got a significant edge. Our group of investors are many of the heavyweights in the industry, and one of our industry investors said to us, look, they'd come across what they believed was the best tennis bettor currently in the world, and said, you've got to meet this guy. He's a young guy, ultra impressive, and he's got a real edge. And coming from him, I thought, well, gosh, that's something. So I reached out to Tom and had an initial call with him, and was blown away by the call. But even though my dad's been a professional bettor for 45 years and has never had a losing quarter, it's not my area of expertise. So I talked to my dad about him, I talked to other investors that were professional bettors in the space to get a gauge and see if they'd heard of Tom. I spoke to many bookmakers to get an understanding of, had they heard of him, was he winning, did he consistently win, was he betting at the top of the market, how was he transacting with them. And then, when Tom mentioned that he'd worked for Tony Bloom, and Tony Bloom is obviously an industry legend who owns Brighton football club and is probably the world's best soccer bettor, he very generously sang high praise of Tom and couldn't have been more complimentary. That gave us great confidence, along with the way Tom interacted and the way he handles himself. We were very keen to progress in doing a deal with Tom. And Tom, I'm very pleased to say, you've exceeded all expectations.
Why go all-in on tennis?
Michael: Cheers, Tom. I'm pretty interested, Tom Dry, in just understanding the process. So when you go out on your own, you mentioned you were betting on a couple of sports. What made you go, let's go all in on tennis?
Tom Dry: Yeah. So when I left Starlizard, I kept up my snooker, and I'd been so long building this snooker model. I was really lucky that I approached somebody I'd never met on Twitter, who I've subsequently met in person, who gave me probably the richest snooker data set I've ever seen. I'm going to blow his cover now, but basically ball by ball scoring of snooker frames, and he just gave it to me for nothing, just for the joy of sharing it. And that gave us a massive edge in snooker for a period. At this point I didn't have a particularly large bankroll, and so snooker was the world for me, because I could easily bet up to what I wanted on this one sport and do great. And it got to a point where I wanted to bet more, but the snooker markets couldn't really take it. It's not a massive international sport. It's basically just the UK and China, and you can't really access the Chinese market. And so, tennis was a sport I'd played since I was a kid. I'd grown up with my brother watching Federer lose heartbreaking slam finals to Djokovic, and between us, we were probably watching 80 or 90 per cent of Federer's matches in certain seasons, and we played so much against each other, this is my brother Harry. I had never modelled tennis when I was at Starlizard, but I just had a hunch that if there was a big sport with strong liquidity that I could excel at betting on, then tennis might be it. And tentatively, I kind of learned from what Tony had done, and what I'd realised on snooker, which was that the foundation of a winning model is unique, exclusive data that's been crafted by people who really understand the sport. And that's what I set about trying to build on the tennis side. So it's a bit of a ramble, but I needed to move up the levels. Snooker was not quite it. I continued to bet on it, but I was not going to make a magnificent living on snooker, and so I moved on to tennis.
When do you put a model into practice?
Michael: Yeah, sure. And then, I guess, you decide to move on to tennis. How do you go about deciding when to graduate from, say, paper trading or betting small amounts, to basically implementing the model in full? Do you decide that you need an edge that the market doesn't understand, or multiple edges? What's the level of confidence that you need to get to before you start to do that?
Tom Dry: A good question. I probably proceeded with more haste than your average person. I just felt like I would learn more quickly losing real money than losing imaginary money. And I felt like, look, if I'm betting on Betfair and I'm paying the commission, worst case I might lose at 2 or 3 per cent if I've just got no edge, I'm clueless, and my bets are meaningless and don't have any bearing on what the true probabilities are. That's still a scenario I can live with. And I found that preferable to back-testing, which, to be honest, to do that properly, you need to be extremely rigorous with how you construct your models, and it's very, very easy to think you've got an edge when you don't. So I'd just rather jump in, have probably quite a few bad bets, and look at how the market actually responds to them in live mode, and figure it out from there.
What's the approach to data?
Michael: Yeah, okay. And we've had a few questions come in already. One of them that's pretty common is people asking about data sources. Now, you obviously can't share the secret sauce or anything like that, but are you able to talk more generally about what your approach is? Maybe not even just on tennis, but in general. Are you looking at more sort of player statistics, or is it more alternative things that maybe other people don't think about? Is it more creative than just pure stats?
