Turcotte Analysis - NHL draft combine data
Sean Collins
December 9, 2015
Data analysis performed for Turcott’s physiology of exercise lab report
# Load necessary R packages
library(dplyr) #For data manipulation##
## Attaching package: 'dplyr'
##
## The following objects are masked from 'package:stats':
##
## filter, lag
##
## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(reshape2) #For data manipulation related to plotting
library(ggplot2) #For plotting
library(pastecs) #For descriptive statistics## Loading required package: boot
##
## Attaching package: 'pastecs'
##
## The following objects are masked from 'package:dplyr':
##
## first, last# Set working directory - mac file system
setwd("~/physExLab/NHLcombinedata")
# Load data
nhldata<-read.csv("data.csv")
#Create weight in kg variable
nhldata<-mutate(nhldata, Weight_kg = Weight_lb/2.2)
#Create aerobic test peak watts / kg variable
nhldata<-mutate(nhldata, VO2max_watts_kg= VO2max_watts/Weight_kg)
#Create variable for how much the HR dropped after the Wingate
nhldata<-mutate(nhldata, rHR_wingate = WingateHR_End - WingateHR_1minPost)
#Create dataframe with only the variables we are interested in
nhldata<-select(nhldata,Weight_kg, WingatePeak_Watts, WingateMean_Watts, PeakWingate_W_kg, VO2max_l_min, VO2max_ml_kg_min, VO2max_watts, VO2max_watts_kg, WingatePercentDrop, WingateHR_End, WingateHR_1minPost, rHR_wingate)
nhldata <- nhldata[complete.cases(nhldata), ]
str(nhldata)## 'data.frame': 94 obs. of 12 variables:
## $ Weight_kg : num 95.1 80 87.4 82.1 81.2 ...
## $ WingatePeak_Watts : int 1136 1017 1171 933 1089 1223 1210 917 1047 1191 ...
## $ WingateMean_Watts : int 939 72 971 764 877 921 921 718 782 899 ...
## $ PeakWingate_W_kg : num 12 12.7 13.4 11.4 13.4 14.1 13 12.8 13.9 14 ...
## $ VO2max_l_min : num 4.75 3.99 4.94 5.64 4.92 4.41 5.19 4.17 4.39 5.45 ...
## $ VO2max_ml_kg_min : num 50 49.9 56.5 68.7 60.6 50.6 55.8 57.9 58 63.8 ...
## $ VO2max_watts : int 520 400 520 520 520 520 520 440 480 520 ...
## $ VO2max_watts_kg : num 5.47 5 5.95 6.33 6.41 ...
## $ WingatePercentDrop: num 54.5 52.5 36.7 39.5 45.2 55.3 54.4 45.8 57.4 49.3 ...
## $ WingateHR_End : int 185 185 192 191 180 200 192 194 179 192 ...
## $ WingateHR_1minPost: int 180 174 175 176 151 168 178 172 172 181 ...
## $ rHR_wingate : int 5 11 17 15 29 32 14 22 7 11 ...summary(nhldata)## Weight_kg WingatePeak_Watts WingateMean_Watts PeakWingate_W_kg
## Min. : 69.68 Min. : 784 Min. : 72.0 Min. : 9.00
## 1st Qu.: 80.49 1st Qu.:1050 1st Qu.: 809.8 1st Qu.:12.80
## Median : 84.95 Median :1142 Median : 868.5 Median :13.40
## Mean : 84.73 Mean :1134 Mean : 859.6 Mean :13.41
## 3rd Qu.: 89.07 3rd Qu.:1195 3rd Qu.: 924.0 3rd Qu.:14.00
## Max. :109.41 Max. :1502 Max. :1130.0 Max. :15.90
## VO2max_l_min VO2max_ml_kg_min VO2max_watts VO2max_watts_kg
## Min. :3.640 Min. :44.90 Min. :400.0 Min. :4.396
## 1st Qu.:4.383 1st Qu.:51.77 1st Qu.:480.0 1st Qu.:5.649
## Median :4.725 Median :56.30 Median :520.0 Median :6.064
## Mean :4.773 Mean :56.41 Mean :505.5 Mean :5.981
## 3rd Qu.:5.185 3rd Qu.:60.58 3rd Qu.:550.0 3rd Qu.:6.334
## Max. :6.280 Max. :68.70 Max. :640.0 Max. :7.556
## WingatePercentDrop WingateHR_End WingateHR_1minPost rHR_wingate
## Min. :33.70 Min. :163.0 Min. : 72.0 Min. :-15.00
## 1st Qu.:44.85 1st Qu.:185.2 1st Qu.:164.0 1st Qu.: 13.00
## Median :49.60 Median :190.5 Median :172.0 Median : 17.50
## Mean :49.42 Mean :189.7 Mean :170.9 Mean : 18.87
## 3rd Qu.:54.20 3rd Qu.:194.0 3rd Qu.:178.0 3rd Qu.: 22.75
## Max. :77.60 Max. :208.0 Max. :196.0 Max. :120.00As you can see we have gone from 100 subjects to 94 subjects - so 6 subjects had incomplete data in the variables of interest.
