You can’t teach speed: sprinters falsify the deliberate practice model of expertise

Study 1: Biographies of Olympic champions

We examined the biographies of Olympic champions because becoming an Olympic champion shows unambiguous evidence of expertise. Moreover, because there is often great interest in sprint champions, biographies have been written about many of them. These generally include detailed information on the sprinter’s athletic development, making them ideal for addressing the predictions of interest.

Although our main focus was testing the two predictions discussed above, we also explored whether champion sprinters had participated in organized sports prior to beginning their sprinting career. This was important because it could be argued that exceptional sprinting performance prior to formal sprint training reflects skill transfer from other sports (Baker, Côte & Abernathy, 2003; Smeeton, Ward & Williams, 2004).

Results and discussion

All 15 Olympic champion sprinters were recognized as being exceptionally fast relative to their peers before or coincident with their initiation of formal training. There was no indication in any biography that any sprinter was initially unexceptional. We condensed key information in Table 1 and summarized relevant passages from each biography in Table S2.

Table 1:

Male and female 100 m and 200 m Olympic gold medal winners for which autobiographies or biographies published in print were available.

Athlete	Sex	Olympic games	Events won	Superior sprinting speed documented as youth or teenager	Years of DP to reach world class statusa	Reference
Jesse Owens	M	1936	100 m, 200 m	Yes	4	Baker (1986)
Helen Stephens	F	1936	100 m	Yes	3	Hanson (2004)
Wilma Rudolph	F	1960	100 m, 200 m	Yes	2	Smith (2006), Mallon (1995), Rudolph (1977), Schraff (2004)
Bob Hayes	M	1964	100 m	Yes	2	Hayes (1990)
Wyomia Tyus	F	1964	100 m	Yes	7	Davis (1992), Carlson (1995), Tyus (2010)
		1968	100 m
Tommie Smith	M	1968	200 m	Yes	3	Smith & Steele (2007)
Evelyn Ashford	F	1984	100 m	Yes	1	Davis (1992), Hornbuckle (1995b)
Florence Griffith Joyner	F	1988	100 m, 200 m	Yes	2	Davis (1992), Hornbuckle (1995c)
Carl Lewis	M	1984	100 m	Yes	NA	Lewis & Marx (1990)
		1988	200 m
			100 mb
Ben Johnson	M	1988	100 mb	Yes	3	Christie (1988)
Gail Devers	F	1992	100 m	Yes	6	Hornbuckle (1995d)
		1996	100 m
Gwen Torrence	F	1992	200 m	Yes	7	Hornbuckle (1995a)
Michael Johnson	M	1996	200 m	Yes	5	Johnson (1996)
Marion Jones	F	2000	100 mc	Yes	1	Jones (2004), Gutman (2000)
Usain Bolt	M	2008	100 m	Yes	4	Bolt (2010), Cantor (2011), Irving (2010)
		2012	200 m,
			100 m,
			200 m

DOI: 10.7717/peerj.445/table-1

Notes:

a Defined as representing their country in international competitions (e.g., Olympic Games, World Championships).

b Awarded the gold medal because Ben Johnson was disqualified as winner for using performance enhancing drugs.

c Disqualified as winner for using performance enhancing drugs.

The biographies reported that adults (e.g., teachers, coaches) initially recognized the superior sprinting ability of nine sprinters (five women) and encouraged them to begin formal sprint training or competition. For example, the superior abilities of Wilma Rudolph, Helen Stephens, and Wyomia Tyus were discovered while they played basketball (Table S2), whereas Bolt (2010) and Hayes (1990) were discovered while they played cricket and baseball, respectively. In five cases (two women), sprinters reported recognizing their superior sprinting ability beginning in childhood. For example, Marion Jones reported that she was “always fast” and excelled at multiple sports (Jones, 2004) and Tommie Smith reported that he excelled at all sports as a schoolboy (Smith & Steele, 2007).

Sprinters required one to seven years of training to reach world class status (men: median = 4 years, mean = 4.6 ± 2.0 years, n = 7; women: median = 2.5 years, mean = 3.1 ± 2.4 years, n = 8; Table 1, Fig. 1). In fact, eight sprinters qualified for the Olympics as teenagers (Table S2).

