Value Driven Data Science: Boost your impact. Earn what you’re worth. Rewrite your career algorithm.

[00:00:00] Dr Genevieve Hayes: Hello and welcome to Value Driven Data Science, the podcast that helps data scientists transform from technical implementers to strategic experts by contributing to the decisions that really matter. I'm Dr. Genevieve Hayes. Today I'm joined by Professor Steve Stern. Steve is a professor of data Science at Bonds University on the Gold Coast and is the official custodian of the Duckworth Lewis Stern Cricket Scoring System, which he also revised.
[00:00:31] In this episode, we'll explore exactly what's involved in transforming technical data science solutions. Into globally adopted standards. So get ready to boost your impact, shape decisions, and own your expertise. Steve, welcome to the show.
[00:00:48] Prof Steve Stern: Thank you very much. Thank you for having me. Genevieve.
[00:00:50] Dr Genevieve Hayes: Back in my PhD days at the Australian National University, I dreamt of making an impact on the world through my work, but struggled to imagine how that was even possible from a city as small and remote as Canberra down the hall from me, one of my PhD supervisors, a statistician, had recently started analyzing cricket data.
[00:01:12] Although this was without a doubt the most interesting use case being explored by any of my colleagues getting an international sports body like the ICC to notice this work seemed like an impossible dream. Fast forward over a decade to about a month ago when I was doing a newspaper quiz with my family and up popped a question about the Duckworth Lewis Stern Cricket scoring system.
[00:01:40] Hey, is that the stern you used to work with? Ask my mom. I was very proud to be able to say yes. Yes it is. But in the days following that conversation, I found myself wondering how does a statistician working on a technical problem in a small city in Australia achieve the seemingly impossible goal of having their solution adopted by sports governing bodies around the world?
[00:02:07] And today I get to find out. So let's start at the beginning. What exactly is the Duckworth Lewis Stern or DLS method beyond just a cricket scoring system? And how did you first get involved with it?
[00:02:22] Prof Steve Stern: That's a great question. So, the system itself is designed to deal with a very specific problem, which at least as far as I'm aware, is relatively unique. In sports, and that is that cricket is not a dynamic back and forth sport like most in terms of you know, soccer or rugby or a FL where.
[00:02:43] The offense and the defense change throughout the match dynamically. And Cricket one team bats and scores as many runs as they can. And then the other team comes and tries to score more than them. So there's only one offensive and defensive period. And because of that, if the game is affected by rain.
[00:03:01] It's typically affected asymmetrically for the two sides, if a soccer game is affected by rain or an a FL game is affected by rain, you'd normally just say, well, look, the rest of the time the teams would've probably had the ball equally anyway, and you don't need to readjust for that.
[00:03:15] But in cricket it's crucial because, if one team has already finished their entire scoring innings and then rain cuts the other team's innings in half. Well then obviously you have to adjust for that somehow. So it's a sports problem, but it's also a fundamental mathematical problem.
[00:03:31] And I think that's the unique aspect of it and it's what got me involved. I mean, When I first came to Australia, one of the things that is the most common thing to do to Americans who immigrate to Australia is to see if you can make a fool out of them. By making them try and understand all the esoteric laws of cricket.
[00:03:49] But I am pretty pedantic when it comes to these things. And so they couldn't do that. And a lot of what happened to me is incredibly serendipitous. It just so happened that I moved to Australia in 1994.
[00:04:01] Which was just when the DL system was becoming known. And so when my friends discovered they couldn't, trick me into silly things about actual cricket rules, they said, well then you need to tell us what this new thing is. 'cause even we don't understand what the heck it is. And so it interested me because I've always been a sports fanatic, but as I said it actually seemed to me to be a real mathematical problem.
[00:04:24] As well as a sports issue. And so I started looking into it and it's interesting you point out, I didn't have the idea at the time that this was going to catapult me somehow into the international spotlight. It was more just that it was a fascinating problem to work on.
[00:04:41] And it was a problem that was different than a lot of other problems. So it meant that there was a lot of room to move. And to try different things.
[00:04:50] Dr Genevieve Hayes: So I'm guessing that Duckworth Lewis and the cricket governing bodies weren't initially aware of your work.
