I’ve just returned from a month-long trip to the USA and UK, where I gave a talk (in various guises) called “The Talent Equilibrium”, my take on the strategic and scientific approaches to talent identification and management (because one without the other is useless).
Below is one of those presentations – this one given at the BASES 2014 Conference at St George’s Park on 26 November. It’s a shorter version of a previous talent management talk I’ve posted on before, which you can see here.
But first, a summary of the talk, going through its seven major themes. If there’s one thing to remember, you stop reading here, it’s this: It pays to ask “WHY?” before you move on to “WHAT? and HOW?”. I often feel that scientists begin at the second step, and the result is the gap between theory and its application in the real world. Unless the purpose is very clear, then it doesn’t matter how sound the theory is, it won’t fit the realities it is meant to confront. That has held back the talent discussion in the past. Purpose is only discovered if you know why you’re doing something – so ask why Talent ID and development need to happen.
The devil is in the details below. And in the presentation – if you’d like to hear the full talk, please send me an email and I’d love to give it a conference or an event, or consult on these subjects at some stage in the future!
Enjoy
Ross
Top
1. In the real world, talent ID and management are strategic & tactical, then operational. In that order
The most important point I’d make about Talent is that it is primarily a strategic and tactical function, rather than an operational one. This may seem semantic, but if you want to understand the real-world confusion around the issues then you have to recognize that when science debates talent, it does so from an operational perspective – the tests, the criteria for ID, things like age-effects (Matthew effect), biological development, PHV – are all operational.
The conflict between science and strategy begins with the definition of what talent is, and certainly there are conceptual problems with this. Ask a group of coaches or managers in sport to define talent and you’ll get a range of different perspectives, which can be problematic. There’s some overlap, but enough differences that you can easily to get trapped, and if that happens you never move beyond the minutiae and the details.
Conceptually, I believe the most difficult aspect of defining talent is that you don’t know which direction you’re facing – talent is evaluated today, with a view to tomorrow, but is based on things that happened yesterday.
But honestly, none of this really matters, because, in the real world, coaches and managers are identifying talent for practical reasons anyway. We have to remember what is happening in the real world and speak to that, not obscure concepts that only serve to confuse. Complex simplicity is what we are after…
The most simple reduction of talent ID is to ask “where is the person who in 5, 10 or 15 years will represent the country?”
This is followed by questions of management and talent development: “What should they do, when, and with whom?” This is where science and research guide and inform the answers, but the questions remain strategic and tactical.
That is, talent is a concept crucial to obtaining a strategic advantage over competitors. Whether I’m the Patriots or Broncos, Manchester City or Chelsea, the Stormers or the Bulls, I am in the business of finding better players, who I will then develop better than you, with the goal, ultimately, of beating you.
All this happens in the context of a “zero sum game” – if you pick one person, you can’t pick another, and neither can your rival team pick that person. Understanding this is crucial to appreciating the “race to the bottom” that exists in the major sports around the world, and why decisions are made the way they are. That in turn changes what a viable solution is, as we shall see shortly.
But second, and perhaps even more important, talent exists for budgeting reasons – it is tactical, and serves the purpose of guiding how limited resources are allocated to a selected group of individuals. Point here is that resources are always limited. For some (Olympic sports in SA) more than others (football and NFL in the UK and USA respectively), but coaching resources, money, facilities and even exposure to competition cannot sustain everyone.
So, what Talent ID exists to do is to guide the selection, effectively providing a filter that takes say 300 people into 30 (a 90% filter), so that those 30 get access to our best resources. In fact, this is how I would define Talent ID – the process of identifying individuals with the likeliest probability of success so that limited resources can be preferentially allocated to them.
Remember, the concepts: Strategy (competitive advantage) and tactics (budgeting), because those explain why the scientific principles are often overlooked in favor of the strategic prerogative.
Top
2. A risk management perspective on Talent ID
Practical talent ID is about filtering in order to optimize resources, which is not unlike deciding which shares to buy when you have limited funds and a world of them to choose from. You can buy the mature, high performing ones, but they’re expensive. That’s like buying Lebron or Ronaldo or Rory. Talent ID involves buying shares in the “start-ups”, long before they become the Apples and Googles. That means you’re inviting some kind of risk, which makes Warren Buffett more similar to you that might have thought.
With respects to Talent ID and Development, there are three risks in the share decision:
The inclusion error – you identify the wrong person, and spend time, money and people on them. What we should be asking, wearing our investor hats, is what kind of efficiency we’re after, and how great this cost can become before it’s unacceptable?