For 10 of 15 sprinters there was no evidence that they had participated in organized sports of any kind prior to the recognition of their superior sprinting ability or their initiation of deliberate sprint practice.

The results of this study clearly contradict the DPM: sprinters were consistently fast prior to formal training, achieved world class status in much less than ten years, and, in most cases, their exceptional development cannot be attributed to skill transfer. Nonetheless, this study has two possible limitations. First, the sample size of 15 is modest. Second, many of the individuals became Olympic champions several decades ago. Because world class sprint performances have continued to improve (Seiler, DeKoning & Foster, 2007), this raises the question of whether our results would differ if we used a more contemporary sample of sprinters. Study 2 was designed to address these limitations.

Study 2: Biographies of fastest 20 U.S. males of all time

In Study 2, we examined the development of the 20 fastest male U.S. 100 m sprinters. This is an excellent sample because the U.S. has been one of the strongest sprinting countries since the onset of modern international competition and record keeping (Lawson, 1997). This is revealed by the fact that 14 of 20 of these men won at least one individual World Championship or Olympic sprint medal (100 m, 200 m, or 60 m indoors); four of the others have won at least one relay medal at the World or Olympic championships. Moreover, all of these men achieved performances that meet contemporary standards of world class performance, including the 2012 Olympic A Qualifying Standard (i.e., 10.18 s automatically qualifying them to participate in the Olympic Games; http://www.usatf.org).

We again examined whether these sprinters were exceptional prior to initiating formal training and how long it took for them to reach world class status. We also searched for evidence indicating that these men were unexceptional relative to their peers prior to their beginning formal sprint training.

In addition, we documented the trajectories of performance improvement, particularly the percentage of improvement after age 19. The DPM makes no quantitative claim regarding the magnitude of improvement among regularly training adult athletes. However, the “talent matters” framework implies that once athletes have reached physical maturity and done some formal training, subsequent improvements will be relatively modest.

Study 3: Surveys of collegiate sprinters

In Study 3, we recruited individual sprint qualifiers for the 2012 National Collegiate Athletics Association (NCAA) national championships to complete an online survey. We asked sprinters about their speed relative to their peers as children and adolescents. To address the specificity of their athletic ability, we also recruited a control group, collegiate throwers (e.g., shot put, discus, javelin) who qualified for these meets. The “talent matters” framework predicts that sprinters generally will recall being faster than their peers as children and adolescents than will the throwers. To further address specificity, we also asked about physical strength and overhand throwing ability. We predicted that throwers would recall being stronger and having better overhand throwing ability as youths than would sprinters.

The surveys also allowed us to obtain systematic data on sprinters’ performances in their first season of high school competition, which was generally coincident with their onset of formal training. Again, the “talent matters” framework predicts that sprinters will be much faster than most of their peers even at this early stage in their careers, whereas the DPM does not.

Results and discussion

Contrary to the DPM, collegiate sprinters recalled being faster relative to their peers than did collegiate throwers (Table 3; Fig. 2). This difference was significant and substantial for recollections of 6–10 and 11–15 years of age, and the differences held within men and women (Table 3). In fact, 90% of sprinters reported they were faster or much faster than their peers at 6–10 years of age and 80% reported they were faster or much faster at 11–15 years of age. As we predicted, throwers recalled being stronger and having better overhand throwing ability relative to their peers than did sprinters, and these differences held robustly for both age ranges and within men and women (Table 3). These results corroborate Studies 1 and 2 by showing that expert sprinters consistently recalled being faster than their peers as children. Furthermore, these recollections were at least somewhat specific to sprinting and so cannot be dismissed as a manifestation of general athletic ability.

Table 3:

Recollections of childhood and adolescent athletic abilities of sprinters and throwers who qualified for the 2012 U.S. collegiate track and field outdoor championships.

