Read in .txt file
chess_elo <- read.csv("https://raw.githubusercontent.com/sserrot/DATA607/main/Project%201/tournamentinfo.txt")
#display first two player rows
head(chess_elo, n=9)## X.........................................................................................
## 1 Pair | Player Name |Total|Round|Round|Round|Round|Round|Round|Round|
## 2 Num | USCF ID / Rtg (Pre->Post) | Pts | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
## 3 -----------------------------------------------------------------------------------------
## 4 1 | GARY HUA |6.0 |W 39|W 21|W 18|W 14|W 7|D 12|D 4|
## 5 ON | 15445895 / R: 1794 ->1817 |N:2 |W |B |W |B |W |B |W |
## 6 -----------------------------------------------------------------------------------------
## 7 2 | DAKSHESH DARURI |6.0 |W 63|W 58|L 4|W 17|W 16|W 20|W 7|
## 8 MI | 14598900 / R: 1553 ->1663 |N:2 |B |W |B |W |B |W |B |
## 9 -----------------------------------------------------------------------------------------Split out rows
Here, we’ll split out each row delimited by dashes and then commas to get a list of all the players data
chess_elo <- chess_elo %>% str_split("-----------------------------------------------------------------------------------------")
chess_elo <- chess_elo[[1]]
chess_elo <- chess_elo %>% str_split(",")
chess_elo_players <- chess_elo[2:length(chess_elo)]
head(chess_elo_players, n=2)## [[1]]
## [1] "\""
## [2] " \" 1 | GARY HUA |6.0 |W 39|W 21|W 18|W 14|W 7|D 12|D 4|\""
## [3] " \" ON | 15445895 / R: 1794 ->1817 |N:2 |W |B |W |B |W |B |W |\""
## [4] " \""
##
## [[2]]
## [1] "\""
## [2] " \n\" 2 | DAKSHESH DARURI |6.0 |W 63|W 58|L 4|W 17|W 16|W 20|W 7|\""
## [3] " \" MI | 14598900 / R: 1553 ->1663 |N:2 |B |W |B |W |B |W |B |\""
## [4] " \""Loops for creating vectors
We need to loop through each player in the data and extract each of the key values we want to store:
- Player Number
- Player Name
- Player Points
- Player pre tournament chess rating
- Player average opponent rating
Here is an example:
“,” 1 | GARY HUA |6.0 |W 39|W 21|W 18|W 14|W 7|D 12|D 4|“,” ON | 15445895 / R: 1794 ->1817 |N:2 |W |B |W |B |W |B |W |“,”
# empty list for instantiation
player_number <- c()
player_name <- c()
player_points <- c()
player_state <- c()
pre_tournament_rating <- c()
post_tournament_rating <- c()
#iterate through players and parse out each game and append each value to a list
for(player in chess_elo_players) {
first_line <- player[2] %>% str_split("\\|") %>% unlist()
#check if any of the games are NA
if(any(!is.na(first_line)) == TRUE){
player_number <- c(player_number, first_line[1] %>% str_extract("[0-9]+"))
player_name <- c(player_name, first_line[2] %>% trimws())
player_points <- c(player_points, first_line[3] %>% trimws() %>% as.numeric())
}
second_line <- player[3] %>% str_split("\\|") %>% unlist()
player_rating <- second_line[1:2]
player_state <- c(player_state, player_rating[1] %>% str_extract("[a-zA-Z]+"))
player_chess_rating <- player_rating[2] %>% str_split("/") %>% unlist() %>% str_split ("->") %>% unlist() %>% trimws() %>% str_extract("[0-9]+")
pre_tournament_rating <- c(pre_tournament_rating, player_chess_rating[2])
post_tournament_rating <- c(post_tournament_rating, player_chess_rating[3])
}
#remove NAs
pre_tournament_rating <- pre_tournament_rating[!is.na(pre_tournament_rating)] %>% as.numeric()
post_tournament_rating <- post_tournament_rating[!is.na(post_tournament_rating)] %>% as.numeric()
player_state <- player_state[!is.na(player_state)]
tournament_df <- tibble(player_number, player_name, player_state, player_points, pre_tournament_rating, post_tournament_rating)Now we have a dataframe of all the original data:
head(tournament_df, n=9)## # A tibble: 9 x 6
## player_number player_name player_state player_points pre_tournament_r~
## <chr> <chr> <chr> <dbl> <dbl>
## 1 1 GARY HUA ON 6 1794
## 2 2 DAKSHESH DARURI MI 6 1553
## 3 3 ADITYA BAJAJ MI 6 1384
## 4 4 PATRICK H SCHILLING MI 5.5 1716
## 5 5 HANSHI ZUO MI 5.5 1655
## 6 6 HANSEN SONG OH 5 1686
## 7 7 GARY DEE SWATHELL MI 5 1649
## 8 8 EZEKIEL HOUGHTON MI 5 1641
## 9 9 STEFANO LEE ON 5 1411
## # ... with 1 more variable: post_tournament_rating <dbl>Tournament Analysis
However, we want to calculate an average of all the player’s opponents chess ratings. There are many different ways to approach this issue. I chose to simply
- Go through all the games played which is represented by a player number
- Replace the player number with the player ranking
- Calculate the average rating for those games
We have to be sure to ignore NAs for those players that played less than 7 rounds.
