Player/Game Data

[travisendicott]

Hey, I'm new to soccer data analysis but is there anyway that you could post/include player data from individual games? I'm trying to see which USL Championship players are taking higher xG shots and which are taking (potentially) more lower xG shots. Additionally, I would like to see which teams are giving up better (higher xG) shots and against which players. In my mind at least, I think I can do this with player/game type data. I understand that I can take the xGF/xGA divided by the total number of shots taken/conceded, but I'd like to do a bit more with the game stats. Any help you could provide would be greatly appreciated!

[mattyanselmo]

You can use the date filter on the xG app page to go week by week and get player data (or same on the team xG page). Admittedly that is a somewhat tedious task. If you use our "itscalledsoccer" packages in R or Python, you can programmatically loop through dates and make those pulls. The get_player_xgoals functions have start_date and end_date inputs that could be part of a loop.

See: date filter

[travisendicott]

Thank you that worked! Is there a reason why the goals added variable is a list and nested within another variable? I'm having a hard time getting that information out so I can play with that data. Do you all have some code to unnest or is there a problem in the way that I'm exporting the data?

[mattyanselmo]

Assuming you're working in R? Here's some code I used recently...

schema <- "mls"
# Get field player g+ ####
gplus <- asa_client$get_player_goals_added(
  leagues = glue("{schema}"),
  season_name = seasons,
  split_by_seasons = TRUE,
  stage_name = c("Regular Season", "NWSL Challenge Cup Group Stage", "MLS is Back Group Stage")
)

gplus_data <- lapply(gplus$data,
                     function(temp){
                       temp %>% 
                         select(-c(count_actions, goals_added_raw)) %>% 
                         pivot_wider(names_from = action_type, values_from = c(goals_added_above_avg))
                     }) %>%
  bind_rows()

gplus <- gplus %>%
  select(-data) %>%
  bind_cols(gplus_data)

[travisendicott]

I am working in R, sorry for not saying that upfront. Thank you for the code, I think that works but I'm still getting an error because there is a null value in one of the lists. I have tried some code to turn it into NA but I'm having trouble since it's a list. Have you encountered this before and if so how did you fix it?

Sorry for all the questions, I'm new to R and soccer data analysis. Most of the data analytics that I've done has been in Stata before and I'm transitioning to R since it's a free resource. But there have been a lot of syntax issues that I've run into that I can normally figure out by searching on Stack Overflow or something similar. Unfortunately, I haven't been as lucky with this variable list issue with the G+ data.

[mattyanselmo]

No problem! I think it would help if you shared your code so that I can replicate the error. I can't remember having this issue off the top of my head.

[travisendicott]

Absolutely! Hopefully this copies in correctly. I'm starting out by looking at Monterey Bay FC in the USL Championship and then I'll move onto other teams. But I wanted to get data for the first week first.

### Load Libraries
library(ggplot2)
library(dplyr)
library(tidyverse)
library(itscalledsoccer)
library(data.table)
library(lubridate)