[00:04:57] Prof Steve Stern: No. So when I first started working they absolutely weren't. But the first paper that I prepared which was a pretty typical mathematician's view of another mathematician's work, which is, that's great, but I can do it better, and I decided that it would just be polite to send a copy to Duckworth and Lewis in telling them that, I was planning on sending this for publication and just to let you know, and, never really expecting anything other than a Well, thanks for letting us know.
[00:05:24] Good luck. You know, the typical academic response to these things. But I'll never forget for the rest of my life, I sent the thing off to Frank Duckworth on one day and I got an email response the very next day, which said, thank you for your paper.
[00:05:40] My wife Jeanie and I are just about to take a trip to the US to visit the Grand Canyon, and I've been looking for something to read on the plane. So Frank Duckworth is the most generous academic I've ever met in my life. So he's just recently passed, unfortunately I miss him every day.
[00:05:57] We hardly ever met in person, but we used to communicate by email at least once a week. And so I've taken it as sort of, a badge of honor to continue his tradition. Every time someone writes to me about DLS and a lot of people do I do my absolute level best to respond to everyone, and then I'm looking for the next person.
[00:06:15] Because essentially what happened was obviously Duckworth and Lewis didn't agree completely with what I was doing. And interestingly enough, and this is probably a good thing to keep in mind, their response was, look, we don't disagree with you mathematically, but remember, this has to be something that's a practical tool as well as a mathematical exercise.
[00:06:34] And my initial approaches, they felt, and I now have come to agree with them, were just a little bit too complicated to actually be implemented. As a rule in an international sport. But it led to a back and forth where, there was just many, as I said, serendipitous things.
[00:06:51] And so I think Frank and Tony became aware of me about two or three years after I had started, 'cause that was about how long it took me to get the data involved and write and come up with the idea. And then in almost a year or so after that, the ICC became aware of me because Frank and Tony.
[00:07:06] Who were of an older generation than I was, they had written a program to implement their method, but it was written for dos, if you can remember the age of dos and the ICC, it was just moving along with the rest of the world into the Windows era.
[00:07:22] And Frank and Tony didn't have the expertise to write a program and the ICC who claimed they did kept peppering Frank and Tony with questions that. Made it clear to them that they probably didn't, most of their IT people were actually web designers, not actual programmers. So I wrote a program in Java, which is now the program that's used.
[00:07:41] And I, was initially seen as just the guy who was the technical support for Duckworth and Lewis, but because the program has to be updated. Every time they update the system, based on changes in the pattern of play in cricket, I got to be involved in all of their discussions because I then had to implement them.
[00:08:01] So I was always sort of being carried along, along the way, and then they decided to retire. I mean, they were both in their mid seventies. In the 2000 tens. And they decided that they had had enough, they were tired and someone else had to take over. And so, I didn't set out to do that on the journey, but it just sort of happened.
[00:08:20] And I'm absolutely ecstatic that it did, obviously.
[00:08:23] Dr Genevieve Hayes: What I find fascinating about that is by the time you got to this point where you were doing this work, you must have been at least an associate professor, if not a professor, and this reminds me of the time that I spent as a research assistant early in my career,
[00:08:37] Prof Steve Stern: Yeah, look, I talk to people all the time these days about, oh, I wanna be , a sports analytics specialist like you. Should I do? And I tell them, look, the first thing that you have to do is you have to be a good analyst, there's no substitute for understanding your discipline.
[00:08:53] And so you're right. I think I started thinking about the DL stuff when I was just applying for promotion to associate. But I didn't become the custodian until I was a full professor. And I guess I have to be honest and tell you that there were points in my academic career where I do believe it was a mark against me when I was applying for promotions.
[00:09:15] 'cause people thought, well, that's not true academic work. Fortunately, I've had the ability now to not only. Become a professor. I'm now on bonds promotions committee. It's one of the things that I asked specifically to be on because I think that, there's more than one way to do academic work and the traditional Ivory Tower way is, there's not, have nothing against it, but it's not the only way that people can make impact.
[00:09:41] And I think that's the true measure of whether you're doing research that's important or not, is is it impacting the world? And, I guess you can argue that fixing a problem in the sport of cricket is not exactly curing cancer, but it is making impact in the world, that's for sure.
[00:09:55] Dr Genevieve Hayes: Oh you have your own Wikipedia page.
[00:09:57] Prof Steve Stern: Yeah. My kids still are amazed by that. They thought I wrote it, by the way, but you're not allowed to write your own Wikipedia page.
