“And Samson took hold of the two middle pillars upon which the house stood, and on which it was borne up, of the one with his right hand, and of the other with his left.
And Samson said, “Let me die with the Philistines”. And he bowed himself with all his might; and the house fell upon the lords, and upon all the people that were therein. So the dead which he slew at his death were more than they which he slew in his life.”
Judges 16: 29-30
Independent of the ostensible absence of religious or moral tone, in this story Samson applied his architectural knowledge of the function of the pillar columns that support the structural elements of the house above. Assuming that there was equal composition and spacing of the pillars the load would have been distributed evenly. By compromising the pillars of the Philistine temple, Samson would therefore corrupt the load distribution and lead to some or part of the temple to fall. Now I am not sure whether legend is an accurate representation of events, whether Samson did indeed exist, or whether he would have been able to take out a single temple pillar let alone two (Although I do recall witnessing the late Jon Pall Sigmarsson, leaning against promotional car after a strong man display at my home town Bournemouth Athletics club. The warp-warp sound of the driver’s side door panel bowing under the mass of this Viking mountain, was not only a great source of amusement to Sigmarsson but also to the promoter who’s car it was!) without some seriously shoddy buildings work (see www.ratedpeople.com to avoid a similar slewing) or a some rubbing of the African and Arabian tectonic plates. Nevertheless this is my attempt at an introductory metaphor for the components of a structure. Equally I could have called upon similar metaphors; roots and branches of a tree, constructs of a mainframe, pieces of a pie, chapters of a book, sections of an orchestra, etc.
These are determinants, discriminators, constituents, factors, constructs, contributors, components, elements, correlates, predictors, the combination of which make up the whole. For example, we all know that lactate threshold, is a more influential ability for marathon running than it is for 800m running. However, the lazy amongst us will label an event, say ‘endurance’ and then measure and report with an equal emphasis upon the various components without discrimination as to the proportional contribution to the event. A typical example of this is in a physiological test report that often lists from top to bottom, name, event, DOB, mass, stature etc, all the way down the typical grid of the stage test which will no doubt feature all the headline measurements for each stage. For the 800m runner for example, the lactate response is important because of the underpinning fitness and the relation to relative intensities that can be derived for training purposes etc. But it is not the most crucial parameter such as VO2max or anaerobic capacity. Nevertheless, the bog-standard physiologist will produce a mammoth blood lactate vs speed graph, that seems to shout “look athlete I can measure blood lactate AND plot it on a graph, check me out”. In this instance the tone and tenor of the report is not in relation, in context, in proportion to the demands of the event.
So how do we know what is indeed important? Well there are several ways to cut an event up?
· The energetic model – aerobic, anaerobic capacity, anaerobic power etc. (example)
· The parameter model – VO2max, economy etc. (example)
· The performance model – start, turns, finish etc. (example)
Some events lend themselves more so to each one of the models. Swimming for example is tricky to get actual physiological measurements during, but is bound by the walls of the pool – so it lends itself nicely to the performance model. Middle distance running, with the pack running dynamics is trickier to track actual distance performance, so an energetic or parameter model works better. Cycling is infinitely measurable so can often be sliced up by any of these ways and often is by way of extensive SRM measurement.
My first formal foray into trying get a handle on event demands was with the rowing team. Inspired by Tim Noakes’ papers that drew associations between peak treadmill velocity and some other physiological measures with long distance running performance, I set about drawing up similar associations. With the assistance of Greg Whyte, Kate Jones and the bio-statistician Alan Nevill we developed a mathematical model of determinants of ergometer rowing performance. . Beyond the results there were a few surprise results. The maximum power test, usually performed almost as a pre-test ‘chest thump’ was found to be as effective as discriminating the fast from the slow as any other measure, in a predominantly aerobic event. Crucially using methods such as backward step-wise linear regression (don’t worry I am not going to go too ‘statty’ on you), you can also derive beta-coefficients that give you a weighting as to the relative importance (i.e. if you multiply by that factor it can be more influential upon the criterion, in this case ‘speed’). This way you can not only let the mathematics do the choosing for you, find out what is important, but then also find out how important each element is. In this case vVO2max came out top.
To round off the story about my own research, I can tell you not to read it. The linear relationship does not make sense, I’ll tell you why if you haven’t already nodded off at the mention of ‘beta-coefficients’. As an engine (read physiological capacity) increases, speed does indeed increase but not in a straightforward way. Just as your miles per gallon drops inexorably as you go increasingly above 65 mph, as your engine increases, speed increases and so will aero/hydro-dynamic drag. Effectively the faster you go, the better you need to be at pushing air/fluid out of the way. This is often overlooked when you hear those incessantly predictable commentaries “it is now down to who wants it most”. For a start I have never heard anyone tell me they didn’t fancy it that much. Secondly the small margins between athlete performances are magnified by disproportionately larger differences between physiology (according to cubic function – just in case anyone fancies squeezing that little nugget into the 100m final commentary). Hence our second paper, which ostensibly says, “you know that last paper, it was wrong – read this one”.
So not only does determinants of performance allow you to construct a model of performance, it can also allow you to find out which is the most influential performance factors and therefore allow you to develop a profile of strengths and weaknesses for each individual athlete. I am not going to go into the debate about whether you should improve strengths further vs bringing weaknesses up to scratch, but what you could compute is whether a weakness is become a hindrance to performance. i.e. a variable might not be able to develop because another is limiting it or if you remove a pillar from a building – run.
A word of warning, models come and go. After we undertook our middle distance running analysis, showing the importance of VO2max and economy, a number of female athletes ‘moved up’ from 400m to 800m and not surprisingly those athletes didn’t have marvellous VO2maxs and economies. Nearly all of them had been previously been told that their VO2max “was rubbish” (their interpretation, not the actual diagnosis), yet they could run under 2 minutes. The model doesn’t predict accurately because the model was not constructed with them included.
If you don’t fancy doing the math or you haven’t got access to reams of data to play with then at the very least assemble your thoughts, categorise your thinking, develop your model of performance – it will be your reference in the hard times and an outlet in the good times. More importantly gaining an understanding of the determinants of performance, will give you a story to tell. For this reason I put the determinants of performance in my top 10 applications of sports physiology.