We can also take the player opponent chess ratings and calculate the expected outcome Using the formula from the Wikipedia Elo page, we can calculate the expected value as below:
`Player 1 Expected Rating = 1 / 1 + 10 ^ [(OppRating - Rating) / 400]
elo_calc <- function(opponent_elo, player_elo) {
expected_outcome <- 1 / (1+ 10^((opponent_elo - player_elo)/400))
return(expected_outcome)
}
#test function
elo_calc(1400, 1000) # expected outcome of a player with rating 1000 against rating 1400 (win = 1, draw = .5, loss = 0)## [1] 0.09090909## redo players for just games to get average since we got all the prechess ratings linked to players now
average_opponent_rating <- c()
expected_points <- c()
for (player in chess_elo_players) {
rating_list <- c()
games_played <- player[2] %>% str_split("\\|") %>% unlist()
player_number <- games_played[1] %>% str_extract("[0-9]+") %>% as.numeric() # for elo calculation
games_played <- games_played[4:10]
slice <- games_played %>% str_extract("[0-9]+") %>% as.numeric()
rating_list <- tournament_df$pre_tournament_rating[slice]
rating_list <- rating_list[!is.na(rating_list)] %>% as.numeric() # list of opponent ratings
average_opponent_rating <- c(average_opponent_rating, mean(rating_list))
# for elo calculation - expected outcomes
expected_points <- c(expected_points, sapply(rating_list, elo_calc, player_elo=tournament_df$pre_tournament_rating[player_number]) %>% unlist() %>% sum() %>% format(digits = 2) %>% as.numeric())
}
# round elo and remove na at the end
average_opponent_rating <- average_opponent_rating[!is.na(average_opponent_rating)] %>% round()
expected_points <- expected_points[!expected_points==0]
tournament_df <- tournament_df %>% add_column(average_opponent_rating)
tournament_df <- tournament_df %>% add_column(expected_points)
glimpse(tournament_df)## Rows: 64
## Columns: 8
## $ player_number <chr> "1", "2", "3", "4", "5", "6", "7", "8", "9", "~
## $ player_name <chr> "GARY HUA", "DAKSHESH DARURI", "ADITYA BAJAJ",~
## $ player_state <chr> "ON", "MI", "MI", "MI", "MI", "OH", "MI", "MI"~
## $ player_points <dbl> 6.0, 6.0, 6.0, 5.5, 5.5, 5.0, 5.0, 5.0, 5.0, 5~
## $ pre_tournament_rating <dbl> 1794, 1553, 1384, 1716, 1655, 1686, 1649, 1641~
## $ post_tournament_rating <dbl> 1817, 1663, 1640, 1744, 1690, 1687, 1673, 1657~
## $ average_opponent_rating <dbl> 1605, 1469, 1564, 1574, 1501, 1519, 1372, 1468~
## $ expected_points <dbl> 5.2, 3.8, 1.9, 4.7, 4.4, 4.9, 4.6, 5.0, 2.3, 1~tournament_df <- tournament_df %>% mutate(expected_points_difference = as.numeric(format(player_points - expected_points, digits = 2)))
tournament_df <- tournament_df %>% mutate(rating_change = post_tournament_rating - pre_tournament_rating)
tournament_df <- tournament_df %>%
mutate(rating_change_type = ifelse(rating_change > 0, "increase", "decrease"))
tournament_df <- tournament_df %>% mutate(rating_comparison = pre_tournament_rating - average_opponent_rating)
tournament_df <- tournament_df %>%
mutate(rating_comp_type = ifelse(rating_comparison > 0, "above", "below"))Rating Change and Rating Comparison
tournament_df %>% ggplot(aes(x = player_number, y = rating_change, color = rating_change_type)) + geom_col() +ggtitle("Rating Change") +theme(plot.title = element_text(hjust = 0.5), axis.text.x=element_blank())
tournament_df %>% ggplot(aes(x = player_number, y = rating_comparison, color = rating_comp_type)) + geom_col() +ggtitle("Rating vs Opponent Average Rating") +theme(plot.title = element_text(hjust = 0.5), axis.text.x=element_blank())
Extra Credit
Expected points difference
tournament_df[which.max(tournament_df$expected_points_difference),] %>% select(player_name, player_points, expected_points, rating_change, expected_points_difference)## # A tibble: 1 x 5
## player_name player_points expected_points rating_change expected_points_diff~
## <chr> <dbl> <dbl> <dbl> <dbl>
## 1 ADITYA BAJAJ 6 1.9 256 4.1However based off of actual rating change it would be:
tournament_df[which.max(tournament_df$rating_change),] %>% select(player_name, player_points, expected_points, rating_change, expected_points_difference)## # A tibble: 1 x 5
## player_name player_points expected_points rating_change expected_points_d~
## <chr> <dbl> <dbl> <dbl> <dbl>
## 1 JACOB ALEXANDE~ 3 0.043 699 2.96CSV Output
Output results to csv
tournament_df %>% write.csv("Tournament Players.csv", row.names = FALSE)