Case study 1 · Tue 27 Oct · David Mann

Models of talent development

The Dutch tennis federation wants more players, playing longer and having fun, and four singles players in the world top 100. Which talent-development model should it adopt?

≈ 185 min prep 11:00–12:45 NU-3A67 · 13:30–15:15 NU-4A51 Canvas page

Step 1

Guess first

Answer before you read. Guessing wrong is fine: a wrong guess makes the right answer stick better.

Guess 1guess

Ericsson claimed that skill comes down to accumulated deliberate practice. In a meta-analysis across domains, how much of the differences in performance did deliberate practice explain?

Show the answer

Answer: C. About 12% (Macnamara et al., 2014). It was higher in predictable games like chess (about 26%) and lower in sport, education and professions. Practice matters, but far less than Ericsson claimed. The mini-lecture covers this.

Guess 2guess

Researchers traced the development of 256 elite athletes from 27 sports. What share followed a straight, linear path from junior to senior level?

Show the answer

Answer: D. Only about 7%. Most moved up, down and sideways, and some switched sports (Gulbin et al., 2013, cited in Savelsbergh & Wormhoudt). This is why FTEM has no fixed ages and allows detours.

Guess 3guess

Lowering the net in junior tennis (to 0.65 or 0.52 m instead of 0.91 m) changes how 10-year-olds play. What do you expect?

Show the answer

Answer: C. More winners with no extra errors, more shots from inside the baseline and more first serves in (Limpens et al., 2018). In a related study children enjoyed the scaled net more (Timmermans et al., 2015). Both are cited by Savelsbergh & Wormhoudt.

Step 2

The case

About 30 minutes for the full case.

You play Kiki de Vries, a freelance sport psychologist. The KNLTB (the Dutch tennis federation) has asked her to recommend one talent-development model and to bring a figure that summarises the plan. She has to defend it in a meeting.

What the KNLTB wants

Participation: more than a million Dutch people play tennis. The KNLTB wants more members who play for longer and have fun. It also runs padel and pickleball.
Funding: high participation shows the government and NOC*NSF (the Dutch Olympic committee and sports federation) that tennis keeps people healthy. NOC*NSF also requires a well-defined talent-development model before it funds a federation.
Elite: four singles players in the ATP/WTA top 100, out of about 130 competing countries, many much bigger than the Netherlands.

The problem

Mees Röttgering was the world's top-ranked junior boy in early 2025. But many junior stars never succeed as seniors, and many "fall out of love" with tennis. Eastern European players often come through early, high-volume training, while Western European and American players more often sample other sports first. NOC*NSF is not impressed by federations that use LTAD, but the Dutch cycling federation (KNWU) does use it.

Exhibits at a glance

A, world top 10: the men are mostly from Western Europe, North America and Australia; the women are mostly from Eastern Europe and the former USSR. Ages run from about 19 to 33, averaging about 26.
B, Ericsson's violinists: the best students and the professionals had each accumulated about 10,000 practice hours by age 20, more than "good" students and future teachers.
C, DMSP: three pathways (see the models below).
D, LTAD: seven stages with sex-specific age bands.
E, KNWU cycling pathway: LTAD-based, but its age bands shift with the timing of each rider's growth spurt (early, average or late).
F–H: figures of FTEM, specialised sampling and the DMGT.

Questions Kiki has to answer

Can recreation and elite goals be combined in one model?
Should the KNLTB encourage children to play other sports, even if some leave tennis?
How much weight should injury, burnout, boredom and dropout get?
Should the model be holistic or tennis-specific? An existing model or a new one?
Does the model fit Dutch culture?
At what ages should players specialise and enter a talent pathway, if ever?

Watch

Federer on playing many sports (1 min): He'd tell juniors to play all sports for fun until their mid-teens; focusing on one sport too early risks burnout and turning it into a job.
Federer: how other sports helped his tennis (30 s): Table tennis, badminton, squash, basketball and football: "there's got to be a ball somewhere". Ball sports built his hand–eye coordination.

This is a summary in our own words. Read the full case before class: you're expected to know the text, the exhibits and the data.

On Canvas

First reactions

What would you recommend right now, before the readings? Note your gut call, then see whether the evidence changes it.

Step 3

Readings

For each reading: skim the summary, read the paper, then do the recall check without looking.

Framework paper · Journal of Sports Sciences · ≈ 40 min

Gulbin et al. (2013): An integrated framework for the optimisation of sport and athlete development (FTEM)

Gulbin, J. P., Croser, M. J., Morley, E. J., & Weissensteiner, J. R. (2013). An integrated framework for the optimisation of sport and athlete development: A practitioner approach. Journal of Sports Sciences, 31(12), 1319–1331.