The exclusion error – you fail to identify the right person. That person might have been Usain Bolt, Lionel Messi or Michael Phelps, but they slipped through, were overlooked (more likely the former) and ended up selling vacuum cleaners, or worse, playing for your big rival. Again, we need to decide how great is this cost, because make no mistake this “opportunity cost” can be enormous – do you want to pass up that kid from Rosario who was so small, but turned out to be Barcelon’s number 10?
The environmental error – you have the right group of people but you don’t provide the environment. This is obviously an enormous area – it involves coaching, competition, medicine, science, pretty much everything. Not a focus of the above talk or this short post, but it’s where the big investments should come, for the reason that once the person is in the system, they are a sunk cost, and the only aspect that the manager or system can control is its own optimization of its new “asset”.At this stage, it almost doesn’t matter whether the person is the right one, the only modifiable aspect is the environment, and that’s why so much time and energy, particularly when dealing with children and young adults, should go into developing coaches, and the understanding of how to develop young sports people.
There is actually a fourth risk – failing to provide a good environment for the non-selected individuals, but I’ll address this later.
Top
3. A physiological basis for talent – predisposition
The next section of the talk discusses the various lines of evidence that suggest that certain individuals are predisposed to success in certain sports. This concept refutes the ideas of Outlers (Gladwell), Bounce (Syed) and the 10,000 hour rule (Ericsson), by clearly recognizing that there is abundant evidence that some people
Respond faster to training and reach different ceilings than others, and;
Have physical attributes that training cannot overcome but which impact performance either favorably or negatively
In the talk, I mention only three – the trainability in aerobic capacity (which has been linked to genetic polymorphisms, or variants of genes – some have them, others don’t), vision (hardware drives software acquisition. This one via David Esptein), and testosterone. I do however list a number of these physiological factors that clearly have an influence on performance, including:
Sex – obvious, but true, and the simplest illustration of all that genes affect performance. Entirely.
Height – it takes 294,831 gene variants to account for 49% of height. Don’t expect a single gene to explain Usain Bolt.
VO2max
Trainability
Muscle fiber type (and the related hundreds of biochemical characteristics)
Vision
Anthropometry – an obvious one for basketball, but equally crucial for weight lifting, swimming, distance running etc
Skill acquisition
Injury predisposition – some people literally cannot do the volume of training because of genes that increase risk of injury
If you want to read more on this, I published a paper evaluating the relative contribution of genes and training to performance with geneticist Prof Malcolm Collins, and I highly recommend David Epstein’s “The Sports Gene”, which does the best job you’ll ever see of deconstructing, explaining and storytelling these important points.
The bottom line here is that talent ID exists to identify and then capitalize on these predispositions. If you believe (and I really hope this is obvious) that any of them – from height to injury-risk – have a bearing on performance, and that any of these factors are genetic or innate, then you have to also recognize that an advantage exists for that person/team who can identify them best. That brings us back to the strategic and tactical drivers of Talent ID and Development in Point 1 above.
Specifically, consider trainability and performance ceilings are a result of these genetically influenced traits. If you have $1 million, would you rather invest it in ten athletes who are highly trainable and can succeed within 5,000 hours, or would you prefer to adopt a 10,000 hour, “anyone can succeed with enough deliberate practice” model (Syed), and spend your $1 million on five athletes who do take that long? The hourly cost will always be the same, but your return on investment is significantly better if you can capitalize on the obvious innate differences.
I know what I’d be choosing – if Talent ID and Development serve to direct resources, then this physiological basis of talent ID is absolutely crucial to the business and risk-taking decision.
Top
4. The 10,000 hour concept is not benign, and it’s unnecessary too
In the presentation below, I offer a quote which I pulled from the discussion forums to a website piece by Daniel Coyle in which he argues, as the scientific evidence does, in favor of more generalists. That is, to delay specialization. The quote, which you can read in the presentation, offers the counter argument, is very much relies on this 10,000 hour fallacy to support it.
Effectively, what is being said here is that if a child does NOT specialize, then it is impossible for them to accumulate the necessary training by the time they enter into the professional cut-throat world of sports. 10,000 hours, at 2 hours per day, takes just under 14 years. It’s a long-term, massive commitment, and it means they must start young, cannot focus on anything but the deliberate practice in a chosen sport, and reinforces this enormous competitive pressure.
It’s for this reason that the 10,000 hour concept, or the idea that practice, practice, practice is actually more than just a rallying cry or a feel-good message. There is a downside. The downside is that it’s now becoming increasingly apparent that early specialization is not only unnecessary (see the budgeting aspect for why this matters – it spreads cost to diversify), but it’s also potentially harmful. A name to look out for is Neeru Jayanthi, who is quoted in this Wall Street Journal article on children’s specialization in sport, because he has studies now showing that young athletes who specialize are 1.5 times more likely to become injured than those who do not.
