Open Operational Research

Previously: I found the OpenPowerlifting dataset and made some comparisons.. Since Sept last year I’ve put 10kg on my bench press, 10kg on my deadlift and 20kg on my squat.

There’s going to be the 3rd annual Garage Gym Competiton. (Why you ask, Reddit I reply.) Partially so I can feel like I’m not going to collect the wooden spoon I’ve grabbed the 2019 results and thrown them into a spreadsheet.

There’s a couple of bad values, and Gray Matter thinks in lbs so I’ll turn his table into kg.

James <- tribble(
  ~Handle, ~Squat, ~Bench, ~Deadlift,
  "James", 170, 100, 170
) %>%
  mutate(Total = Squat + Bench + Deadlift) %>%
  pivot_longer(-`Handle`, values_to = "Weight")

garagegym <- garagegym %>%
  filter_if(is.numeric, ~!is.na(.x)) %>%
  rename(Handle = `G Handle`) %>%
  select(`Handle`, Gender, Squat, Bench, Deadlift, Total) %>%
  pivot_longer(-c(Handle, Gender), values_to = "Weight") %>%
  mutate(Weight = Weight / 2.2) 

Some messing around with ggplot2:

First, density plotting everyone’s total, excluding me since I didn’t take part, but including a vertical line where I am.

garagegym %>%
  filter(name=="Total") %>%
  ggplot(aes(x=Weight)) + geom_density() + geom_vline(data=filter(James, name=="Total"), aes(xintercept = Weight)) +
  ggthemes::theme_tufte() + 
  labs(
    title="Density plot of all 2019 competitors, JR highlighted",
    x="Weight (kg)"
  )

Facet wrapping the above for the other plots

garagegym %>%
  ggplot(aes(x=Weight)) + geom_density() + geom_vline(data=(James), aes(xintercept = Weight)) +
  ggthemes::theme_tufte() + facet_wrap(~name, scales = "free") +
  labs(
    title="Density plots of all 2019 competitors, JR highlighted",
    x="Weight (kg)"
  )

The comp has no age categories, but there is self-submitted Gender, so comparing male with me:

garagegym %>%
  filter(Gender=="Male") %>%
  ggplot(aes(x=Weight)) + geom_density() + geom_vline(data=(James), aes(xintercept = Weight)) +
  ggthemes::theme_tufte() + facet_wrap(~name, scales = "free") +
  labs(
    title="Density plots of all male 2019 competitors, JR highlighted",
    x="Weight (kg)"
  )

Very similar, 163 male competitors to 41 female competitors.

Out of interest, what were the female numbers like:

garagegym %>%
  filter(Gender=="Female") %>%
  ggplot(aes(x=Weight)) + geom_density() +
  ggthemes::theme_tufte() + facet_wrap(~name, scales = "free") +
  labs(
    title="Density plots of all female 2019 competitors",
    x="Weight (kg)"
  )

Anyway, in this dataset my everything is just below the mode, except squat. What sort of ranking would I be looking at?

(ha, I thought insta handle would be a unique identifier. I was wrong. I want my data wide again and can’t do it. Need to reload the data.)

James <-pivot_wider(James, names_from = name, values_from = Weight)


garagegym  <- read_csv(here::here("static/data/GarageGym/garagegym2019 - Sheet1.csv")) %>%
  select(Handle = `G Handle`, Squat, Bench, Deadlift, Total) %>%
  filter_if(is.numeric, ~!is.na(.x)) %>%
  mutate_if(is.numeric, ~(.x/2.2))

garagegym <- bind_rows(James, garagegym)

Then I’ve worked out my percentiles if my current 1RMs were in last year’s competition.

garagegym %>%
  mutate_if(is.numeric, ~percent_rank(.x)) %>%
  filter(Handle=="James") %>%
  mutate_if(is.numeric, scales::percent) %>%
  knitr::kable()