University of Groningen Staying on track Stoter, Inge

(1)

Staying on track

Stoter, Inge

DOI:

10.33612/diss.113131465

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2020

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Stoter, I. (2020). Staying on track: the road to elite performance in 1500m speed skating. https://doi.org/10.33612/diss.113131465

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

537381-L-sub01-bw-Stoter 537381-L-sub01-bw-Stoter 537381-L-sub01-bw-Stoter 537381-L-sub01-bw-Stoter Processed on: 22-1-2020 Processed on: 22-1-2020 Processed on: 22-1-2020

Processed on: 22-1-2020 PDF page: 1PDF page: 1PDF page: 1PDF page: 1

STAYING ON TRACK

The road to elite performance in

1500m speed skating.

Inge Stoter

See also the three minute animated video about this dissertation on

www.stayingontrack.nl

(3)

Netherlands.

PhD training was facilitated by the research institute School of Health Research (SHARE), part of the Graduate School of Medical Sciences Groningen

Printing of this thesis was financially supported by

• University of Groningen

• University Medical Center Groningen

• Koninklijke Nederlandsche schaatsenrijders bond

• International Skating Union

• Thialf

• Innovatielab Thialf

Paranymphs: Ruby Otter Jöran Stoter

Cover design: CreativeMonkey

Layout and printed by: Gildeprint – The Netherlands ISBN: 978-94-034-2384-5

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic and mechanical, including photocopying, recording or any information storage or retrieval system, without written permission from the author.

(4)

Staying on track

The road to elite performance in 1500m speed skating

PhD thesis

to obtain the degree of PhD at the University of Groningen

on the authority of the Rector Magnificus Prof. C. Wijmenga

and in accordance with the decision by the College of Deans. This thesis will be defended in public on Monday 2 March 2020 at 14.30 hours

by

Inge Klasina Stoter

born on 5 November 1987 in Eindhoven

(5)

Prof. C. Visscher Prof. F.J. Hettinga Assessment Committee Prof. K.A.P.M. Lemmink Prof. G.J.P. Savelsbergh Prof. R.P. Lamberts

(6)

General introduction

7

Performance development

19

Creating performance benchmarks for the future elites in speed skating.

Stoter I.K., Koning R.H., Visscher C., Elferink-Gemser M.T. (2019). Journal of Sports Sciences, 37 (15), 1770-1777.

Underlying performance variables:

37

pacing, technique and muscle fatigue

Pacing strategy, muscle fatigue and technique in 1500m speed skating and cycling time-trials.

Stoter, I. K., MacIntosh, B. R., Fletcher, J. R., Pootz, S., Zijdewind, I., & Hettinga, F. J. (2016). International Journal of Sports Physiology and Performance, 11(3), 337-343.

Pacing

55

Development of 1500-m pacing behavior in junior speed skaters: a longitudinal study.

Wiersma, R., Stoter, I. K., Visscher, C., Hettinga, F. J., & Elferink-Gemser,

M. T. (2017). International Journal of Sports Physiology and Performance, 12(9), 1224-1231.

Technique

71

Changes in technique throughout a 1500-m speed skating time-trial in junior elite athletes: differences between sex, performance level and competitive seasons.

Stoter I.K., Hettinga F.J., Otten E., Visscher C., Elferink-Gemser M.T. (submitted)

General discussion

91

Appendices Summary 106

Dutch summary 109

About the author 112

Previous SHARE dissertations 115

Dankwoord (Acknowledgements) 117

3

Chapter

1

Chapter

2

Chapter

4

Chapter

5

Chapter

6

Chapter

(7)

(8)

General

introduction

Chapter 1

(9)

8

GENERAL INTRODUCTION

Performance in sport is about giving it the best you have. Elite sport performance is about that the best you have is better than the best of your competitors. Many people can compete in a sport, but obviously only few can make it to the elite level. A long road with many years of training and competitive experience precedes elite performance. Proper guidance during that period is of great importance for later success. The ever returning questions in this regard are whom to guide from a young age onward and what goals to work on at different stages of development. To answer these questions, the present thesis aims to unravel the road to elite performance for the 1500m in speed skating. Previous elites will be studied to gain knowledge on speed skating performance, performance development and on the underlying mechanisms of performance. Speed skating is a time-trial sport in which the ultimate performance variable is the time needed to cover a certain distance on a 400m ice-track. Individual races are skated in pairs, with each individual skating in a separate lane. Speed skaters are able to reach velocities up to 60 km/h by adopting a crouched aerodynamic position and by pushing-off in a sideward direction (figure 1). Olympic individual distances range from 500m to 10.000m. The middle distance in speed skating is the 1500m (≈2 min) at which there is a comparable contribution of the anaerobic and aerobic system to the total energy needed for the race (van Ingen Schenau, de Koning, & de Groot, 1990). The 1500m is also known as the key distance in speed skating, as both endurance

Figure 1. Start and finish of a 1500m speed skating competition on a 400m ice-track with illustrations of push-off angle in the frontal plane and knee angle in the sagittal plane.

(10)

1

9

and sprint athletes can prosper at this distance. Therefore, the 1500m will be the focus of the present thesis. Official 1500m competitions start at the age of 13 years, at the age of 19 speed skaters enter the senior competition and the age of winning Olympic gold medals is on average 26 years.

Based on previous research on youth athlete development towards elite sport performance, the International Olympic Committee (IOC) consensus statement encourages evidence-informed approaches to youth athlete development. However to do so, more scientifically based data on sport-specific tests per development phase are needed. Using ecologically valid methods, the road to elite 1500m performance will be studied in the present thesis in order to enable such evidence-informed guidance of the future elites in speed skating.

Testing youth athletes in speed skating

Generally, research on youth athlete development towards expertise focuses on testing various underlying, individual performance characteristics, such as anthropometric, physiological, psychological, tactical and technical characteristics (Elferink-Gemser, Jordet, Coelho-E-Silva, & Visscher, 2011). In literature, research done on performance and underlying individual performance characteristics in competitive youth speed skaters is to our knowledge limited to seven studies. Four out of seven did longitudinal research (de Koning, Bakker, de Groot, & van Ingen Schenau, 1994; Elferink-Gemser et al., 2013; Elferink-Gemser et al., 2015; Noordhof, Mulder, de Koning, & Hopkins, 2016). Performance development of high performing junior speed skaters over one year was found to be related to the psychological skills reflection, intrinsic motivation, and goal orientation (Elferink-Gemser et al., 2013; Elferink-Gemser et al., 2015). Physical and anthropometrical characteristics were studied in relation to four year performance development of the Dutch National speed skating selection, but no relations were found in a research group with 24 speed skaters (de Koning et al., 1994). The fourth longitudinal study focused on the variability in junior performance and the effect of altitude and open or closed ice rinks, but did not consider the relation with individual performance characteristics (Noordhof et al., 2016). Technical and tactical characteristics were not taken into account in the previous longitudinal studies on junior speed skaters. Technique and tactics during a 1500m race were studied in one of the three cross-sectional studies on junior speed skaters (de Koning, Foster, Lampen, Hettinga, & Bobbert, 2005). However, the study did not focus on differences in performance level, but on the experimental evaluation of the power balance model in speed skating (de Koning et al., 2005). The two other cross-sectional studies showed that better performance in high performing junior speed skaters was related with several technical, psychological, anthropometrical and physical parameters. Though, the small and homogeneous groups made it hard to draw solid conclusions (De Greeff, Elferink-Gemser, Sierksma, & Visscher, 2011; Van Ingen Schenau, De Koning, Bakker, & De Groot, 1996). As such, to better

(11)

10

understand performance and performance development of youth speed skaters towards elite performance, more research is needed with longitudinal data and larger sample sizes.

