Good books about sports science and statistics are as rare as rocking horse dung. I have read two, back to back, that were both readable and informative: ‘Everest’ by Harriet Tuckey and ‘How to Read Numbers’ by Tom Chivers and David Chivers.
One of the problems with sports science is that it has disappeared down a cul-de-sac of its own making whereby a combination of desperation to ‘Publish or Perish’ and bad writing makes relevant and accessible information beyond the reach of the people who need it the most: the sports coach.
This results in thousands of research papers being unread and coaches abdicating any form of ‘fitness -training’ to a crew of eager but inexperienced undergraduates who lack ‘context.’ This might involve fitness testing players and disappearing with the results or inflicting gym training sessions that are easy to measure but have little to no transfer to the competitive arena.
I am talking about the 95% of the sporting world, not the rarefied atmosphere of Olympic and top-flight professional sport. Both of the books that I have summarised below offer insights into how communication and understanding can be improved between coaches, athletes and support staff.
Everest: The First Ascent by Harriet Tuckey
Fascinating and well written
I can’t remember who recommended this to me but I am glad that I made a note of it. It is the detailed account of the work done by Griffith Pugh, a Royal Army Medical Corps (RAMC) Doctor turned physiologist.
The headline is his work supporting and leading the first successful climb of Mount Everest in 1953. I have no interest in mountaineering as a sport but found this fascinating. The author is Pugh’s daughter and she gives a warts and all account of how Pugh was both thorough, insightful and driven, as well as irritable, aloof and absent-minded.
The book covers part of Pugh’s war efforts helping develop Mountain Warfare equipment and training programmes prior to his Everest expeditions. It then shows how his research helped with:
Cold water survival strategies (Royal Air Force).
Altitude training and acclimatisation (Mexico Olympics).
Heat exhaustion (endurance runners).
Hypothermia and exposure (Duke of Edinburgh award scheme).
All of the above still use protocols developed and suggested by Pugh decades ago,
Pugh’s work as a researcher (not as an overall human being) should be recognised and posted in mind for all scientists working to support others. Sir Arnold Burgen said of Pugh,
‘He has an extraordinary facility for dealing with quite fundamental matters of human physiology in simple terms, applying general physics to them and working them out with little in the way of specialised equipment.’
This all might sound a bit dry and geeky but Tucker brings the man to life and adds her own personal feelings that shed a spotlight on this ‘restless sharpshooter’.
Thoroughly entertaining and enlightening.
How to Read Numbers: A Guide to Stats in the News (and knowing when to trust them) by Tom Chivers & David Chivers.
Superb examples
Don’t panic: this is a book of words with just a few numbers. One of the problems with interpreting ‘science’ is understanding how statistics work. The COVID Pandemic has resulted in a deluge of numbers that threaten to drown us: we either sink or swim to safer, more familiar, shores and allow others to give us a summary.
This is partly due to bad writing and partly due to deliberate ‘massaging’ of the stats to suit a pre-determined narrative. I have got an ‘O’ level in Stats, did Stats in ‘A’-level maths and did a module on research methods (taught atrociously) in my MSc. I still find it hard to understand what is being written.
In 22 succinct chapters, the authors summarise, explain and illustrate the most common statistical methods and flaws that we are likely to encounter. It is eminently readable and enjoyable. It is a classic example of true experts understanding that, ‘You haven’t taught until they have learned.’
I learned a lot.
I couldn’t put the book down (a sentence I never thought that I would apply to a stats guide).
One chapter covers ‘Goodhart’s law’; ‘When a measure becomes a target, it ceases to be a good measure.’
I have seen countless physical measures become targets for poor athletes suffering at the hands of an ‘S&C coach’ or researcher. Often done without the sports coach intervening because they have been bamboozled by numbers and pseudoscience.
Sports coaches: do yourself a favour and buy a copy (or borrow from the library) of this book and start questioning the methods used by support staff.
‘When reading about (or reporting on) targets, metrics and statistics, remember that they’re proxies for the thing we care about, not the thing itself.’
British Fencing once told all its fencers that in order to represent GB they would have to be able to do a side plank for 90 seconds! An example of a measure becoming a target!
Years ago, I saw Jack Blatherwick tear apart the research papers linking weight lifting to 10m sprint performance. The researchers had amalgamated different gender/ age groups and drawn a regression line through the data. They had eliminated the confounding variables that might also affect sprint performance: men run faster than women and adults run faster than teenagers (on average).
Thanks to this book, I can now see what was happening and draw a critical eye on research papers (Dr Robin Williams wrote about dodgy sports science stats ).
I have added this book to my ‘Recommended reading for teachers and coaches’ list. It is a worthwhile investment to help you get a better understanding of how important numbers are in our everyday lives.
The benefits and pitfalls of using technology to monitor training.
No single number can measure your performance
There are many different ways of measuring the work you do in training.
If you are doing a single discipline sport like swimming or running or cycling then you could use a simple metric such as distance covered.
But using just one ‘easy to measure’ number would then mean that your 10km ‘easy’ run on a Sunday would then look a lot tougher on paper (or screen) than your 6 sets of 60-metre hill sprints at ‘full speed‘ on a Monday even though the purpose of each session is different.
With all the recent developments in training technology, much of which is becoming affordable (just) for the recreational athlete, it is easy to get lost in the Data Smog.
In this blog, I will outline a simple measuring tool that allows you to measure your overall training load across all your sessions. It will then look at how to use that information on a weekly and monthly basis to plan your training and reduce the chances of overtraining or illness.
I include the Case Study of a Modern Pentathlete who has to juggle 5 disciplines and her supplemental training and how this tool helps her.
Why measure training load?
The underlying premise when coaching athletes is to plan and measure progress to achieve a goal. A combination of training and adequate recovery allows the body to adapt and become better at doing what it has been practising (1,2).
Too much work and too little recovery lead to staleness, illness and potentially burnout or injury. Too little work and too much recovery mean your performance either stays the same or regresses.
Measuring the work accurately allows each athlete to gain a better understanding of what works for them as long as performance is measured too. Just measuring work, without any idea of performance has little relevance.
A simple example is trying to lose weight. Usually, more work done equals more calories burnt. But an example of unexpected outcomes that occur when just focussing on work was the well-reported study from the University of Pittsburgh that looked at fitness trackers and weight loss (3).
