Star 61, you make two great points very pertinent to the topic at hand
“I haven’t really seen any indication that, for one specific athlete compared only to himself, when that athlete improves from say 10.5 to 9.8 secs in the 100m, that his GCT initiates any further under his body or terminates any closer to his body at toe off.”
“And its always been my interpretation of the convential wisdom that increasing frequency via training is extremely difficult, as opposed to increasing stride length, which can be improved.”
There should be a clear distinction drawn between cross sectional studies which measure/describe differences between stratified groups and longitudinal studies which measure individual changes in qualities across time. Additionally relevant of course are intervention studies which explore improvement due to specific methods (this study type is especially relevant to the latter part of Mike’s presentation).
In evidence or research based practice, all study designs are not of an equal level, and some types of studies are not feasible, most notably well controlled long-term intervention studies in elite athletes. For many reasons, levels/hierarchies of evidence universally list intervention studies at the top. Importantly too, a lack of higher quality evidence does not increase the validity of the lower quality evidence available. Ralph Mann’s data presented in his book, as well as the studies of Weyand (and others) are almost all simple cross sectional studies, and do not really examine the effects of training nor individual changes over time. Mike’s blog asks us what we can learn and apply and I think we probably need longitudinal data and/or intervention studies rather than cross-sectional data and expert opinion to address this topic adequately.
In evidence based practice, a simply descriptive cross-sectional study represents a very low level of evidence, and would very rarely be used to confidently guide clinical practice, and only ever in a very cautious manner in lieu of better quality evidence. With regard to the use of “expert opinion” as a means of evidence, this is usually either regarded as the lowest form of evidence available or disregarded altogether. One could argue that expert opinion in medicine generally involves a much larger sample size and much simpler intervention/questions than that of coaches involved with complex training programs of elite sprinters.
While the opening chapters of Mann’s “The Mechanics of Sprints and Hurdling” do discuss addressing limiting factors in training, the actual data presented in the book refers to observed differences between “poor”, “average” and “good” performers in a cross sectional sense (the performance levels associated with these categories are not exactly defined though you can get the gist from the data). The book does not explore which factors change and to what extent (and how consistently between subjects) as athletes move from one category to the next (if indeed they can at the level of Mann’s subjects).
A very major component of considering the evidence base is relevance or generalisability. This is obviously a huge issue in sprinting as studies on less trained populations may present problems when extrapolated, and tightly controlled, long term training intervention studies on elite athletes (with a reasonable sample size) are often not feasible. However, it must conversely be noted that Mann’s subjects are at a very high level likely approaching a genetic ceiling, e.g. for men the mean Max V values are “poor” = 11.25m/s, “average” = 11.90m/s, “good” = 12.55m/s (p.92), which if held for 10m give splits around 0.89, 0.84 and 0.80. It must be questioned as to how close athletes in the poor or average group are to their genetic, drug free potential. This has implications for the relevance of the observed differences to training practices, as many aspects of the differences may not be very trainable. Mann has succeeded to a large extent in “understanding the characteristics of great sprinters” as his blurb states but the data doesn’t really concern how they got there.
Regarding the question of “when that athlete improves from say 10.5 to 9.8 secs” we must consider that (with the assumption that any athlete can run 9.80 with 0.0 wind drug free) any athlete running this time was likely faster than 10.5 in high school or otherwise very quickly after starting serious training, so we have pubescence and the level/presence of structured, formal training as major confounding factors.
A possible middle ground between non-feasible intervention studies and relying on crosssectional data/expert opinion is longitudinal monitoring of changes in components/qualities and associated concurrent performance changes over time (as most coaches themselves attempt to implement). Additionally, to observe whether particular components were consistently identified as important across the sample or whether significant individual differences exist would be enlightening. I would have to think Ralph Mann has some longitudinal data for individuals, and I would be very interested to know how many (if any) went from “poor” to “good” as senior (> 20 y.o.) athletes. There is a statement in “The Mechanics of Sprinting and Hurdling” which reads as follows :
“The research on sprinting indicates unequivocally that improvement in stride rate is the means by which the better sprinters improve their performance”
However the data in The Mechanics of Sprinting and Hurdling, appears only cross sectional in nature and would have limited inference regarding individual performance improvement(no reference is provided in the book regarding the specific research in question), and as such I don’t think the above statement is directly supported in the text. I think this underlines the issue regarding what constitutes high quality evidence which can confidently be used to directly guide training practices. In summary, in think there are two major concepts at play for coaches looking to guide training/improve performance:
1. The fact that higher quality evidence (well designed intervention studies in a relevant population) is absent or involves unfeasible study designs does not give cross-sectional data or expert opinion increased validity/inference.
2. Longitudinal monitoring of changes in components/qualities, and associated concurrent performance changes over time represents a much more relevant study design for the question of which factors should be addressed in training.
Mann states that “Currently the inability or unwillingness for a coach to actively teach (change) their athletes is the weakest aspect of the current development of US athletes. On the other hand, at this stage of the High Performance Program, it is felt that the “what to teach” area has been comprehensively investigated” (p.24)
I would suggest that the key factors to be focused upon in training cannot really be properly identified from Mann’s (or others’) cross-sectional data, though some longitudinal data must surely exist, and that may suggest otherwise. Again, there seems to be confusion regarding cross sectional data and what can logically be concluded and applied to training.
The other obvious elephant in the room when it comes to breaking speed barriers and the performances discussed in Mike’s presentation is doping. That is probably best left for another time but for me it is almost certainly the most fundamental factor in a discussion of current world records, and the impact of some very interesting recent developments will be interesting to watch.