Biomchanics and Medicine in Swimming, Jyväskyla, Finland June 1998

Abstract

Key words: Swimming, Freestyle, Disability Sport, Paralympics, Stroke Rate, Stroke Length

introduction

In competitive swimming for persons with a disability a functional classification system is used in an attempt to achieve fair and credible competition. The system strives to relate degree of physical impairment to potential swimming performance. In fact, swimmers with varying impairments compete against one another in one of 10 classes according to scores on muscles testing, range of motion, co-ordination tests and/or on level of amputation. A separate classification is made for crawl (like) strokes (S classes) and breaststroke (SB classes). This functional system which was introduced in the 80’s has evolved and been refined further through discussion among classifiers, coaches, swimmers and officials. It has, however, rarely been the subject of biomechanical study (Pelayo et al., 1996).

Since 1988 video recording have been used to conduct race analyses during Olympic swimming events. Through collaborative efforts with Olympic researchers at the 1996 Atlanta Games, the opportunity presented itself to make video recording during the Paralympic swimming events which followed shortly afterwards. The information gained could not only be used to provide advice to swimmers and coaches, but also for monitoring the classification system. The purpose of this study was to examine certain race components and stroking variables during the 100m freestyle event for men at the 1996 Paralympic Games. More specifically, it was to examine which of these factors contribute to the end race result (ERR), to see how the contributing factors may be different among the classes and between Olympic and Paralympic swimmers, and to determine the implications of these findings for the “fairness” of swimming classification. The 100m freestyle event was of particular interest because it is, next to the 50m freestyle, the only event in which all classes from the most (S10) to the least (S1) functional compete at the Paralympic level.

methods

To compare the performances among groups (Olympic swimmers and Paralympic classes) all times were converted to a point score using the following formula:

Pts = C x (100m TIME)^(-3)

The world record for each group studied was assigned 1000 points and the constants (C) were then calculated (VanTilborgh et al., 1984).

Conversion of performances to a point score was also needed to obtain the normal distribution of data necessary for some forms of statistical analysis. For this purpose, the ERRs for all groups were given a separate point score based on the world record for class S10. Furthermore, CSS, ST, TT and FT could also be assigned a point by extrapolating the time for that race component to a complete 100m distance. Following these transformations a number of the variables measured were combined into indexes relating ST, TT, FT or point score to ERR or CSS.

Descriptive statistics, Spearman correlations and ANOVA’s were computed using the SAS statistical package (p<0.01). Group means were compared using the Student-Newman-Keuls multiple range test.

results

In Table 1, the means and SDs for the variables measured are given for Olympic and for all classes of Paralympic swimmers. Only one swimmer competed in class S1, therefore no results are provided. The significant differences among groups are indicted in Table 2. As expected, the values for all pure performance variables (ERR, race mean CSS=the mean for the four race sections, ST, TT and FT) decreased systematically from Olympic to Paralympic swimmers and then with decreasing functional class. Olympic swimmers were significantly faster than Paralympic swimmers for all these variables, but in no case were the differences always significant between subsequent Paralympic classes. In fact, the most functional classes S10 and S9 never differed from one another, and S8 and S7 only differed in TT. Class S6 could generally be considered as a separate class, while classes S5 and S4, as well as classes S3 and S2, tended to show similar performances for all race components.

difficult (or easy) to reach the final in each of the classes in question. Comparison of race components among equally competitive groups could be more meaningful, whereas, differences between unequal groups could be due more to the differences in performance level than to a fundamental difference between the classes.

The mean class point score as well the values of the eighth place finisher were highest in Olympic swimmers. Nevertheless, the men’s 100m freestyle event for Paralympic classes S10 to S7 appeared to be nearly as competitive as the Olympic 100m freestyle. In Paralympic swimmers, however, the lower classes showed less depth. In all Paralympic classes the drop off in point score of the B finalists was greater than for Olympic swimmers.
Spearman correlations between the CSS for each of the four race sections, as well as ST, TT and FT with ERR and race mean CSS were high in all Paralympic classes. When the first 25m of the race were not considered the correlations were always higher than 0.80. The correlations were lower and not always significant for Olympic swimmers probably because of the smaller range of performance. In all groups there was a slight tendency toward higher correlations between ERR and a race section CSS in later sections of the race. The correlations between FT and ERR (.60 - .88) were always lower than those of the final race section CSS with ERR (.77 - .94). The final race time may in general be decided before the actual finish, although the difference between first or second place individuals could still be determined in the last 7.5m.