Even though the research on the development of youth speed skaters is limited, there is a wide variety of research on performance and multi-dimensional performance characteristics in adult speed skating. Previous literature on speed skating mainly focused on physiological, technical and anthropometrical characteristics and less on psychological and tactical characteristics (Konings et al., 2015). The present thesis aims to add to the body of literature on elite athlete development by focusing on in-competition measures. The start will be studying 1500m performance development, followed by studying the underlying mechanisms of performance like pacing, technique and fatigue during the 1500m.

Elite performance development (chapter 2)

One of the basic principles in unraveling the road to elite performance is to define what elite performance is. In literature however, there are more than eight definitions of elite or expert performance and even within each definition, the level which defines elite differs (Swann, Moran, & Piggott, 2015). This way the skaters classified as “sub-elites” in the one study could be the same as “elites” in another study, making results hard to compare or interpret. Both from the perspective of science as from practice, there is a clear need for a standard measure to define and interpret performance levels.

A second challenge in studying elite performance development is that only a few athletes make it to the top (i.e. Olympics or World Championships) in a specific sport. Statistical analyses over the long road of performance development is difficult with relatively small and homogenous groups. In order to expand the period of longitudinal analysis, combining information over different generations might be a solution to increase sample size. However, caution should be taken as a sport evolves over time. The evolution of speed skating is illustrated in the improvements of the world records (de Koning, 2010; Kuper & Sterken, 2003; Talsma, 2013). For example, the first speed skater to go under the two minutes on a 1500m speed skating time trial, was Ard Schenk in 1971. In the 70’s this was an exceptional performance of which many thought it would never be matched again. However, the world record is currently 1 minute and 40 seconds, 20 seconds faster than in 1971. Moreover, in the 2018-2019 season over 200 Dutch male and 18 Dutch female speed skaters finished the 1500m within two minutes. So, whether a certain end-time on a specific distance can be considered of the elite level is dependent on the era in which the athlete is performing. As such, the evolution of a sport should be taken into account when aiming to provide the future elites with scientific insight in expected performance development.

(12)

1

11

Furthermore, reaching elite performance is a long-term goal and needs substantial deliberate practice (Ericsson, Krampe, & Tesch-Römer, 1993; Gladwell, 2008). To reach a long-term goal, it is advised to add specific short-term goals (Kyllo & Landers, 1995). For a junior athlete, the long term goal might be to break the world record or to become an Olympic champion. However, there is no literature on what short-term goals a junior athlete should achieve in order to stay on track for this long term goal. In order to enable evidence-based guidance of the future elites, short-term performance goals realistic for the age category and era in which the athlete competes are essential. In chapter 2, these short-term goals will be provided in relation to the long-term goal of becoming an elite speed skater in the future.

2) Creating performance benchmarks for the future elites in speed skating.

Chapter 2 introduces a method to define performance that is independent of calendar year and makes it possible to compare different generations. Using this method, elite performance will be defined. By analyzing the performance development of those who made it to the elite level, age-related elite performance benchmarks and goals will be provided for age 13-26 years.

Underlying performance characteristics – pacing, fatigue and technique (chapters 3-5)

Studying longitudinal performance development of elite speed skaters gains insight in the road towards elite performance and can provide short-term age-related performance goals. The follow-up question is how to reach these goals. In many individual time-trial sports such as speed skating, an optimal energy distribution within a race, so called pacing, is essential for successful performance (Abbiss & Laursen, 2008; Foster et al., 1993). Before finishing the race, all available energy stores must be used, but not so early in a race that a meaningful slow down occurs (Foster et al., 1993). This pacing behavior can be characterized by the velocity profile during the race. During middle-distance events in various sports of a similar duration to the 1500m speed skating (≈ 2min), a fast start followed by a decrease in velocity towards the end of the race is commonly observed (Foster et al., 1993; Foster, Schrager, Snyder, & Thompson, 1994; Foster et al., 2004; Muehlbauer, Schindler, & Panzer, 2010). However, how fast this fast start should be in a 1500m speed skating time-trial cannot be unambiguously concluded based on previous studies (Hettinga et al., 2011; Muehlbauer et al., 2010). Biomechanical models showed that starting faster than skaters are used to would be more beneficial for their performance, although practice shows otherwise (Hettinga et al., 2011; Muehlbauer et al., 2010). Observation of elite senior speed skaters even showed that those starting relatively slower, seem to perform better (Muehlbauer et al., 2010). The biomechanical models were also used in cycling, a sport relatively similar to speed skating concerning velocity reached, body posture and muscles used. In contradiction to the speed skaters, cyclists were able to pace their race close to the predicted optimal pacing strategy by the theoretical model (Hettinga, de Koning, Hulleman, & Foster, 2012). It might be, however, that the biomechanical models were not sport-specific enough for analyses in speed skating.

(13)

12

The body position in cycling is constrained and supported by the bicycle and will therefore be less affected by fatigue than in speed skating, where athletes carry their own body weight. When speed skaters fatigue they often show an increase in body angles towards a less crouched position (de Koning et al., 2005). It might be that the faster start in speed skating creates an earlier onset of muscle fatigue and therewith an earlier increase in body angles, probably affecting performance. No previous studies have been done on muscle fatigue in speed skating. Muscle fatigue is often defined as an exercise-induced reduction in the force-generating capacity of the neuromuscular system (Bigland-Ritchie, Johansson, Lippold, & Woods, 1983) and is generally measured by changes in maximal voluntary contraction. Decrease in maximal voluntary contraction can be caused by both central (at or proximal to the motor neuron) and peripheral (distal from the motor neuron) mechanisms (Gandevia, 2001). These mechanisms might also play a regulatory role in pacing (Roelands, de Koning, Foster, Hettinga, & Meeusen, 2013). To better understand pacing and why speed skaters are not able to benefit from a faster start, the interaction between pacing, muscle fatigue and technique in speed skating as well as in cycling will be studied. It might be that due to the different demands of the two sports, other pacing behaviors are more optimal for speed skating than for cycling.

3) Pacing strategy, muscle fatigue and technique in 1500m speed skating and cycling time-trials.

In chapter 3 the effect of an instructed faster and slower start on pacing strategy, muscle fatigue and technique in both speed skating and cycling is investigated.