Tom Dry: Yeah, let me think about how to answer that. I'll put it like this, I don't want to be too specific, but I think a lot of people feel unduly constrained. They would approach this problem thinking, right, what data is out there? We can scrape this site, or we can purchase from this provider, and maybe these providers supply bookmakers, or there are basic statistics on there. And fine, it's nicely formatted, it's JSON, and you can access their API and everything, but it's not really going to help you solve the problem you're working on. You want to work out what you think matters in tennis, or in any sport, and then go about collecting the data to test that hypothesis. So I guess I'm saying it's more of a carte blanche, where you make your own data.
What unique insights feed into the model?
Michael: Yeah, that makes sense. And I know, again, you can't really go into all the specifics and all the secret sauce, but I know, for example, Tom mentioned his dad has been a professional punter for a long time, very data driven, but there are still human elements you take into account that feed into the modelling. For example, you might not want to back a jockey when they're going through a divorce or something like that. Are there human elements that you look to take into account, or even just deep knowledge of the sport that you and the team take into account, and then feed back into the model?
Tom Dry: Yeah. I mean, feeding it back into the model is a little bit of wishful thinking. There are so many things people have mentioned to me where I'm thinking, wow, I'd love to test that, and in practice, the 11 and a half month annual grind of betting tennis obstructs that process. But without doubt, human elements. Look, there are going to be other groups who do great with more expansive models. I rely on a little bit more of a bread and butter model, which I supplement with lots of human elements and expert opinions and my own take. Just to give you one example which you can grab on to. One of my team is a French guy who has what I think of as a kind of crazy infatuation with China. He just absolutely loves China. He thinks the CCP are saviours and the West is completely screwed, and maybe he's right about that. But anyway, he basically packed up his bags and went to Taiwan. He's decent now in Mandarin, so he was based out there while we were working together, and he can access the Chinese internet. And basically, he found, and this was during Covid, so all of the Chinese players hadn't been playing on tour because there'd been ultra stringent lockdowns in China, he found that they were playing sort of shadow matches, Chinese league stuff. I would never have found any of this. And he tells me, okay, look, the top two Chinese players at the time, Wu Yibing and Zhang Zhizhen, they've played each other like 10 times in the Chinese national system, and Wu has beaten Zhang nine times out of 10. They finally emerge after Covid, playing a Challenger tournament in Italy or somewhere, I can't remember exactly, and just by complete luck of the draw, Wu and Zhang are drawn to play in the first round. Zhang is a bit more of an established player, so the bookies had him at probably 2 to 5, in their pricing. And we know that Wu has just owned the guy for the last year. He's literally beaten him nine times out of 10. I can't remember what price we made it, but it's just an example of collecting information from different people, factoring stuff in, having a really great network, and just being lucky to work with so many good people who pass these things along. And Wu did actually win that match in straight sets, I think, so it was a big day for us.
Michael: Yeah. Can you also share that story you mentioned to me, about the guy who plays exceptionally in the wind?
Tom Dry: Well, there are a few players. The French player Richard Gasquet famously said he was born in the wind, and if you actually watch him play, you'll kind of understand, because he has these arcing groundstrokes off both sides, massive net clearance, and he can really manipulate the spin of the ball. So he said he was born in the wind. This was when he was way down in the rankings, sort of one or two years ago. He said if the tour was played in the wind, he'd still be top 10. I think that's somewhat well known amongst tennis gamblers now, but there are lots of funny scenarios like that, where a player is going to be way above or below their regular level because of certain circumstances.
How do you actually get the bets on?
Michael: There's also another factor in running the operation, right? It's not just being able to price and pick what you think the odds are, but it's also getting those bets on. And that's been a question we get. It's probably been the most common question we've got before the webinar. Are you able to just talk about, I guess, challenges when you look to actually implement the model in practice? Are there challenges between how you'd be going about things theoretically versus in practice?
Tom Dry: Let me think. Look, anyone who's done this at a certain level would know that trust is extremely important. If you want to scale a betting operation, you're not going to be able to do that purely from your geographic territory, because you'll saturate the market there too much. So it's important to build a bit of a network, and try and access different liquidity. Tennis is lucky in that respect, because most countries in the world will host tennis tournaments, so that made it very different to something like snooker. If I had to rank my priorities, I'd say number one is A-grade credit settlement. The worst thing is to sweat the bets twice, where you have to hope your players win, and then you also have to hope to get paid. So the top one, by a long way, would be very good friends, or people I've worked with for a long time, who have always done the right thing, and who I rate ethically to the absolute highest level. I'm lucky to work with lots of people like that. I don't know, Mike, could you expand a bit more? I know you were asking about getting on.