On Canvas

A pathway built by practitioners at the Australian Institute of Sport: 10 phases from first movements to lasting elite success, no fixed ages, and room for detours.

Question

Can one framework cover active lifestyle, sport participation and elite success, and still match how athletes really develop?

Approach

Not an experiment. Practitioners combined existing research models (reviewed and criticised in their Table I) with decades of running talent programmes.

Findings

Foundations. F1: basic movement skills and physical literacy. F2: refining movement through sampling, with real instruction, because free play alone is not enough. F3: committing to a sport or to competition. You can stay at F3 for life.
Talent. T1: potential is spotted (outperforms at least 90% of peers), but only as a first judgement. T2: talent is verified over months by coaches in real training, looking at trainability, commitment and self-management. T3: practising and achieving, the biggest group of future elites but the most vulnerable, because funding flows to the ends of the pathway. T4: breakthrough and reward, such as a scholarship, academy place or squad.
Elite and mastery. E1: senior international, or the top level of a professional sport. E2: podium success, or professional success, which for tennis means a top-100 ranking. M: E2 sustained over 8–10 years.
Non-linear. No fixed ages. Athletes can skip phases, move back (for example after switching sports) and be in several sports at once.

Conclusion

A flexible scaffold for sports and national systems. Table II links each phase to drivers such as the athlete, coach, daily training environment, camps and competitions, and research.

Limitations

Little validation at publication: the authors say the research to test it is "underway".
Built for elite sport. The early stages get little detail; the mini-lecture makes the same point.
With no age guidance, it is harder to plan resources.

Use it in class

Arguing for flexibility: late bloomers, transfers and non-linear paths all fit.
The KNLTB's goal is FTEM's tennis benchmark for E2: top 100.
Criticising other models: Table I says LTAD is prescriptive about ages and content, has little evidence, favours physiology over skill and is commercially driven. The DMSP is tied to age ranges, based on a few Canadian and Australian sports, and never defines "elite".

Remember

Ten phases: F1–F3, T1–T4, E1–E2, M.
T2 = verification over months in real training; T3 = the vulnerable middle.
Tennis E2 = top 100.

Recall check

Close the paper, answer in your head, then check.

What happens in T2, and why does FTEM insist on it?

Show answer

Talent is verified over months by coaches in real training (trainability, commitment, self-management), because one-off T1 tests, often physiological, predict poorly.

Did you get it?

Why is T3 called the vulnerable phase?

Show answer

It holds the largest group of future elite athletes, but funding and support go mainly to the ends of the pathway (foundations and elite), not the middle.

Did you get it?

Give two of Gulbin's criticisms of LTAD.

Show answer

Any two of: prescriptive about ages and content; little empirical evidence; physiology and conditioning at the expense of skill; commercially driven.

Did you get it?

How does FTEM handle an athlete who switches sports as an adult?

Show answer

They can re-enter at an earlier phase (for example T1) without starting from scratch. Phases have no ages and can be skipped or revisited.

Did you get it?

My notes on this reading

Quotes, numbers, disagreements, how you'll use it.

Position paper · Movement & Sport Sciences · ≈ 25 min

Savelsbergh & Wormhoudt (2019): Creating adaptive athletes, the Athletic Skills Model

Savelsbergh, G. J. P., & Wormhoudt, R. (2019). Creating adaptive athletes: The athletic skills model for enhancing physical literacy as a foundation for expertise. Movement & Sport Sciences – Science & Motricité, 102, 31–38.

On Canvas

Become an athlete first and a specialist later: build broad movement skill through varied, challenging, playful practice, and use other sports on purpose.

Question

How should movement programmes be designed so children build physical literacy, a base for both lifelong activity and expertise?

Approach

A position paper by the model's own creators. It draws on ecological dynamics (the constraints-led approach) and selected studies.

Findings

Phases, not ages: a mover having fun, then a versatile good mover, then an athlete, then a specialist.
Road 1, concentric approach: develop one skill through all its variants. For tennis hitting, that means hitting in baseball, table tennis, golf and the volleyball smash.
Road 2, challenging environments: shape the task and environment instead of giving instructions. In junior tennis, lowering the net led to more winners without more errors and more first serves in (Limpens 2018). Scaling court and net gave a faster, more adult-like game, and children enjoyed the scaled net more (Timmermans 2015).
Road 3, transfer: multi-sports share no basic movement skill with your main sport; they build broad literacy, lower injury risk and dropout, and help you find the sport that suits you. Donor sports share at least one basic skill: futsal improved football passing, and indoor climbing helped on ice.

Conclusion

The ASM can serve recreational and elite pathways alike, balancing performance, fitness and health.

