Matthew Benham: How data, culture and smart bets built Brentford’s success

How concerned were you last summer?

(Last summer, Brentford lost their Head Coach Thomas Frank, his coaching staff and star trio Brian Mbeumo, Yoane Wissa and Christian Nørgaard).

Matthew Benham: I was a little bit concerned, but not as concerned as other people. We like to think in percentages. So we expected there was a chance that one of our two star strikers (Mbeumo and Wissa) would leave.

We didn’t think that both would and a lot of people were saying, ‘If you lose both of them that’s it, you’re 100% relegated.’ Whereas we saw it more like, ‘If we keep both of them we’ve got a 10% chance to go down. If Brian leaves and Wissa stays, it’s 15%. If they both go it’s a 20% chance to go down.’

Which is not ideal, but we still thought we were more likely to stay up than go down.

(As it turned out, Brentford are having the best Premier League season in their history and sit seventh in the table).

A lot of that hinged on our centre-forward (Igor) Thiago, who’s having an amazing season. We thought he had an awful lot of promise, but he missed most of last season with a knee injury and we weren’t sure his knee could withstand the demands of a full season. But we felt if he had a good season we were in a good position.

Start of your love affair with football

I just loved everything about it. I actually grew up in a non-football household. We were cricket fans. So I didn’t get into football until I was eight-years-old, but then I just became absolutely obsessed about it. It wasn’t on TV much in our day, but any time it was, I’d make sure I could watch it. I was always reading books about the history of the World Cup.

So back in the day, if you asked me what was the score between USA and Chile in the 1950 group stages, I could tell you Chile won 5-2 for example.

Setting up Smartodds

I think it was just around the turn of the century and the internet was taking off. Before then, if you wanted to bet you had to go to a High Street bookie and there were only like four of them and they operated like a cartel. You could only bet really small stakes and at horrible odds.

Suddenly you had all these overseas internet bookies, so you could bet bigger and at better odds, and the market was super inefficient in those days. I mean, we talk about expected goals – you didn’t actually need that in those days. A very very basic model with goals as your only input was enough to beat the market. It was so inefficient back in the day.

It was a very inefficient market that very few of us were playing in. There was certainly not any genius on the part of any of us. It was just that games were priced very very poorly, so a really basic model could beat them. There’s no use having a model if you’re just going into High Street bookies, because you can only bet a tenner and even then if you’re any good they close you down.

Taking over Brentford

The first few years of being financially involved, (were) really from 2005 until about a year after, when I took over fully. I didn’t really see a serious long-term prospect. It was more a labour of love. It’s the team I supported as a kid. This is a bit of fun and can I put enough money in so that they can carry on being a third or fourth division team?

Around 2007, we got to our lowest point. We were second bottom of the fourth division. We lost 7-0 at Peterborough. And then we had to sack our manager unfortunately, who was England legend Terry Butcher. It was the start of the rise after that.

But it was really only in around the summer of 2013 that I thought it could be a bit more than just a labour of love or fun. That was just after we failed to get promoted to the Championship. We had an infamous game where we got a penalty in the 95th minute of the last game of the season against Doncaster and if we scored we get promoted.

Our club captain and legend, Kevin O’Connor, was a designated penalty taker and instead we had a kid on loan from our deadliest rivals Fulham and he grabbed the ball, insisted that he took it and missed, which was pretty devastating at the time.

Although, funnily enough, the following season we were desperately in need of a striker. We had nowhere to turn, so he came back to us again and he did get us promoted that following season. Marcello Trotta.

Until then, the conventional wisdom was that unless you’re a huge club with a huge following and a massive stadium, then you’re always going to be a lower league team. But it got to the stage where the television deal in the Premier League was so big that it seemed that you didn’t need this traditional support to make us a viable business.

Obviously, if we got to the Premier League, then we’d have that lovely TV money. But even when we weren’t in the Premier League, we could get some of the benefit because all that money was washing around. So the Premier League would  buy the top players in the Championship and then Championship teams would buy the top players in League One and so on.

So we realised there was potentially a route: we could climb up the food chain via the transfer market. Basically if we can make up for a lack of traditional revenue by operating at a profit in the transfer market.

And a bit of an irony – the one  transfer that caught my eye as, ‘Oh my god, that looks inefficient,’ was in the summer of 2013, when Crystal Palace bought Dwight Gayle for six million quid from Peterborough. I thought, ‘that seems mad.’ And the reason there’s a bit of an irony is that seven years later we paid six million quid for a striker from Peterborough, Ivan Toney, who was the guy who got us promoted into the Premier League.