How do you manage the bankroll?
Michael: I think that's a good answer, to be honest. How do you think about bankroll management, and how much you would maybe risk on a given match or tournament? Just, how much are you allocating to a bet?
Tom Dry: Yeah. I'd say two years ago, I was betting very aggressively. I thought I had a very big edge, and the market could absorb those bets. Even if that edge was as strong today, and the market could withstand the same level of betting, I still don't think I'd do it like that again, because, I don't know, there's a quote, it's better to sleep well than to eat well. There is just the stress of any given day, or say if there are tournaments in the US or South America, and I don't stay up all night to watch them, and you wake up in the morning and you've lost a massive chunk of your bankroll. It's not even the losing, well, losing is awful, but it's the angst and anxiety around maybe losing it, and checking your phone thinking, man. And then on top of that, you have that regret, like, okay, were these bets even good? And so I've tried to be a bit more conservative in the last few years. I think it's just down to your personal taste to some extent. If you're following Kelly literally, which no serious professional would ever do, then you could end up with enormous swings on a bet. If you make something a 60 per cent true probability and you can back it at evens, Kelly is going to tell you to back 20 per cent of your bankroll, which no sane person would do. You'd be dead within a week, not just because you lose a few bets and you're virtually dead, or ground down to just pips, you'd be dead from the stress. So you've got to manage the whole thing, which is really your life, not just a financial simulation.
Michael: Yeah, sure. There was a question that came through from Brendan. It sounds like you don't let a model completely determine the allocation of your bets.
Tom Dry: I'll just come in on that. I'm much more fluid on things like that. Firstly, it can come down to luck. I could ask a partner for a bet, and maybe because the price is at the right point, or we have some opposition somewhere, we could just get filled for a lot of money, and you don't always have complete control of what's going to come back. Secondly, there's definitely no algorithm dictating this. It could be as simple as me thinking, okay, I've personally watched the last two or three matches of these players and I'm very confident this price is wrong. Obviously the model thinks that, I think that, my team is backing that up. It could depend on, okay, do we have one random piece of information, maybe it's an injury read, where we think we have a bit of a jump on the market here. You're just trying to assess the relative strength of your bet, and also manage the psychology. But yeah, none of those things are at the mercy of an algorithm.
How do you handle downswings and maintain an edge?
Michael: Yeah, okay. That's surprising. I actually had this idea in my head that it was very numbers driven, so it's interesting to learn that. There are obviously periods where you have a bit of a downturn or a downswing in the bankroll, which could be due to purely getting unlucky, or other reasons. How do you determine, when things aren't going your way, whether that is in fact just bad luck, or there's something that needs to be changed?
Tom Dry: Good question. Look, I think there are always things that need to change, even when you're winning. One thing, and the brain trust at Starlizard would probably come for me for this, but I was somewhat frustrated there that they would wait so long, literally tens of thousands of bets, before they'd consider a bias in the model being the issue. The issue was always variance, until basically Tony had had enough. So they were very, very patient and trusting of their models, and look, rightly so. They've delivered unfathomably good results over such a long period of time, and variance is a devastating thing, and it can play with you massively. My approach was always a bit different. If I went on a losing run, it's not like I'd break the model, I'd just be thinking, right, what could I be doing better? Does this make sense? And I would always be going back into it. Obviously, it's more natural to do it when you're losing. That's when you've really got the motivation. And just apply that process of trial and error. I think Taleb said you need at least a thousand IQ points to beat trial and error, so I kind of played by that, and it's got me out of a few holes, probably, that process. And it's not like I'm data mining or something, I'm just reapplying myself. It's important even just to make a few cosmetic changes, just to snap out of the rut psychologically, and just think, okay, I've changed one or two things. Look, nothing's probably wrong, and it's also good to make these checks to reaffirm that nothing is wrong, or possibly there's variance. But yeah, I didn't like losing, and I'd really knuckle down when, as happens every single year at a certain point, the bets don't go your way.
Is it difficult to recruit in this space?
Michael: I'd love to move on to some of the challenges and opportunities before we open up broadly for questions. One of the questions that was really top of mind for me was just hiring in this space. It seems like a very specialised, niche skill set. Is it difficult to hire people to come and work with you?