Limitations

Written by the model's authors.
The transfer evidence is mostly near transfer (futsal to football), and the multi-sport evidence is correlational.
Claims such as "injury risk decreases significantly" are stated without data.
Little on talent identification or elite stages, and little scientific publication on the model itself.

Use it in class

It is Dutch: developed at VU and the KNVB (Dutch football association) and used at Ajax and the KNVB campus.
It has direct tennis evidence: scaled courts and nets.
Suggest donor sports for tennis, such as throwing a frisbee for the backhand or a volleyball smash for the serve, or padel and squash for racket skills.

Remember

Three roads: concentric approach, challenging environments, exploiting transfer.
Multi-sport = no shared basic skill; donor sport = at least one shared basic skill.
Only 7% of elite athletes followed a linear path (Gulbin et al. 2013, 256 athletes, 27 sports).

Recall check

Close the paper, answer in your head, then check.

What is the difference between a multi-sport and a donor sport?

Show answer

A donor sport shares at least one basic movement skill with your main sport (judo's fighting and falling for a footballer). A multi-sport shares none and builds general physical literacy (bouldering for a footballer).

Did you get it?

Name the ASM's three roads to physical literacy and expertise.

Show answer

The concentric approach (one skill through all its variants), challenging environments (constraints-led design), and exploiting transfer (multi-sports and donor sports).

Did you get it?

What happened when the net was lowered for junior tennis players?

Show answer

More winners without more errors, more attacking play from inside the baseline, and more first serves in (Limpens et al., 2018).

Did you get it?

What's the main weakness of the ASM's evidence?

Show answer

Mostly near-transfer studies and correlational multi-sport data, presented by the model's own authors; little peer-reviewed testing of the model itself.

Did you get it?

My notes on this reading

Quotes, numbers, disagreements, how you'll use it.

Mini-lecture · 70 min · ≈ 70 min

Mann (2025): Models of talent development

Mann, D. L. (2025). Models of Talent Development. TID mini-lecture, Vrije Universiteit Amsterdam.

On Canvas

Seven models, two meanings of "talent", and how to choose: elite or recreational goals, type of sport, resources and how fast you need results.

Question

Which talent-development model fits which situation?

Approach

A tour of Bloom, deliberate practice, LTAD, DMGT, DMSP, FTEM and the ASM, followed by real federation models (NOC*NSF, KNWU cycling, Dutch handball, Athletics Canada).

Findings

Two meanings of talent: potential that predicts the future ("raw talent", giftedness), versus skill developed by now (Gagné, FTEM).
Deliberate practice explains about 12% of performance differences overall, about 26% in predictable games, and less in sport. "Doesn't apply within sport" as a 10,000-hour rule.
LTAD: the evidence for critical periods is weak, but adapting training to maturation is a good message. Early specialisation only suits sports that peak around maturation (gymnastics, diving, figure skating).
DMSP: there is no "recreation via early specialisation" pathway: early specialisers who fail tend to drop out.
NOC*NSF rule of thumb: podium success comes about 8 years after talent is identified and confirmed, and sideways entry is possible.

Conclusion

Choose by four things: elite and/or recreational goals; type of sport (in skill-based sports talent is harder to spot early than in speed or strength sports); money and resources; how quickly you need the talent.

Use it in class

Mann likes the DMGT, especially its chance factor, and calls its 10% cut-offs arbitrary.
He criticises FTEM for thin early stages.

Remember

Four choice criteria: goals, sport type, resources, time.
Tennis peaks well after maturation (about 26), so by LTAD's own logic it is a late-specialisation sport.

Recall check

Close the paper, answer in your head, then check.

What are the two meanings of "talent" the mini-lecture contrasts?

Show answer

Potential that predicts future excellence (raw talent, giftedness, often seen as fixed) versus skill that has been developed and can be measured now (Gagné, FTEM).

Did you get it?

Which DMSP pathway does not exist, and what does that imply for a federation that cares about retention?

Show answer

Recreational participation after early specialisation. Early specialisers who don't reach elite level tend to drop out, so early specialisation threatens retention.

Did you get it?

What are Mann's four criteria for choosing a model?

Show answer

Whether you want elite and/or recreational athletes; the type of sport; your money and resources; how quickly you need the talent.

Did you get it?

What is NOC*NSF's rule of thumb on the time from talent identification to podium?

Show answer

About 8 years from when talent is identified and confirmed, with sideways entry possible.

Did you get it?

My notes on this reading

Quotes, numbers, disagreements, how you'll use it.

Step 4

The models

Make a call on each model before opening its for and against. Then check whether you missed something.

Deliberate practice · Ericsson et al., 1993

Skill is the result of accumulated deliberate practice: effortful, not inherently enjoyable, with immediate feedback. Start early and pile up hours ("10,000 hours").