Noisy boys ahead

Welcome to this year’s PremPredict, where there are twenty teams, twenty-one players and a massive £105 up for grabs.

But what’s worse is that the noisy boys are on top: Yes, the old man, the brother and the best man. Please make it change!

 

Latest standings

Here’s what you really want to know:

##              Names Scores Bonus       WorstClub WorstCost
## 1      Les Penfold    470   -50         Chelsea       169
## 2     John Penfold    506   -50 AFC Bournemouth       121
## 3         Joe Wood    682   -50        Brighton       256
## 4    Paula Penfold    704     0         Chelsea       225
## 5  Mathew Saunders    736   -50        Brighton       169
## 6    Robin Penfold    772   -50        Brighton       169
## 7    Hannah Harrop    802     0        Brighton       225
## 8     Laura Finnis    860     0         Burnley       196
## 9     Beth Penfold    866   -50         Burnley       196
## 10  Andrea Laporta    880     0         Chelsea       196
## 11   Marion Finnis    886   -50         Chelsea       256
## 12     Vera Finnis    890     0        Brighton       196
## 13    Peter Finnis    896     0        Brighton       225
## 14   Neil Waterman    898   -50        Brighton       196
## 15    Alan Butcher    908     0        Brighton       196
## 16     Alan Finnis    948   -50        Brighton       256
## 17   Imogen Finnis    950     0         Chelsea       324
## 18     Liz Penfold    954     0         Burnley       196
## 19     Mike Finnis    958     0        Brighton       225
## 20     Luke Finnis    980     0         Chelsea       324
## 21    Scott Harrop   1096   -50         Chelsea       256


Fortunately, there’s still a long way to go!


Collective expectations

But what are we collectively expecting from the Premier League this season? Are we expecting it to be a repeat of last season?

By the look of our picks, we seem to side with the bookies. As a group, we predict Manchester City to finish highest on average, with the recently-promoted teams struggling.

 

And we collectively feel that Sheffield United will come last. But will we be more accurate than before?!

Appendix

For completeness (and reproducibility), here’s the code that I used to calculate what’s above.

library(ggridges)
library(premPredictor)
suppressMessages(library(tidyverse))
  
# player_info <- 
#   "https://www.dropbox.com/s/uin6zk4w5cyk2m1/PremPredict-19-20.csv"
# player_data <- get_player_data(url_value = player_info)

player_data <- read_rds('data/data_input_2019-20.rds')

get_latest_standings(
  data_input = player_data, 
  use_saved_data = T, 
  data_file = 'data/201920_week1.rds'
  )


… and …

data_input1 <- as_data_frame(player_data)
averageView <- round(rowMeans(data_input1[, -1]), 2)
views <- cbind(data_input1[,1], averageView)
  
data_input2 <- data_input1 %>% 
  gather(key = "Player", -Club, value = "Prediction") %>% 
  left_join(views, by = "Club")
  
ggplot(
  data = data_input2,
  mapping = aes(
    y = reorder(Club, -averageView), 
    x = Prediction, 
    fill = averageView, 
    color = averageView
    )
  ) + 
  geom_ridgeline(
    stat = "binline", 
    bins = 20, scale = 0.95, 
    draw_baseline = FALSE
    ) + 
  scale_x_continuous(
    breaks = c(5, 10, 15, 20), 
    labels = c(5, 10, 15, 20)
    ) + 
  labs(
    y = "", x = "", 
    title = "\n Our collective predictions for this season \n"
    ) + 
  scale_fill_gradient(
    low = "green", high = "red", 
    guide=FALSE
    ) + 
  scale_color_gradient(
    low = "green", high = "red", 
    guide=FALSE
    ) +
  theme(
    title = element_text(size = 10), 
    axis.text.y = element_text(size = 6)
    )


(Finally, here are your team-by-team predictions, alongside the latest team standings.)