Not only this, but performance doesn’t seem to this pre-occupation with volume and early starts, and may actually be favoured by a delay in high training volumes that invariably accompany this all-in approach. In the presentation, I describe one such study by Moesch et al, which found that athletes who went on to become great (top three in Europe and top 10 globally) actually DELAYED high training volumes compared to those who became merely “good”.
Point is, it’s neither necessary nor optimal for the health or performance of the child to drive this message of accumulating as much practice as possible to get ahead. The better approach is to wait, at least until adolescence, and then use common sense and an informed approach to select those sports (or a single one, in some cases) where the likeliest predisposition lies. And a big part of that lies with the child/athlete themselves. We are not picking cars out of a catalogue here and so the emotional/human side is just as important.
Early specialization for children is for adults to pick a sport, or to drive a sport, in the hope that the child is Tiger, Serena or Stefan Holm (see Sports Gene). Those cases are likely the exceptions, in a big way. Rather be patient, be prudent, and expose children to a range of different sports, called “sampling”, for as long as possible, before the physiological cards they have been dealt start to appear more clearly. Then choose, and target those resources.
Top
5. Adolescence – the watershed, and the example of SA rugby
In the Moesch study comparing great to good athletes, the watershed moment is adolescence. There are a few different reasons for that, and they probably complement one another. I would suggest that the root cause is likely physiological capabilities that only appear at this stage. In other words, a person is not a good athlete until they have gone through the change of adolescence and remain a good athlete. This is because many of the attributes that predispose to sporting success only appear at this stage.
An example of the practical outworkings of this are found in South African Rugby, which has long had three age-group competitions – an Under-13 tournament, an Under-16 tournament and an Under-18 tournament. Teams are selected on merit (with a few fineprint details which SA readers will understand). That makes for a convenient journey where we can monitor how many people who enter the system at the age of 13 appear on the other side at 18.
Remember that rugby is a game that favours size, strength and speed. Therefore, at the age of 13, children who are biologically more mature or relatively older than their peers have an advantage. Bigger children stand out. There’s a picture in the presentation, used twice, that shows two Under-13 players at a Cape Town school match, standing either side of the adult referee. That picture captures perfectly the theoretical problem with choosing for ability in a sport where biology is so important. Who would you choose?
In any event, back to the data – 349 boys represent their district teams at the age of 13 in 2005. Of those, just under one in three continues to represent a district team by the age of 16. Some appear within a year – these are precocious youngsters, very mature and able to compete against 16-year olds at the age of 14. Others appear later, and by 2008, 110 have also appeared in the same merit-based system at Under-16 level.
So the journey continues to Under-18 level – here, we see that of the 110 who appeared in both Under 13 and Under 16 teams, 84 appear in the Under-18 teams. So from Under-13 to Under-18, the ultimate efficiency is 349 becomes 84, or 24%. One in four.
From Under-16 to Under-18, however, it’s much higher – of this cohort of 110 who made it from 13 to 16, 84 also made it to Under-18. That’s 76%, three in four, and what it suggests is that the accuracy of your investment is significantly greater if you wait to the age of 16. The reasons are clear – if success in the sport is the result of physical characteristics, then you cannot know who will possess them until after this crucial biological watershed.
There is a catch here, one which the authors of that paper missed. The cohort who go from 110 at Under-16 to 84 at Under-18 have also come through the Under-13 system, so they’re particularly ‘durable’. They are not necessarily the same as the people who enter at the age of 16 – the conversion here is likely lower, but not as low as the 24% conversion from Under-13 to Under-16. I’m sure there is data on this somewhere.
The point remains that the earlier the selection, the worse it is likely to be. In fact, prior to puberty it’s a guess. Between puberty and adolescence, it improves, but it’s only once growth and full development occur that we can really say who has the right stuff, physically, to be a viable athlete. Note that this does not account for any psychological, emotional or other developmental factors, but it takes care of the physiological argument.
And finally, before we move to a strategic solution as a conclusion, you might now start to ask yourself why, if the accuracy of your decision is so poor prior to adolescence, are teams all around the world in the big sports hunting for and recruiting 6-year old children who display “talent”? Their chances of success are about the same as Warren Buffett’s if he closed his eyes and threw a dart at the financial indicators. But this question is key to understanding the rules of engagement, and your best response strategy, which is I emphasized it previously.