This well-designed study that monitored 471 subjects over 2 years found that those subjects who used the trackers actually lost less weight compared to those with no trackers. The researchers thought that by providing more data on work done, those subjects would be encouraged to do more. The opposite happened.
This study had far more subjects and was conducted over a far longer time than almost all of the studies looking at athletes and training loads. The key message is that the measuring can be a distraction from the performance (weight loss in this case), which is dependent on many other factors.
What about measuring GPS, Heart Rates, Lactate Thresholds or Power Outputs?
I often see people measuring things that appear to have little relevance to performance. They measure things because they can be measured, rather than thinking “Will this help me go faster?”
No one ever won an Olympic Medal for having the best Heart Rate, Lactate or Power outputs. People win medals by crossing the line first, jumping further, lifting heavier things or throwing things further.
Whilst “marginal gains” has become a popular mantra, this only applies to a very few people. Instead, most gains can be made by focussing on one or two big, important, variables that really impact your performance. Measuring those and then manipulating them so that your body adapts will lead to a performance improvement.
There are more than one or two things that affect your performance, and it is easy to get distracted by measuring minor influences because your friend uses this fancy gizmo.
As creatures of habit, many athletes get stuck with a very similar training load each week, or sometimes each day. Having a variety of training, alternating hard and easy days, and adjusting training from week to week are sound training principles that allow sustained progress and adherence.
Monotony in training can lead to injury
If an athlete just trains on feel or believes that every session has to be hard, then burnout, staleness and injury can occur. By monitoring training load, the athlete can see if they have been following the plan, or whether they have got stuck in a rut.
Problems with measuring training loads
Before reviewing some training load options, it is important to understand that the body is a complex organism and many factors outside of training will affect how each athlete adapts. These include:
gender, age, training history, illness, stress, diet, sleep, work, travel, climate, hydration and exams.
If we have two athletes with similar performances in a 5km run of 16 minutes, then a similar training programme may have different results due to the above factors. Monitoring training load is very useful but is only part of the process.
There are three main problems with measuring training loads:
Reliability: Is the measurement accurate at all times and with all users? A set of scales should be easy enough to use by different people from different backgrounds and give accurate results for mass.
Is the same accuracy present when it analyses your body fat percentage? The more complex the measurement, the more room for error. Do pedometers on phones measure distance covered as well as they do steps? (In a sidebar, why measure steps at all?) Be careful about relying on apps.
2. Validity: Is the measurement relevant to your sport?
Is body fat percentage important to you as a cyclist or runner? If so, then those with the lowest body fat percentage would be the fastest. But, is Mo Farah’s bf % low because he trains so much, or can he run fast because he is lean? It is obvious that no runner with 25% bf is going to make an Olympic final, but would 6% vs 7% bf be the deciding factor as both runners are very lean?
A more obvious example would be using a heart rate monitor to measure your gym session. It is irrelevant for what you are trying to achieve unless you are doing a cardiovascular circuit.
3. Transfer: Many of the measurements are modality-specific. Heart Rate is great for endurance work, but this changes according to whether you are running, swimming or cycling. 168bpm means different intensities for all 3 of these activities.
Measuring kilometres covered is highly relevant for cycling, but useless for run speed sessions. 10km would be a tough swim session, but only a warm up for cycling.
Trying to use the same tool to measure different things is common because it is easy. The alternative is often to have different tools for everything, but then you are laden down with data and it all becomes meaningless.
Common Ways to Measure Training Load
Here are some measurement tools with their respective advantages and disadvantages.
To save repetition, remember that these tools are designed to measure a singlecomponent of fitness. Adjusting your training to improve these results, rather than what counts in a race: crossing the line first, is the most common error amongst many athletes.
These have some use for the ‘team-skill sports’ but they do not measure the quality of play nor the effectiveness of your play. I.e. you could produce great numbers on the pitch but be as effective as a deckhand on a submarine.
Distance would be an example of this. It is a great tool, but if your training changes so you add more miles to get better scores, but neglect variations of pace, intensity and terrain, you will limit what performance changes you make. This is human nature.
Heart Rate: A very simple method which requires no equipment except a watch. Heart Rate rises with a corresponding increase in effort and work. Therefore, if all else is equal, the workout with the higher heart rate has required more effort and work done (4).
Very useful for comparing like for like workouts over time such as running a fastest mile in 5:30 with a heart rate of 180 beats per minute(bpm). You can then try and match that intensity in sub sets of 800m or 400m. Or try and run the same mile at the end of the training block at the same pace and see if your heart rate is lower or higher, indicating that your heart has got stronger.
“Heart Rate should be an indicator not a dictator.” Bryan Fish.
There are several disadvantages, some of which are due to a misunderstanding of application. Use Heart Rate to help you pace and judge how you feel based on times and distances. The main error is in estimating your maximum heart rate and then planning sessions around percentages of this fictitious maximum.
Instead, use something like running your fastest possible mile to get a closer approximation of what your maximum heart rate is.
Climate, hydration and stress are three of the factors that can influence Heart Rate and therefore it should not be used as the sole indicator (5) of training load.
Distance: A simple and effective measure, made easier with technology. Great for single discipline sports such as running or cycling. Less effective when comparing across disciplines, useless in the weights room (load in kg would work better).
Lactate: Less common now, but very popular twenty years ago when portable lactate testers became accessible and they are still used in swimming. However, there have been many flaws found partly due to outside factors such as carbohydrate ingestion, preceding exercise and muscle damage affecting results (6). Also, the measurement error that arrives from a pinprick of blood outweighs and potential changes in exercise intensity, so you are looking at very flawed data.
GPS: Useful for measuring distance and changes of pace and speed. Very useful to assess and monitor how you change according to terrain and difference portions of the session (7). Do you run even 1km splits or start slower and finish faster? Disadvantages include interpreting this data and using it adjust your subsequent sessions. If you are doing short sprints or change of direction, this data is less accurate (8).
Power Output (Watts): Used extensively by cycling now, but has zero transfer to other sports. Can quantify the work done in each session and is useful in conjunction with distance covered and speed. It allows the cyclist to see if they are adapting to the training.
A simple but effective alternative
This would be a 10 on the sRPE scale
A simple measurement tool that I use is the Session Rating of Perceived Exertion (sRPE).
This was first ventured in 1998 by researchers in Milwaukee when trying to quantify training load and identify correlations with overtraining (9).