[00:10:05] Dr Genevieve Hayes: Oh, okay. So obviously sending this paper to Duckworth and Lewis must have helped you build credibility and trust in their eyes. How did you go about building credibility and trust in the eyes of the cricket boards and also in the eyes of the cricket supporters who are probably harder to convince?
[00:10:25] Prof Steve Stern: They are. And look, it's still hard to build that credibility for people who are new to the system. I guess the first thing that I'd have to say is that I had the benefit of Duckworth and Lewis already having laid the groundwork. For mathematics being a proper solution to this problem.
[00:10:42] And so they, had many, many discussions with the International Cricket Council and also the Cricket Board in England, and learned lessons, which they passed onto me about the way in which you have to approach these things. The most crucial aspect, is you have to build that trust and that trust comes from listening.
[00:11:01] They're going to have questions and some of them, you're going to think yourself but that's completely off the point of what's happening here. But that can't be your attitude, you have to try and bring them along and make them realize that the approach that you're bringing is one that actually is, well, I guess the thing I'd say the most is all the previous efforts to solve this problem before Duckworth and Lewis were very anecdotal based. In other words, they would sort of take a certain scenario. Had already occurred and said, this is how you would fix that problem. And that is exactly what caused the major catastrophe that led to Duckworth and Lewis starting their work, which is the rule at the time in the 1992 World Cup was devised because 90% of the time, up to that point when rain came and disturbed a match, it happened.
[00:11:54] Right in between the middle of the two innings. So you'd have one team which didn't have to deal with the rain, and another team, which did. Right from the start of their innings. And so the method that was used at the time was perfect for that. And indeed, interestingly enough, the World Cup where the match that caused all the controversy occurred, the method that was used was used twice before earlier in the tournament without anybody having any concerns.
[00:12:18] Because in both those instances, the rain occurred during the lunch break. But it just so happened that in this case, the rain occurred. At the very end of the match and they had just never envisioned what to do in those circumstances. And so the thing I think as a mathematician, I was able to do, and also not just as a mathematician, but as an educator, so, understanding how people learn things and how they take things in is to show them that the system we were building was holistic.
[00:12:49] It wasn't just designed. To fix specific problems that had occurred. It took the problem as a holistic thing and said, here's the general way to fix it. That's pretty crucial, I think, and to give them a perspective to hang onto. To sort of say, these are the principles. In fact, probably the paper that I think is the most influential that I've written about this is the most recent one from about 2016, which is essentially a paper about what are the principles, forget that DLS exists or that there might be competitors or whatever.
[00:13:19] What would the principles of any legitimate rain rule need to be? And so I wrote out four core principles, and I think that's the thing which has, impressed them the most is you're absolutely right. They still probably don't understand the details, but once they saw the principles, they understood the application.
[00:13:39] The analogy I always try and use is you can go out every morning and go to work by driving your car. You don't have to understand how the engine works. For you to drive your car. Now, if your car breaks down, you need a mechanic, you have to go to somebody who understands the engine, that's fine, but you don't have to understand the details.
[00:13:57] And so to understand how to implement DLS and use it appropriately, you don't have to understand the mathematics that built it. You just have to understand the basic principles on which it rests. And that, I think has stood me in good stead throughout, three different.
[00:14:14] Regimes at the ICC in terms of different CEOs and different heads of the, cricket Competition Committee and so forth. And most of the time when I get emails from people and of course most of the time they start off with How dare you make my team lose?
[00:14:29] And so you have to have thick skin and you have to kind of say, look, no, no, no. This is the principle that went on. And get them to see through their own preconceptions. About what the model is actually supposed to be doing. The most common type of critique that I get is that the model doesn't account for something that people think that it should.
[00:14:50] One of the more controversial matches that occurred was a match in which New Zealand and South Africa were playing each other. And Ab DERs, who was the best batter in the world at the time was batting when rain came. And so all of the South Africans said, well, but you don't account for the fact that we have the best batter in the world.
[00:15:07] So the rain stopped the best player from having his time to bat and, and look, I admitted readily to them that that's absolutely true. But at the same time, you can't have a rule which is built around personalities, you can't say. The rule is one thing. If ab DERs is batting, but it's another thing of anybody else's, it's not the way sport works.