Ability		Sprinters (n = 20)		Throwers (n = 44)		df a	t b	Cohen’s d
		Mean	SD	Mean	SD
6–10 years
Sprinting
	Total	4.3c	0.8	3.5	1.2	62	2.72**	0.79
	Men	4.3	1.1	3.3	1.3	23	1.78	0.83
	Women	4.2	0.6	3.6	1.0	37	2.00	0.73
Strength
	Total	3.3	0.8	3.9	0.8	62	3.21**	0.86
	Men	2.7	0.5	3.9	0.9	23	3.27**	1.64
	Women	3.5	0.8	3.9	0.7	37	1.58	0.52
Throwing
	Total	3.1	0.7	4.2	0.9	61	4.80**	1.23
	Men	3.1	0.4	4.0	1.0	23	2.25*	1.17
	Women	3.0	0.8	4.3	0.9	36	4.32**	1.49
11–15 years
Sprinting
	Total	4.2	0.7	3.2	1.0	62	3.88**	1.10
	Men	4.0	0.8	3.1	1.0	23	2.22*	1.04
	Women	4.2	0.7	3.3	1.0	37	3.07**	1.10
Strength
	Total	3.5	0.9	4.3	0.8	61	3.22**	0.87
	Men	2.7	0.5	4.1	1.0	23	3.37**	1.73
	Women	4.0	0.7	4.4	0.6	36	1.76	0.58
Throwing
	Total	3.1	0.7	4.6	0.6	62	8.63**	2.25
	Men	3.0	0.6	4.5	0.7	23	4.98**	2.31
	Women	3.2	0.8	4.7	0.5	37	7.06**	2.24

DOI: 10.7717/peerj.445/table-3

Notes:

* p < 0.05.

** p < 0.01.

a Degrees of freedom differ because some participants did not complete all items.

b Student’s t-test.

c Values in table represent scores on surveys on a five point scale with higher scores indicating higher self-rated ability; see Study 3 Methods for scales.

Table 4 provides information for each sprinter regarding their background, onset of training, and best performances. Seventeen of 20 sprinters reported at least one best performance in their first season of high school competition, and only two of these reported they had begun serious training prior to this. Of the 15 sprinters who reported first season high school performances and no prior serious training, 13 of 15 were age 15 or younger at the end of this first season, supporting our decisions regarding age-appropriate benchmarks (see Methods). All 27 performances recalled by these 15 sprinters were faster than 95th percentile benchmarks. Moreover, seven of these sprinters recalled at least one performance faster than the 99th percentile benchmarks, and two of the others recalled performances that were within 0.5 s of 99th percentile benchmarks. These results represent more objective evidence that, relative to their peers, these sprinters were exceptional prior to the accumulation of substantial training.

Table 4:

U.S. collegiate sprinters’ recollections of their onset of training and best performances.

Sex	NCAA division	Age	Began training	Age first year high school	First year high school best performance			Lifetime best performance
					100 m	200 m	400 m	100 m	200 m	400 m
F	I	18	*17a	14	12.7	26.3	NAb	NA	NA	51.1
F	I	22	18	14	NA	NA	59.Xc	11.4	23.4	56.X
F	I	24	*21	*17	NA	NA	NA	11.1	23.1	NA
F	II	23	15	14	NA	26.X	58.X	NA	24.1	53.0
F	III	21	*13	14	13.X	28.X	NA	11.9	26.0	62.2
F	III	22	*15	15	12.7	27.2	NA	12.0	25.4	NA
F	III	23	*19	*15	13.1	26.X	60.0	NA	25.3	56.4
F	III	20	*16	*15	13.5	26.8	59.9	12.4	24.5	55.5
F	III	22	*17	*17	13.X	27.X	59.X	12.0	24.3	55.2
F	III	21	*16	*15	13.X	27.X	61.X	13.X	26.0	56.8
F	III	20	*13	*14	15.X	27.X	62.X	12.4	24.6	58.1
F	III	19	*15	14	NA	28.X	63.X	NA	24.9	54.8
F	III	21	18	NA	NA	NA	NA	12.9	26.0	56.5
M	I	20	*15	14	11.0	22.5	52.9	10.2	21.1	49.2
M	I	20	18	*16	11.3	23.5	53.4	10.6	20.9	46.0
M	I	20	*16	*15	11.2	23.5	54.X	10.9	21.2	46.3
M	II	22	NA	NA	NA	NA	NA	NA	21.5	47.1
M	II	21	*17	15	11.3	22.8	49.9	10.6	21.1	46.1
M	III	22	14	14	NA	NA	57.X	10.6	21.3	48.8
M	III	21	17	*16	NA	22.2	49.4	NA	22.1	47.4

DOI: 10.7717/peerj.445/table-4

Notes:

a * Indicates age estimated from reported grade (e.g., 9th grade = 15 years).

b NA, no data provided by athlete.

c Performances including an “X” after the decimal indicate uncertainty about exact time.