These researchers faced the same problems already identified about measurement, only some of the tools and technology have changed since. They were trying to identify how much training could be done before an increase in illnesses occurred within athletes.
They found that each athlete had a “training threshold” unique to them and that if they trained above it, illnesses were far more likely to occur. Retrospectively, 84% of illnesses could be explained by a preceding spike in Training Load (TL) above the individual threshold.
Subsequent research has refined the detail of the sRPE.
Her is a quick look at 3 of the variables that were collected from their research and that I have used (the 4th Training Strain I have yet to find useful, but it may be for others).
1.Training Load (TL) = sRPE x duration (mins) of the session. Measured by individual sessions and a daily/ weekly total.
2. Standard Deviation (SD) = how much difference there is between the sessions compared to the average session.
3. Training Monotony (TM)= Average daily training load/ SD
Training Load uses a modified scale of Borg’s Rating of Perceived Exertion as the basis of sRPE (10, Table1). Athletes can gauge quite well how hard their session was with only a small amount of practice; it is best done 30 minutes after a session has finished, to allow a more reflective and accurate approach to be taken.
This needs to be measured after every training session and recorded. A daily and weekly total then needs to be calculated.
Table 1 Modified Borg scale of Session RPE
Session RPE scale
0
Rest
1
Really Easy
2
Easy
3
Moderate
4
Sort of Hard
5
Hard
6
7
Really hard
8
9
Really, really hard
10
Just like my hardest race
Standard Deviation is a statistical tool that is necessary to allow us to calculate the Training Monotony. I was a bit concerned about the Maths at the start, but setting it up on the spreadsheet was easy enough using the inbuilt formulae. Standard Deviation was the only part that needed refreshing in my memory, having last used it 30 years ago.
Training Monotony is a very important number that shows how much or how little variation occurs within the training. The dangers of monotonous training were first found in racehorses but have since been found in endurance athletes too (11-13). There is a psychological component to Overtraining and doing the same type of training too often with little variation appears to be a big factor.
The TM figure should be as close to 1.0 as possible. This shows that you have lots of variation between your days. On paper, you may plan a variety of sessions, but each day and potentially each week could end up being very similar in Training Load and therefore you have Training Monotony.
The advantage of sRPE is that you can compare effort across different types of sessions: running, swimming, weight training and cycling. This allows an overall look at the total work done in a week, rather than adding up different forms of data from individual sessions and trying to make sense of it all. This is especially useful in multi-discipline athletes as you will see in the case study.
Case study
Modern Pentathlon: shooting, riding, swimming, fencing and running.
A 23-year-old Modern Pentathlete who has recently started full-time work.
Previously she was able to rest in between sessions and manage her week around training. Now she has to train before and after work and at weekends. Concerned with how this may add to the overall load, we decided to try using sRPE to monitor load and variety.
Here are her results from two consecutive weeks of training, and comments about how we have adapted as a result.
Week 1
Day
Session
Duration
SRPE
Session TL
Daily TL
Monday 5th
swimming
50
4
200
470
weights
40
6
240
shooting
15
2
30
Tuesday 6th
swimming
45
5
225
395
running
35
4
140
shooting
15
2
30
Wednesday 7th
swimming
45
4
180
390
weights
60
3
180
shooting
15
2
30
Thursday
swimming
60
7
420
620
running
40
4
160
shooting
20
2
40
Friday
swimming
35
5
175
355
riding
60
3
180
Saturday
running
45
4
180
180
Sunday (rest day)
riding
90
3
270
270
Summary
Total TL
2060
Average Daily TL
294.29
SD Daily TL
140.65
Training Monotony
2.09
Week 2
Day
Session
Duration
SRPE
Session TL
Daily TL
Monday
running
40
6
240
620
weights
50
7
350
shooting
15
2
30
Tuesday
swimming
50
7
350
605
shooting
15
2
30
running intervals
45
5
225
Wednesday
swimming
50
9
450
730
weight lifting
50
4
200
Speed drills
20
4
80
Thursday
running
40
6
240
490
riding
50
5
250
Friday
running
40
4
160
180
shooting
10
2
20
Saturday
swimming
60
3
180
660
fencing (competition)
120
4
480
Sunday
0
200
riding
50
4
200
Summary
Total TL
2995
Average Daily TL
427.86
SD Daily TL
222.26
Training Monotony
1.93
The main difference between Week 1 and Week 2 was the fencing competition on the 2nd Saturday which she won. The idea was to use this as a “tough” training environment, but in the end she won comfortably.
Other differences can occur with seemingly small changes. For example, by adding 10 extra minutes to her Monday weights, with a small increase in intensity, the actual training load increased by 46%!
Week 1 Monday weights
40
6
240
Week 2 Monday weights
50
7
350
A similar thing happened on the following morning’s swim session, resulting in an increase of 55%!
Week 1 Tuesday
swimming
45
5
225
Week 2 Tuesday
swimming
50
7
350
What should happen is that looking at these two increases, a corresponding decrease should take place on the Wednesday. Instead, she carried on as normal and actually increased the workload due to a tough swimming session, going from TL 390 to TL 730 a whopping 87% increase.
But, going into the competition she did reduce training somewhat over the Thursday and Friday in week 2 compared to week 1 and got the desired result.
Riding a horse is not ‘rest’.
The training monotony is too high for both weeks (1.93 and 2.09); we need to move this closer to 1.0. There is only one really hard day with TL over 700, but several in the 600s, none in the zeros, 100s or 500s.
The “rest day” of riding in week 1 was not resting enough with TL 270 due to the amount of time spent on the horse. We shall look closely at how to add more variety, and I will reinforce the need to “rest” and that easy means easy.
There are three key points for readers to note from this case study:
1. Small individual changes can make a big difference in total over the week.
2. The results must be recorded immediately and looked at; so that changes can be made to the following day’s training.
3. Training Monotony creeps up on you if you are not careful. Whilst individual sessions are different, the TL needs to differ day-to-day too.
Summary
Technology is developing rapidly and so is its availability to recreational athletes. It is easy to get caught up in measuring things without understanding if and how they affect your overall goal. Using sRPE may be a simple tool that allows you to get an overall picture and ensure that you have plenty of variety in your training.
This should be used to compare results on an individual basis, rather than using it to compare athletes. There is no “one size fits all” training plan. Athletes respond differently and so the loads will differ for optimal training.