Research on pacing behavior is mainly done on senior athletes and it remains unknown whether pacing behavior changes during adolescence. As athletes mature, their energy systems, muscle power and anthropometrics change (de Koning et al., 1994; Malina, Bouchard, & Bar-Or, 2004), probably influencing the profile of pacing that is optimal for 1500m performance (Abbiss & Laursen, 2008; Hettinga et al., 2011). Furthermore, pacing is a goal-directed process, for which reflection, planning, monitoring and evaluation might play an important role (Elferink-Gemser & Hettinga, 2017; Smits, Pepping, & Hettinga, 2014). These psychological skills have been found related to youth athlete development at the highest performance levels (Jonker, Elferink-Gemser, de Roos, & Visscher, 2012; Jonker, Elferink-Gemser, & Visscher, 2010). To further understand the road to elite performance, the development of pacing behavior in junior speed skaters over four years will be investigated. Chapter 4 will retrospectively analyze junior speed skaters, who are at the top of their age category at age 17-18 years. High performing youth athletes, close to entering senior competition, likely develop a profile that is optimal for performance and might already push boundaries of human performance at a young age. Studying these high performing juniors is of additional interest for understanding the concept of pacing.

(14)

1

13

4) Development of 1500-m pacing behavior in junior speed skaters: a longitudinal study. Chapter 4 studies the development of 1500m pacing behavior of elite junior male speed skaters from age 13-19 years with respect to the development of sub-elites and non-elites. Speed skating distinguishes itself from most other time-trial sports with the sideward push-off and maintaining a static crouched position on one leg during the gliding phase. Technique influences the velocity of a skater and therewith performance as well as the pacing behavior (Konings et al., 2015; Noordhof, Foster, Hoozemans, & de Koning, 2013). Small knee angles towards 90 degrees reduce aerodynamic resistance and extend push-off length (de Boer et al., 1987; Konings et al., 2015; van Ingen Schenau, 1982; van Ingen Schenau, de Groot, & de Boer, 1985). However, these small knee angles, together with high quasi-isometric muscular forces during the gliding phase in speed skating, cause a restriction of blood flow (Foster et al., 1999). This results in less oxygen delivered to the working muscles (Hettinga, Konings, & Cooper, 2016). By increasing their knee angles, skaters can actively decrease the restriction of blood flow and therewith increase oxygen delivery to their leg muscles (Foster et al., 1999). However, this is disadvantageous for the aerodynamics of the skaters (de Koning et al., 2005; Noordhof et al., 2013). It might be that athletes regulate the trade-off between oxygen delivery to the working muscles and aerodynamics by changing their knee angles during the race. Previous studies on the 1500m in speed skating showed an increase in knee angles as well as push-off angles for junior (de Koning et al., 2005) and senior speed skaters (Noordhof, Foster, Hoozemans, & de Koning, 2014). The study done in juniors was limited to 8 athletes and no distinction was made between sexes (de Koning et al., 2005). To gain more insight in the possible regulation of changes in knee and push-off angles for both male and female junior speed skaters, an extensive study will be done on technical changes during the 1500m. This in relation to their pacing behavior and development towards the senior level.

5) Changes in technique throughout a 1500-m speed skating time-trial in junior athletes: differences between sex, performance level and competitive seasons.

In Chapter 5 changes in technique, knee and push-off angles, during a 1500-m time-trial will be investigated in elite junior speed skaters in relation to sex and performance level. Additionally, the longitudinal development of technique will be explored to provide perspectives on the development of elite junior speed skaters towards senior level.

The various studies in the present thesis all serve to define the road to elite 1500m performance and understand the underlying mechanisms of 1500m performance for junior speed skaters. Research will be done by monitoring the previous elites, in order to enable evidence-informed guidance for the future elites. The studies in this thesis apply methods that are ecological valid and useful for researchers studying performance development in other skating distances and sports.

(15)

14

References

Abbiss, C. R., & Laursen, P. B. (2008). Describing and understanding pacing strategies during athletic competition. Sports Medicine (Auckland,

N.Z.), 38(3), 239-252.

Bigland-Ritchie, B., Johansson, R., Lippold, O. C., & Woods, J. J. (1983). Contractile speed and EMG changes during fatigue of sustained maximal voluntary contractions. Journal of

Neurophysiology, 50(1), 313-324.

de Boer, R. W., Cabri, J., Vaes, W., Clarijs, J. P., Hollander, A. P., de Groot, G., & van Ingen Schenau, G. J. (1987). Moments of force, power, and muscle coordination in speed-skating.

International Journal of Sports Medicine, 8(6),

371-378. doi:10.1055/s-2008-1025688 De Greeff, M. J. W., Elferink-Gemser, M. T.,

Sierksma, G., & Visscher, C. (2011). Explaining the performance of talented youth speed skaters. . Annals of Research in Sport and Physical

Activity, 1, 85-99.

de Koning, J. J. (2010). World records: how much athlete? How much technology? International

Journal of Sports Physiology and Performance, 5(2), 262-267.

de Koning, J. J., Bakker, F. C., de Groot, G., & van Ingen Schenau, G. J. (1994). Longitudinal development of young talented speed skaters: physiological and anthropometric aspects. Journal

of Applied Physiology (Bethesda, Md.: 1985), 77(5),

2311-2317.

de Koning, J. J., Foster, C., Lampen, J., Hettinga, F., & Bobbert, M. F. (2005). Experimental evaluation of the power balance model of speed skating. Journal of Applied Physiology (Bethesda,

Md.: 1985), 98(1), 227-233. doi:10.1152/

japplphysiol.01095.2003

Elferink-Gemser, M. T., De Roos, I. M., Torenbeek, M., Fokkema, T., Jonker, L., & Visscher, C. (2015). The importance of psychological constructs for training volume and performance improvement: a structural equation model for youth speed skaters. International Journal of Sport

Psychology., 46(6), 726-744. doi:https://doi.

org/10.7352/IJSP.2015.46.726

Elferink-Gemser, M. T., De Roos, I. M., Torenbeek, M., Idema, W., Jonker, L., & Visscher, C. (2013).

Training behaviour in relation to performance development in talented speed skaters: a descriptive study. In M. J. Coelho-E-Silva, A. Cupido-Dos-Santos, A. J. Figueiredo, J. P. Ferreira & N. Armstrong (Eds.), Children and

Exercise XXVIII. The Proceedings of the 28th Pediatric Work Physiology Meeting (pp. 263-267).

London: Routledge.

Elferink-Gemser, M. T., & Hettinga, F. J. (2017). Pacing and Self-regulation: Important Skills for Talent Development in Endurance Sports.

International Journal of Sports Physiology and Performance, 12(6), 831-835. doi:10.1123/

ijspp.2017-0080 [doi]

Elferink-Gemser, M. T., Jordet, G., Coelho-E-Silva, M. J., & Visscher, C. (2011). The marvels of elite sports: how to get there? British Journal of

Sports Medicine, 45(9), 683-684. doi:10.1136/

bjsports-2011-090254 [doi]

Ericsson, K. A., Krampe, R. T., & Tesch-Römer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological

Review, 100, 363.

Foster, C., deKoning, J. J., Hettinga, F., Lampen, J., Dodge, C., Bobbert, M., & Porcari, J. P. (2004). Effect of competitive distance on energy expenditure during simulated competition.

International Journal of Sports Medicine, 25(3),

198-204. doi:10.1055/s-2003-45260 Foster, C., Rundell, K. W., Snyder, A. C.,

Stray-Gundersen, J., Kemkers, G., Thometz, N., . . . Knapp, E. (1999). Evidence for restricted muscle blood flow during speed skating. Medicine and

Science in Sports and Exercise, 31(10), 1433-1440.