[00:15:25] So, there still is an element of luck involved, an element of randomness I often say that rain ruins, cricket matches DLS saves some of them, but not all of them, but. If the match was ruined, it was the re's fault. It's just the DLS couldn't save it.
[00:15:44] Right. So,
[00:15:45] Dr Genevieve Hayes: But the rain does not have an email address. You do.
[00:15:48] Prof Steve Stern: very true. That is very true. Although I did once get an email from someone who said, do I get royalties when it rains? And I think my response to them was from who?
[00:16:00] Dr Genevieve Hayes: Yeah. One thing I really like about this principles approach is it reminds me of the principles of fitting machine learning models, because what you described, pred Duckworth Lewis, was basically an overfitting approach. You take the specific situations and fit to those specific situations, whereas this principles based approach that you've described, it's creating a.
[00:16:27] Set of general rules, which is what you're trying to do when you're fitting a statistical model. Anyway.
[00:16:32] Prof Steve Stern: Exactly right. Exactly right. And interestingly, so the rule that was in place as I said during that famous, 1992 Match itself was sort of a conglomeration of previous ideas with minor modifications here and there. And a very good friend of mine at the ICC described it to me in a wonderful way, which I agree, which is what he called error whack-a-mole. So every time your model doesn't do something right. You tag on a little bit, and, there's nothing wrong with that as an idea to start with, but after you've done it for the sixth or seventh time, you've created this Frankenstein's monster. And you really should start from scratch,
[00:17:11] and the fundamental. Aspect of Duckworth and Lewis is they said, look, let's stop all this individualizing. Let's get some principles and let's create a holistic approach to this. And you're right, that's absolutely the way that you have to do statistical models.
[00:17:25] And indeed, one of the disappointments I have with some of the more modern machine learning, approaches is that they don't do that. They've developed models and then they've seen that they haven't done quite what they want. And rather than going back on a principal's basis and going, well, what are causing these problems?
[00:17:41] They've tried to fix them with front end, little bits and pieces. All of these chatbots, with generative ai, they fix their problems by secretly hiding bits into the prompt that tell the thing. Don't do this and don't do that. And that might work if there's only two or three things you don't want it to do.
[00:17:58] But as soon as the seventh or eighth thing comes along, it just doesn't cope with that. And someone needs to go back and say, look, why isn't the model doing what we want it to do? Is there some fundamental component that we are missing? And I think the answer to that is an absolute yes.
[00:18:13] I think that most of the large language models. That these things are based on, have an optimization focus, which is different than the focus that they're being employed for. And that's the problem if you fit a logistic regression. You know that the technique was designed to minimize the error between your predictions and the outcomes, but the large language models are just designed to make an output, which sounds realistic. It's not designed to be truthful or to be logical or any of those things. So there's something of a disconnect. I think the current term they have in the field is alignment. So there's something misaligned about the optimization focus.
[00:18:52] The models are using to come up with their outputs and the focus of the users who want a certain thing from it.
[00:19:01] Dr Genevieve Hayes: What are you optimizing for? And what decision are you trying to solve?
[00:19:04] Prof Steve Stern: Exactly right. And I think that's why sometimes these things will give answers that are very strange. And my personal belief is that most of those chatbots were released. With an attempt to monetize them about a year, maybe two or three too early.
[00:19:20] And the evidence I give is that, when they were first put out and they were then tested by journalists and various other people, and they said weird things. I think one journalist said that he was having a conversation with it and all of a sudden it told him that it should leave his wife and come and live with it.
[00:19:34] I mean, you can understand that these things aren't perfect, so they're going to make mistakes. The thing which struck me about that scenario though, is that the journalist said, he wrote back to the company that had produced this chatbot and said, why do you think it did this? And their response was, we just don't know. Now, if that's the case, my response is then it's too early to put that out to market, if you don't understand why your thing is going to make. Rather strange and notable errors, you don't understand it well enough. And I think you're right.
[00:20:01] It's because the principles involved in this have been lost in the way they're fixing it because it's a quick fix to monetize. Is this error whack-a-mole? I've read some important I think work, which indicates that in some sense we've gone about as far as we can towards artificial intelligence with large language models.
[00:20:21] If we want to get a real artificial intelligence, something that is actually being logical and is optimizing towards being true, we're gonna have to use something different. We have to start from scratch. Large language models were an interesting experiment. They're very fascinating things they can do a lot of really interesting things in terms of organizing existing information.