Alternating easier and harder days is a fundamental training principle, with hard being hard and easy being easy. This will allow you to train effectively over longer periods of time which then leads to better results.
Last month I attended Vern Gambetta’sGAIN conference in Houston, Texas. A great mix of practical sessions, seminars and informal idea sharing, it is my annual chance to take time out and immerse myself in learning.
I shall be sharing some of the ideas and insights learnt this year. The act reviewing what happened and disseminating that into a hopefully useful blog post is part of my ongoing learning.
Today I start with Peter Weyand’s second seminar which was a great overview of the scientific process and how things stand in this millennium.
Sorting Sport Science in the Digital Era
In the last millennium, there was little or no information available to sports coaches. Peter said that much or most of what is available now is “shaky”.
Here are his 5 “Drivers of Disinformation”:
Proliferation of Information Outlets (Instagram, Facebook, YouTube, Podcasts and Twitter).
Volume of data and literature being produced (wearables and new technology).
Poor quality research training.
Pressure to publish (anything).
Self-Promotion (Not all bad, helps share ideas, but often results in self-citations).
This results in “literature pollution” and disinformation. Peter said that “laziness is the default intellectual condition”.
It is hard to filter what is good or useful in this age. In fact, “Computers don’t reduce work, they create more of it” (Peter Taylor, 1994).
So how can busy sports coaches develop a filter and understand what will work best for their teams and athletes?
The Scientific Method
Two years ago I was asked to present a CPD event to physiotherapists in Exeter. I gave my thoughts and observations on using motor skills learning in rehabilitation so that patients are working towards useful (and interesting) outcomes. At the end, one physio asked “Yes, but what about the science?”
“The science”? As if there is one thing that is all encompassing, this from a person with a science based degree showed a lack of understanding of the scientific process. Many coaches have no formal scientific background, but can still follow the scientific method.
Peter laid it out very well, and these principles will help you as a coach develop a filter.
Get an idea or question.
Make observations.
Analyse observations.
Idea supported: Yes/No?
Peter suggested that good researchers ask good questions and then look to first principles for answers.
Step 1: The research question must be good.
Step 2: The hypothesis must be testable. The design of the study must yield data that will “get out of the noise”.
Step 3: Analyse the observations in the right way. Peter used several examples to illustrate what works/ doesn’t work.
Step 4: Proving and disproving: how well does data support the idea?
An interesting point was that an idea can never be proven true! Instead, the scientific method can only disprove. It only takes one outlier or piece of data to disprove a theory: the exception.
For example, Peter was studying sprinters in action and a common hypothesis was that symmetry between limbs was needed. One sprinter had a big asymmetry and yet was very fast. This one individual therefore disproved the symmetry hypothesis. Other factors must be important in sprinting.
Degrees of Uncertainty
In the past I have often got confused about what is presented as “research” compared to “theories”. This is especially true in ideas like Long Term Athlete Development (LTAD), where many papers are published stating that this latest version is the definitive answer.
Peter helped me understand better the hierarchical language of degrees of certainty.
1: Hypothesis (an idea).
2: Model (LTAD is an example).
3: Mechanism.
4: Law (Gravity). Hard to argue with this.
(Peter may yet to have dealt with “Mum Chat” or “Bloke down the pub” which trumps all of the above! No matter what I do to try and help educate parents, they prefer to listen to their friends).
Conclusions
This presentation really helped me understand the scientific method (much more so than a whole module of “research methods” at Brunel University whilst studying for my MSc).
“If you cannot explain the conclusion in 1-2 sentences, you will never reach a general audience”. I would add that if you cannot explain the conclusion succinctly, you may be unclear yourself as to what is happening.
Isaac Newton
Peter used Isaac Newton as an example of making a big subject very simple. Newton expressed his 3 laws in simple terms and then came up with a very simple equation F=Ma.
When doing research (that includes looking at your own teams) it is important to “Get the big stuff and keep moving” (so much for “marginal gains”). Find out what matters most and look at that.
When reading research “It’s critical to be critical”.
Check the scientific method of the paper:
1: Is the idea supported Yes/ No and does it have a value?
2: Is it testable?
This will then help you decide whether to try and implement some of the ideas into your own practice.
Peter’s whole talk was illustrated with examples of his research and that of his colleagues. I was impressed with the detail he goes into, how much work and effort is required and also how he explained it.
Said Dr Mike Joyner at the head of his 2nd seminar at GAIN. His talk covered four key questions we need to ask before implementing a new scientific find in our training, as well as interesting insights that he has found useful.
In a discussion the evening before, Dr Joyner had revealed that “40% of medical evidence turns over every 15 years, but certain fundamentals don’t turn over”. That means that every 15 years 40% of what was “evidence” changes!
The fundamentals that are constant are: Don’t smoke; access to clean water; don’t get fat and be physically active amongst others.
As coaches it is easy to get distracted by new things and ignore the fundamentals. Similarly we may feel obliged to chuck out what is working because something new is found and published, even if that is later to be found false. (Naseem Taleb talks about this in Antifragile, he calls it neoism).
4 questions we should ask of sports science
Dr Joyner is an expert on athletic performance and is based at the Mayo clinic. His talk was extremely useful and was an example of critical thinking. (This is supposedly taught at Universities, but yet many recent graduates blindly regurgitate “facts” based on “research” without appearing to question it). Dr Joyner went through the following questions we should ask and gave examples of each.
Is it measureable?
Max Oxygen uptake is measurable. However in a laboratory setting research needs to show a 1-5% improvement for the study to be valid. If you win a 10km race by 1% you win by 100m! Coaches are often looking for the 0.1 -1% Science can explain the big picture, but it sometimes misses the detail and often the context.
McNamara’s fallacy
During the Vietnam war, the USA decided to try and measure winning by counting body bags of US troops versus the Viet Cong. This became known as “McNamara’s fallacy” where this became the focus of politicians rather than a meaningful political-military strategy (Assuming that anyone can actually “win” a war).
Is it meaningful?
There is a good correlation between a runner’s Lactate Threshold (LT) and their Marathon speed. Therefore LT is both measurable and meaningful for Marathon runners. (I have seen this extrapolated to Judoka who have been told to “improve their LT” by running on a treadmill more. Here the sports scientists were getting the tail to wag the dog).
