The Athletic Combine Part 3: Results, The Most Important Thing in Athletic Performance Testing
The stats are in! With data from over 200* athletes, 12 sports, 12 tests, and an age range of nearly 30 years there have been a lot of numbers to crunch.
May 9, 2018
The stats are in!
With data from over 200* athletes, 12 sports, 12 tests, and an age range of nearly 30 years there have been a lot of numbers to crunch.
So what did we find and what does it all mean for you?
Before we get to the results, a quick (2 minute) crash course in statistics.
*not all our athletes tested, we did not let testing interrupt tournament and major competition preparations. Certain rehabilitation athletes also didn't test.
**This is part 3 of a three-part series on testing, reading the first two parts are not crucial before reading these stats but may provide some context. Part 1, Part 2.
Causation, Correlation and P-Value.
Causation means one variable caused another variable to change. For example, spending more time practising a musical instrument leads to an improvement in playing ability. Doing more exercise will increase your fitness levels. Thing A caused thing B to change.
In our combine we cannot pull any direct causations, To establish correlations you need to track, test and monitor athletes change performance over time and watch to see if improvements in one metric lead to changes in another over time. Hopefully, we can see some causative stats in the next combine.
This tells us the relationship and connection between two factors.
In our data the strongest correlation was between 5m sprint and 10m sprint (R=0.92), followed by height and wingspan (r=0.9) then vertical leap and peak velocity (R=0.8).
The closer the R is to 1.0 or -1.0 the more perfect the correlation.
Having a higher correlation means a more perfect relationship, but not necessarily a stronger one, that's where P-value comes in.
P-value shows you the strength of the correlation, so while R= gives the perfection of a relationship, the p-value gives us the strength of the relationship.
P-value is shown as a decimal score between 1.0 and 0.0. with P= 0.99 meaning the results are 99% down to chance (potluck, completely random), while P = <0.01 means there is only a 0.1% chance the results are a fluke. The lower your P-value the stronger your findings.
In most scientific work having a P value under 0.05 is considered good, while a score of <0.01 would make your findings incredibly strong.
Of the 78 correlation pairs we ran, 67 had a P value of under 0.01. Most of the 11 that didn't meet the 0.01 involved correlations with handspan, which it turns out doesn't effect push-ups, lower body strength or speed all that much.
Ok, with that out of the way let's get to the big headline:
Relative* Strength is King
Whether you want to improve your speed, jumping ability or agility and change of direction, getting stronger is without doubt the best thing you can do. And the stats prove it.
*Relative strength is how much you can lift relative to your bodyweight.
While some outliers can clearly be seen, the general trend it clear. Stronger people jump higher
Strength had a huge impact on vertical, with a correlation of R = 0.58, meaning 58% of the variation in vertical leap is linked to an athlete's relative strength.
When we compared the best jumpers (top ten) to the next best (jumpers ranked 11-20) and the bottom ten some interesting stats showed up:
On average, the best jumpers had an extra 15 months of training experience than the worst jumpers (not only were both older and had put in more time in the gym).
Not a single athlete jumped over 60cm without a strength score of at least 0.75x
62% of athletes who jumped over 50cm had a relative strength score of 1.0 or above.
The bottom 20 jumpers (8 male, 12 female) had an average strength of only 0.56x
For the guys:
No one jumped higher than 68cm without having a strength score of at least 1.12x (average of 1.55x)
The male jumpers ranked 11-20, or the 'almost-elites' (67-62cm) were on average 30% weaker (1.08x) than the best ten jumpers
For the girls
For the six females who cleared 50cm they had an average strength of 1.21x
For the 13 females to jump between 45-49cm the average strength was 0.98x. 19% weaker than the girls who scored in the 50cm+ category
Not a single female jumped higher than 46cm without having a relative strength of at least 0.8x BW
There is no doubting it, stronger athletes jump higher. You can read more about all things vertical leap and what makes good jumpers here.
Despite the intricacies and complexity involved in change of direction and agility, the effect of strength can clearly be seen in the graph above. With an R of 0.5, 50% of the variance in agility can be explained by an athlete's lower body strength.
Given this was a modified test we had no normative data to compare with prior to testing. The elite males and females helped us develop two high performing categorisations.
For the guys: the Sub 4.6s club:
The average strength of the males who broke 4.6s was 1.40x bodyweight
seven of the nine athletes who ran under 4.6s also have vertical leaps above 60cm. The other two were 54cm and 59cm
The best of the best five agility scores (all under 4.55s) had strength scores of 1.47x and above
The almost-elite athletes who just missed out on running sub 4.6s (4.60-4.69s) were on average 31% weaker
Of the males who went under 4.8 (35), only 4 had a relative strength score under 0.75
For the girls: the Sub 5s club:
The average strength of the females who broke 5.0s was 1.14x bodyweight.
The next 11 athletes (5.0 - 5.15s) were on average 32% weaker
six of the seven females to break 4.95s, had strength scores above 1.0 (the other is 0.91)
For both agility and vertical leap, it's clear that even amongst the best 20 performing athletes significant differences in strength exist between the elite and the almost-elite, and between the top 20 and the bottom 20 athletes there exists a chasm in strength. While there are always outliers, increasing strength is one of the most valuable things an athlete can do to improve their agility and change of direction speed.