Foster, C., Schrager, M., Snyder, A. C., & Thompson, N. N. (1994). Pacing strategy and athletic performance. Sports Medicine (Auckland, N.Z.),

17(2), 77-85.

Foster, C., Snyder, A. C., Thompson, N. N., Green, M. A., Foley, M., & Schrager, M. (1993). Effect of pacing strategy on cycle time trial performance.

Medicine and Science in Sports and Exercise, 25(3),

383-388.

Gandevia, S. C. (2001). Spinal and supraspinal factors in human muscle fatigue. Physiological Reviews,

(16)

1

15

Gladwell, M. (2008). Outliers: The Story of Success. San Francisco, CA: Little, Brown and Company;. Hettinga, F. J., de Koning, J. J., Hulleman, M., &

Foster, C. (2012). Relative importance of pacing strategy and mean power output in 1500-m self-paced cycling. British Journal of Sports Medicine,

46(1), 30-35. doi:10.1136/bjsm.2009.064261

Hettinga, F. J., De Koning, J. J., Schmidt, L. J., Wind, N. A., Macintosh, B. R., & Foster, C. (2011). Optimal pacing strategy: from theoretical modelling to reality in 1500-m speed skating.

British Journal of Sports Medicine, 45(1), 30-35.

doi:10.1136/bjsm.2009.064774

Hettinga, F. J., Konings, M. J., & Cooper, C. E. (2016). Differences in Muscle Oxygenation, Perceived Fatigue and Recovery between Long-Track and Short-Track Speed Skating.

Frontiers in Physiology, 7, 619. doi:10.3389/

fphys.2016.00619 [doi]

Jonker, L., Elferink-Gemser, M. T., de Roos, I. M., & Visscher, C. (2012). The role of reflection in sport expertise. The Sport Psychologist, 26(2), 224-242.

Jonker, L., Elferink-Gemser, M. T., & Visscher, C. (2010). Differences in self-regulatory skills among talented athletes: the significance of competitive level and type of sport.

Journal of Sports Sciences, 28(8), 901-908.

doi:10.1080/02640411003797157 [doi] Konings, M. J., Elferink-Gemser, M. T., Stoter, I. K.,

van der Meer, D., Otten, E., & Hettinga, F. J. (2015). Performance characteristics of long-track speed skaters: a literature review. Sports

Medicine (Auckland, N.Z.), 45(4), 505-516.

doi:10.1007/s40279-014-0298-z [doi] Kuper, G. H., & Sterken, E. (2003). Endurance in

speed skating: The development of world records.

European Journal of Operational Research, 148(2),

293-301. doi:10.1016/S0377-2217(02)00685-9

Kyllo, L. B., & Landers, D. M. (1995). Goal setting in sport and exercise: A research synthesis to resolve the controversy. Journal of Sport &

Exercise Psychology, 17(2), 117-137. doi:10.1123/

jsep.17.2.117

Malina, R. M., Bouchard, C., & Bar-Or, O. (2004).

Growth, maturation and physical activity. (2nd ed.).

Champaign, IL: Human Kinetics.

Muehlbauer, T., Schindler, C., & Panzer, S. (2010). Pacing and performance in competitive middle-distance speed skating. Research Quarterly for

Exercise and Sport, 81(1), 1-6.

Noordhof, D. A., Foster, C., Hoozemans, M. J., & de Koning, J. J. (2013). Changes in speed skating velocity in relation to push-off effectiveness.

International Journal of Sports Physiology and Performance, 8(2), 188-194.

Noordhof, D. A., Foster, C., Hoozemans, M. J. M., & de Koning, J. J. (2014). The association between changes in speed skating technique and changes in skating velocity. International Journal of

Sports Physiology and Performance, 9(1), 68-76.

doi:10.1123/ijspp.2012-0131 [doi]

Noordhof, D. A., Mulder, R. C., de Koning, J. J., & Hopkins, W. G. (2016). Race Factors Affecting Performance Times in Elite Long-Track Speed Skating. International Journal of Sports Physiology

and Performance, 11(4), 535-542. doi:10.1123/

ijspp.2015-0171 [doi]

Roelands, B., de Koning, J., Foster, C., Hettinga, F., & Meeusen, R. (2013). Neurophysiological determinants of theoretical concepts and mechanisms involved in pacing. Sports Medicine

(Auckland, N.Z.), 43(5), 301-311. doi:10.1007/

s40279-013-0030-4; 10.1007/s40279-013-0030-4

Smits, B. L., Pepping, G. J., & Hettinga, F. J. (2014). Pacing and Decision Making in Sport and Exercise: The Roles of Perception and Action in the Regulation of Exercise Intensity. Sports

Medicine (Auckland, N.Z.),

doi:10.1007/s40279-014-0163-0 [doi]

Swann, C., Moran, A., & Piggott, D. (2015).

Defining elite athletes: Issues in the study of expert performance in sport psychology doi:https://doi.

org/10.1016/j.psychsport.2014.07.004

Talsma, B. G. (2013). Performance Analysis in Elite

Sports (Doctoral dissertation). Retrieved

from http://www.rug.nl/research/portal/files/ 14500448/Performance_Analysis_in_Elite__1. pdf

van Ingen Schenau, G. J. (1982). The influence of air friction in speed skating. Journal of Biomechanics,

(17)

16

van Ingen Schenau, G. J., de Groot, G., & de Boer, R. W. (1985). The control of speed in elite female speed skaters. Journal of Biomechanics, 18(2), 91-96. doi:0021-9290(85)90002-8 [pii]

Van Ingen Schenau, G. J., De Koning, J. J., Bakker, F. C., & De Groot, G. (1996). Performance-influencing factors in homogeneous groups of top

athletes: a cross-sectional study. Medicine and

Science in Sports and Exercise, 28(10), 1305-1310.

van Ingen Schenau, G. J., de Koning, J. J., & de Groot, G. (1990). A simulation of speed skating performances based on a power equation. Medicine

and Science in Sports and Exercise, 22(5), 718-728.

(18)

(19)

Relative season best time

(20)

Adapted from:

Stoter, I. K., Koning, R. H., Visscher, C., & Elferink-Gemser, M. T.

(2019). Creating performance benchmarks for the future elites in

speed skating. Journal of Sports Sciences, 37(15), 1770-1777.

Elite performance

development

Creating performance benchmarks for the

future elites in speed skating.

(21)

20

Abstract

Sports performance benchmarks useful to select and guide future elites are limited in literature. The present study introduces a method to enable comparison between sports performance of different generations and creates performance benchmarks for the future elites in speed skating. 1500 m Season Best Times (SBT) of Dutch skaters (1043 females, 1812 males, age 13–26 years), who competed in at least six seasons between 1993 and 2013, were corrected for the prevailing world record (WR): rSBT= (SBT/WR)*100%. Regression analyses showed that the calendar year affected SBT (p<0.01), but not rSBT (p>0.05). Based on rSBT, performance groups were defined: elite (rSBT<110%), sub-elite (110%<rSBT<115%), high-competitive (115%<rSBT<120%), medium-competitive (120%<rSBT<125%) and low competitive (rSBT>125%). Benchmarks were based on the slowest rSBT per age of the elite group. Of the total skaters performing within the elite benchmarks, the elite performance group represented <20% up to age 16 and <50% up to age 21. An out of sample group (n=299) confirmed the usability of the benchmarks. So, by correcting time-trial performance for the prevailing WR, elite performance benchmarks can be made based on multiple generations of elite skaters. The benchmarks can be used to select and guide future elite skaters from age 13–26 years.

