The Science of Football: Inside the 2026 World Cup

0:28This episode is brought to you by Google Chrome. and AI give football teams an edge. In short, can science win the World Cup? That's what we're getting into on today's special episode of The World, the Universe and Us from New Scientist, recorded live at the Royal Society in London. I'm Rowan Hooper. I'm a journalist at New Scientist, and we've got some fantastic guests on today's show, and I'm going to introduce them now. Dr. Charlotte Cowie is Chief Medical Officer at the British Olympic Association, and from 2016 to 2026. She was Chief Medical Officer for the British Football Association. Professor Holly Bridge.

1:04Holly is Professor of Neuroscience at the University of Oxford and is part of the Football on the Brain team exploring the connections between neuroscience and football. And Dr Ian Graham. Ian was Director of Research at Liverpool FC from 2012 to 2023. He's the author of How to Win the Premier League and he runs the sports advisory business Ludo Nautix. So welcome all of you guys. I wanted to start by asking about the current World Cup which I'm sure everyone's been watching. Let's start with you Charlotte. Is there a team you can point to where you've seen an especially interesting use of science or medicine or data that has struck you?

1:47Yeah thanks. Obviously I I have a little bit of a bias, so I'm going to go with the England team. And actually their preparation started a long time out from the World Cup. You know, this involves a lot of preparation for sports medicine, sports science teams. This time last year, this is not a trade secret because it has been published in the media, but they got the squad to swallow heat sensing pills so they could measure in real time their core temperature and put them in heat chambers at different temperatures, different humidities, to test their reaction to heat, how they dealt with it, and to experiment with different ways of managing individually their response to heat as well.

2:22So it was a great scientific intervention, but also helped to start the conversation with players about what it might be like this summer and create some anticipation of some of the difficulties we know they will face in this World Cup. Because it was already anticipated to be a hot World Cup, which we're going to get into. So there were mini thermometers, effectively, that went in. And did that lead to any changes in how recommendations for what players should do to keep cool? So, yeah, I think it's helpful to have an understanding of how individual players respond, because what they were also recording was their reported fatigue levels.

2:57And it's amazing how different players perceive heat and the level of fatigue relative to their actual fatigue when you measure their heart rate and other physiological responses. So I think it helped, A, to understand individual responses, be for players to understand what their responses were and try and mitigate those in advance. But also, as I say, it created just that expectation of a month spent or hopefully more spent in that extreme heat and how they needed to start, you know, acclimatising even before they got out there in terms of their own exposure to heat and starting to get comfortable with it.

3:32Holly, we've already seen some big teams go out on penalties already. very sadly Germany's gone out and the Netherlands very sadly what about the this dreaded penalty shootout is there you know can you give us some insight into that what's going on there when when a players face that situation? So obviously I don't work with players directly I'm very much a sort of brain person but if you think about what's going through a player's brain while they're standing there deciding what to do so theoretically they should have made up their mind what they're going to do at a point when they're practicing but clearly no one goes to take a penalty that's not going to be in a corner and there's going to be at an easy level to to save so if you think about when people are making a decision they're going to be using information their sort of prior knowledge which is going to be sort of incorporated along with their current sensory information now I think so obviously last night that goalkeeper was very tall so you know in terms of reach and where can he actually save the ball, it's greater than for other people who are smaller.

4:40So Germany previously are known for being very psychologically strong and irrespective of what's going on, they're going to win their penalty shootout. So this was a complete turn of events where actually that didn't happen this time. So I think it's the first time they've not won a penalty shootout. So you've got your prior experience of taking penalties, but you've also got your recent experience of what's going on on the pitch. now the sort of expectation that you're going to beat a team gives you a sort of expectation that you're going to play well you're going to be able to pass the ball around and because potentially because what they've done didn't live up to those levels it makes it much more difficult then to sort of take your penalty in a very sort of grounded way so you need to kind of reset if you can before what about when the goalies like what was that goalie who said put it in the corner there and really mess with the player's mind.