Descriptive summary of the variables of interest:
options(scipen=100) #supresses scientific notation of output
options(digits=2) #limits output to 2 significant digits
stat.desc(nhldata, basic=F) #runs basic stats## Weight_kg WingatePeak_Watts WingateMean_Watts
## median 84.955 1142.50 868.50
## mean 84.725 1133.59 859.63
## SE.mean 0.647 13.15 12.46
## CI.mean.0.95 1.285 26.10 24.74
## var 39.387 16243.43 14586.41
## std.dev 6.276 127.45 120.77
## coef.var 0.074 0.11 0.14
## PeakWingate_W_kg VO2max_l_min VO2max_ml_kg_min VO2max_watts
## median 13.400 4.725 56.300 520.00
## mean 13.415 4.773 56.412 505.53
## SE.mean 0.119 0.056 0.557 5.70
## CI.mean.0.95 0.236 0.111 1.106 11.32
## var 1.330 0.292 29.174 3057.24
## std.dev 1.153 0.540 5.401 55.29
## coef.var 0.086 0.113 0.096 0.11
## VO2max_watts_kg WingatePercentDrop WingateHR_End
## median 6.064 49.60 190.500
## mean 5.981 49.42 189.723
## SE.mean 0.065 0.71 0.859
## CI.mean.0.95 0.128 1.41 1.706
## var 0.393 47.34 69.385
## std.dev 0.627 6.88 8.330
## coef.var 0.105 0.14 0.044
## WingateHR_1minPost rHR_wingate
## median 172.000 17.50
## mean 170.851 18.87
## SE.mean 1.468 1.33
## CI.mean.0.95 2.916 2.64
## var 202.623 166.44
## std.dev 14.235 12.90
## coef.var 0.083 0.68Univariate distributions with histograms
qplot(VO2max_ml_kg_min, data=nhldata, geom="histogram")## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.
qplot(WingatePercentDrop, data=nhldata, geom="histogram")## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.
qplot(WingateHR_1minPost, data=nhldata, geom="histogram")## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.
qplot(WingateHR_End, data=nhldata, geom="histogram")## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.
qplot(rHR_wingate, data=nhldata, geom="histogram")## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.
Does a higher VO2 max result in a greater drop in heart rate after the Wingate?
qplot(y=rHR_wingate, x=VO2max_ml_kg_min, data=nhldata) + stat_smooth(method=lm, formula=y~x)
Looks like there is no relationship.
Does a higher VO2 max result in lower 1 minute post heart rate after the Wingate?
qplot(y=WingateHR_1minPost, x=VO2max_ml_kg_min, data=nhldata) + stat_smooth(method=lm, formula=y~x)
Looks like there is no relationship.
How does post wingate HR relate to % drop in power during the wingate?
qplot(y=WingateHR_End, x=WingatePercentDrop, data=nhldata) + stat_smooth(method=lm, formula=y~x)
Looks like there is not much of a relationship. But does their peak watts influence the result?
But does their peak watts influence the result?
This plot is a repeat of the above with Wingate peak watts influencing the size of the point. From the points it looks like people with a higher peak watt have a greater drop off.
qplot(WingatePercentDrop, WingateHR_End,data=nhldata,geom=c("point", "smooth"),method="lm",size = WingatePeak_Watts)
Do players with a higher peak watts on Wingate have a higher drop off during the test?
qplot(y=WingatePercentDrop, x=WingatePeak_Watts, data=nhldata) + stat_smooth(method=lm, formula=y~x)
Seems to be the case. To me this suggest that % drop is not due to poor aerobic capacity; but due to such high power output that players cannot sustain it.
How does wingate average power fit in to the above relationship?
qplot(WingatePeak_Watts, WingatePercentDrop,data=nhldata,geom=c("point", "smooth"),method="lm",size = WingateMean_Watts)
It looks like mean watts may be related to peak watts (makes sense - a higher peak leads to a higher mean), but that it is not related to percent drop as the scatter of point size along the x axis is fairly consistent.
Here is a plot of mean watts and percent drop to confirm those suspicions:
qplot(y=WingatePercentDrop, x=WingateMean_Watts, data=nhldata) + stat_smooth(method=lm, formula=y~x)
And here we see the relationship between wingate mean watts and peak watts - relationships don’t really get much better than this one….
qplot(y=WingateMean_Watts, x=WingatePeak_Watts, data=nhldata) + stat_smooth(method=lm, formula=y~x)
Thoughts -
Peak watts and mean watts are highly related; and peak watts and percent drop are also related. It makes me think that % drop which is supposed to be the fatigue index is really driven by having high peak power. In other words, it’s not as much that someone cannot sustain their power; it is that people with really high power cannot sustain it - possibly because it is, but it’s nature, unsustainable.
The fact that vo2 max does not relate to recovery heart rate is a bit perplexing to me….there is just a lot of Vo2 max variability and not much HR recovery in one minute following the Wingate - so there is little variability in the Wingate HR 1 minute post.