Below is an expanded version of an article I wrote for a South African newspaper on Monday this week. It concerns Matthew Syed’s and Dave Brailsford’s mythical concept of “marginal gains”, which is also known as common sense, and which includes elements of sports science, in the high performance sports world. I also fixed a little mistake from the original (my bad!)
“I think it’s a load of rubbish, if I’m honest. A lot of people made a lot of money out of it and David Brailsford used it constantly as his calling card. But I always thought it was a load of rubbish.”
So says 2012 Tour de France champion Bradley Wiggins, about the concept of marginal gains, during a public appearance last week. Wiggins, it must be said, speaks with some personal venom, fueled by his perception that has been sacrificed by his former team in the sport’s latest doping scandal. He is not, however the first person to cast doubt on the mythical marginal gains concept, and nor is he likely to be the last.
Marginal gains, for the uninitiated, was a philosophy of innovation that Sky’s director, David Brailsford, made famous. The idea was that hidden advantages could be summed together to produce an overall improvement (a faster cyclist) by finding 1% gains in these up-to-then undiscovered areas.
It’s best heard from the mouth of one its great cheerleaders, Matthew Syed. So here’s a video on the mythical, revolutionary marginal gains:
So marginal gains, in the case of Sky, meant doing things like providing riders with their own pillows and mattresses at hotels, adding pineapple juice to water to make it more palatable, warming down after stages, testing bikes in wind tunnels to make them more aerodynamic, and using antibacterial hand gels to reduce the risk of infections.
Apparently Brailsford also revolutionized training using complex algorithms, getting a jump on decades of sports science by thinking in exactly the way that decades’ worth of sports scientists (and coaches) are trained to think.
Each was meant to contribute to performance, importantly, in ways no other team had discovered. This allowed Sky’s dominance of the Tour de France, starting in 2012, to be attributed to this philosophy and the unique processes they had in place. I say “in ways no other team had discovered” because this tenet was critical to the idea of marginal gains as a competitive advantage.
Before the Sky began falling down, “marginal gains” was spoken of with mythical reverie. Matthew Syed, star of YouTube videos, and a British journalist and chief marginal gains evangelist described it as a “doctrine”, having “revolutionized some sports”. It also revolutionized his bank account, spawning books and corporate speaking gigs that were not marginal in cost.
Revolutionary? No, just repackaging righteousness
As for whether it was revolutionary to sports, the answer is definitively no. For as long as competitive sport has existed, coaches and athletes have understood the need to optimize all the elements that produce performance. It’s a big part of the reason why positions like “high performance director” exist, and is a cornerstone of my own profession of sports science, which aims to understand how A+B+C+D…= Z, and then work towards optimizing the letters through systematic measurement. Syed portrays himself as a man of science, so he must recognize this.
I suspect that in many other domains and industries, the concept is even older, and has been widely used. But because Syed was ignorant of all this, he’s vulnerable to ideas dressed up with “bells and whistles”. So he portrays it as a revolution beginning with Sir Dave, because he boarded the train at that station – where the train had been before is irrelevant to him.
In his pursuit to craft and pad his reputation as an intellectual, he’s actually peddling basics to experts without any awareness or sense of irony. Granted, there are enough non-experts who need to hear certain elements of the philosophy, and as I’ll touch on below, it’s not a ridiculous notion. But in elite sport, Syed (as a salesman for Brailsford/Sky) is selling trinkets dressed up as gold, and that’s important because it’s the application of marginal gains to elite sport that is in question here.
I’ve been lucky to work with some very good coaches in a range of sports – rowing, triathlon, cycling, kayaking, rugby and cricket – and I can assure you that the idea of breaking down the sport “into all of its component parts” is so common it is what you’d find a mediocre coach doing. Every one of those sports was doing it before I ever arrived, and will keep doing it once I’m gone. What separates a good from a mediocre coach is the ability to see both the detail and the big picture simultaneously, to not get lost in a few pixels of a giant picture (the same is true of good science). That, and people leadership.
But the notion of optimizing a number of processes and systems is so old that it doesn’t bear mentioning. In fact, I started out in elite sport with a marginal gains mindset, but I didn’t think of it in those words – it was just sports science to me, our fundamental reason for existence.