5:32That was really cunning, wasn't it? Yeah. I mean, so one of the things that you really, really have to do is make sure that you can block out all the sort of noise. And that takes a lot of practice. So you can train your brain to do that, but it takes practice. And if you are, so for example, if you are already in stress situation, blocking those things out is much more difficult. So given that they had already underperformed, I think blocking out the stress is much more difficult. Ian, so you pioneered the use of data in the Premier League. We'll get on to that in a bit more, perhaps. But are there any teams in the World Cup that you've seen or that you know of are making particularly clever use of data?

6:18Or are they all doing it now? I would say that very few are doing it. Really? I think it probably takes longer to permeate the national game than it does the club game. I'll stick on the penalties subject. I'm Welsh, so it's very difficult for me to praise England. But through gritted teeth, I'll say England were the first team to take penalties seriously. I think it was Glenn Hoddle said penalties are a lottery. Well, of course, they're not a lottery. The psychological and physiological aspects are super important. but there's also a whole lot of statistics and data that you can bring to bear on it.

6:53So one example would be if we know that Messi always shoots to the left, it's very easy to save his penalty because you know which way he's going to shoot. But Messi knows that you know that he always shoots to the left, so he should think he should shoot to the right. But you know that he knows that you know that. This is game theory. And so statistics, what players have done in the past with their penalty shootouts, plus game theory, knowing that I know what you've done in the past and you know that I know what you've done in the past, that can inform you where best to place your penalty in the future.

7:27And the best strategy tends to be a mixed strategy. So you want to randomise where you're going to put the ball. It gets a little bit more complicated because you can have different strategies. So some of the best players will make a little stutter in their run-up to get a trigger action from the goalkeeper and just pass the ball into the opposite corner. So that's called a keeper-dependent strategy. The best goalkeepers will fake a trigger to try and goad the player into shooting in that direction. So a good friend of mine, Ignacio Palacio Suerta, works with the England team since 2018 to go through this game theory approach to say, against this goalkeeper and in this situation, you should kick this way unless the goalkeeper does this, in which you kick this way.

8:16From a lottery, you don't get to 100 % we're going to win the penalty shootout. But if you can get to 70%, I'd take that over 50-50 any day. Yeah, yeah. Well, sometimes you see the goalies with, they look at their water bottles, don't they? Yeah, and that all comes from deep stats research. And just the data of he shot there and the keeper dived there, the data's not granular enough. You want to see if there's a stutter in the run-up and if there's a trigger move with the goalkeeper as well. To Holly's point as well, if you have done that homework, then you feel more confident in that situation as well.

8:51So you're more likely to be able to get over those nerves and focus. Just one factoid on this sort of randomising your strategy. Messi, Maradona, Ribery, Cantona have all said, when I run up to take the penalty kick, I don't know which direction I'm going to kick the ball in. I think you really need to be a top player to do that. but that kind of says something about their psychology but also how they've optimised their strategies. So they can make a really split-second decision probably subconsciously as to what the goalie's doing and then they decide where they put it. Wow. I thought I'd ask each of you to give an example from your work individually that is something that's made a difference perhaps over the course of your career.

9:36Something that's really changed from basically when we were kids watching football back in the day. Again, obviously I've got an England team's bias here, but I think the thing that I would say has made a massive difference from a sports medicine science point of view is I think everybody's thinking there'll be some specific cutting edge technology or something like that. But in actual fact, the most important things in terms of creating a really successful player are not the one percenters, they're the 99 percenters. It's the things like good nutrition, good sleep, good recovery, good injury prevention, and great strength and conditioning to build athleticism.

10:10And you can't decide to have a good diet or do a strength and conditioning programme the night before a World Cup final. You have to do that every day, every week for years and years. And so the dual combination of St George's Park National Football Centre, which is now somewhere where potential England players come from the age of 14 and start to learn what it's like to become an England player or how it's like to compete in an under-17s World Cup and under 19s and under 21s and start to engage with our sports medicine and sports science from an early age and then to parallel that in club football the elite professional player program eppp which is a mandated program where academies have to provide education mental health support strength and conditioning means that the group of players that we have now for England that are coming through are not just talented but they are so athletic they're so tuned in they are educated you can have a conversation with them about their nutritional strategy about their recovery strategy and that's produced this almost embarrassment of incredible you know articulate players such that you know we're in a situation where Cole Palmer and Phil and are not in an England squad, which I think is just massive tribute to those steady, inching away medical and sports science and academic achievements of those players.