[00:20:41] But in terms of creating new content they're not as reliable as they could be and probably never will be as reliable as we want them to be. There needs to be a new direction. Now, what that new direction is, is anyone's guess, and I think that's a disappointment as well, is that because all of the time spent.
[00:20:59] Monetizing the current things is time that could have been spent coming up with a new approach, but as far as I'm aware, there's not been a lot of new research beyond. Large language models. They're essentially creating ensembles, instead of one large language model, they now put seven or eight together,
[00:21:16] but at their core, they're still just very deep neural networks and so we haven't made any major structural advances. Since then and so I don't think we're going to get much further than we are at the moment at you.
[00:21:31] Dr Genevieve Hayes: Returning back to Duckworth Lewis Stern. What I'm really interested in is you mentioned the principles that underpinned it. Can you give us an example of some of those principles?
[00:21:41] Prof Steve Stern: Sure. So they break themselves into two categories. They're sort of structural. And then and this, I think is pretty crucial again along with your query about creating that trust. With the ICC is that you need to have principles which maintain the integrity of the game.
[00:21:56] There are certain things about cricket that are fundamental to its ethos, if you will, and so the first two are basically rules which I think anybody would come up with, but they still need to be stated. And that is that if you are batting and scoring runs, then if you score more runs.
[00:22:16] You shouldn't be penalized for that, so if you compare two situations. And they're identical except one team scored 10 more runs. Well then the target for their opponent has to be higher. And you'd think, well, who would ever create a system that didn't adhere to that?
[00:22:30] And you would be right. No one does so intentionally, but this is the point of holistic perspective as opposed to anecdotal. If you create a model and you don't think about what it does at the extremes, and I suppose it's based on polynomials. Well, as we know, polynomials get very wild at the edges and.
[00:22:46] Who knows what they'll do and so the second one is the same, if you're a bowling team, you should not be penalized for taking more wickets, you shouldn't be asking the team to intentionally drop catches but then the more sort of.
[00:22:58] Cricket based ones. The one that I like the most is and it's interesting 'cause it mathematically has some fascinating implications and that is that in a cricket match without any interruptions, the team who scores the most runs wins. And nobody asks the question, when did they score those runs?
[00:23:15] Did they score them early in their innings or late in their innings? Did they score them with five wickets down, or only two? So no one comes off the field and says, well, yes, you scored 201 runs and we only scored 200, but you scored all your runs at the beginning, and that doesn't count as much,
[00:23:28] so now that may not seem like a big deal, but it's fundamentally important to the nature of cricket, and it translates itself in a rain delay in the following way. You can't have a situation where a team scores a particular number of runs, but whether they win or not depends on the proportion of runs they scored before the interruption versus after the interruption,
[00:23:49] because that's suddenly giving preference to runs scored. When during an innings. And interestingly enough, and I won't bore your listeners with the details of the mathematics, that rules out what they call probabilistic methods. And so one of the most I think, intuitive approaches to the problem, which turns out not to work, is to say, well, why can't you say when the rain came each team had a percentage chance of winning and then when they restart, we want the percentage chance of winning to stay the same.
[00:24:18] Certainly sounds a reasonable principle, but there are two fundamental problems with it. One is that, well, if one side is washed out completely, well one team has to go to a hundred percent and the other team has to go to 0%, and the other is, as I said, it immediately implies that whether you win or lose depends on how many runs you had before and after, because the percentage chance of winning when rain comes, depends on how many runs you currently have.
[00:24:41] So those are the sorts of principles that I came up with. And it just so happens that as far as I'm aware, DLS is the only major attempt at the problem that satisfies all those fundamental principles, which is I think why it's had its longevity and why when people ask me, well, why did it do this?
[00:24:58] Or Why did it do that? I can always give them an answer. And I think that's another crucial bit, that trust that I have with the ICC at the moment it didn't happen overnight, they kind of took me on and gave me the benefit of the doubt on Duckworth and Lewis' recommendation.
[00:25:14] But slowly but surely. They see you helping them answer questions when people have queries and they start to be able to trust in that, well, yeah, if there's a problem, we'll just write to Steve and he'll give us an answer.
