Table 2 Accuracy of predicting male performance at the international, national, and regional levels using KsA values and other approaches.

From: Special endurance coefficients enable the evaluation of running performance

Dataset

C1

C2

C3

C4

C5

C6

 

Area/contries

World

Europe

GBR

GER

BLV1

SHLV2

 

Performance level

international

national

regional

 

Number of runners

 

9

6

24

77

21

31

 

100m: mean time (s)

400m: mean time (s)

9.89

10.15

10.37

10.52

10.69

10.89

 

44.36

45.40

45.54

46.38

48.27

48.37

Mean ± SD3

100m → 400m

KsA

− 1.47

− 1.20

0.63

0.24

− 2.13

− 0.50

1.03 ± 0.71

Prediction error (%)

VDOT

not applicable for short distances

 

Riegel

not applicable for short distances

 

LOG

not applicable, times for two distances are necessary

Number of runners

42

59

60

100

27

60

 

1500 m: mean time (sec)

211.17

213.74

218.28

220.32

227.80

233.52

 

3000 m: mean time (sec)

450.38

459.96

466.10

473.02

495.26

505.32

 

5000 m: mean time (sec)

775.54

791.10

805.37

815.78

858.04

881.32

Mean ± SD

1500m → 3000m

prediction error (%)

KsA4

1.56

0.66

1.44

0.89

− 0.37

0.10

0.84 ± 0.58

VDOT5

1.47

0.88

1.27

0.84

− 0.66

− 0.06

0.86 ± 0.49

Riegel6

− 2.53

− 3.04

− 2.60

− 3.18

− 4.09

− 4.02

3.24 ± 0.68

LOG7

− 0.73

− 1.16

− 0.57

− 0.90

− 1.12

− 0.54

0.84 ± 0.27

3000m → 5000m

prediction error (%)

KsA

0.09

0.21

− 0.25

− 0.06

− 0.52

− 1.18

0.39 ± 0.42

VDOT

− 0.38

− 0.34

1.01

− 0.82

1.00

1.22

0.80 ± 0.36

Riegel

− 1.62

− 0.94

− 0.46

− 0.67

− 0.52

− 0.33

0.76 ± 0.47

LOG

− 0.66

− 0.08

− 0.73

− 0.60

− 0.91

1.09

0.68 ± 0.34

  1. 1BLV, Badischer Leichtathletik Verband and 2SHLV, Schleswig-Holsteinischer Leichtathletik Verband: Regional track and field federations of Germany; 3The calculation was performed using the absolute values of the error, regardless of whether the prediction was overestimated or underestimated; 4The median KsA values, derived from the annual best times of German male runners between 1980 and 2022, of the respective neighboring (Table 1) or non-neighboring distances were used for the calculations; 5V dot 02 method; 6Riegel´s formula; 7logarithmic power law method. Calculations of the prediction error for the different methods are given in the Materials and Methods section. Underlying dataset/original data: C1–C6 (Tab. S02).