A possible limitation of Study 3 is that the response rate of college athletes was low. However, a low response rate is reasonable because we did not provide athletes with incentives to participate and because we attempted to contact them after the academic year had ended. More importantly, to minimize response bias, we constructed the survey questions to appear neutral to the DPM or the “talent matters” framework.

Elite sprinters are initially remarkable

The first DPM prediction is that elite sprinters should have generally been unremarkable prior to training. Contrary to this, the biographical materials examined in Studies 1 and 2 indicated exceptional initial ability for all 26 world class sprinters for whom we were able to obtain relevant information. Study 3 corroborated this pattern in national qualifying collegiate sprinters, showing that they recalled being faster or much faster than their peers as children. In addition, these collegiate sprinters reported achieving performances in their first season of high school competition that would have exceeded 95–99% of their peers despite the fact that most had begun formal training that same season.

A limitation of these studies is that the use of biographical materials relies on the retrospective recall of information from many years earlier, and this information may be inaccurate or biased (Shiffman et al., 1997), although studies have demonstrated moderately high correlations between information obtained by retrospective recall and that found by examining diaries (Ericsson, Krampe & Tesch-Römer, 1993; Sloboda et al., 1996; Baker, Côte & Deakin, 2005; Ward et al., 2007). Most empirical studies addressing the DPM framework use systematic methods, such as requiring participants to maintain regular training diaries (Ericsson, Krampe & Tesch-Römer, 1993; Sloboda et al., 1996; Baker, Côte & Deakin, 2005; Hodges et al., 2004). Such studies have not been conducted for sprinting, and they seem impractical. This is because, to our knowledge, there are no formal training programs or sports academies that endeavor to train “typical” children or adolescents so that they develop into elite sprinters. Apparently formal, dedicated sprint training is only taken up by individuals who are recognized as being exceptionally fast prior to formal training.

We also note that there are several reasons why inaccurate or biased biographical materials cannot provide a satisfying explanation for our results. First, if many sprinters in Studies 1 and 2 were not exceptionally fast prior to formal training, it would seem that at least some coaches, competitors, or peers would attempt to report the truth to biographers and journalists. For example, if Gwen Torrence and Evelyn Ashford had not, as untrained high school students, each beaten their school’s star (male) football player in a race, as their biographies attested (Hornbuckle, 1995a; Davis, 1992), we might have expected someone to dispute these or the many similar claims in other biographies. Similarly, most biographical accounts of extraordinary youth sprinting ability are corroborated by publicly documented timed performances. Moreover, objective facts such as race results are more accurately recalled than are subjective states (e.g., recalling the amount of effort put into a particular practice session) (Brewer, 1998).

A second point is that Study 3, based on national qualifying collegiate sprinters, fully corroborated Studies 1 and 2. Although it was based on self-reports, the responses were anonymous, and this should have minimized self-presentation bias that might have occurred in the biographies (e.g., champion sprinters desiring to portray themselves as being innately gifted.)

A final point is that DPM proponents have used biographical materials similar to ours to support their key claims. Most notably, Ericsson, Krampe & Tesch-Römer (1993) and Howe, Davidson & Sloboda (1998) reviewed biographies and retrospective studies of a variety of eventual experts (e.g., musicians, painters, chess players) and argued that these indicate that the experts’ initial performances were consistently unexceptional and that many years of deliberate practice always preceded their emergence as experts. Studies 1 and 2 demonstrate beyond doubt that this pattern does not hold for the biographies of most (and perhaps all) expert sprinters. There is no reason to dismiss sprinters’ biographies as highly inaccurate or biased while accepting the veracity of other biographies.