Is it actionable?
Bud Winter quote
Referring to Bud Winter’s book “Relax and Win”, Dr Joyner said that relaxation is a trainable effect. Therefore we can use it in our sessions.
“If you just turn your training sessions into exercises and suffering, you’re missing the point.”
In swimming, every turn counts, so it is important to work on each move in a meaningful way. If this (and the dive) are ignored or paid lip service to, then the performance will suffer.
Is it durable?
Dr Joyner showed a list of diets and the research that shows if they affect weight loss. Guess what? The Atkins diet, the Zone diet, weight watchers and the Ornish diet ALL work. They work IF they are followed. The problem is that the really restrictive diets that stop people living normally like eating as a family or choosing from a restaurant menu are simply unsustainable.
Any training programme or new piece of research must be durable and last beyond 6 weeks (the length of many studies) in order for it to be effective in the long term. Think accumulation of training rather than blitzing.
What sports science can do for us
lactate threshold testing
So after quite a critical look at some urban legends and poor examples, Dr Joyner then gave some examples of what we can learn mainly for endurance type activities (where his interests lie). Lactate Threshold in untrained subjects is about 60% of their maximum effort. In trained subjects it ranges from 75-90% of their max.
LT is highly trainable. The increase in mitochondria means more pyruvate is oxidised and less is shunted to lactate. “Almost anything you do that has frequency, intensity and duration” will make a difference to your Vo2 max and LT. For example:
3-5 minute repeats will help VO2 Max (see yesterday’s blog on The Volume Trap)
200m repeats will help improve LT.
Running economy in the other hand is highly variable (up to 30%) and it is unclear how trainable it is. (I always question the research on this: it is often done on treadmills and the “interventions” bear no resemblance to exercises that I do with runners to improve their technique. Conclusions are then drawn that it doesn’t work, rather than “we don’t know how to coach in a lab”.)
Dr Joyner then looked at the recent attempt to run a sub 2 hours Marathon by Nike. What did they do to try and get this time? They looked at all the small factors added together. The course, the temperature, fuelling the runners, as well as manipulating the running economy with drafting, pacing and of course the shoes.
What was interesting here was the effect of drafting (something cyclists in a peloton know) with 8% of the total energy cost of a 5km race coming from having to overcome wind resistance. In the 100m sprint this rises to 16%!
Training in a fasted state
Trained fasted state every morning at GAIN
Sports science can help us identify potential limits to human performance too. Much research has been conducted on training in a fasted stated. However, Dr Joyner made the point that so many gels are used nowadays that people rarely train in a truly fasted state.
People can fatigue from having low glycogen in the muscles or from Neuroglycopenia (Neuro= Brain, Glyco= Sugar, Penia =deficiency. So, low brain sugar). People who fast and go low on Carbohydrate (CHO) down regulate their enzymes. When they return to a normal diet, their enzymes are less able to process this food.
The impression I got from this was that that maybe we should just try to eat normally. Especially when sprinting and doing high intensity exercise: you need CHO.
The Scientific Process
I haven’t really given Dr Joyner justice due to my poor notetaking and poor grasp of physiology. However, please take away the thoughts on questioning research and what you are measuring.
I asked him at the end about “Science” which now seems to be only valid if published, versus the “scientific process” which we should all be doing as coaches wanting to improve our athletes’ performance.
He mentioned the “Citizen science” project which is about sharing ideas that work and testing them. I suggested we have an aide memoire or checklist to help us validate what we do or discard practices that are defunct.
His final words were “A lot of sports scientists are just data acquisition people and analysers”. We were in agreement that it is what we do with this information with real people that counts.
Dr Mike Joyner is a faculty member of the Mayo clinic specialising in human performance physiology.
I met him early on the Wednesday morning when he was attempting to roll around on the floor and get up despite his very long levers. What impressed me was his effort and concentration in attempting a new task, no matter how difficult.
We then had a great conversation over breakfast about long term athlete development, fundamental tumbling skills and education for those from a less than ideal background. Fuelled by his enthusiasm and some pancakes and coffee, I was primed to learn his thoughts on sports science.
Here is a summary of his key points
Lactic Acid Makes Me Sore. Lactic acid is removed in 40/50 minutes post exercise. Active recovery does help this process, but ALL lactic acid is gone within 24 hours. Soreness after training is due to muscle damage.
Sports drink and glucose are necessary. There is no effect of glucose ingestion until after 60 minutes on steady state exercise. Longer duration bouts of exercise may require some. There are many different variables including; the duration and intensity of the bout factors, and the nutritional status before exercise.
Most studies are conducted early in the morning when the athletes are fasted, so extrapolating this to afternoon exercise may be tenuous. The 2% reduction in bodyweight due to dehydration DOES impact performance, so hydration matters. Rinsing out the mouth with sugar can affect performance positively: it is like “brain candy”.
It must be genetic.Size is the obvious example where genetics matter (I would say gender matters more) but there are only a few examples of what Dr Joyner calls “O. Athletes” (Knockout).
Christian McCaffrey
An example of breeding would be Christian McCaffrey (drafted by Carolina Panthers) whose dad was Ed McCaffrey (Giants, 49ers, Broncos) and his maternal grandfather was Dave Sime who was an Olympic silver medallist in the 100m in 1960.
Otherwise, studies have found little evidence for a “talent gene” except for some with ACTN3 and ACE genotypes for speed. There is no evidence for gene testing in young people to “predict talent”.
Dr Joyner said there is a lot of “lazy thinking” about genetics. He then showed a slide with the headline “There are more mile/ 1500m world record holders from Kansas than Kenya”!
The DNA variables would need to explain: Energy systems, muscle fibre type, superior coordination, body composition, motivation, psychology and trainability. They would NOT explain social factors.
To stretch or not to stretch? There is a vast amount of evidence on this, and it is all contexts specific. I made the point that a lot of the studies are asking the wrong question. “Does stretching before exercise prevent injury?” and then tested on military recruits before doing a 20 mile route march with kit in boots. Stretching is obviously the least important factor in that context.
Altitude Training. The 1968 Olympics played a key role in the development of this research as for the first time athletes would be competing at altitude on a big scale. There is a need to compare the short vs long term effects due to the initial reduction in training quality.