Keywords

Speed skating, talent development, world record, sports performance, time trial and athletic performance

(22)

21

2

2 Introduction

The goal of talent development programs in sports is to support and foster the future world champions throughout their youth. Though it is hard and probably impossible to predict the later world champion out of thousands of youth athletes, one thing is certain; the future world champion needs to reach an exceptional high level of performance in order to participate at the world championships to at least have a chance of winning. The question is how these elite athletes get to this high level of expertise and whether we can use their performance development as a benchmark for the future generation.

Speed skating is a time trial sport in which time needed to cover a certain distance is the ultimate performance variable. The key distance in speed skating is the 1500 m (~2 min) at which both endurance and sprint athletes can compete (Foster et al., 2004). Official competition starts at age 13 years, and the age of winning Olympic gold medals on the 1500 m is on average 26 years (Alles Met Sport, 2014). Paradoxically, although the road to expertise is long, longitudinal studies in sports are scarce. One of the few longitudinal studies in speed skating showed that from age 14 to age 18 years no more than 59% of the best performing speed skaters remained at the top of their age category (Wiersma, Stoter, Visscher, Hettinga, & Elferink-Gemser, 2017). Furthermore, performance at age 17 explained only part (9% for female and 36% for male) of performance 3–4 years later (de Koning, Bakker, de Groot, & van Ingen Schenau, 1994). Thus, even within four years, the best adult performers are not necessarily the best in earlier years, highlighting one of the great challenges for talent development programs (Barreiros, Cote, & Fonseca, 2014; Elferink-Gemser, Jordet, Coelho-E-Silva, & Visscher, 2011).

The lack of longitudinal research over more than four years can be attributed to the fact that over time athletes transfer out or quit a sport (Cobley, Schorer, & Baker, 2012). Additionally, as only a few can make it to the top, study groups become too small when the study period increases. To overcome this limitation of group size, multiple generations could be taken into account to gain insight into the longitudinal development of elite athletes. One challenge, however, is that the sport usually evolves over time, for example, by technological innovations or better training techniques that lead to improved performances (de Koning, 2010; Kuper & Sterken, 2003; Talsma, 2013). This evolution of a sport is clearly reflected in the improvement of world records (de Koning, 2010). In 1500 m speed skating, the world record for males has been improved by 11 s from 1993 to 2013 (International Skating Union, 2017). When studying performance development over multiple generations it is important to take this into account.

The present study introduced and tested a method to compare speed skating performance over multiple generations in order to provide more insight into those few who reach the elite level. To

(23)

22

this aim, we (1) compared speed skating performance over 20 calendar years by correcting the time needed to finish the 1500 m by the prevailing world record, and (2) defined performance benchmarks within which all elite athletes perform from age 13 years onward. Finally, we (3) determined how many athletes, including those not in the elite group, performed within the elite benchmarks, and whether performance and performance development can distinguish the elites from the other athletes performing within the elite benchmarks. The ultimate goal of the present study is to provide talent programs with a data-driven and science-based tool to select and monitor those athletes who have the potential to make it to the top and, perhaps even more important, to not miss any potentials by deselecting them at a younger age.

Materials and methods

All official 1500 m speed skating results between 1993 and 2013 were obtained from the Royal Dutch Speed Skating Association (KNSB). Only the results on Dutch speed skating rinks were included to limit the influence of altitude (Koning, 2005). For each speed skater, all Season Best Times (SBT) were obtained. The study was approved by the ethics committee of Human Movement Sciences at the University of Groningen, in the spirit of the Helsinki Declaration. Only those skaters who were in competition for at least six seasons between 1993 and 2013 and had at least one measurement at age 16 or younger were included. As speed skaters start competing in 1500 m time trials at age 13 years, and the average age of Olympic 1500 m champions is 26 years (Alles Met Sport, 2014), the age-range for inclusion of a season was set on 13–26 years. This resulted in 2993 (1102 female, 1891 male) individual skaters with 16,574 SBT’s (9117 for female, 16,295 for male).

Compare performance

The present study comprises performance data over two decades. To make performances from different calendar years comparable, the present study introduces a simple method to correct for evolution in a sport (de Koning, 2010; Talsma, 2013). All SBTs were related to the prevailing world record (WR) of the corresponding sex. Meaning the official world record at the date the athlete skated his SBT. The corrected SBT will be referred to as relative Season Best Time (rSBT) and presented as percentage of the world record (see Equation 1).

(24)

23

2

The effect of the method was analysed using regression analyses before and after the correction, with SBT or rSBT as dependent variables and calendar year as an independent variable. The effect of the method was tested for the elite in age categories 16, 17 and 18 years separately, in which most skaters were represented.

Following the correction for the prevailing world record, skaters were allocated to the elite group when they have at least one SBT within 10% of the world record (rSBT<110%), meaning that they either once or multiple times reached this level, at any age. The 10% limit was based on the performance during the world cup on 10 December 2016 in Heerenveen, the Netherlands, where all competitors skated the 1500 m within 9.6% of the world record (range female 4.1– 9.6%, male 4.0–8.0%) (International Skating Union, 2016). Note that due to altitude effects, the officious low-land world records are above the official world records. In 2017 this was 2.2% and 2.3% above the world record for female and male skaters, respectively. Therefore, rSBT’s below 102% is not expected (Schaatsstatistieken.nl, 2017). To compare the development of elite speed skaters with their competitors, three more performance groups were defined based on the best rSBT per skater: sub-elite (110%<rSBT<115%), high-competitive (115%<rSBT<120%), medium-competitive (120%<rSBT<125%) and low-medium-competitive (rSBT>125%). Table 1 presents the female/male distribution and the number of observations for each performance group.

Table 1. The number of speed skaters and observations for each performance group per sex.

Elite Sub-elite High-comp. Medium-comp. Low-comp.

Female Individuals (n) 63 116 209 212 502

Observations (n) 651 1059 1812 1827 3768

Male Individuals (n) 100 292 418 409 672

Observations (n) 1054 2651 3933 3587 5070

Out of sample validation group

To validate the results of the present study, 10% of each performance group was randomly selected and excluded from the initial analyses and served as an “out of sample validation group”. Using random allocation of numbers per group per sex, 6 female and 10 male elites were excluded. For the sub-elite, high-competitive, medium-competitive and low-competitive there were, respectively, 12, 21, 21 and 50 females and 29, 42, 41 and 67 males excluded from initial analyses and assigned to the out of sample group. The out of sample validation group comprises a total of 299 skaters with 2379 observations.