### Get USL-Championship Game Data
knitr::opts_chunk$set(echo = FALSE)

asa_data <- AmericanSoccerAnalysis$new()
uslc_names <- asa_data$get_players(
  leagues = "uslc")
uslc_teams <- asa_data$get_teams(leagues = "uslc")
uslc_managers <- asa_data$get_managers(leagues = "uslc")
uslc_goalies <- asa_data$get_goalkeeper_xgoals(leagues = "uslc")
uslc_gamesh <- asa_data$get_games(leagues = "uslc")
uslc_gamesa <- asa_data$get_games(leagues = "uslc")
uslc_games1 <- merge(uslc_teams, uslc_gamesh, by.x=c("team_id"), by.y=c("home_team_id"))
uslc_games1 <- merge(uslc_games1, uslc_managers, by.x=c("home_manager_id"), by.y=c("manager_id"))
setnames(uslc_games1, new = c('home_manager_name', 'home_manager_nationality'),
         old = c('manager_name', 'nationality'))
uslc_games1 <- uslc_games1[, c(2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 1, 24, 25)]
uslc_games2 <- merge(uslc_teams, uslc_gamesa, by.x=c("team_id"), by.y=c("away_team_id"))
uslc_games2 <- merge(uslc_games2, uslc_managers, by.x=c("away_manager_id"), by.y=c("manager_id"))
setnames(uslc_games2, new = c('away_manager_name', 'away_manager_nationality'),
         old = c('manager_name', 'nationality'))
uslc_games2 <- uslc_games2[, c(2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 1, 24, 25)]
setnames(uslc_games1, new = c('home_team_name', 'home_team_short_name', 'home_team_abb'),
         old = c('team_name', 'team_short_name', 'team_abbreviation'))
setnames(uslc_games2, new = c('away_team_name', 'away_team_short_name', 'away_team_abb'),
         old = c('team_name', 'team_short_name', 'team_abbreviation'))
home_teams <- subset(uslc_games1, select = c("game_id", "home_team_name", "home_team_short_name", "home_team_abb", "home_manager_id", "home_manager_name", "home_manager_nationality"))
uslc_games = left_join(uslc_games2, home_teams, by = 'game_id')
names(uslc_games)[1] <- 'away_team_id'
uslc_games$date <- date(uslc_games$date_time_utc)
uslc_games <- uslc_games[, c(5, 32, 14, 15, 16, 17, 7, 8, 13, 19, 20, 21, 22, 9, 26, 27, 28, 29, 30, 31, 1, 2, 3, 4, 23, 24, 25, 10, 11, 6)]
names(uslc_games)[18] <- 'home_manager_id'
rm(uslc_games1, uslc_games2, uslc_gamesa, uslc_gamesh)

### Game 1 - Monterey Bay FC
uslc_g1c <- asa_data$get_player_xgoals(
    leagues = "uslc",
    start_date = "2022-03-11",
    end_date = "2022-03-13"
)

uslc_ga1 <- asa_data$get_player_goals_added(
    leagues = "uslc",
    start_date = "2022-03-11",
    end_date = "2022-03-13"
)

uslc_xp1 <- asa_data$get_player_xpass(
    leagues = "uslc",
    start_date = "2022-03-11",
    end_date = "2022-03-13"
)

uslc_g1t = left_join(uslc_names, uslc_mbfc_g1c, by = 'player_id')
uslc_g1 = left_join(uslc_teams, uslc_mbfc_g1t, by = "team_id")
rm(uslc_g1c, uslc_g1t)
mbfc_gm1 <- subset(uslc_g1, team_name=="Monterey Bay FC")
mbfc_gm1$date <- c("2022-03-13")
mbfc_gm1$date <- date(mbfc_gm1$date)
mbfc_gm1 = right_join(uslc_games, mbfc_gm1, by = 'date')
mbfc_gm1 <- subset(mbfc_gm1, away_team_short_name=="Monterey Bay")

uslc_ga1 = left_join(uslc_names, uslc_ga1, by = 'player_id')
uslc_ga1 = left_join(uslc_teams, uslc_ga1, by = 'team_id')
mbfc_ga1 <- subset(uslc_ga1, team_name=="Monterey Bay FC")
mbfc_ga1$date <- c("2022-03-13")
mbfc_ga1$date <- date(mbfc_ga1$date)
mbfc_ga1 = right_join(uslc_games, mbfc_ga1, by = 'date')
mbfc_ga1 <- subset(mbfc_ga1, away_team_short_name=="Monterey Bay")
mbfc_gm1 <- left_join(mbfc_gm1, mbfc_ga1)

gplus_data <- lapply(mbfc_gm1$data,
                     function(temp){
                       temp %>% 
                         select(-c(count_actions, goals_added_raw)) %>% 
                         pivot_wider(names_from = action_type, values_from = c(goals_added_above_avg))
                     }) %>%
  bind_rows()

[mattyanselmo]

I'm hitting on error on what ends up being line 58 uslc_g1t = left_join(uslc_names, uslc_mbfc_g1c, by = 'player_id'). I don't see the object uslc_mbfc_g1c.