Top
6. The race to the bottom
The answer to that question should be clear by now. The major sports teams have engaged in a progressive race to the bottom because of competition between themselves, and also between sports. This latter battle cannot be ignored – Real Madrid can lose out to Barcelona when a potentially great football player chooses the Catalans, or they could lose out to rugby if that player decides to stay in Argentina to play for the Pumas. Similarly, track and field loses to basketball, rowing to rugby, rugby to football, triathlon to swimming, and so on.
And so is created a competitive market where supply is very limited, but demand is enormous. The cost doesn’t appear until much later, however, when the player reaches maturity and the ‘bet’ made by the team has actually come to maturity. That is, Messi was a bargain at 8, he was getting costly by 15, and by 21, priceless.
Therefore, what the major sports have done is assess their desire for efficiency, and given how much money they have, realized that it doesn’t actually matter if they waste $100,000 on 100 players (cost of $10 million), because if the 101st player is Lionel Messi, then they’re way ahead of the game. That’s not to say they don’t care at all – if they could find a Messi once every 50 rather than 100 players, then they save $5 million, but the cost of NOT identifying him is enormous. That’s the opportunity cost I spoke of earlier, and to wealthy sports and teams, this is the driver, not expense.
The other factor you have to consider (it’s a cynical world) is that if there’s doubt (there always is), but a player seems to have prospects, then it’s better for a team to buy him young when it’s cheaper because it denies opposition the same player. In South Africa, this happens out of school – teams will buy three or four of the same position, not only for back-up resources, but because it means their rivals must make do with fifth and sixth choice players. This is a source of tremendous inefficiency, but is the result of the competitive landscape in which athletes are prospected for as assets.
Ultimately, you have a race to the bottom, and it is reinforced by the fact that the career pathway becomes clear to the aspirant player. The pathway ends with a professional contract, but one step below that is getting into a very good representative team (district, college, school). That in turn means going to an excellent high school, which usually means getting into a good primary school, or an academy aimed at this age of athlete, and the result is that the journey to professionalism begins at about six years old, and the incentives driving this come from both directions. A push up the system by the aspirant child, and the suction provided by the triangle of resource limits, supply and demand.
Top
7. Viable strategic solutions and viable athletes
In this HP world, a system-wide approach has to recognize these forces and concede that in many instances, they may have too much inertia to change. Certainly, they cannot be instantly altered – a super-tanker, however slow moving, is a difficult proposition to turn around. If a strategy is offered to do that, it fails. 100% of the time, and it’s not viable. The viable strategy is the one that recognizes where insurmountable inertia exists, and where opportunities do, and then nudges the system in the right direction. Alternatively, it accepts the inertia in one area, but realizes that this may create opportunities in another, and those are the two approaches I’d be seeking were I in charge of an entire system.
Very gradually, one can change behavior by changing the incentives, especially at the level of competition, which is where poor coaches often find themselves most compromised. That is, any coach is going to respond to the incentives placed before him or her, and so if those incentives drive early selection, early specialization, early high training volumes – all the things we’ve said are NOT optimal for the young athlete or the team – then that’s what they’ll do anyway.
If winning an Under-13 school tournament is highly rewarded, then the coach will pick and prepare players to win. If winning at age 13 is rewarded less than developing talented and capable players at age 16, then the coach may adjust the approach, and that’s the lever you want to find if you’re able to.
This brings me to the LTAD model, which I’m sure most are familiar with, and was proposed and developed by Istvan Balyi. It’s sound on paper, theoretically meritorious, but as a viable strategy, there are very few sports that I’ve seen where it would actually work. One of the weak links is the competition structure – if a sport does not change its competition structure at the time it adopts LTAD, then it will simply create confusion for coaches who cannot meet both the LTAD principles and pressure provided by the school.
Any large rugby playing nation in the world illustrates this – the stage where LTAD proposes exploration and learning new skills, the competition systems of rugby promote the first steps of specialization and a highly-competitive winning environment. Later, at 18 years of age, LTAD proposes the commitment of the player to their chief sport and the willingness to invest time and energy into that sport. That’s already happened at 13 years of age in the big rugby playing nations, and as I explained in Point 6, it’s symptomatic of the race to the bottom that these sports engage in. The adoption of LTAD is thus doomed unless it is accompanied by competitive structure changes, but those are usually made by different organizations, and I think the inefficiency here is too large to overcome.
The second solution is to recognize where inertia exists, and to accept that it will be inefficient but that this can actually provide the opportunities for other sports to capitalize on.