11:37I wonder what the pundits make of it, because they came through a time where, you know, you'd go out drinking and smoking after, or maybe before a match in some cases. And now they're seeing something completely different. it must fry their minds in some ways. I think that's where some of the coaches who are currently successful and will be more successful are those coaches who can speak to players, the new players, not the old players. And that is a slightly different mindset. Jim Hicks is in the audience and I know him from Millwall Football Club, which is my first ever football club. And yes, certainly before the match as well as after the match was not the nutritional or lifestyle choices that we probably choose nowadays.

12:25So, yeah, I've seen that. Yeah, Holly, what about from your background point of view? So I think actually I would probably look forward because, you know, over the past sort of two years or maybe two to three years, actually what's happened is people have started to ask us about, you know, what is the role of neuroscience in football? So sort of moving away from the psychology but saying, you know, what are the things that are going on in people's brains that allow them to be a professional or a successful footballer. And I think people don't quite appreciate what an impact the brain has and that, you know, Charlotte was talking about education.

13:01And that's really important because obviously you're using your brain for education, but you're actually using it when you're playing football. You're using, I mean, people talk about, you know, you're using 10 to 20 percent of your brain, which is completely not true because you're using a massive proportion of it. if you're playing football because you're using all your senses you know you need to see what's going on you've got visual spatial awareness where is everybody else you know you have to listen for things you have to talk all of these actions are required for football and you need your brain to be working well and I think people are really interested in what can you do to sort of boost decision making what can you do to boost creativity and all of these are ways in which your brain can learn you know your brain can learn these things so looking to optimize how you coach people um both sort of on the pitch but also in the classroom is going to be really important so i'm looking forward but um so when when people talk of like oh he's got a good footballing brain is that what they mean like like someone like alonso at liverpool or say who you know you can see what's happening in the match and pick people out and and sort of is that what they mean like All these different senses and inputs are being integrated very rapidly and from a high level.

14:15So, I mean, when you're talking about a football brain, I mean, you've got to integrate what's going on at that point. So you're taking in sensory information, but you're also integrating with your past experience. So in order to make a decision, you have to know what happened last time. If it went wrong, you need to learn from that mistake. So those people who've got a good football brain have probably made quite good predictions initially. then they've corrected them and that cycle of making a mistake, correcting it to improve, may well just work faster in them. But I guess the argument would be that everyone can learn because brains are wired to learn, whether it's football or whether it's education.

14:52Ian, what about the use of data? Data is a really easy answer because 20 years ago when I started my career, data about football just didn't exist. So the data revolution in football is seen as a new thing and it is a new thing, but it's because the raw material, the data is now available. So if you look at baseball or cricket for hundreds of years, ball by ball, you could see who the bowler was, who the batter was, what the outcome of the shot was, and the outcome of that shot, a six or a wicket, means something for the game. Football data until 20 years ago was the full-time score, the attendance and the name of the referee.

15:32You would know if a player was sent off. you might know the team sheets like who played as well very very little data to go on then maybe 15 years ago and touch by touch data you knew where the ball was who was doing what with the ball so that went from like one row of information per game to about 2 000 rows of information and now using optical technology so there's cameras in every premier league stadium every World Cup stadium, 25 frames per second for 22 players plus the ball, plus the referee. Although the Premier League scrubbed out the position of the referee when I worked for a Premier League club, so we couldn't complain about the referee not being able to see the action.

16:14You can see, yeah, there's literally nowhere for the players to hide. So even when you've got some information that I passed the ball backwards, okay, that's not a very good thing to do. now we can see exactly the array of options that a player had maybe passing backwards was the best i could have done in that situation or maybe i turned down a great pass to you in the box so the amount of information we can wring out of the data available today is huge 20 years ago i was i was building very simple pass on goal scoring rates for teams because that's literally all of the data that we had and to try and say something this player played for this team while they were scoring a lot of goals.