For speed, while there were more outliers than with vertical leap and agility, the strength of the relationship is still clear: Over both the 0-5m and 0-10m distance, strength has an impact on sprinting and acceleration, accounting for 30-40%* of the difference in performance.
This lower correlation compared to vertical leap and agility makes sense. Sprinting and starting mechanics can be quite complex and require high levels of technical ability, elasticity and power making speed such a multifactorial physical characteristic.
But isn't agility even more complex?
Agility is in a lot of ways far more complex and technically demanding than straight line sprinting but the huge forces encountered during the acceleration and deceleration during the change of direction make strength such a vital determining physical quality.
No point* being technically sound with your cutting mechanics in an agility test if you simply can't accelerate and decelerate powerfully enough out of the corners to overcome your own inertia.
*There actually is plenty of point, efficiency, mechanics and coordination are vital and we spend heaps of time teaching these to all our athletes.
Speed on the other hand, has a single straight line acceleration, the start. Once you are up and running, technique and elasticity become just as important as strength as deceleration is not important.
Some interesting deeper stats when it came to speed:
The 10 fastest athletes who all ran 0-5m in under 1.25s had an average strength of 1.29x
The differences in strength between the top ten athletes and the those ranked 11-20 were negligible.
The top ten athletes (both genders and distances) were on average 33% stronger than the bottom ten athletes. Strength is still vital**
Interestingly, 0-10m speed had a stronger correlation than 0-5m speed with strength, agility and vertical leap. This contradicts most research, but may have been caused by variations in starting technique, population factors, or the fact that both distances are effectively measuring the same thing.
**Maybe there exists a threshold of strength for improving speed, that beyond a certain level of strength, say 0.75x or 1.0x bodyweight, technique and genetics start to play a big factor.
Anthropometry: Introducing the Wings Award.
Lankiness and length are valuable commodities in nearly every sport, from slam dunking a basketball, holding off an opponent in soccer to reaching a drop shot in tennis, having longer arms and bigger hands are of huge benefit to an athlete.
The most common measurement of this is your wingspan to height ratio, dividing your wingspan in by your height. We have decided to include handspan from pinky to thumb into the equation to give a complete picture of an athletes lankiness.
Michael Phelps (Team USA swimmer) (203+29)/193 = 1.20
What was Missing
Upper body power and fitness.
Push-ups are a great test of general upper body strength and in some cases muscular endurance, but they don't tell us anything about explosive power. We have already started working on fixing this so watch this space.
Fitness is trickier to test at the scale we are trying to do, to run over 200 athletes through a Yo-Yo test in a gym can be quite intrusive on other athletes trying to workout. The bike test is a good general indicator of fitness and anaerobic threshold, but how that translates to running fitness is harder to tell. Ideas for shorter duration tests of anaerobic fitness are welcomed in the comments below!
The Big 5 Take-Aways
To sum things up, we have put together the biggest most important five correlations and stats from all our athlete's data.
1. Power in one direction goes with power in the others.
Better jumpers also tended to be better sprinters, better at agility and had better peak velocity and any which way you slice it if you are good at one test you tend to do well at the others. Here's why:
Strength. Relative strength was a massive factor in all these tests
Training age. Athletes who have been training with us longer are stronger, more explosive and usually have better technique
Age. Training teenagers there are wide-ranging differences in physical maturity and development, most physical characteristics don't peak until your 20s so the sky is the limit young ones.
Genetics. Type II muscle fibres, tendon quality, limb length and nervous system excitability are huge factors in how fast we can develop force.
2. Strength is the lead domino to improved test scores.
Improving your strength is the best thing you can do to improve your athleticism. Focusing energy, time and effort on getting stronger in the weights room will pay huge dividends on the court, field or track. If your sport can be decided by being half a second or half a centimetre to late then let strength be your secret weapon.
The fastest ten athletes were 30% stronger than the slowest ten athletes.
3. Even in the highest performing athlete's, strength separates the great from the good.
One key stat/quote:
The top ten (1-10) athletes, both male and female, for both vertical and agility were 20-40% stronger than the next ten (11-20) athletes.
4. There seems to be a minimum threshold of strength necessary for high performance on tests of power and speed.
Whether this number is 0.75x or 1.0x bodyweight it is pretty clear that weak athletes are much rarer in the top 10-20 performing athletes than strong ones.
No females broke 4.95s on the agility test without a strength scores of at least 0.91x
No male broke 4.66s without having a relative strength of at least 0.75x
No one jumped higher than 68cm without having a strength score of at least 1.12x
Not a single athlete jumped over 60cm without a strength score of at least 0.75x
All six females who cleared 50cm they had an average strength of 1.21x
Not a single female jumped higher than 46cm without a relative strength of at least 0.8x BW
Every athlete who ran 0-5m in 1.28s or faster had a relative strength of at least 0.71x bodyweight
Every athlete who ran 0-10m in 2.02s or faster had a relative strength of at least 0.71x bodyweight
The weakest athlete to break the 3:00 mark scored 0.71x bodyweight for relative strength
*It's worth noting the athlete who scored 0.71x is also 202cm tall. What a beast to be that athletic at that height. We would love to take credit but he is just getting started with us.
5. Handspan is as much about flexibility as it is about bone length
It was really interesting watching athletes do the handspan test, while you would think it is purely genetic and bone length based, it really is not. Finger spread and the ability to stretch the thumb away from the index finger played a huge role in handspan score. Maybe hand stretching and mobility work may be in order before the next combine.