The presentation ends with my offering a hypothetical model for a 4% efficient system, which takes in 100 young athletes at the age of 13, and then gradually filters them down to 4% at senior level. At each stage, the filter spits out athletes who had something, were identified, but failed to clear the next bar. I’ve called these athletes the ‘ghosts’, and they offer real potential for the other sports to capitalize on. Their presence in the system suggests some level of athleticism and perhaps even the psychological and emotional attributes to succeed, but for whatever reason they have not parlayed that into next level success.
That is, often, a failure of environment, not the athlete, and so another team in the same sport might look at these “ghosts” as real prospects provided they can deliver an environment conducive to that particular individual’s development. Or, another sport might target them, and this is what effective talent transfer has done – let the groundwork be done elsewhere, and then offer the opportunity when the ‘bet’ is much more secure.
The secret here is to keep the athlete viable for as long as possible. Because talent selection is zero-sum, the inclusion of one person means the exclusion of another, and with that comes a disparity in how resources are allocated. That excluded athlete cannot be allowed to disappear forever. They must be allowed to remain viable.
What does this mean? Well, you may recall that I mentioned three key risks earlier (Point 2). Well, here’s the fourth risk I mentioned – failing to provide the right environment to the excluded people. What you should be doing is trying to understand where those people go after exclusion, and then making sure that they continue to receive the minimum coaching support and opportunities for competition that will keep the revolving door spinning just long enough that some time in the future, they can get back in. That door need not lead back to the same sport, but could be a different one altogether (the UK has done this exceptionally well).
Point is, sometimes you might want to consider where your best coaches are – is it possible that your overall efficiency would increase if your best coaches worked with the excluded “ghosts” rather than the early identified superstars? Also, can you set up systems that include those athletes who may be disadvantaged by their biology? For example, the Dutch have recently set up an academy exclusively for children born later in the year – the relatively young ones. The Swiss Football Federation did the same by imposing quotas for relative age – keep them viable, keep them in competition.
The viable athlete this produced is thus characterized by three things:
Physiologically, they possess the minimum athletic requirements to succeed in some sport. Which that will be is up to the initiative of the individual sports, the physiology and the desire of the athlete
Environmentally, the viable athlete continues to play and sample sports for as long as possible. The same sport, new sports, and always with good access to coaching and competition. That’s not necessarily leagues and trophy competition, but competition within a training environment where they can be monitored, tested, evaluated, for all the attributes that may make them future success stories.I mentioned above that these “ghosts” often fail not through their own inadequacies, but rather because the environment is not quite right for them. There’s no reason why a different context can turn the same individual into a better athlete, and this is a high leverage area for sports teams to explore. These ghosts may well be failing because they’re at 80% of their capacity – your challenge is to find the strings that pull to get them to 95%, 100%. That would likely see them surpass those who survived, and then they become yet another narrative case-study of the “one that got away”. In fact, they didn’t, they were always identified as having something to be there in the first place, it just wasn’t optimized. I’m sure you can think of many examples of such athletes.
Psychologically, the viable athlete possesses a high degree of self-awareness (which makes them easier to teach), they are resilient (they need this to withstand the inclusion that placed in their situation, but it may be the thing that brings through and beyond their peers), and them something I’ve left as vague as I could, “real character”. I think that in the next decade or two, this will be understood far better, and we’ll realize what it means for young athletes, and even adults, to possess the “soft-skills” necessary to succeed and make the transitions to new sports, or returns to old ones. In fact, it will only day match the physiological attributes I mentioned earlier.Practically speaking, however, these soft-skills are identifiable only in the context of their continued participation in competitive environments. You can’t judge ‘real character’ or even ‘resilience’ in a classroom on a piece of paper, and so you must expose those viable athletes to the environment that allows them to demonstrate what they possess. That, in turn, means outstanding coaches, who can recognize these attributes and then reward them by putting that athlete back on the pathway they may have left, or onto a new one.Therefore, the ultimate solution, from the system, is this: Be aware of these possibilities, recognize the opportunities, and make damn sure you have people on the ground who will not miss them. Coach education, and the quality of the coaches at the exclusion level are going to make the difference between that hypothetical system of 4% that I showed earlier, and one that goes to 8% because of good people, with character and awareness.
And that is that. The presentation is below, I hope it makes some sense now that I’ve explained the key themes.
As I mentioned, this is actually a summary. A long one to read, sure, but the talk covers them more engagingly. It’s a talk I’d love to give more often, and so if you’re reading this, and feel there’s value to it (and a lot more that I haven’t said here), please do email me and we can perhaps make a plan!
Ross
The Science of Sport - Scientific comment and analysis of sports and sporting performance