afltables.rmd

---
---

The easiest way to get the results of past AFL matches in bulk is from [afltables.com](https://afltables.com/afl/seas/2018.html). They aren't in a format particularly amenable to analysis, so a little bit of work needs to be done in wrangling the online tables into data frames:

```{r,message=FALSE}
library(tidyverse)
library(rvest)
season <- read_html('https://afltables.com/afl/seas/2017.html')
```

```{r}
process_match_description <- function(x){
  home_team <- x[1,1]
  away_team <- x[2,1]
  home_score <- x[1,3]
  away_score <- x[2,3]
  metadata_string <- x[1,4] #includes date, time, attendance, venue
  venue <- str_extract(metadata_string,'Venue:.*$') %>%
    str_replace('Venue: ','')
  attendance <- str_extract(metadata_string,'Att: [0-9,]+') %>% str_replace('Att: ','') %>% str_replace(',','') %>% as.numeric()
  date <- str_extract(metadata_string,'[0-9]{1,2}-[A-Z][a-z]{2}-[0-9]{4}') %>% lubridate::dmy()
  tibble(home=home_team,away=away_team,home_score=home_score,away_score=away_score,venue=venue,attendance=attendance,date=date)
}
```

```{r}
get_season_results <- function(year){
  url <- paste0('https://afltables.com/afl/seas/',year,'.html')
  season <- read_html(url)
  results <- html_table(season,fill=TRUE)
  results <- results %>%
  keep(~all(dim(.x)==c(2,4))|all(dim(.x)==c(1,2))|all(dim(.x)==c(1,1)))
  i <- 1
  output <- list()
  round <- ''
  for (result in results){
	  if (all(dim(result)==c(1,2))){
		  round <- result[1,1]
	  }
	  if (all(dim(result) == c(1,1))){
		  round <- result[1,1]
	  }
	  if (all(dim(result)==c(2,4))){
		output[[i]] <- process_match_description(result)
	  	output[[i]]$round <- round
	  	i <- i+1
	  }
  }  
bind_rows(output)
}
```

```{r}
results <- map(2010:2018,get_season_results) %>%
  bind_rows
as.data.frame(results)
```