16:56It doesn't really say if they're a very good player or not. So when you have all this massive amounts of data now, is it a human going through it and saying, it looks like you shouldn't have done this part? Or can you get an AI to do this for you? Yes, so AI is a good technique. We try to build models based on physics and on reasoning, first of all so that we've got a chance to understand why the model is making the suggestions it does so for example when we've got the information about where all of the players on the pitch are there's a concept called pitch control which is who controls what part of the pitch that's based on where you are on the pitch but also which direction you're running in if i'm already in motion i control a bit more of the pitch ahead of me and a bit less behind me a bit less behind because i have to decelerate and turn around if i want to go back in that direction so you can basically see who's in charge of which parts of the pitch and you can put a basic physics model a physics passing model to say i can hit the ball with this accuracy at this velocity and predict whether that pass with that trajectory will get intercepted or not that gives you an idea of passing lanes and what the chance of a pass being completed is and really you know players abilities are can they execute a pass better than the average player so Stephen Gerrard is one of my favourite players his speciality was 40 yards diagonal passes that our model would say you know your typical player's not going to make that pass so we could say it's a bad decision because we think the pass has got a low chance of success but if you've got the skill of Stephen Gerrard for him it's a good decision so we can find and it's only today that we can start to say something about the decision-making behaviour of players.

18:48We just couldn't see what they were doing before. I want to talk about youth versus experience. One of the stars of this World Cup already is Lamine Yamal of Spain. He's only 18. And then at the other end of the spectrum, we've got Messi, 39, Ronaldo's 41. Can we talk about the relative benefits of, you know, raw youth and a wise old head? so as i've already alluded to the brain is essentially built on experience so the more you the more you do something the more you experience situations essentially the the more efficiently your brain will work so if you are older and you've been playing football for many more years the likelihood is that you will have experienced the situations more often your brain will be wired to to respond quickly so you so i mean you see this in youth football all the time if the ball's up in the air and the little kid tries to head it they're nowhere near it but you don't see that in adult football so much and you certainly don't see it in the professional game and that's because the brain becomes very very good at predicting what's going to happen it's very good anticipation so in those older players you might argue that actually what they've done is they've sort of moved from that sort of the being the prime physical sort of specimen to actually building their brain so they can make up for that time that where they might be a bit slower but what Isn't there something to be said for the lack of fear that you have as a kid, as an 18-year-old?

20:14You see them attempt things, perhaps, that your older player wouldn't bother. I mean, that's like driving a car recklessly, right? So we know that the adolescent brain is wired to take more risks. So there's a sort of balance between your risk-taking areas, which give you great reward if you're successful, and your prefrontal cortex, which is sort of the voice of reason, where it says, I'm never going to make that pass, or I'm 99 % certain that I'm not going to make that pass. So there is this idea that as an adolescent, you can take more risks. And the brain, that prefrontal cortex, is not fully mature until about 25.

20:54So a lot of footballers are within that age range where they will be tempted to take risks. Maybe that's what makes football more fun to watch. Yeah, that's really interesting. One of the things you see in, you know, Winger is the position where young players get a lot of opportunity because, you know, a lot of explosive speed is needed. The number of dribbles and one-on-one attempts that players make, that's the first thing to really drop off. And you kind of wonder, is that because they've got a little less physical capacity or is it because they now have got a better judgment of whether it's the right thing to do or not?

21:33So that 25 years old certainly rang a bell with me. When you look at a wide range of players statistically, on average, players will peak at about 27. So that's the sweet spot between you're still in your physical prime, but you've built up that level of experience that you know the right thing to do in every moment. It's quite interesting the way that varies across position. Goalkeeper, doesn't matter if you're 35 or 40 even, you can still be a really, really good goalkeeper. a winger peaks earliest because it requires that explosive pace a lot of wingers kind of change their role and position slightly and then interestingly like defensive midfielder and centre-back where those positions it's a lot of off-ball positioning and judgment rather than your skill on the ball they peak a lot older centre midfielder is one where you know seeing the picture of the game and knowing the right thing to do in this situation is as important or more important than your physical capacity or your technical skill.

22:35With Messi in particular, you know, Messi was the best player in the world, aged 21, and just stayed there. And so Jamal, if he stays as professional as Messi did, could go the similar way, similar sort of thing with Ronaldo or Modric. But you have to bear in mind that those players are very much the exception. They're so skillful that Messi, Messi's also allowed his teammates to do a lot of his work for him, which allows him to look as good if he had to do the same physical effort as he had to do age 25. I'm not sure he would look as good, but he's so skillful that, you know, it's the right thing for his teammates to do his running for him.