(25)

24

Elite performance benchmark

To define the characteristics of performance development of the elite group, a performance benchmark was defined based on the maximal rSBT (slowest performance) per age and per sex of the elite group at age 13–26 years. Before doing so, outliers were excluded using box-plot analyses per age category and per sex for the rSBT’s of the elite group, using the interquartile range. This is the range between the upper quartile (median of upper 50%) and lower quartile (median of lower 50%). The values higher than 1.5 times the interquartile range above the median of the elite rSBT’s were identified as outlier. Outliers who appeared to be the first season of a skater and/or a relative poor season for the individual skater, suggesting injury, illness or a decrease in training, were excluded from further analyses, as they were not representative for the general development of elite athletes. In total 42 outliers, 9 female and 33 male observations, were excluded from the elite performance benchmark. Benchmarks were defined for females and males separately, based on the maximal rSBT per age. As it might be that some athletes will reach the 110%WR at an early age, after which they decrease performance in the years following, the elite performance benchmark was pre-set to be monotone. A monotone elite performance benchmark means that with every successive max rSBT lower than the previous, the benchmark will decrease towards the value of this rSBT, but with every successive max rSBT higher than the previous the benchmark will remain at the same value.

Elites vs other performance groups

The total number of speed skaters who performed within the elite performance benchmarks was defined for every age category. As the number of skaters within the elite performance benchmarks can differ per age, further analyses were done for every age category separately. The percentage of the total group within the benchmarks that are represented by the different performance groups was defined for every age category. Furthermore, one-way analysis of variance (ANOVA) was performed to define differences in performance (rSBT) and performance development of the preceding year (delta_rSBT) per group for every age. To determine differences between elites and skaters from the other performance groups, post hoc analyses with Bonferroni correction were performed.

The reproducibility of the results was tested with the out of sample validation group by defining the percentage elites within the elite performance benchmarks, and by repeating the ANOVA’s with the out of sample validation group.

For all tests, significance was set at p < .05. Effect sizes for ANOVA are presented as small (η 2_{= .01), medium (η}2_{= .06), large (η}2_{= .14)(Cohen, 1988).}

(26)

25

2

2 Results

Compare performance

Regression analyses revealed that calendar year had a significant impact on SBT. Standardized regression coefficients for female elite speed skaters were −.569 (p < .001), −.614 (p < .001) and −.588 (p < .001) for age 16, 17 and 18 years, respectively, with calendar years explaining 32.3%, 37.6% and 34.6% of SBT. For male elite speed skaters, standardized coefficients were −.753 (p < .001), −.787 (p < .001) and −.761 (p < .001) for age 16, 17 and 18 years, respectively, with calendar years explaining 56,8%, 62.0% and 57.9% of SBT. For female and male athletes, SBT decreased (meaning improved performance) with increasing calendar years (see Figure 1). After correcting for the prevailing world record, calendar year had no impact (p > 0.05) on performance (rSBT) for female and male athletes. rSBT remained constant over the calendar years within the separate age categories (16, 17 and 18 years). Figure 1 shows an example of the distribution of SBT and rSBT of the elite at age 16 years.

Elite performance benchmark

The maximal rSBTs per age, excluding the 42 outliers, are represented by the dotted black line in Figure 2. The solid black line represents the monotone elite performance benchmarks. Finally, the grey area’s in Figure 2 shows the range of the elite performances within the elite performance benchmark for female and male speed skaters, with the grey solid line representing the mean rSBT per age for the elite group. The benchmark rSBT values per age are presented in Table 2 for female and Table 3 for male.

Elites vs other performance groups

Tables 2 and 3 show per age and per performance group the number of speed skaters within the elite performance benchmarks for female and male skaters. Up to age 16 years, the later elites represent less than 20% of the skaters performing within the elite performance benchmarks. From age 21 years and older, elites represent more than 50% of those speed skaters performing within the elite performance benchmarks.

Differences in performance

Table 4 shows the mean and standard deviation for rSBT for all skaters performing within the elite performance benchmarks, per age and for female and male skaters separately.

Main effects of performance groups for rSBT were found for female age 13–24 years (P < .01, η2_{> .14) and for males age 13–25 years (p < .01, η}2_{> .14). Post hoc analyses revealed differences}

between elites and the other performance groups (p < .05), with better performance for the elite group, for both sexes at all ages, except for elite vs sub-elite females at age 13 and 14 years.

(27)

26

Figure 1. The relationship between calendar years and season best times (upper graph) and relative season best times (lower graph) on the 1500 m speed skating for elites age 16 years.

(28)

27

2

Differences in performance development

Table 5 shows the mean and standard deviation for development in rSBT over one year (delta_ rSBT) for all skaters performing within the elite performance benchmarks, per age and for female and male skaters separately. On average, elite approached the prevailing world record with 1.9% (SD = 4.2) per year, when getting older. No differences in performance development between elite and the other performance groups were found, except for females aged 24 years (p = .016, η2_{= .18 elites vs sub-elites),males aged 14 years old (p < 0.01, η}2_{= 0.03,}

elite vs low-competitive) and 25 years old (p = .032, η2_{= .16 elites vs sub-elites) with more}

improvement for the sub-elite and low-competitive groups. Do note that the sub-elite groups consisted of a limited number of four (female 24 years) and three (male 25 years) skaters.

In the out of sample validation group, 100% of the performances of the female elite and 95% of the performances of male elite were within the elite performance benchmarks. The 5% of the male elite performances, which were worse than the benchmarks, were on age 13–16 years and at age 22 years. Based on the out of sample group, the percentage of the total skaters within the performance benchmarks represented by elites was not higher than 20% up to age 17 years for female and 16 years for male. From age 21 years onward elites represented 50% or more of the total number of skaters performing within the elite performance benchmarks.

ANOVA of rSBT showed main effects at all ages for female and male speed skaters (p < .05). Post hoc analyses showed that female elites performed better than all other groups at age 19 years old (p < .05). Female elites performed better than high competitive and medium-competitive speed skaters from age 14–22 and 24 years (p < .05). Additionally, female elites performed better than medium-competitive speed skaters at age 23 and 25 years old (p < .05). No differences between groups were found at age 13 years old (p < .05). Male elites performed better than all other performance groups at age 17, 19–23 and 25 years (p < .05). Male elites performed better than the high competitive, medium-competitive and low-competitive at age 14, 16, 24 and 26 years (p < .05). Additionally, male elites performed better than low-competitive speed skaters at age 13 (p < 0.05) and better than medium-competitive speed skaters at age 15 years (p < .05).

ANOVA of rSBT_development with post hoc analyses did not show any differences between the elite performance group and the other groups (p > .05).

(29)

28 Female Male 16 20 24 16 20 24 100 110 120 130 140 150 Age (years)

relative Season Best Time (% WR)

Figure 2. Performance development of female (left) and male (right) elite speed skaters on the 1500 m from age 13 to 26 years, with the black line representing the monotone elite performance benchmark, the grey solid line the mean rSBT, and the dotted line representing the max rSBT excluding outliers. The grey area represents the range of the elite performances excluding outliers.

(30)

29

2

Table 2. For female, the total individuals per age group, benchmark rSBT values of the elite group and the total number of speed skaters performing within the elite performance benchmarks are presented. Per performance group the number and percentage of total speed skaters within the elite performance benchmarks are presented.