Tom Dry: Good question. I mean, the team I've assembled, and virtually all of them work on tennis itself, a ton of tennis experts, is absolutely outstanding. There are guys in that team who have taught me unbelievable amounts. It is difficult, but these people, I don't know, do they find you or do you find them? So I posted on a few tennis forums, where people like me hang out, and I think I posted there in about 2019, and just left my email address and a tiny bit about what the job was, and it's evolved over time. So over the years, I've had lots of people reach out with applications, and we'd sit down together on a Zoom call and watch some tennis together and discuss what we're seeing on the stream. I'd ask if they play tennis. Some of them would be college players in the US, some of them would be parents who are coaching one of their children to a very high level. I was really looking for people who either knew the sport at a professional level really well, or understood the game and the structure and fabric of points extremely well. And some of the guys are just mind-boggling. I could name a lot of them. There are obviously a few guys that have worked with me all of this time and really stand out, so a massive shout out to them. It is difficult to hire, because it's something so peculiar and specific, and you can't really train people for what I'm asking them to do. They basically need thousands of hours of tennis experience in the bank. But I'm massively grateful to them, and I'm really pleased that we've lasted this long.
Will AI change professional betting?
Michael: We've got a lot of questions on AI. Is that a development that you see impacting the world of professional betting, for good or bad? Is there any sort of near-term impact to the business, or do you think it's maybe a trend you can take advantage of better than others?
Tom Dry: Sorry, Mike, I'm just thinking about this one. I don't think sports betting is bang in the wheelhouse of AI, as I see it evolving over the next year or so. I think it's slightly pointless to speculate beyond that, because we've already been surprised so much in the last two and a half or three years by what's happened. I guess what I'm getting at is, to bet effectively, you simply need to aggregate information from too many disparate sources, and that particular function is a little bit difficult for AI. You also need some sort of domain knowledge to apply within your modelling. There will be random, spurious results that come out of pieces of analysis, which someone who watches tennis can basically call out and say, look, this may be true statistically in the past, but I wouldn't expect it to carry on because of X, Y and Z. An AI that is just left to optimise on data can fall into that trap and produce a poor model in the out of sample case. But look, you can play this back to me in two years' time, when I'm completely gone. That's just my take at the moment, that it's not quite there. Let's see.
Will you diversify into other sports?
Michael: Yeah, sure. We had a question from James, who said, do you see yourself diversifying into other sports, or is tennis the main focus for the foreseeable future?
Tom Dry: Someone asked me at dinner today about cricket, and they were saying, oh, you could do the same thing on cricket. And I said to him, my top score as a cricketer, as a batsman, was 13. That's 13 runs, my best ever score at the crease. So I couldn't tell you, watching a cricket match, if the guy has top edged it for six or he's just middled it for six. I don't know. So look, the only other sport that I have played to a good enough standard to interpret properly is football, and I think we all know how challenging that market is, and I have no intention of getting involved anytime soon. Look, there's a line in a J. Cole song, he's a rapper from the US, where he's basically saying he's not just trying to survive, he's so ambitious that his goal is to buy his mum the best cars ever. I'm not like that. I just want to survive, and tennis is more than enough for me at the moment.
Michael: We've got a related question from Kingsley. He asked, do you think the same modelling process would be successful on horse racing, or do you put your success down to deep knowledge of tennis?
Tom Dry: Yeah, I think he's guessed from what I've said already. The domain knowledge is extremely important. Just to expand on that a bit, there came a point in 2022 when I started betting on the women's game. So 2020 and 2021 was purely the men's, because that's what I'd been watching growing up, and all the players were very familiar to me. I started betting on the women's, and it was a brutal experience, actually, for a few months. I've never really looked back to see just how bad those bets were, but I didn't have that domain knowledge to help me filter out bad spots and correct the model. And I had guys who'd worked with me for years coming to me going, Tom, I'm not doing the women's anymore. And that's because the domain knowledge was non-existent. Now, going back to where we were at the very beginning, I said I'd rather just lose money and work it out than paper trade and pretend I'm a genius by tweaking all the parameters and making money on a spreadsheet. So in the end, the way out was just to stream the WTA every hour of the day and just absorb it, suck it up, listen to my team. It took me maybe a few years, but I think it's at the same level as the men's now. Domain knowledge is crucial.
