-
Playbook: Exploring decathlon competition data. Part 2
By Mladen Jovanovic on 18/05/2014Playbook: Exploring Decathlon Competition Data Click HERE to read part 1 Clustering What we might be interested next is similarities between athletes, or in other words, which athletes have similar profiles. For that purpose we can use Hierarchical Clustering and Principal Component Analysis (PCA) which we are going to cover later HCWard <- hclust(d = dist(decathlon.normal),...
-
Playbook: Exploring decathlon competition data [Part 1]
By Mladen Jovanovic on 17/05/2014Playbook: Exploring Decathlon Competition Data Data set Decathlon data set comes from FactoMineR package and represents two competitions: Decastar and Olympic Games. For this example we will explore only Olympic Games competition, so we need to subset the data. # Load the needed packages library(FactoMineR) library(ggplot2) library(reshape2) suppressPackageStartupMessages(library(googleVis)) # Load the decathlon data data(decathlon) # Subset the...
-
How to Analyze Movement Screen Tests? [Addendum]
By Mladen Jovanovic on 28/04/2014How to Analyze Movement Screen Tests? Continuing on the previous post I had an idea to make further analysis. Please note that this is only a playbook. The idea is to transpose the data, and instead of clustering the athletes, we cluster the metrics or tests. What we are looking for is to find tests/metrics that are similar...
-
How to Analyze Movement Screen Tests?
By Mladen Jovanovic on 28/04/2014How to Analyze Movement Screen Tests? In the recent post I shared the movement screen we designed and implemented. The question now is how to analyze the data and make real life decisions on it? What do we need to get from the analysis in the first place? In my (current) opinion we need these: Identify groups of athletes...
-
How to (Pretend to) Be a Better Coach Using Bad Statistics
By Mladen Jovanovic on 25/03/2014How to (Pretend to) Be a Better Coach Using Bad Statistics Here is a simple scenario from practice: Coach A uses YOYOIRL1 test and Coach B uses 30-15IFT (for more info see paper by Martin Buchheit, which also stimulated me to write this blog) to gauge improvements in endurance. Coach A: We have improved distance covered in YOYOIRL1 test...
-
Analysis of Metabolic Power Data Using Power-Duration Profile in Team Sports
By Mladen Jovanovic on 06/03/2014This is the idea I got from the Training and Racing with a Powermeter book by Hunter Allen and Andrew Coggan. It is an excellent and must read book on cycling, but also great book about endurance training in general…
-
Continuing with Statistical Power simulation in R
By Mladen Jovanovic on 13/02/2014In the last blog post I created a simple simulation of statistical power (probability to identify effects when they are really there) calulation depending on the sample size and effect size (Cohen’s D using Will Hopkins effect levels).
-
“Power” to detect statistically significant effects based on sample size and magnitudes of effects
By Mladen Jovanovic on 12/02/2014I was going through magnitude-based inferences materials by Will Hopkins and I am playing with R simulations. I wanted to see how many times I am able to detect statistically significant effects (p<0.05) depending on magnitude of effects (expressed as Cohen's D, and using Will Hopkins levels) and sample sizes.
-
Playbook: Understanding MODERATION Through Simulation
By Mladen Jovanovic on 31/12/2013Playbook: Understanding MODERATION Through Simulation Introduction I recently spoke with my college professor regarding the understanding of statistics, and I remarked that I learn the statistical concept the best (and with full comprehension) through simulation. He remarked that I might be in the minority of students (p<0.01 – see what I did here?). Not sure if this is true...
-
How to Visualize Test Change Scores for Coaches [Part 2]
By Mladen Jovanovic on 26/12/2013To continue the previous How to visualize test change scores for coaches blogpost, here is another way to visualize individual change scores without falling for group averages. This way we can easily see individual ranks in both pre- and post- test, along with change score (which is also color coded). Quite easy to identify the outliers.