[00:25:25] Even sometimes if they're not completely convinced by my answer, at least there's always an answer which has a logical framing. And, they talk to me about all kinds of different scenarios
[00:25:36] some people think that rather than just change the number of runs a team has to score after rain, they should also lose a number of wickets. But again, I take on board the early conversations I had with Frank and Tony in that regard, and that it's conceivable that you could do that, but it's just too complicated to implement,
[00:25:55] especially if there's multiple interruptions during the game. And so principles have to be the thing that you base your approach on. But yeah, it's not just a principle of sports analytics. And that's why I say to people, look, become an analyst. Understand the principles of analytics before you worry about applying it to one or another of the situations.
[00:26:14] This is true of not just sports. If you wanna be a. Health informatics specialist. You need to know a bit about the workings of the health industry and understand a bit about human physiology. But before any of that, you need to be a good data analyst and understand the principles of what makes a good model and what makes a bad model and how to decide whether the circumstances you're in are appropriate for one type of model or another.
[00:26:39] Dr Genevieve Hayes: Yeah, and what I'm hearing is, so the. DLS system. Your approach was to basically map all of the mathematics that underpins it to these fundamental principles, which are things that can be expressed in regular language and that anyone can understand even without a statistics PhD. And that's what builds trust in it, as opposed to the Cricket GPT approach where you put in the information.
[00:27:07] Cricket, GPT spits out an answer and no one has a clue how it got there.
[00:27:11] Prof Steve Stern: that's exactly right. That's exactly right. I think there are places in the world where black boxes are becoming more acceptable than they used to be, but they shouldn't be. And so you're right, the fact that it's not a black box and that it is based on fundamental principles.
[00:27:26] Is absolutely crucial. Absolutely crucial. I guess the last thing then to put there is the principles are set on top of a databased analysis the principles only define the general structure of what a rule should be, but then it has to be tailored to a particular situation.
[00:27:43] Really, the s in DLS came about. Primarily because when Duckworth and Lewis were first introducing their system in the mid nineties, scores in cricket matches were rarely more than 250. And so while Duckworth and Lewis probably realized, in fact, I think later on in discussions with them, they were cognizant that things were probably going to have to change.
[00:28:07] If the scores ever got to 300 or 350, there just was no data on which to base. A model that would be realistic so they had the principles, but they didn't have the data. And again, serendipity, I came along and really started getting heavily involved at the same time as T 20 crickets started taking off.
[00:28:27] And, the teams started to realize, we don't have to finish. With like five wickets, we could be nine wickets down. As long as we scored lots of runs, nobody's going to care. And so they realized that, they could score a lot more runs if they use different strategies in their approach to batting.
[00:28:44] And so as Frank and Tony were winding down and I was more and more taking over, suddenly there also became this reservoir of very high scoring data matches. That allowed me to now have an actual dataset on which to apply those principles and see where the extensions to the base dl, which worked really well for 2 42 50 had to be slightly modified to deal with 3 40, 3 50.
[00:29:13] And people ask me all the time, so when is the next. Big change, one thing that people don't realize, I think is, I reanalyze data every 12 months to see if there needs to be slight tweaks to the parameters of the model. Usually 12 months isn't a big enough time, but after about 24
[00:29:30] so there's a tweak every couple of years. But structurally speaking, over a 30 year period, there's only been three well, two structural changes. There was the original one. And then two structural changes. And so the question is will there need to be a third structural change? And I can't answer that question definitively, but my response would be if there is a sufficient change to the way cricket is played, if someone discovers a new strategy that somehow allows teams to score 500 runs, then probably there'll have to be a structural change in the sense that the model will have to be.
[00:30:06] Its flexibility will have to be increased to cope with those wider range of extrapolations.
[00:30:12] Dr Genevieve Hayes: The same with all statistics. If the world changes, the model has to change.
[00:30:15] Prof Steve Stern: Exactly right. Exactly right. So in that regard. Sports' no different than anything else. It really isn't.
[00:30:20] Dr Genevieve Hayes: So if data scientists listening today could focus on just one thing to increase the chances of their work achieving real world impact, what should it be?
[00:30:29] Prof Steve Stern: Well, I think that idea that your work needs to be based on principles and that you should start by picking a problem that you think is really important. And it doesn't have to be important to the world. It just has to be important to you. Most of those things end up translating if you have enough passion in them one way or another.