So maybe Palace paying six million quid for Dwight Gayle was a decent transfer as well.

FC Midtjylland and specialist coaching

(Benham became the majority owner of the parent company of Danish club FC Midtjylland in July 2014. Bennett asked him about the club’s pioneering use of specialist coaching, including a Set Piece Coach, Throw-in Coach, Kicking Coach and even a a Philosophy Coach.)

Not quite (a Philosophy Coach) – he wasn’t teaching people about Wittgenstein or anything! He (Flemming Pedersen) was Head of Footballing Philosophy, which is similar to what we now call a Technical Director.

I loved everything about the Midtjylland adventure. The thing that struck us first when we went over there was just that the culture was fantastic – just really amazing that people were so friendly and open. The culture in English football at the time was a bit different – not so friendly, not so open

And not just in terms of accepting ideas, but just in the way that people behaved towards one another. It was quite a surprise for us walking past players who were really friendly and welcoming (at Midtjylland), whereas in England you walk past players and they’re saying stuff behind their hands.

I should say we managed to bring a lot of that wonderful Danish culture over to Brentford. That was one of the huge achievements of Thomas Frank. He left an amazing culture at Brentford – the way everyone interacts with each other, just friendly, open, not so many cliques, not too much bitching, and that was what really struck us at Midtjylland, just how nice, friendly and open the people were.

Creating a philosophy at Brentford

I’d like to put a bit of a caveat before that, because in 2015, when myself, Phil Giles and Rasmus Ankersen really stuck into Brentford and really tried to make a lot of changes, we were maybe a bit caught up in our own hubris . ‘We’ve got all these new ideas, we’re going to change the world.’

We actually wrote up a blueprint or philosophy and we called it Dogma 95, after the Danish film movement.

At the time, we were a bit drunk on we’re going to change everything and Phil Giles did say to me recently that he found this document and a lot of it was quite embarrassing, about how we’re going to be so different, we’re going to disrupt everything.

We have introduced a few new things to an extent, but a lot of the success has just been in running things efficiently, good management, everyone being on the same page and especially culture.

Appointing Keith Andrews and what you look for in a leader

I’m a big believer in the collective making decisions. Not in a group think way – I don’t like the idea of me saying, ‘Oh, I think we should go for Keith,’ and they all say, ‘Yes boss, yes boss.’

Myself, Phil Giles, who’s a Director of Football, Lee Dykes, the Technical Director, and Ben Ryan, Head of Performance, the first thing we did was go away separately and think about who could be interesting. Then we come together to talk.

All four of us thought Keith seemed like a really, really strong candidate. He is clearly, as you’ve seen, an amazing communicator, fantastic leader. We knew that he was very good with the players and also with the existing staff. We realised we were in a bit of a high-risk scenario anyway, with all the turnover of staff and players.

So we could have brought in a big name from outside who comes with an entourage. That would have had a lot of risk in itself.

We gave it a little bit of thought, but not an enormous amount of thought. It just seemed that Keith was the clear logical candidate. And maybe we were in a bubble, because then when it got announced there was all this uproar and we were like, ‘Fucking hell, what have we done?!’

But from where we were sitting, it just seemed that Keith was a clear outstanding candidate. Usually, when a player or a staff member leaves our first thought is, ‘Is there someone internally who already knows the club who can come in?’

I used the film director Paul Thomas Anderson as an example when I spoke to Keith. I said that’s someone who I absolutely love and who’s probably going to win the Oscar. But that’s someone who’s an auteur, who’s the director but also the producer and the screenwriter and will be heavily involved in the editing and the cinematography.

Traditionally an old-style football manager was like that. Fergie and Cloughy did everything – recruitment, medical department, philosophy.

You didn’t have sport science in those days, but if you did, then Fergie or Cloughy would have run it. For us, we talk about having a Head Coach not a manager and it was really important for us to have Keith as a fantastic leader but also for the Technical Director and Head of Performance to be able to have their input as well.

Why didn’t it work out for Thomas Frank at Tottenham?

Oh, that’s a bit of a tricky question. I mean, I’m not entirely sure what’s been going on internally at Tottenham. But I think potentially it was a really really difficult job for anyone. Anyone could have come in and struggled and it’s not as if they’ve had a huge bounce and improvement since he’s left as well.