Our studies are also notable because they ruled out two alternative explanations for sprinters’ initially exceptional abilities. In Study 1 we addressed the transfer hypothesis, whereby remarkable initial performance in one domain, such as sprinting, might be due to previous training in another, such as football or baseball (Howe, Davidson & Sloboda, 1998). We showed that a transfer hypothesis is not viable because the biographies of Olympic champions revealed that two-thirds of them did not participate in organized sports prior to beginning sprinting. A DPM proponent might protest that perhaps informal sports participation was crucial for sprinting development. This argument has some validity: informal (and difficult to measure) experience (e.g., play) might be crucial. However, accepting this argument would still entail abandoning the underlying premise of the DPM, that expertise must be based on formal, deliberate training.

With Study 3 we addressed the possibility that remarkable initial sprinting ability might be merely indicative of unusual general athletic ability or early physical maturation. Contrary to this, the collegiate sprinters generally recalled being faster or much faster than their peers as children, whereas another group of similarly elite athletes, throwers, did not recall being as exceptionally fast. That the opposite pattern occurred for physical strength and overhand throwing shows that sprinting ability is not merely a manifestation of general athletic ability; it is specific, at least to some extent, consistent with definitions of talent (Howe, Davidson & Sloboda, 1998). As we will discuss below, however, speed is crucial for many other sports so it is not surprising that many champion sprinters excelled in sports besides track and field.

There are other possible concerns about our tests of this prediction, but none seem compelling. One concern is that, although we found evidence of initially outstanding speed in all three studies, the sample sizes were modest (i.e., n = 26 for Studies 1 and 2 combined; n = 20 sprinters for Study 3). This does not seem like a major weakness, however, because genuine experts are, by definition, rare, which is why many DPM studies use modest samples. For example, the seminal paper by Ericsson and colleagues (1993) included data from only 54 musicians, only 22 of whom were considered expert.

A second, related concern is that our results are based on expert sprinters and perhaps may not apply to the development of sprinting abilities in a general population of athletes. In other words, for most athletes initial performance might explain little or no variation in the attainment of sprinting ability. This point is largely valid: studies of the initial performance and development of sprinting ability with proper training in broad range of individuals are certainly desirable. Nonetheless, this point does not mitigate the challenge our results pose to the DPM. This is because DPM proponents have consistently stressed the study of genuine experts and they have, with no apparent exceptions, assumed that the same principles apply to all individuals, i.e., there is no meaningful distinction (e.g., talent) between expert and non-experts besides in their training (Ericsson, Krampe & Tesch-Römer, 1993; Howe, Davidson & Sloboda, 1998). In addition, we note that Study 3 included many athletes whose best performances to date are far from world class (Table 4); given the base rate occurrence of world class performances and the typical patterns of world class sprinter development (Fig. S1), most of these sprinters never will achieve world class performances. Nonetheless, their initial performances in childhood and high school were generally outstanding. This indicates that the patterns we documented do occur generally.

Elite sprinters break the “10 year rule”

Our studies also contradict the DPM’s prediction that at least 10 years of deliberate practice are necessary to achieve expert level performance. The results of Study 1 showed this because the median time to reach world class status for 15 Olympic champions was only three years. Study 2’s results indicated a median of 7.5 years to reach world class status for the fastest 20 men in U.S. history. This might seem roughly consistent with the 10 year rule, but it must be noted that our estimate of the initiation of formal training was highly conservative. For example, if someone began participating in track and field competitions at age 8 (Table 1), we considered that as the beginning of formal training. However, we found no indication that, as children, any of the sprinters in Study 1 or Study 2 engaged in anything remotely similar to the demanding, time-intensive training that has been documented for future elites in other sports such as tennis, swimming, and gymnastics (Bloom, 1985).

Moreover, similar to other previous studies (Starkes et al., 1996; Kalinowski, 1985; Helsen, Starkes & Hodges, 1998; Hodges & Starkes, 1996) our measure of the duration until reaching world class status confounds training with physical maturation. For example, if a child begins training at age 10 and reaches world class status at age 20, one interpretation is that 10 years of deliberate practice were necessary for this improvement. An alternative interpretation is that elite performance can only occur after physical maturation and that a talented individual could have reached the same performance level at age 20 if they had only begun training at age 18. Although this might seem far-fetched, in the course of our research we discovered two documented cases of men beginning formal sprint training as adults and reaching world class status within one year. The athletic biographies of these sprinters, Dave Sime and Delano Meriwether, are impossible to reconcile with the DPM, and we have summarized their biographies in Table S3.