23:13I think the other thing I would say, you know, it's a team sport, right? So obviously you ideally want that combination of some people who will run and run and are really enthusiastic and haven't experienced the bitter failure that some of the older players have. But you need, you know, Jordan Henderson is there, you know for a number of reasons but it's not that he is absolutely the most able player in the squad but he's experienced a series of major tournaments it's been a really important influence to younger players including Jude Bellingham so I think that mix of youth and experience is probably your ideal combination.

23:48Yeah okay so you've mentioned Bellingham and I wanted to ask in particular about him but there was a moment in the game England against Panama when, you know, just me watching from my armchair, we hadn't been playing very well. And it seemed like suddenly Bellingham just said, I've had enough of this. I'm just going to go and win this game now and seemed to turn it on. And even though he's surrounded by other pro footballers, he seemed somehow different. And what was going on there? Ian, do you want to talk about that? I mean, I'm going to pour cold state of water over this post hoc narrative fallacy you've just explained.

24:31Go for it, yeah. So yeah, narrative fallacy and post hoc fallacy, which is we won the game, therefore we must have been better, therefore it was Bellingham deciding to do this. It's a disease of football and one that's very hard to fight against because our storytelling brains are hardwired to love that sort of story and causal explanation. so think like taking a scientific or a data approach you have to have a hypothesis so what I just want to do is go back in time with you to before the game and say well what do you think Bellingham is going to do in this game would you have made that prediction and you can ask you know if he was able to do that in the 80th minute why didn't he do it in the 20th minute he's a really good save us all the bother of yeah that being said Panama's not a great team they were very well organised but with some fatigue they're probably less fit than England I don't know if that's true or not they probably are and they probably have worse quality substitutes than England so towards the end of the game favourites tend to score more heavily from a statistical point of view the problem with football and the beauty of football is there are a hundred explanations and ten of the factors probably have some effect so we can all argue about you know which effect is the most important.

25:52I'll always go for the boring statistical answer. I know, that's why you're here. We want that now. So a sort of alternative explanation would be the fact that in terms of, so you could ask anyone, what motivates you? And the problem is if you've got a team that's almost expected to win and you've got a team that are perceived to be weaker, the team that are stronger sort of expect to score. So our brains are wired to sort of reward us when we get a sort of unexpected surprise or unexpected good things happen. So what that does is your expectation reward will motivate you more or less. So because scoring against Panama to an England player would essentially be much less rewarding than Panama scoring against England, that sets things up differently in the sort of minds of the players.

26:42So there's a dopamine has a lot to play in that as well. Is that about similarly when you have a minnow playing a big team and no one, well, often the minnows will just park the bus or park two buses in front of the goal and it's hard to do anything. You know, they won't let you play or that's what people say, you know. The level of motivation is going to be high. And, you know, what constitutes a reward if you're playing against England? It may well be the fact that you've tackled G Bellingham. You know, the rewards that you get are not necessarily scoring a goal. they can be anything that sort of gives you success.

27:15A spirited loss is kind of a success for Panama. You know, we went toe-to-toe with one of the favourites. Yeah.

27:48I want to talk about the role of the manager, particularly in a World Cup, because we're used to managers really imprinting their style onto the team. And, you know, we have to mention Jurgen Klopp because you work very closely with him. And then you get this sort of heavy metal football style. or you can have someone like gareth southgate who is is more defensive minded and perhaps a bit more cautious but what i was thinking of with with all your use of data is um is when you get an ai like like an alpha go right deep minds alpha go this was trained it was trained to play go over and ago millions and millions of times but it didn't have any baggage from the human knowledge of thousands of years of playing Go, it just learnt itself and it became better than any human.

28:38And it makes me wonder, like, sometimes, well, should we really just follow the data and not have this baggage that we might bring to us from being a defender when we were a football player or something? No, you shouldn't just follow the data is the short answer. It's a game played by people and okay AIs can help with psychology and mood and cohesiveness and so on but the manager has got access to a lot of privileged information about if the player's going through a difficult time you know there's a lot of data if they're in a physically good condition the data about how mentally a good condition they're in is much harder to access at Liverpool because Demis from DeepMind is a Liverpool fan.

