It’s a great day in sports analytics! Today I’m going to briefly discuss the use of Non-Parametric Correlation as a method of sports analytics.
What is Non-Parametric Correlation?
Before we get into how non parametric correlation is used in sports analytics, we first must understand non parametric correlations in a textual concept.
Non parametric correlation is a correlation technique that can be used with any variables that can be transformed into ranks. This means that data used for non-parametric correlation methods is often ordinal and does not rely on numbers. Instead it focuses on ranking, relation or some order of some sorts during analysis.
However, for non parametric correlation to hold, there are two basic assumptions:
- Variables to be measured should be measured at an interval or ratio scale or at an ordinal level.
- There should be a monotonic relation between the two variables to be used.
Electronically, softwares like SPSS or Minitab are used to compute non parametric correlations of variables.
Types of Non-Parametric Correlation
There are two types of non-parametric correlation: Spearman’s Correlations and Kendall’s Correlation
The Spearman’s correlation, often denoted by the symbol ₨ or the Greek letter ρ, is an analytical method that measures the direction and the strength of association between two variables which are measured on an ordinal scale. For example, you could use spearman’s correlation to know the relation between time spent during training and actual performance of players.
Kendall’s correlation, commonly referred to as Kendall’s Tau, measures the ordinal association between two or more variables to be computed.
Non Parametric Correlation as it Relates to Sports
Now that we’ve established the meanings and types of non-parametric correlation, let’s see how they relate to the sports industry. Using the non-parametric method of sports analysis, analysts have and can improve performance of athletes and sustain winning streaks.
Case Study: 2005 Monaco Grand Prix
During the 2005 Monaco Grand Prix, non-parametric correlation method was quickly employed to enable the McLaren team to win. Here is how it happened:
Well into the race, Kimi of McLaren team was leading with a little gap. The third driver, Schumacher accidently smashed into the second driver and both cars ended up needing repair. Other following drivers approached the turn that was debris filled and the race marshals deployed the safety car.
The ideal thing for drivers to do during the safety car period was to pit, change tires, refuel and continue. However, the McLaren team radioed Kimi, asking him not to pit. Obeying, Kimi fired in a few quick laps and increased his lead with 35 seconds. On the forty second lap, he pitted and came out still with a 13 seconds lead and won the race.
With quick thinking, the McLaren team applied non-parametric correlation to make a decision and win the race. They related variances like fuel, time of tires to last and lap time variances in their ranks to ensure success.
Advantages of Non-Parametric Correlation
- Non-parametric method of sports analysis can be done quickly. As seen in the case study above, the method was applied during the course of the race and still ensured victory.
- Can be computed without relating numerical variables. In non-parametric correlation, analysis can be done without having to relate numerical variables. Events carried out can be used in its computation.
- No need for multiple data. Using this method, few data are required to carry out an analysis.
Summarily, with the use of non-parametric correlation methods in sports analytics, sports analysis can be done quickly and still result to an expected performance.
How mature is your team’s analytics program? Take the Sports Analytics Maturity Assessment.
Learn about the Groundbreaking Sports Analytics Model coaches and sports analysts are talking about!
Learn all about Sports Analytics here.