Moreover, the biographies of Sime and Meriwether patently contradict the DPM’s claim that an athlete who starts deliberate practice at a relatively late age would not be able to “catch up” to an athlete who started training earlier (Ericsson, Prietula & Cokely, 2007; Baker, Cobley & Fraser-Thomas, 2009) and the complementary view that extensive experience in a sport is necessary to reach world class status (Williams & Ford, 2008). We also note that rapid development of sprinting expertise was not limited to men. Female Olympic gold medal winners Helen Stephens, Wilma Rudolph, Evelyn Ashford, and Marion Jones all achieved world class sprinting status within three years of beginning training (Table 1).

DPM proponents might argue that achieving expertise in less than 10 years might reflect extraordinarily intense training, so that, in a sense, 10 years of training might be compressed into nine (or fewer) years. However, the biographies of world class sprinters do not fit the claim of unusually intensive training (see Table S2). For example, Usain Bolt’s biographies (Cantor, 2011; Irving, 2010), including his autobiography (Bolt, 2010), document that his often extreme disinterest in training has been very frustrating for his coaches. In fact, Bolt was left off of the Jamaican national team competing at the 2003 IAAF World Championships because of his lackadaisical attitude (Cantor, 2011).

We also note that sprint training requires high intensity efforts on the track and in the weight room several days per week with the remaining days spent in rest and low intensity recovery exercises (Francis & Coplon, 1991; Ward & Dintiman, 2003; Smith, 2005; Edwards, 2012). Even Ericsson, Krampe & Tesch-Römer (1993) and Ericsson (2006) recognized that very intensive sprint training is difficult for long periods even with periods of rest. Consequently, the time that sprinters spend practicing sprinting may be very brief. For example, Carl Lewis, considered by many to be the paragon of professionalism as a track and field athlete, reported that a typical sprint training session consisted of sprinting 200 m six times averaging 23.0 s for 200 m with one minute of rest between sprints (Hurst, 1994). Keith Roberts, Grand Valley State University’s sprint coach, (pers. comm., 2013) estimates that collegiate and professional sprinters typically spend 600–700 h per year training on the track and in the weight room combined and that high school sprinters would typically spend considerably less time. At the professional rate, it would take between 14 and 17 years to accumulate 10,000 h of deliberate practice.

The DPM has been falsified for sprinting

Is there any way to reconcile our results with the DPM? We do not think there is. As reviewed here, several lines of evidence contradict the DPM, and even the data that seem consistent with it (e.g., some sprinters “needing” ten years of deliberate practice) are easily accommodated into the “talent matters” framework. Furthermore, in our review of biographical materials and discussions with coaches and experts, we did not encounter even one account of an elite sprinter that prototypically fits the DPM (e.g., a sprinter who was mediocre compared to a general population of their peers and who engaged in deliberate practice for many years and eventually became elite). The absence of such accounts, together with the data in our three studies and many complementary physiological and genetic studies (Bouchard et al., 2011; Costill, Fink & Pollock, 1976; Ahmetov & Fedotovskaya, 2012), allow us to conclude with confidence that the DPM has been falsified in the domain of sprinting.

Sprinting is an authentic expertise domain

Although the evidence indicates that the DPM has been falsified for sprinting, it might be argued that this does not seriously weaken the model because sprinting is an inauthentic or highly unusual expertise domain. One version of this argument is that sprinting is inauthentic because performances are highly constrained in that the goal is the same for every performance, to run as fast as possible. By contrast, prototypical domains, such as chess or music, demand far greater flexibility in decision-making and/or motor skills. This argument is unconvincing, however, because DPM researchers have frequently assumed the relevance of other highly constrained domains. For instance, Ericsson, Krampe & Tesch-Römer (1993) discussed how unremarkable adults can be trained to achieve prodigious performance in specific memory tasks, and Ericsson, Nandagopal & Roring (2009) noted that, with proper training, otherwise unremarkable adults can greatly improve their endurance running or complete thousands of pushups per day. Thus, it seems likely that if DPM proponents could demonstrate that most healthy adolescents can achieve outstanding sprint performances (e.g., <11.5 s for 100 m; see Study 2 and Table 3) with only a few months of deliberate practice, they would do so and cite this as evidence supporting the DPM. The fact that such a demonstration is apparently impossible must, therefore, count against the DPM.