29:23We had a cooperation with them to make, we started off saying we're going to make the AI coach. Then we thought Jürgen would get a bit upset about that. So we said the AI assistant coach, it's just going to help him, right? But it's a super hard thing to do the more human side of the job. So from a tactics and a selection on player skill level and on how those players cooperate with each other, yes data can take you a long way but there's still a sort of i wish it was all data but there's still a human element yeah but what about for example i'm just speaking as an absolute armchair you know fan watching the japan game right japan brazil you think right they've got a chance here what while you've got while you're one nil up keep going get another one but they didn't and we've seen this with england as well just sat on a one nil and and it's not gonna you know you might get away with it but you might not and Japan didn't and you just think don't be so defensive and surely so I mean in in football in general the impact of a coach is overrated I will make that to Jürgen I will make an exception of Jürgen so the the sort of rule of thumb is getting a really good coach versus a really bad coach is similar level of impact in terms of your points this season of getting one extra good player versus one extra bad player.

30:47Longer term, it's more of a problem because if you've got a bad coach, the results continue to be a bit subpar. The players don't respect the coach. The coach maybe want to transfer in some bad players as well. So Jürgen's kind of like the Messi of coaches. So we give him a free pass on this one. That being said, it's because what coaches bring is really difficult to measure in data. So they bring the kind of emotional philosophy of the team can't measure that very easily they bring the tactics which you can measure but that's kind of like it's up to the coach and it's a negotiation between the owners of the team and the coach what style do you want to play picking the best players that's something you can measure in data and i think one of the things with england was you had the best attacking talent for a generation or for more generation it was world-class attacking talent for the first time in many, many years at England.

31:42The defenders weren't as world-class as they have been in the past. I don't know Gareth Southgate at all. I would guess he's quite a conservative person. His response to not having the best defenders and having a great set of attackers was to put more defenders on the pitch and to play in a more defensive style, which kind of minimised England's chance of success, in my opinion. but again from the outside not knowing what his plan was you know that that's this just my opinion of what my guess to his plan you have to be on the inside to really know how much of a difference they made so what was the real answer well I guess the only thing I would say is you know we're talking about earlier on is you know you're sort of characterizing that in a you know a narrative of he wasn't very successful when in actual fact he was really successful um so I guess you know there's lots of ways in which what he did has has worked extremely well and got the squad to the point where we expect them to reach the final stages of a tournament um which was not the case when he first started with them so i guess it you know it depends isn't it i i do think that the the non-ai bit of coaching is probably something that he's really good at and jrgen as well which is around what in the old days people would have called man management um but i guess involves a lot is a lot around culture is a lot around the interaction between different players and the way you conduct yourself and where you sort of get over some of the inevitable hurdles that will happen over the course of a major tournament with lots of people who don't need to know each other very well living together in very close quarters not able to see their family not able to escape and coming under increasing intensity of physical and psychological pressure so I think you know those are the bits it's much harder to measure and also if you take sort of average data even from average sort of say professional player data these players are such outliers that I think it's really hard to sort of average anything in elites for because there are still relatively few data points in terms of what makes a truly great player or a truly great team.

33:50So I guess that's my sort of caveat of the AI and data world. This is a very hot World Cup, as we already mentioned earlier, and we've got these ad breaks or hydration breaks in between. What do we know about that from a science point of view? I mean, people are upset that it's breaking up the flow of the game. But what do we think of it scientifically? And is it good to have these little mini breaks take on some more fluids? I mean I would say if it's hot yes but obviously you know some of these hydration breaks have been taken when it's not even really that hot so I guess there's some controversy about what the hydration break means in terms of flow in terms of coaching and what advantages that might bring but effectively once the wet bulb globe temperature you know temperature plus humidity reaches a certain point it is extremely fatiguing and it is fair to players to have some kind of a break or opportunity to cool down if it starts to get to a point where it might be dangerous to their health and also detrimental to their performance so in theory yes a good idea but you know we all saw it was raining and they still had a hydration break so I mean in terms of heat if you look if you look at sort of brain function because the blood gets drawn to the extremities you get an increased sort of cardiac output but But a lot of that actually, instead of going to the brain, goes to the rest of the head so you can cool down.