In Olympic swimmers the ST appeared to be the most important part of the race (.74), while this was never the case for Paralympic swimmers, especially in the four highest classes. In class S6, however, the correlation for ST with ERR was exceptionally high (.93). The TT and FT tended to be equally important for the ERR in all groups studied. Class S6, however, was again the exception with a high correlation for TT with ERR (.94).

Tables 1 and 2 also show results for the indexes relating start, turn and finish speed to swimming speed. There were in fact no significant differences to speak of. The Olympic swimmers did start significantly faster than class S3, and Olympic swimmers and class S9 started relatively faster than class S2. It is, nevertheless, interesting to note that only Olympic swimmers all turned faster than they swam and thus gained an advantage. In classes S10 and S9, based on the values of the SD, it could also be said that a majority of swimmers gained swimming speed by turning. In class S8 and lower, however, a good number of swimmers gained no advantage by turning. The high SD for this index, as well as that for the start indicated the large diversity of starting and turning ability within the lower classes. There were, however, almost no significant correlations between these indexes and ERR or CSS.

L
astly, Table 1 lists the race means for SR and SL. In Figure 1, the mean SR per group is plotted against the race mean CSS per group with SDs also indicated. In general, classes S2 and especially S3 had slower stroke rates than the other groups. The generally increasing range of SR with decreasing functional class should be noted. SL decreased systematically from Olympic to Paralympic swimmers and with functional class. Almost no significant correlations were found for SR with CSS while SL showed higher (significant) correlation values in more classes.

DISCUSSION

These finding generally support the logic of the functional classification system for this freestyle event. The performances decreased with functional class as expected for all race components. With the exception of class S6, no race component appeared to be more important for success than any other. In class S6, however, the correlation between ST and ERR was exceptionally high (.93). In this mixed impairment class which includes swimmers who can take a diving start with swimmers who start in the water, the good starters appeared to have the advantage. Class S6 was also the exception with a high correlation for TT with ERR (.94). As with the start, those who could push off better had the advantage.

A more complete explanation of the impairment profile of each functional class is not possible here. In general, athletes in classes S10 to S7 have at least two completely functional arms. This may partially explain the fact that there are only small differences in swimming performance among these classes. Most swimmers in the higher classes may also provide for their own transport to and from the pool. Many can train together with able-bodied swimmers, thus training and competing more often. The 100m Freestyle is also the most popular event and together with the 50m freestyle the only one swum by all classes. This all indicates that the crawl stroke most often (but not exclusively) used in freestyle events is easy to learn for all impairment groups and functional classes. This may, furthermore, partially explain why the competitiveness in this event for men, in the highest Paralympic classes, is comparable to the Olympic 100m Freestyle event. The potential for recruitment of Paralympic athletes is, however, obviously smaller than for the Olympics.

Such findings, nevertheless, must be confirmed in women who swim the same event, as well as for both men and women in longer freestyle distances, backstroke and butterfly events. These all take place under the same classification system. The present database is being used for just this purpose. Furthermore, a continual monitoring of the Paralympic and other international disability swimming meets should take place. The diversity of function found in each Paralympic class also makes a closer look at individual results, as well as at the results of groups of swimmers with similar impairments within the same class, essential. This must, of course, be done within the limits of the rules of medical ethics put down by the international governing bodies.

REFERENCES

1. Arellano, R., Brown, P., Cappaert J., & Nelson, R.C. (1994) Analysis of 50-, 100-, and 200-m Freestyle Swimmers at the 1992 Olympic Games. Journal of Applied Biomechanics, 19, 189-200.

2. IOC Subcommission on Biomechanics and Physiology of Sport (1996) Competition Analyses of Swimming Events: Olympics Games, Men’s Events. (available from D.J. Smith, Faculty of Kinesiology, University of Calgary, CANADA)

3. International Paralympic Committee SAEC-SW (1995) Swimming Classification Manual. International Paralympic Committee.

4. Pelayo, P., Sidney, M., Wille, F., Moretto, P., Randhaxe, P. & Chollet D. (1996) Stroking parameters in top level disabled swimmers. In P. Macconnet, J. Gaulard, I. Margaritis, & F. Tessier (Eds), Proceedings of the First Annual Congress of the European College of Sport Sciences: Frontiers in Sport Science, The European Perspective (pp. 156-157) Nice, France: University of Nice Sophia-Antipolis.

5. Van Tilborgh, J., Daly, D., Vervaecke, H., & Persyn, U. (1984) The evolution of some crawl performance determinant factors in women competitive swimmers. In J. Borms, R. Hauspie, A. Sand, C. Suzzanne,& M. Hebbelinck (Eds.), Human Growth and Development (pp. 666-676) New York, N.Y.: Plenum Press.