Female Bench-mark Within benchmark

Age

(yrs) Total (n) (%WR)rSBT Total (n) E (n) SE (n) HC (n) MC (n) LC (n)

13 539 139.8 292 30 (10%) 66 (23%) 93 (32%) 67 (23%) 36 (12%) 14 771 131.4 345 43 (12%) 84 (24%) 121 (35%) 80 (23%) 17 (5%) 15 943 125.8 352 51 (14%) 96 (27%) 126 (36%) 74 (21%) 5 (1%) 16 1006 122.2 292 53 (18%) 97 (33%) 118 (40%) 24 (8%) 0 17 992 119.8 255 55 (22%) 95 (37%) 105 (41%) 0 0 18 966 117.5 182 55 (30%) 87 (48%) 40 (22%) 0 0 19 798 117.5 148 50 (34%) 76 (51%) 22 (15%) 0 0 20 648 116.6 125 50 (40%) 60 (48%) 15 (12%) 0 0 21 488 115.9 83 44 (53%) 38 (46%) 1 (1%) 0 0 22 378 113.8 50 37 (74%) 13 (26%) 0 0 0 23 269 113.8 34 29 (85%) 5 (15%) 0 0 0 24 196 113.8 31 27 (87%) 4 (13%) 0 0 0 25 148 113.8 24 23 (96%) 1 (4%) 0 0 0 26 111 113.8 19 18 (95%) 1 (5%) 0 0 0

E = Elite; SE = Sub-Elite; HC = High-competitive; MC = Medium-competitive; LC = low-competitive.

Table 3. For male, the total individuals per age group, benchmark rSBT values of the elite group and the total number of speed skaters performing within the elite performance benchmarks are presented. Per performance group, the number and percentage of total speed skaters within the elite performance benchmarks are presented.

Male Bench-mark Within benchmark

Age

(yrs) Total (n) (%WR)rSBT Total (n) E (n) SE (n) HC (n) MC (n) LC (n)

13 1125 144.2 611 47 (8%) 184 (30%) 226 (37%) 110 (18%) 44 (7%) 14 1439 138.9 759 67 (9%) 212 (28%) 252 (33%) 121 (16%) 107 (14%) 15 1668 125.8 596 79 (13%) 218 (37%) 226 (38%) 72 (12%) 1 (0%) 16 1774 121.4 544 90 (17%) 236 (43%) 202 (37%) 16 (3%) 0 17 1741 117.7 410 88 (21%) 226 (55%) 96 (23%) 0 0 18 1701 116.1 320 89 (28%) 201 (63%) 30 (9%) 0 0 19 1390 114.9 219 85 (39%) 134 (61%) 0 0 0 20 1088 114.9 171 78 (46%) 93 (54%) 0 0 0 21 861 113.7 127 73 (57%) 54 (43%) 0 0 0 22 654 113.7 94 61 (65%) 33 (35%) 0 0 0 23 473 113.7 69 51 (74%) 17 (25%) 0 0 0 24 374 113.7 50 39 (78%) 11 (22%) 0 0 0 25 262 113.7 28 25 (89%) 3 (11%) 0 0 0 26 230 113.7 21 20 (95%) 1 (5%) 0 0 0

(31)

30

Table 4. Mean and SD for rSBT per age for the female and male skaters performing within the elite benchmarks, presented per performance group.

rSBT (%) (mean ± SD)

Female Male

Age

(y) Elite Sub- elite High- comp. Medium- comp. comp.Low- Elite Sub- elite High- comp. Medium- comp. comp.

Low-13 126.6 127.6 130.6* 133.2* 135.1* 127.7 131.9* 134.7* 136.6* 139.8* ± 6.0 ± 5.3 ± 4.8 ± 3.7 ± 3.5 ± 6.3 ± 6.1 ± 5.3 ± 4.6 ±3.9 14 121.3 122.9 125.0* 127.4* 129.4* 121.8 124.4* 127.0* 129.0* 134.8* ± 5.2 ± 4.1 ± 3.5 ± 2.6 ± 1.5 ± 5.5 ± 4.3 ± 3.8 ± 3.2 ±2.9 15 117.9 119.4* 121.6* 123.8* 125.7* 117.0 119.6* 121.6* 123.6* 125.1* ± 4.3 ± 3.0 ± 2.5 ± 1.5 ± 0.1 ± 4.0 ± 3.0 ± 2.5 ± 1.4 ±0.0 16 115.1 117.2* 119.0* 121.2* 113.7 116.6* 119.0* 120.6* ± 3.5 ± 2.4 ± 1.7 ± 0.7 ± 3.2 ± 2.2 ± 1.6 ± 0.4 17 113.3 115.6* 117.8* 111.1 114.5* 116.7* ± 3.2 ± 2.0 ± 1.3 ± 2.4 ± 1.5 ± 0.8 18 111.4 114.3* 116.5* 109.8 113.8* 115.7* ± 2.9 ± 1.4 ± 0.7 ± 2.6 ± 1.4 ± 0.3 19 110.2 114.2* 116.0* 109.2 113.0* ± 2.7 ± 1.6 ± 0.7 ± 2.4 ± 1.4 20 110.3 114.0* 115.8* 108.9 112.8* ± 2.5 ± 1.5 ± 0.5 ± 2.8 ± 1.3 21 109.7 113.5* 115.1* 108.7 112.4* ± 2.8 ± 1.4 ± 0.0 ± 2.4 ± 1.1 22 108.8 112.1* 108.4 112.2* ± 2.5 ± 1.2 ± 2.2 ± 0.9 23 108.0 112.6* 108.8 112.0* ± 2.5 ± 1.3 ± 2.5 ± 1.0 24 108.2 112.7* 108.5 112.7* ± 2.8 ± 1.5 ± 2.4 ± 1.0 25 107.9 112.7 108.1 112.2* ± 2.5 ± 0.0 ± 2.3 ± 1.6 26 108.1 111.9 107.8 110.7 ± 2.6 ± 0.0 ± 2.8 ± 0.0

* indicates significant difference with the Elite Group (p<.05)

Discussion

The ultimate goal of the present study is to provide talent programs with a data-driven and science-based tool to select and monitor those athletes who have the potential to make it to the top and to not miss any potentials by deselecting them at a younger age. This was

(32)

31

2

done by first correcting the season best times (SBT) on the 1500 m by the prevailing world record, which neutralized the effect of the evolution of speed skating times over 20 years. This simple method enabled us to compare the performance development of elite athletes over different generations. In total 163 elite speed skaters were included and formed the basis of the elite performance benchmarks. Up to age 16, the elite group represented less than 20% of the skaters, skating at or below the elite performance benchmarks. From age 21, the elite group represented more than 50% of the speed skaters performing within the benchmarks. In general, later elite skaters have a better performance than their age-matched competitors performing within the benchmarks. The elite group also remained faster over the years following, by improving performance similar to their competitors. The out of sample validation group confirmed the results, indicating that the elite performance benchmarks can be used as a reliable benchmark to monitor performance development of a small group of skaters in their pathway to elite performance.