This was in 2006, long before Brailsford, which shows again that Sky’s philosophy was a repackaging rather than an invention. What I discovered immediately is that elite performance is much more “resolute” than people think – finding 1% is an enormous challenge, and pineapple juice, pillows and mattress don’t cut it. That’s in part because everyone is already doing it, or some variation of it, in the quest to find every possible advantage.
It was a humbling experience for me to enter that world full of wonderful theoretical ideas, absolutely convinced that my initiative and unique discoveries and learnings were going to revolutionize the sports, only to discover that a) everyone had already tried them, often before I was even born, and b) that many of them had been discarded because they don’t even work!
That’s a dose of humility through knowledge that a few could do with.
So a lot of what was being promoted as a gain was an irrelevant sideshow, and it trivialized the true contributions made by sports science and high performance management.
The other objection I had to marginal gains was its implicit arrogance. By attributing Sky’s success to marginal gains, some in the media propagated the idea that Sky were the only ones invested in aerodynamic bikes, fuel and hydration strategies, health and recovery, when in fact some of the best minds in sports science had been there for decades.
I bet Syed has no idea that the stuff he’s arguing as “revolutionary” is in fact twenty years, sometimes thirty years old. Testing aerodynamics of a bicycle? They didn’t put Chris Boardman into the superman position because it looked sexy. NASA has been involved with bicycle design for certain Tour champions. Training optimization – thousands of research articles and coaches who could claim ‘algorithms’. Sleep strategies? Diet? All done, ad nauseum, in elite sport.
Had this reality been acknowledged, then marginal gains as a source of competitive advantage evaporates. This again trivialized the real challenges in high performance sport, and by extension, sports science. We do not, as I’ve joked, study for ten years to help win medals by providing fluffy pillows and teaching adults how to wash their hands!
Objection to context, not concept
So it was the way it was framed, the context, more than the concept, that was objectionable. As a concept, there is no doubt that most businesses and individuals would improve performance in most normal domains by adopting a mindset that seeks to add the effect of changes in various areas of improvement.
That’s assuming they’re getting the important stuff right first, of course. I think there are downsides, or negative consequences to the aggressive pursuit of marginal gains that should probably be acknowledged. As a cost-benefit exercise, for instance, it’s unlikely to be where you get good return on investment, and if it comes at the expense of sound leadership and basic organizational functions, it’s a problem. In a sporting sense, I’ve been involved in teams led by very innovative coaches who can easily “over-innovate” and tinker so much that they a) neglect the basics, and b) create instability in the team environment. So that’s one way that this philosophy can undermine performance, and it’s rarely acknowledged.
However, generally speaking, if you are a runner, would your performance improve if you changed your diet, your sleep strategy, and your training programme by 2% each? Of course. But that’s so obvious you can read about it in Runners World, literally every single month.
Does an elite runner, who has already been optimizing preparation with scientific, medical and coaching support acquired over decades of experience and research, get the same competitive benefits (that is, relative to his opponents who are doing the same?). No. And thus half of “marginal gains” doesn’t exist, literally, because there’s no competitive gain. It’s “Marginal
The other half disappears if you understand that the equivalent of much of what Brailsford claimed as a gain is not. If you are an elite runner, would changing the way you tied your shoelaces make a difference to your half marathon performance? That’s the equivalent of what is offered by many of the marginal gains fallacies that Brailsford and Syed (and a few others) have propagated. So, “
So is marginal gains a complete write off? Of course not. Every person, business, organization, stands to improve by questioning the “inputs” that produce the performance output, and asking whether they can be improved. In most, given that many of these processes are not normally the focus of attention, there are probably some healthy gains to be made. So I have no issue with the application of what I’d call an “analytical performance mindset” to every industry, every person, and I’d support this mindset in business leaders who have already optimized the essential fundamental processes.
Marginal gains philosophy can survive in that world, getting people and businesses from X to Y, improving themselves. In the world of elite, competitive sport, however, X to Y is a) a lot more difficult than Syed seems to appreciate, b) not as decisive as he needs it to be, and c) not original or the source of any advantage at all in a world where performance outcomes are complex, multifactorial and already being pursued by everyone.