Dr Joyner says the data on Live High- Train Low is “all over the place”. The long term effects of living at altitude are an increase in lung capacity. But, you have to keep the training quality up. Those who used altitude training successfully did a lot of short intervals to maintain quality.
Some key points he asked us to consider were:
Beware of individual variation; more is not always better; give it time to work; beware of effects on intensity training and volume; recovery is sometimes affected due to a reduction in sleep quality.
My programme is better than your programme! Dr Joyner showed a video clip of one of the Olympic middle distance races (forget which) where the top 3 finishers were very close. All 3 of those runners followed very different training programmes: high mileage or high intensity intervals and so on. Yet, all 3 were effective.
The idea that one programme is inherently better than another is flawed. In strength training research it isn’t so much the number of sets vs reps it’s the training to failure that is important. As long as intensity is involved, gains will be made in strength.
Dr Joyner then showed video clips from the “Miracles of Men” ESPN documentary of the Soviet Ice Hockey team doing some very basic “old school” training in a gym. The imagination and variety of exercises was novel but the players were working hard too. (This clip can also be seen in the Red Army documentary on Netflix).
He also showed the clip of the La Sierra High School training programme of the 1960s
and what 15 minutes a day can do to form the foundation that is lacking in today’s youth.
Today’s athletes are better. More people are competing today, with better financial incentives, so records tend to fall. Doping has also had an impact on some performances too.
However, some of yesteryear’s performances were pretty impressive. Don Lash, in the 1930s, set the 2 mile record of 8:58.4 on a weekly mileage total of 25 miles.
This comparison of Andre DeGrass vs Jesse Owens shows the difference in track and shoes between 1936 and 2011
Dr Joyner also showed how innovation changed standard practices and protocols. Everyone knows about Dick Fosbury, but at the same time Debbie Brill, a 13 year old girl, was doing the same technique. Both of them were able to try this because of better landing surfaces on the other side of the bar.
Summary
In the discussion that followed Dr Joyner summarised with “Get kids out, have fun, spend time with good coaches”. (That sounds a lot like what we are trying to do at Excelsior ADC).
This was a refreshing and engaging discussion, which I have only briefly touched upon. I spoke to Dr Joyner about academics preaching to each other from Ivory Towers without actually coming into contact with real people in the real world. He said “That’s why I practice Medicine one day a week, so I stay in touch”.
Next Up: Steve Myrland’s “Coaching better every day” about creating a culture.
“The average number of readers of a scientific paper is…”
(Answer at the bottom of the page). Sir Martin Rees in his book “Before the beginning: our universe and others” discusses science, evidence and why information fails to get through to the public
University undergraduates are told by their lecturers that they must reference academic journals and that they need to be current. Books are less relevant as they are “out of date”. Naseem Taleb in “Antifragile” (a book) calls this “neomania“: the obsession with something new.
Rees has this to say about journals:
“But these journals- what scientists call ‘the literature’– are impenetrable to non-specialists. They now just exist for archival purposes, largely unread even by researchers, who depend more on informal ‘reprints’, email and conference.”
Does that ring a bell for coaches who are wading through articles?
Information distortion
In the age of the tweet, the soundbite and 24 hr rolling news coverage, Rees explains that information can get distorted. Ben Goldacre talks about this in “Bad Science” where he postulates that science gets bad coverage due to the media being dominated by humanities students.
Rees (the cynic) says “the distortion is even greater because some scientists (and some institutions) are far more effective than others in communicating and promoting their researches“.
In the pseudoscience world, have you ever wondered why “power” is often narrowly defined by the ability to be tested on a force platform? Answer: where does most of the research come from? Which researcher is on the board of the company that makes the force platform?
This power “research” is then disseminated as gospel (negative results are rarely published in journals, skewing the system further).
Even if we see a well designed study, Rees suggests we bear in mind what Francis Crick has to say “no theory should agree with all the data, because some of the data are sure to be wrong!”
Cancer is more serious than sport
The world is going to keep turning if sports scientists publish poor research about the best number of squats to produce better swimmers. But cancer is much more serious. As reported in the New Scientist, an investigation into 23-highly cited papers in preclinical cancer biology found that fewer than half of them could be replicated. This could explain why less than 30% of phase II and less than 50% of phase III cancer drug trials succeed.
The money, effort and lives at stake in this research is huge. Open access of information helps data sharing and replication (or not) of studies to see if they work. This is how science is supposed to work.
However, if information is not shared, then the studies that can not be replicated get cited more and more and become ‘impact’ papers. This can entrench a series of academics into defending their ‘worthy’ study even if lives are at risk.
Why we should ask difficult questions
Of course, we get what we deserve. Francis Bacon said this in “The advancement of learning” (1605).
“For as knowledges are now delivered, there is a kind of a contract of error between the deliverer and the receiver; for he that delivereth knowledge desireth to deliver it in such form as may be best believed, and not as may be best examined; and he that receiveth knowledge desireth rather present satisfaction than expectant inquiry.”
Steve Myrland says that we believe our own fallibility more than the person presenting to us and that “those parts of presentations that are most confusing to us tend to be the parts we question least.”
This then allows the “expert” to carry on building up an awe-inspiring reputation that remains unchallenged.
Pseudoscience and the LTAD Model
I see this a lot in pseudoscience journals from the UKSCA and NSCA: academics who have less coaching experience than our local primary school teachers are given platforms to promote their unfounded theories.
Models are not scientific evidence nor are they laws. Yet, some researchers looking at physical interventions in children and youth populations cite an LTAD model as ‘evidence’ for the basis of their exercise programming! There is no proof that the LTAD model works: no one has taken a group of children through 15 years of that programme and seen the results. It hasn’t been around for 15 years for a start! Second, every child, every school, every town and every sports club are so different that there can not be a ‘Model’ for all.
I should know: I have been coaching this stuff for 20 years and set up an Athletic Development Club to help local children. Things change so much week to week with my own 2 children, let alone term to term and year to year, that planning for 15-year progression is nonsense.
Referencing that model shows a lack of understanding. Unfortunately, once it’s published and then cited, it keeps getting cited by more and more articles until it gains ‘impact’.
I once spent an afternoon trawling through the 150 references in pseudoscience article about sprint starts in swimming. Many of them were generic points about ‘power’ or ‘sprinting’ on dry land. The few that referenced swimming starts were vague and one of them contradicted the recommendations of the author!
What is a coach to do?