The evolution of speed skating between 1993 to 2013 had a significant impact on the 1500 m performance. Calendar year explained 32–38% and 56–62% of differences in SBT over 20 calendar years in the age categories 16, 17 and 18 years for elite female and elite male skaters, respectively. In those 20 years, the world record improved 8.5 s for female and 11.0 s for male speed skaters (International Skating Union, 2017). Correcting SBT’s for the prevailing world record (rSBT) neutralized this effect of calendar year on the performance measure. That the effect of calendar year was excluded entirely after applying this correction was above expectations, as breaking world records is not a continuous process and dependent on individuals, opportunities, new training techniques, and/or innovations like the klapskate which was introduced in 1997 (de Koning, 2010; Talsma, 2013). Previous literature used more complex calculations to exclude the effect of technological innovations and general improvement of speed skating performance over multiple generations (Talsma, 2013). Nevertheless, the statistics and Figure 1 show that correcting for the prevailing world record is a promising method to compare performance over multiple generations. With quantifying performance as a percentage of the WR, the method and results of the present study also have the poten tial to be applied at other distances and in other sports.

The present study used longitudinal data over multiple generations to get more insight into the few athletes who make it to the top. In 20 years, 63 female and 100 male Dutch speed skaters performed within 10% of the prevailing WR, reaching the pre-set “elite” level of expertise. Discussion might be raised about what the elite level is. However, the limited amount of skaters assigned as elite in a country which was world leading in speed skating at that time, shows that 10% above the world record on a low-land ice-rink does select only the best athletes of each generation. Together, the athletes of multiple generations did represent

(33)

32

Table 5. Mean and SD for rSBT_development (1-year change of rSBT) per age for the female and male skaters performing within the elite benchmarks, presented per performance group.

rSBT_development (%) (mean±SD, n)

Female Male

Age

(yrs) Elite Sub elite High comp. Medium comp. compLow Elite Sub elite High comp. Medium comp. Low comp.

14 −6.4 −5.8 −7.2 −7.0 −7.1 −8.5 −8.6 −9.4 −9.2 −11.9* ±3.5 ±5.5 ± 5.4 ± 5.6 ±7.7 ± 7.5 ± 5.3 ± 6.1 ± 5.2 ± 7.0 n = 30 n = 68 n = 95 n = 65 n = 15 n = 51 n = 190 n = 234 n = 105 n = 87 15 −3.6 −4.5 −4.6 −4.9 −7.5 −5.4 −5.2 −5.3 −4.9 −6.3 ± 3.3 ± 3.6 ± 4.7 ± 4.3 ± 3.6 ± 4.0 ± 4.0 ± 3.7 ± 3.4 ± 0.0 n = 42 n = 89 n = 118 n = 68 n = 5 n = 69 n = 207 n = 219 n = 71 n = 1 16 −2.9 −2.5 −3.1 −3.5 −3.5 −3.3 −3.1 −3.9 ± 2.9 ± 2.5 ± 3.0 ± 2.6 ± 2.8 ± 2.6 ± 4.0 ± 2.8 n = 50 n = 95 n = 116 n = 23 n = 81 n = 226 n = 199 n = 15 17 −1.9 −1.6 −2.1 −2.4 −2.0 −2.3 ± 2.2 ± 2.4 ± 2.6 ± 2.4 ± 2.1 ± 2.1 n = 55 n = 94 n = 105 n = 88 n = 226 n = 96 18 −1.8 −1.2 −1.1 −1.6 −1.1 −1.5 ± 2.5 ± 2.0 ± 2.2 ± 1.9 ± 2.1 ± 1.2 n = 55 n = 87 n = 40 n = 89 n = 201 n = 30 19 −1.3 −0.9 −2.7 −0.7 −1.1 ± 2.4 ± 2.7 ± 2.3 ± 2.0 ± 2.1 n = 50 n = 76 n = 22 n = 85 n = 134 20 −0.3 −0.5 −1.9 −0.3 −0.7 ± 2.6 ± 2.0 ± 1.8 ± 1.9 ± 1.8 n = 50 n = 59 n = 14 n = 78 n = 93 21 −0.2 −1.4 −3.0 −0.2 −0.7 ± 3.0 ± 2.1 ± 0.0 ± 2.6 ± 1.7 n = 44 n = 38 n = 1 n = 73 n = 54 22 −0.3 −1.3 −0.1 −0.7 ± 2.6 ± 1.7 ± 2.1 ± 1.6 n = 37 n = 13 n = 61 n = 33 23 −0.9 −1.4 0.1 −1.1 ± 2.7 ± 2.5 ± 2.5 ± 1.7 n = 29 n = 5 n = 51 n = 18 24 −0.2 −4.5* 0.0 0.2 ± 2.1 ± 7.5 ± 1.6 ± 1.0 n = 27 n = 4 n = 39 n = 11 25 −0.5 −1.4 0.2 −2.6* ± 1.9 ± 0.0 ± 1.8 ± 4.2 n = 23 n = 1 n = 25 n = 3 26 −1.4 −2.0 −0.1 −1.2 ± 3.0 ± 0.0 ± 2.0 ± 0.0 n = 18 n = 1 n = 20 n = 1

(34)

33

2

large enough groups to investigate longitudinal performance development of elite athletes over 13 years, which was limited to four years in previous literature (de Koning et al., 1994; Wiersma et al., 2017). Do note that longitudinal statistical analyses could not be done in the present study as multiple generation were included and in competition period could differ per individual from 6 to 13 years. Nevertheless, with the analyses done per age category, the present study was able to build up an indication of the average performance and performance development of the previous elite.

The performance pathway of the elite group was characterized by a fast improvement in performance from age 13–16 years (rSBT_development −6.4 to −2.9%), minor improvement from age 16–19 years (rSBT_development −1.9 to −1.2%), and only a slight improvement from age 19–24 years (rSBT_development −0.1 to −0.6%). The fast development up to 16 years might be explained by the influence of the adolescent growth spurt, or peak height velocity, which occurs around age 11–13 years for girls and 13–15 years for boys (Beunen & Malina, 1988; Philippaerts et al., 2006). During the growth spurt, athletes become taller and stronger very rapidly, which influences their performance. The tempo and timing of the growth spurt varies between individuals (Beunen & Malina, 1988; Philippaerts et al., 2006), which is why performance before age 16 years is unstable and not necessarily representative for later performance.

This can also be seen in Tables 2 and 3, which show that up to age 16 the elite group represents less than 20% of the total athletes performing within the elite performance benchmarks. Though female elite speed skaters, in general, perform better than their competitors after age 15 and male elite speed skaters from age 13 years, the present study recommends talent programs to support all athletes performing within the elite performance benchmarks. This to exclude the possibility that potential elite athletes are prematurely deselected. Based on the elite performance benchmarks a broad group of skaters is suggested to be supported up to age 16 years, with less than 20% reaching elite level of expertise. Nevertheless, this group will narrow down fast towards age 21 years, with more than 50% of the athletes reaching the elite level of expertise once in their career.

The present study is the first that used longitudinal data (over 20 years) to analyse a large group of elite athletes, resulting in a solid benchmark for talent development programs to select the number of athletes to support at different ages. The out of sample group confirmed that the elite performance benchmarks are useful for skaters outside the study group as 100% female elite and 95% male elite measures were within the elite performance benchmarks. Another strong point of the study is that all speed skaters had at least six years of experience, due to which results are less influenced by drop-outs and therewith easier to interpret.