What are the best books for aspiring sports bettors?
Michael: We'll go into the Q&A part now, and we've got a ton of questions. We probably won't get time for them all, but I'll just start throwing them at you, Tom, and feel free to pass if it's not something you can answer, and we'll move on to the next one. We got a question from Jack, who asked, what are the best books or podcasts for aspiring sports bettors? Any recommendations?
Tom Dry: Great question. I didn't see this one coming, sorry, Jack, let me think. Well, the book which changed everything for me, and it doesn't feed into betting exactly, but it's Taleb's The Black Swan. I don't know how to really relate how important that book was to me. There's a quote, I can't remember who said it, and it's not in the book itself, that the purpose of books or literature is that they're like an ice axe to break the frozen sea within us. And Taleb's book, I was reading this at like 19 years old, and it was a revelation, that there was so much nonsense out there peddled as expertise. And you may think, what's that relevant to betting? He'd worked as a trader, there are a ton of very interesting anecdotes, and obviously it's kind of a combination of philosophy and the study of uncertainty and things like that. It doesn't map perfectly onto betting, definitely not, but it changed an awful lot for me. Nate Silver has also written a book, which I read when I graduated. It's quite funny, I probably never finished a history book at university, and then one day, after my last exam, I finished Nate Silver's book in one sitting. I was very green at this point, but I just wolfed it down. It was really interesting. So just start with those two, and keep reading.
Do you bet in-play?
Michael: I've got a question from Warren, who's asking if you bet in-play.
Tom Dry: Yeah, we don't have any in-play models. Basically, all of our reporting and data we are creating based on video downloads and streams that the team will assess after the match has finished, and before the player plays again. So we don't have any architecture where we're piping this data in live. Look, we could bet live based on the same feeds that other people use, but we'd be without our biggest asset, which is our data. There's a little bit I do in Grand Slams, where, honestly, it's just for fun, especially in the men's, where you get these five set matches. I'm honestly not even sure if I have an edge, but I really enjoy doing it. It's probably 1 or 2 per cent of our volume.
When you meet market resistance, do you adjust your ratings?
Michael: We got a question from Shane, who is asking, when you meet resistance in the market, do you adjust your ratings?
Tom Dry: Oh, great question. So, the way I understand ratings is some number, in sort of Tom space, that represents how good a player is in different scenarios. I wouldn't directly adjust those ratings, but what it might mean is there's something about this one particular match that I've misevaluated, and it would definitely give me pause for thought with respect to that particular bet. It depends how you've fared against that particular strand of resistance historically. It depends how predictable you think that resistance is. If you're expecting the market not to know something, then it's very normal to get that resistance, because they're not going to know something and you do. If you think it's a fairly standard scenario, and someone is betting violently against you, then probably you're the person who doesn't know something, and you should back off, or not necessarily stand down, but retreat to take a more favourable number. Personally, for me, it also just depends on confidence. Betting in first rounds can be different to betting in, say, quarterfinals, when you've seen that player play three or four matches already in the tournament, and the chance of some completely exogenous information is much less. But you'd be an absolute fool to plough headlong into resistance, because everyone's out here trying to make money, and there's a very good chance that, if you're not careful, they'll be making it out of you.
Which tournaments do you focus on?
Michael: I've had a few questions on where you allocate your time and your bets, in regards to, say, ITF, Challenger, ATP. Is there any space that you focus on, or is it a bit of everything?
Tom Dry: Yeah. This season has been almost all ATP and WTA. We used to do a lot more Challengers, but I just found it's a better use of my time to have the most thorough process possible for those ATP and WTA bets, and pick off Challenger bets much more selectively. And then it's very hard to bet on ITF, because there's just no serious market for that. Bookmakers are not going to get two-way action, so they're not offering any limits, and there are probably integrity issues. So this season, more so than all the others, it's been really concentrated at the top levels. Yeah, that sort of answers the question.
Michael: Yeah, that's great. Look, we might wrap it there. We've had unlimited questions, which I think we can take as a compliment that what you're doing is very interesting, Tom, but we'll let you go back to it. It's pretty late over there now. So, just on behalf of everyone joining today, thanks so much for sharing your insights, and also, on behalf of the investors, thanks for all your hard work. We're very pleased to have a small share in your hard work and your success. Really, thank you, Tom, for joining.
Tom Dry: Cheers, guys.