It appears from the outside that maybe there are cultural issues inside Tottenham, but I could be wrong.

Capitalising on set pieces

Keith last year was our Set Piece Coach. He wouldn’t want to be identified as a Set Piece Coach. He saw himself as an all-round coach. He made it clear when he spent last season with us as a Set Piece Coach that that wasn’t his ultimate calling.

But he wanted to find a way to join to join the club and that was what was available at the time. But he’s not like Gianni Vio, he wouldn’t see himself as a specialist Set Piece Coach.

And obviously Arsenal have got their Set Piece Coach, Nicolas Jover, who was with us previously. Quite a lot of people have worked for us as Set Piece Coaches and then gone off to bigger clubs.

(The focus on set pieces) wasn’t so much driven by any stats, it was more that I’d always been a big NFL fan. Whenever I’d raised it in football and said, ‘Why can’t they have a playbook and signals and things like that’, it was always, ‘Oh no, that’s way too complicated.’

And I’d say, ‘Well, you get players who barely have a High School education, but they can still memorise a 100-page playbook in the NFL.’

And it seemed that the main resistance to set pieces was that players don’t like it. Players don’t like boring training. Traditionally in English football, players absolutely love someone like Kevin Keegan, because training is just fun. It’s five-a-side and head tennis.

There’s no boring learning or repetition or slow walk-throughs. If you’re going to be good at set pieces or tactical play, there is quite a lot of boring stuff: rote learning, repetition, slow walk-throughs.

A lot of teams weren’t doing it just because players didn’t like that. And traditionally if the players didn’t like it and the manager tried to push it, then the manager might lose the dressing room as they say. So that was really why we tried to push. And for whatever reason it worked a lot better in Midtjylland. It sort of clicked into place really quickly.

Scouting and recruitment

I come back to this thing about the collective, because for me that’s so important.

Everyone goes away, looks at it independently and then comes back and we have a free open exchange of views. I like to use the phrase exchange of views rather than a debate, because when you’re debating someone it’s all about winning and not backing down and not losing face, whereas a free exchange of views that’s more about can we get to the right answer or less of a wrong answer.

(David) Raya is a good example. So we all go away and I came back and was like, ‘No, we shouldn’t sign this guy.’ Now I’d say he’s potentially the best keeper in the world. At the time, I was fixated about having a a keeper who could aerially dominate, because we were pretty weak on crosses. The person who was pushing that (signing) most of all was our Goalkeeping Coach at the time, IIñaki Caña, who’s now Arsenal.

He was adamant about Raya, you who was at Blackburn at the time in I think in League One. He was absolutely adamant this is going to be a world-class keeper.

For keepers, data is pretty tricky. I think for keepers it’s much more about old-style scouting, whereas outfield players it’s a combination of old-style scouting and the data. In terms of what data is important, well, am I going to tell you?!

For a striker for example, receiving, getting in position, is way more informative than finishing. There’s an awful amount of randomness in actual finishing. So for example, if you show me two players at a similar level and they both scored 15 goals last season, but one of them had twice the number of chances, I want the guy who’s had twice the number of chances.

Even though you might say, ‘Oh, but he was less efficient.’ The fact that he was getting in position for the shot, that’s really important for us. So I think when we played Everton at our place late last year, Beto missed three one-on-ones.

Any decent model worth its salt would give Beto an upgrade. Typically a one-on-one, you’re like 50/50 to score. Any player who manages to to make three one-on-ones in a single game, even if he actually happens to miss them, he’s got to get an upgrade on any decent model. Obviously, your man in the street would slate him.

If I go through them (successful signings), some of them were more scouting-based and some were more maths based. Someone like Thiago, the scouts and the models lined up. Ivan Toney, scouts and models. Whereas players like Mbeumo and and Maupee, the models were not that keen on them but but the scouts were were so keen.

I think our models aren’t great for young players, because it’s not like US sport where you’d have data from High School and College. Brian was like 19 playing in the French Second Division. So the model was like ‘Nah, no, not so keen on him.’ But the fact that the the scouts and especially Lee Dykes, our Technical Director, were so positive outweighed the model being so so on him.

Whereas if you’ve got a player who’s mid-20s and he’s played a couple of hundred games, then the model better like him or we’re not going to sign him. But if he’s 19, we know the model’s not going to be that informative.

Any players you regret missing out on?

