The Last Digit of Total Points

Vector of URL’s

current_season = 2018
num_seasons = 4

website = matrix(NA, nrow = 17, ncol = num_seasons)

for (j in 1:num_seasons){
  for (i in 1:17){
    
    year = 2018 - num_seasons + j
    
    website[i,j] = paste0("http://www.espn.com/nfl/schedule?week=",i,"&year=",year,"&seasontype=2")
  }
}

website_vector = as.vector(website)
length(website_vector)
## [1] 68

head(website_vector)
## [1] "http://www.espn.com/nfl/schedule?week=1&year=2015&seasontype=2"
## [2] "http://www.espn.com/nfl/schedule?week=2&year=2015&seasontype=2"
## [3] "http://www.espn.com/nfl/schedule?week=3&year=2015&seasontype=2"
## [4] "http://www.espn.com/nfl/schedule?week=4&year=2015&seasontype=2"
## [5] "http://www.espn.com/nfl/schedule?week=5&year=2015&seasontype=2"
## [6] "http://www.espn.com/nfl/schedule?week=6&year=2015&seasontype=2"
# .
# .
# .
tail(website_vector)
## [1] "http://www.espn.com/nfl/schedule?week=12&year=2018&seasontype=2"
## [2] "http://www.espn.com/nfl/schedule?week=13&year=2018&seasontype=2"
## [3] "http://www.espn.com/nfl/schedule?week=14&year=2018&seasontype=2"
## [4] "http://www.espn.com/nfl/schedule?week=15&year=2018&seasontype=2"
## [5] "http://www.espn.com/nfl/schedule?week=16&year=2018&seasontype=2"
## [6] "http://www.espn.com/nfl/schedule?week=17&year=2018&seasontype=2"

Create Vector of Results

rvest works by looking at the html code of a website, finding a defined tag, and storing something in a list. So in this code chuck, I create an empty variable and fill it as a list of dataframes. Then, I extract the results from each of those dataframes as a string vector.

totals_list = NULL

for (i in 1:length(website_vector)){
  
  temp_list = website_vector[i] %>%
    read_html() %>%
    html_nodes("table") %>%
    html_table(header = TRUE, fill = TRUE)

  totals_list = append(totals_list, temp_list)

  rm(temp_list)

}

# How long is this list?
length(totals_list)
## [1] 235

# What does each dataframe in the list look like?
names(totals_list[[1]])
## [1] "matchup"          ""                 "result"          
## [4] "passing leader"   "rushing leader"   "receiving leader"
## [7] NA


results = NULL

for (i in 1:length(totals_list)) {
  results = c(results,totals_list[[i]]$result)
  
}

head(results)
## [1] "NE 28, PIT 21"       "BUF 27, IND 14"      "GB 31, CHI 23"      
## [4] "STL 34, SEA 31 (OT)" "NYJ 31, CLE 10"      "MIA 17, WSH 10"

Turn Results Vector into Dataframe

# Remove commas "," using gsub
# Split the results vector into a matrix along each space " "
games_matrix = str_split(gsub(",","",results), " ", simplify = T)

# Name the columns
colnames(games_matrix) = c("away", "away_score", "home", "home_score", "overtime")

# Turn the matrix into a tibble
# Make sure numbers are numbers
# Add some calculations
# There was one game that was postponed
games_df  = as.tibble(games_matrix) %>%
  mutate(away_score = as.numeric(away_score),
         home_score = as.numeric(home_score),
         score_total = `away_score`+ `home_score`,
         final_number = (str_sub(score_total, -1,-1)),
         tie_game = if_else(away_score == home_score, "tie", "no tie")) %>%
  filter(away != "Postponed") %>%
  mutate(season = sort(rep(((current_season-num_seasons+1):current_season),256)))
## Warning: `as.tibble()` is deprecated, use `as_tibble()` (but mind the new semantics).
## This warning is displayed once per session.

games_df
## # A tibble: 1,024 x 9
##    away  away_score home  home_score overtime score_total final_number
##    <chr>      <dbl> <chr>      <dbl> <chr>          <dbl> <chr>       
##  1 NE            28 PIT           21 ""                49 9           
##  2 BUF           27 IND           14 ""                41 1           
##  3 GB            31 CHI           23 ""                54 4           
##  4 STL           34 SEA           31 (OT)              65 5           
##  5 NYJ           31 CLE           10 ""                41 1           
##  6 MIA           17 WSH           10 ""                27 7           
##  7 CAR           20 JAX            9 ""                29 9           
##  8 KC            27 HOU           20 ""                47 7           
##  9 ARI           31 NO            19 ""                50 0           
## 10 SD            33 DET           28 ""                61 1           
## # ... with 1,014 more rows, and 2 more variables: tie_game <chr>,
## #   season <int>