Scaling—Which Methods Best Predict Performance?

• Published in: Journal of strength and conditioning research 2014-06, Vol.28 (6), p.1565-1572
• Main Authors: Comfort, Paul, Pearson, Stephen J
• Format: Article
• Language: English
• Subjects: Adult; Anthropometry; Athletes; Athletic Performance - physiology; Body Mass Index; Comparative analysis; Exercise Test; Football - physiology; Humans; Male; Muscle Strength - physiology; Muscle, Skeletal - physiology; Predictive Value of Tests; Rugby; Running - physiology
• ISSN: 1064-8011; 1533-4287
• DOI: 10.1519/JSC.0000000000000394

ABSTRACTComfort, P and Pearson, SJ. Scaling—which methods best predict performance? J Strength Cond Res 28(6)1565–1572, 2014—Athletes with a higher body mass (BM) tend to be stronger, with ratio scaling possibly eliminating this effect. The aim of this study was to compare relationships between sprint performances with scaled measures of strength and power. Fifteen professional rugby league players (age, 26.27 ± 3.87 years; height, 183.33 ± 6.37 cm; BM, 96.86 ± 11.49 kg) performed 1 repetition maximum back squats, power cleans, squat jumps, and sprints (5, 10, and 20 m). Heavier athletes (forward) generated significantly greater absolute levels of power during the squat jump (5,659.11 ± 710.35 vs.4,740.16 ± 558.61 W; p < 0.001); however, when power data were scaled no differences were observed. Squat performance indicated no differences in absolute ability between the subgroups (190.6 ± 14.25 vs. 205.7 ± 18.35 kg), although the lighter group was significantly (p ≤ 0.05) stronger than the heavier group when using ratio and allometric methods (2.1 vs. 1.9 kg·kg and 10.42 vs. 9.87 kg·kg), respectively. Significant relationships with 5-m sprints were only observed for ratio and allometrically scaled power cleans (r = −0.625, p < 0.02; r = −0.675, p < 0.02), with similar correlations between allometrically scaled 10-m sprint and both back squat and power clean performances. Scaled power clean performances were also inversely correlated with 20-m sprints (r = −0.620, r = −0.638, p < 0.02). Where differences in absolute strength are apparent between individuals of different BM, then the use of scaling is required. Because of the similarity between ratio and allometric methods, simple ratio scaling is recommended.