I mean, there’s always going to be ones that you regret. We could have signed (Eberechi) Eze for four million quid I think in 2019. We could have signed (Omar) Marmoushon a free about three years ago.

For four million quid combined. But we didn’t. The summer that we got promoted to the Premier League, there’s actually two players who scouted amazingly well. One was (Mykhailo) Mudryk, which would have been maybe a bit complicated if we signed him, although we were very close to signing him for quite a low fee, I think for 20 million euros, and he eventually went I think for more like 80 million euros although he’s now currently serving a drug band.

And then Michael Elise, his scouting was like unbelievable, just out of this world. But we’d just got promoted, we weren’t used to he crazy agent fees in the Premier League at the time.

And the agent fee for that one was just so insanely high that we stepped away, even though there was part of us thinking, ‘Well, if you combine the transfer fee and the agent fee, it’s kind of not too bad.’

But the agent fee on its own was just so insane that we stepped away. But there’s always going to be ones you miss out on.

Has data made football better?

I personally think so. I know there are all these people who say, ‘Oh, it was better in my day, etc.’ There’s been a lot of articles recently about how how the Premier League is sterile. You actually get, you know, goals per game, it’s a lot higher than it was like 20, 30 years ago. Personally I think it’s is a better watch. I think people can just be nostalgic about the past.

I actually like it personally when there’s not not something on the line so much, because being a club owner, there’s quite a lot of stress. I was speaking to our mutual friend Steve Parish from Crystal Palace before the Euro final in in 2024 and we were both looking forward to that, because we’re like little boys. It’s like you wake up in the morning really looking forward to it, really hoping England win, whereas as club owners, if we’ve got a big game, you just wake up with dread

Like when we’ve had play-off finals, you just wake up with dread. You don’t wake up thinking, ‘God, it’s going to be a great day, I hope we win.’ It’s more like, I can’t wait to get this game out the way.

After the game’s out the way and you’ve won, then it feels pretty good. But in terms of watching for pure enjoyment, that would be watching you know internationals or other games. When it’s watching Brentford, it’s a bit of a stretch to say I enjoy it while the games are going on. Maybe when we’re three or four nil up and there’s a couple of minutes left.

Can Brentford keep improving?

It’s a good question. I think just having everyone super open-minded. There’s always a danger in football – and I’ve seen it so many times – when you’ll have people who’ll be super open-minded until they get some success and praise. 

Then it’s ‘Right, this is my way, take it or leave it, I’m not going to change.’ You just have to always accept that everything you’re doing is inefficient. And your approach should be , ‘We’re never going to be perfect, we just try and be a little bit less imperfect every day.’

And it’s really about everyone having that mindset constantly about how can we improve, what what can we be doing better?

I’m not sure if there’s a there’s a fixed hard ceiling (for Brentford). We just want to want to keep improving. Obviously, our big weakness is the commercial side, especially in terms of brand awareness, recognition.

Our stadium is very small. So, it’s very important for us to try and grow our supporter base outside England. I think that’s the big next step for us.

I wouldn’t say we’re done with inefficiencies, because there’s still got to be other inefficiencies out there, but just probably not as big as 10 years ago.

What’s the next frontier for football?

Well, I could imagine that models and AI models will get better and better and therefore they’ll get more and more prevalent. Even though a lot of people think our recruitment is 99% model driven, there’s still a huge human input as well.

I would presume as models do get better, the human input will be less and less. The possibly interesting/ controversial one would be in terms of coach decisions, in terms of line-ups, strategy, when to make subs, when to change the formation. At the moment, that’s barely model driven, but I’m sure gradually it’ll be more and more so.

In NFL, for many years, all the geeks were saying you should be more aggressive on fourth down, all the coaches are too conservative. Now these days you watch an NFL game and there’ll be a graphic saying the model says here you should go for it.

And so I think it’s now more commonplace when making decisions on fourth downs that coaches are informed to greater or lesser extent by models and I presume that will gradually happen more and more you know in soccer.

Possibly at some point in the future and it’ll be a gradual thing. I’m not saying that in five years time I expect all the Head Coaches to be slavishly following what the model says, but there could  come a point where the model says very very strongly player A and not player B and if the manager does play player B that someone will be saying to the manager, ‘can you take me through your thinking?’

The Head Coach will still have a really really strong role, but possibly he’ll be working more in conjunction with technology and less just based on his intuition.