[00:30:48] Not always, of course, but usually, so I, it's really sort of one of these weird chicken and egg things. If you go in with the attitude that I'm going to change the world, you probably won't. But if you go in thinking, I just want to solve this problem, I'm really intrigued by this problem. And you come up with a solid base solution to that problem, that's your best chance.
[00:31:11] There has to be a bit of luck involved in other people saying, yep, we agree. That's a good problem to solve. And yes, your way is really interesting. Duckworth Lewis and Stern has not been unaided by other people.
[00:31:22] We've had other, colleagues give useful critiques and comments. Probably the most renowned of them is a man by the name of David Kendricks, who is the current ICC. On staff statistician. And he's the man who created their ranking system. Although interestingly enough for you Genevieve, he is by trade an actuary. So.
[00:31:44] Dr Genevieve Hayes: We're infiltrating everything.
[00:31:45] Prof Steve Stern: Yeah. But I think it's not one of those things where, you say, oh, I'm going to work hard enough and I'll change the world. No, you have to start with a slightly more centralized focus, and it's interesting 'cause there are other places where that's not the case,
[00:31:59] if you wanna be a professional athlete, generally speaking, you just put your heart, mind, and soul into it and believe. And, if you work hard enough and you do have the requisite talent, then probably you'll succeed. But,
[00:32:11] Dr Genevieve Hayes: And the requisite genetics.
[00:32:13] Prof Steve Stern: well, yes. Yeah, there are certain, requirements. But I think that for technical stuff you can't sort of force things.
[00:32:19] It has to be a groundswell, but you can put yourself in the position. To take advantage of that when it occurs. And I think passion is as much, you know, I don't think I would have done what I've done if I didn't really love sport.
[00:32:32] So I think that's, you know but that said that almost is like you have to put that in the backseat when you're starting to work on the method and go, I'm doing this because I love sport, but in order to do it properly, I have to kinda put the sport aside initially and think, what are the mathematics of what's going on here?
[00:32:47] What is really the problem? Involved. I get asked all the time if there's another sport I'm going to fix. And unfortunately, I can't think of another, I said Cricket was so unique that the solution is wonderful but it's not broadly applicable to any other sport that I know of.
[00:33:03] So I'm kind of stuck.
[00:33:05] Dr Genevieve Hayes: Is one enough for a lifetime?
[00:33:07] Prof Steve Stern: that's, kind of what I think to myself is like, well, what do you want from me?
[00:33:12] Dr Genevieve Hayes: So for listeners who wanna get in contact with you, what can they do?
[00:33:15] Prof Steve Stern: Thank you. Email me s stern@bond.edu au. I try my level best to answer every email I get. About cricket. So that's the easiest way to do it. There is an upcoming conference that I am co organizing next July, July one, two, and three at the Gold Coast Star Casino which is called the anm Mass Sport conference, which is all about using mathematics in sports in different ways.
[00:33:44] So as I said, my situation involved a nice serendipity which had to do with the fact that I was bold enough to just write to Duckworth and Lewis, if you're interested in becoming a sports analyst, one other thing that you can do obviously is make a network.
[00:33:57] Go to these conferences and learn about what people are doing. Perhaps find a problem by listening to other people's discussions that you really become passionate about. And then you can either. Join with them collegially or move off on your own. But finding other like-minded people is always an incredibly useful aspect.
[00:34:16] Academics aren't always as good as that, that as they should be, but if you can do it, it's amazingly fun to find someone who has like-minded ideas than you, and to fire back and forth off of each other. Because otherwise you can wind up going down rabbit holes every once in a while and turn out not to produce anything that's actually relevant or too complicated or not broadly applicable.
[00:34:39] Dr Genevieve Hayes: And there you have it. Another value packed episode to help you transform from technical implementer to strategic expert. If you enjoyed this episode, why not make it a double next week? Catch Steve's value boost a 10 minute episode where he shares one powerful tip for creating real strategic impact right away.
[00:35:00] Make sure you're subscribed so you don't miss it. And if you found today's episode useful and think others could benefit, please leave us a rating and review on your favorite podcast platform. Thanks for joining me today, Steve,
[00:35:13] Prof Steve Stern: You're very welcome. Thanks for having me, Genevieve.
[00:35:16] Dr Genevieve Hayes: and for those in the audience, thanks for listening. I'm Dr. Genevieve Hayes, and this has been Value Driven Data Science.