A second possible argument is that sprinting is an inauthentic domain because it requires minimal skill. In other words, one might be dazzled by a professional pianist while dismissing a world class sprint performance as something that almost anyone could achieve, albeit more slowly. This argument is also unpersuasive because sprinting does require skill. Elite sprinters exert much effort in physical training (e.g., running, plyometrics, strength training) and technique (e.g., starts, transitions) (Francis & Coplon, 1991; Ward & Dintiman, 2003; Smith, 2005; Edwards, 2012). The resulting improvements may be modest (e.g., 1–6%; see Table 2), yet they can easily make the difference between being a mere qualifier for a championship meet and being the champion.

We also note that sprinting is one of the most popular sports across a broad spectrum of traditional and modern societies (Deaner & Smith, 2013; Gotaas, 2009; Guttmann, 2004a; Guttmann, 2004b; Sears, 2001). Sprinting’s popularity is also revealed by the global fame accorded to Olympic champions such as Jesse Owens (Baker, 1986), Wilma Rudolph (Smith, 2006), Bob Hayes (Hayes, 1990), Marion Jones (Jones, 2004), and Usain Bolt (Bolt, 2010). Furthermore, unlike prototypical expertise domains (e.g., music, chess), no special equipment is needed for sprinting, meaning that, in most societies, it can be undertaken by virtually any individual. These points suggest that sprinting is not merely an authentic expertise domain; it may be an ideal one.

Despite these points, development in other expertise domains, especially those based on decision-making or cognitive skills, could be substantially different than for sprinting and other domains of physical skill (Abernethy, Farrow & Berry, 2003). Nonetheless, there is mounting evidence that much individual variation in achievement in more cognitive expertise domains (e.g., music, chess, educational attainment) also cannot be accounted for by the DPM (Hambrick & Meinz, 2011; Meinz et al., 2012; Meinz & Hambrick, 2010; Hambrick et al., 2013; Wai, 2014).

Sprinting is a foundational expertise domain

We have argued that our results pose intractable problems for the DPM. However, the implications of these results are actually greater than showing that one authentic domain does not fit the model. This is because sprinting is a foundational domain in the sense that elite speed is necessary or at least highly advantageous in many other sports. Coaches and commentators frequently convey this in asserting that “speed kills” and much research supports the point (Huijgen et al., 2009; Little & Williams, 2005). For example, in American football, sprinting speed over 40 yards is a significant predictor of playing ability and highly recruited high school players often possess similar speed to professionals (Ghigiarelli, 2011).

The advantages of superior sprinting speed are particularly well illustrated in cases where elite sprinters take up new sports and reach world class status soon thereafter. For example, an Australian woman recruited to train for skeleton, a winter sliding sport, based on her 30 m sprinting speed reached world class status after only ten weeks of sport-specific training (Bullock et al., 2009). More recently, two American world class sprinters, Lolo Jones and Lauryn Williams, rapidly made the transition to world class bobsled competition. Within several months of taking up the sport, Jones helped the U.S. team win a world championship (http://sportsillustrated.cnn.com/more/news/20130228/lolo-jones-bobsled.ap/). Two years after beginning bobsled training, Jones was named to the 2014 U.S. Winter Olympics team. Her teammate Lauryn Williams, a 2012 Olympic gold medalist sprinter in the 4 × 100 m relay, was named to the 2014 Olympic bobsled team six months after beginning training (http://lauryn-williams.com/).

We also note that while this paper has focused on behavioral data revealing the existence of innate sprinting talent, complementary studies have implicated genetic and physiological mechanisms (Costill, Fink & Pollock, 1976; Macarthur et al., 2006). Studies using similar methods have indicated that characteristics crucial to success in other sports are also partly innate. For example, substantial heritability has been demonstrated for muscle mass and strength, critical contributors to success in power sports (e.g., American football, weightlifting) (Silventoinen et al., 2008; Stewart & Rittweger, 2006) and for untrained VO₂max and VO₂max responsiveness, which are vital for endurance sports (Bouchard et al., 2011; Bouchard et al., 2012). The convergence of this evidence suggests sprinting speed is only one of several innate ingredients that, depending on the particular sport, may be crucial for success.