‘We are drowning in information while striving for wisdom.” E. O. Wilson.
Coach learning (adapted from Grace Golden).
Trusting your eye or instinct is a solution fraught with difficulty: we are all prone to bias in many different forms. We can neither dismiss or accept the body of published work as ‘scientific evidence.’ As seen in the cancer studies, there are some that can be replicated and some that can’t.
I try to be open-minded when reading ‘research’ and I don’t take just the abstract and use that to change anything I do with coaching. I do reflect and review upon the coaching that I do after every session. I also check with my specialist or more experienced peers about new ideas or concepts and get their take.
Finally, I look at my situation and see if the new concept is going to help the athletes I coach get better or to make things simpler for them. If it isn’t, then I don’t use it. If it is, we trial it, observe it and see what results we get.
I was asked on Tuesday by an athlete who is quite new to weight lifting why I would teach cleans which are quite complex, if high pulls also work the triple extension.
The answer is that I have got a lot of time with this athlete, so can afford to work on his technique without sacrificing his work that will lead to strength and power development. The clean will then enable him to perform the jerks without using a rack.
But, the question is an excellent one, and should be asked by Coaches before they do any exercise or series of techniques, instead of doing something because everyone else is doing it.
Some National Governing Bodies specifically want cleans coached – why? If time is limited, then
dumbbell cleans
jump squats
wave squats
high pulls
are all useful alternatives for developing power.
Ben Goldacre’s Bad Science column in The Guardian is a good read and is an example of how to examine wild claims and pseudo science. This type of objectivity is uncommon in a lot of Coaching practice.
It is especially interesting to read how the over complication of diet has led to a new brand of celebrity nutritionists who are being discredited due to their lack of scientific underpinning.
I keep telling coaches and athletes that they should look at what they are trying to achieve, and find tools that do that job most efficiently.
However, many people become attached to the “magic exercise” or “magic food” and then reverse engineer its usefulness to match the aims.
The latest craze in competitive sport appears to be the use of data to aid understanding of, and improvement in sporting performance. This has resulted in a glut of material, each item claiming to have established some new result which may have useful implications in the development and performance of human athletes.
There are often studies conducted with non-athletes as well, and the line between what could be considered medical research as opposed to what is known as sports science is not clearly defined.
I should stress that my knowledge of the field of sports science is limited, the purpose of this article is to question the structure and findings of some typical articles.
A typical paper in this field might take the following form:
1) Design a study with some hypothesis of interest
2) Collect data from subjects (fitness testing)
3) Analyse the data to check for consistency with the hypothesis
4) Draw conclusions.
A good statistician should be able to perform multiple roles.
In my opinion, some of the most important are:
To decide on the real questions at the heart of a problem of interest, not to just churn out results for the sake of it.
To decide if a hypothesis is necessary, and if so to construct one which is of real actual interest. Sometimes it is best to approach a problem with an open mind, in the knowledge that there are likely to be interesting results, but unsure of what they will be.
To employ appropriate methods (typically statistical models) to analyse the collected data (we don’t need to get too technical here).
To explain the underlying reasons behind any results – studies in which results are simply quoted as gospel are of limited interest to me.
To critically review the work, pointing out potential shortcomings and areas for future research.
The final point is perhaps the most interesting. It is often the role of a statistician to dampen (or in some cases pour cold water) on enthusiasm about some exciting results.
Sports Science Statistics must be taken in context.
Conclusions drawn from a study of, say, a weightlifter’s improved performance due to a certain type of training programme should not be used as an automatic basis for a different strength-based sport, such as rowing.
I work in the field of weather forecasting. A modern-day weather forecast involves running a computer model forward in time to produce a single forecast of the atmosphere. Statistics of this forecast (such as the average forecast error) can be calculated at different locations. It is well-known that such statistics vary by location – it is more difficult to predict the weather in Shetland than in the Sahara Desert. We could not, therefore, use statistics derived from one location to predict the average forecast error in another.
In short, statistics is about describing what might have happened in a given context, but didn’t. We can use these findings to issue probabilities of what might happen in the future, on the basis that the context is consistent.
Forget the weather: what about sport science?
The few articles I have read in the sports science field (in all honesty I couldn’t face reading too many!) seem to fall short on many of the above points.
For example, Owen et al. (J Strength Cond Res, 2011) conduct a study of heart rate responses of soccer players when playing in three-sided and nine-sided games. They conclude that the HR of players in three-sided games is consistently higher than for nine-sided games. They also note that three-sided games provide more shooting chances, and encourage players to run more with the ball, whilst the nine-sided games produced more tackles, passes and interceptions.
They draw the conclusions that three-sided games are preferable for fitness training, and suggest that strikers should participate in three-sided games whilst defenders should concentrate on nine-sided games.
I have two main problems with this work from both a scientific and practical viewpoint.
The statistics quoted in themselves should be treated with caution, given the small sample size of fifteen players who participated in only a few games of each type. Without conducting a formal test I cannot be more precise, but these measurements are undoubtedly subject to substantial variation.
What insight does the study really offer us? Aren’t the findings, on which the entire article is based, merely confirmation of the obvious? It is useful here to consider the so-called `pyramid of outcomes’.
This study gives only surrogate measures (the base of the pyramid), but assumes in the conclusion that such surrogates automatically extend in to true performance measures (essentially whether they can be used to increase the probability of winning football matches).
This assumption seems completely without foundation when one considers the practical implications of the study. For example, suppose that on the basis of the study, strikers train in three-sided games whilst defenders train in nine-sided games, in order to provide more shooting opportunities for strikers and more defending opportunities for defenders. Is there really any point in this? Wouldn’t three-sided games just result in strikers shooting from anywhere, and playing (by definition) against less able defensive opposition?Surely the way to improve as a striker is to learn how to play against good defenders?
Frankly, this work smacks of conducting a study for the sake of it, and drawing conclusions based on a few surrogate measurements without paying any attention to the sport of interest.
How to conduct a more informative study.
1) Collect a larger sample of players from a variety of clubs, preferably from different countries.
2) Train different groups of players in different environments, as suggested by the study.
3) Collect surrogate measurements from the different training sessions.
4) Examine if the surrogates had an effect on actual game results (i.e. construct a proper statistical model rather than merely reporting surrogate values).
5) Examine whether a return to previous training routines result in a reversion to previous performance.
A statistical model is essentially the use of surrogate measurements to aid in predicting the value of, and assessing the uncertainty in, measurements at the top of the pyramid. The article mentioned here simply assumes that larger surrogate values immediately imply improved results, an assertion which is without foundation.
Such a study would admittedly be hard to carry out both practically and from a theoretical statistical viewpoint. However, we are dealing with complicated situations – we are essentially trying to model outcomes from the human body, an immensely complicated organism.
This is my overriding point, studies which simply churn out results for the sake of publishing papers are of little practical use. I would go further and suggest that they are actually dangerous in the wrong hands – a statistical model is no good in the hands of an incapable operator.
Conclusion
From my brief consultation of the literature, I have seen many examples of a mis-use of statistics which would not be permitted in a statistics journal.
The typical methods used are likely to underestimate the complexity of the situation at hand. I suspect therefore, that the true value of statistics such as the p-value are somewhat larger than reported.
I feel confident in ascerting that the conclusions of the articles I have read are based on extremely shaky ground in a theoretical sense, let alone their practical shortcomings.
Robin Williams Statistics Phd Student (University of Exeter), England Blind Footballer, 2012 Olympian
First a statement from James Marshall’s book review of my book “Science of Sports Training”
“The book is a bit old now, published in 2001, with most of the research quoted pre-dating that. This would probably disqualify it from being used as an academic text book, but as a Coaching handbook it is very good.”
This made me think:
“How important really is for a coach to have the most up-to-date research?”
I quoted a lot of research papers in this book and in my other books. I did it to back up claims or advice that run contrary to common wisdom (or rather common stupidity…).
Some of the old research I quoted was, and still is valuable no matter whether it was done in 1920 or in 2000. Human physiology (including its expression in human psychology) doesn’t change from decade to decade, not from century to century, hardly from millennium to millennium, so accurate observations of human nature hold true no matter their age. (Think the oldest medical manuals of India and China, or fencing manuals of ages past….)
Valuable studies and experiments are those that reveal truths not likely arrived at by “listening to one’s body” or “paying attention to clues.” Everything else is just fulfilling the academic requirement to publish.
What is important for a coach?
Understanding human body and mind enough to know the relation between input and output, then observing athletes and adjusting the input. In one of my blog posts Training vs Skill Training or More on Super Slow and Similar Approaches, I wrote: “When in doubt, refer to everyday observations. An accurate observation is never wrong.”
Take the most important, in my opinion, principle of sports training: The Principle of Individualization and Accessibility of Training. (When you think of it, all other principles of training are based on that one.) If you apply it, you see that studying the most recent research on exercise science matters much less than observing:
athletes’ mood
movement quality
signs of fatigue
signs of apprehension
and adjusting training process accordingly.
More articles on the practical application of principles of training are here and my observation-based posts are in my blog .
Tom Kurz is the author of “Science of Sports Training.”
is a common flaw we can all suffer from: our System 1 brain retrieves instant information and makes decisions based on currently activated ideas. Unfortunately it fails to allow for information outside of that.
(Those of you who watch the immense self belief and ignorance of people on “The Apprentice” may recognise this!),
Our System 2 brain is better at a more systematic and careful review of evidence, but is heavily influenced by System 1.
This is thinking fast and thinking slow according to Daniel Kahneman in this excellent book.
Danger of Overconfidence
WYSIATI leads to overconfidence. If we make decisions based on our limited knowledge, then we build stories around that knowledge to justify our decision making.
If we have made a decision and we feel good about it, then we can easily ignore any further information that comes our way that may counter our initial decision.
In fact, people who only see one side of the argument (or limited evidence) are much more confident in their decisions than people who have seen all the evidence available.
This is due to our brains being good at pattern recognition and formations. It makes sense from an evolutionary purpose to be able to quickly recognise danger and respond (System 1), compared to having to sit down and do a cost:benefit ratio analysis (System 2).
The Muller Lyer illusion is an example of System1/ System 2 conflicts. Our initial impression is that the lines are different lengths.
It is only when we measure them that we realise they are the same. However, we have to keep reminding our System 1 this is the case!
This video shows a few examples of the System 1 and System 2 thinking and some exercises that will help you understand the concept.
Plausable versus Probable
One of the interesting themes of the book was our inability to understand statistics. This has major implications for our lives as politicians and policy makers also fail to comprehend the difference.
People will ignore a statistical fact “50% of children drop out of sport at 16” but will remember individual stories of girls being give short thrift by their schools compared to boys.
We also place a lot of faith in data collected from small samples (The pseudo sport science world is especially bad at this). The variances that occur in small samples mean that they often appear to be the best and worst cases. For example “people living in rural counties of the USA have the highest incidence of cancer“.
Population density is lowest in these counties so a natural variance around the mean causes a blip. We pay attention to the story and so end up with a world belief that is simpler than it really is.
Worst still, we then use System 1 thinking to explain this “fact”: it must be that rural people smoke and drink more than their better educated urban peers!
Luck pays a big part in a lot of events, and Kahneman covers this in a lot of detail.
He debunks the “hot hand” myth in sports: a good run of scores is likely to end with a regression to the mean. Similarly a good player who is scoring below average will return out of their “slump”.
This has nothing to do with skill, changing their shoes, or even (much as I hate to say it) better coaching: it is a statistical anomaly.
Summary
A tough start to get into this book, but worth the effort. Unfortunately (or perhaps fortunately) it has made me question my ability to make any decisions!
What I had previously thought of sound judgement is probably littered with bias and errors of thinking.
The bottom line is that we need to switch between thinking fast and slow: both are necessary for our well being and to thrive. The trick is to recognise when to use which one. For me I can use fast thinking when coaching in the gym, based on my experience and lots of reading.
When choosing a new car, or buying a house, I need to use slow thinking as I have little knowledge or expertise in these areas. In fact, using System 1 to make these decisions could be very costly.
Of course, we get what we deserve. Francis Bacon said this in “The advancement of learning” (1605).
is a common flaw we can all suffer from: our System 1 brain retrieves instant information and makes decisions based on currently activated ideas. Unfortunately it fails to allow for information outside of that.
In fact, people who only see one side of the argument (or limited evidence) are much more confident in their decisions than people who have seen all the evidence available.
We also place a lot of faith in data collected from small samples (The
Luck pays a big part in a lot of events, and Kahneman covers this in a lot of detail.