Generally, if you're trying to get game-by-game data for players and teams, here's an example snippet. It's a bit ugly because I didn't have the package documentation handy, but I hacked it together. Hopefully it's helpful!

library(ggplot2)
library(dplyr)
library(tidyverse)
library(itscalledsoccer)
library(data.table)
library(lubridate)
### Get USL-Championship Game Data
knitr::opts_chunk$set(echo = FALSE)

## SELECT TEAM ####
team_input <- "MB"

asa_data <- AmericanSoccerAnalysis$new()
uslc_names <- asa_data$get_players(
  leagues = "uslc")
uslc_teams <- asa_data$get_teams(leagues = "uslc")
team_id_input <- uslc_teams$team_id[uslc_teams$team_abbreviation == team_input]
uslc_managers <- asa_data$get_managers(leagues = "uslc")
uslc_goalies <- asa_data$get_goalkeeper_xgoals(leagues = "uslc")
uslc_games <- asa_data$get_games(leagues = "uslc") %>%
  mutate(date = as.Date(date_time_utc))

# Get Monterrey games
selected_team_games <- uslc_games %>%
  left_join(uslc_teams %>% select(home_team_id = team_id, home_team = team_abbreviation),
            by = "home_team_id") %>%
  left_join(uslc_teams %>% select(away_team_id = team_id, away_team = team_abbreviation),
            by = "away_team_id") %>%
  filter(home_team == team_input | away_team == team_input)

# Get team xg results by game
team_xg_by_game <- asa_data$get_game_xgoals(leagues = "uslc")
selected_team_xg_by_game <- team_xg_by_game %>%
  filter(game_id %in% selected_team_games$game_id)

mb_player_xgoals <- data.frame()
for(i in 1:nrow(selected_team_games)){
  mb_player_xgoals <- bind_rows(mb_player_xgoals,
                                asa_data$get_player_xgoals(
                                  leagues = "uslc",
                                  start_date = selected_team_games$date[i],
                                  end_date = selected_team_games$date[i]
                                ) %>%
                                  filter(team_id == team_id_input) %>%
                                  mutate(game_id = selected_team_games$game_id[i],
                                         date = selected_team_games$date[i]))
}

mb_player_goals_added <- data.frame()
for(i in 1:nrow(selected_team_games)){
  mb_player_goals_added <- bind_rows(mb_player_goals_added,
                                     asa_data$get_player_goals_added(
                                       leagues = "uslc",
                                       start_date = selected_team_games$date[i],
                                       end_date = selected_team_games$date[i]
                                     ) %>%
                                       filter(team_id == team_id_input)  %>%
                                       mutate(game_id = selected_team_games$game_id[i],
                                              date = selected_team_games$date[i]))
}

mb_gplus_data <- lapply(mb_player_goals_added$data,
                     function(temp){
                       temp %>% 
                         select(-c(count_actions, goals_added_raw)) %>% 
                         pivot_wider(names_from = action_type, values_from = c(goals_added_above_avg))
                     }) %>%
  bind_rows()

mb_player_goals_added <- mb_player_goals_added %>%
  select(-data) %>%
  bind_cols(mb_gplus_data)

[travisendicott]

Thank you so much. This will do everything that I need! I can't express how much easier you made my hobby of analyzing these data.

[travisendicott]

I'm running into an issue trying to replicate this with the xpass data. My goal is to get the xpass data per game for MB. I have been unable to do this by game with code and I can do it while searching on the website while using the date range function. When I just to use the similar code for xg, I encounter errors saying that the "selected_team_games$date" isn't available. Any suggestions on how to force the search by date?

[mattyanselmo]

I think something like this is what you're looking for! This would get each game for all teams for the month of April, 2022.

df <- data.frame()
for(date in seq.Date(as.Date("2022-04-01"), as.Date("2022-04-30"), "days")){
    date <- as.Date(date, origin = "1970-01-01")
    df <- bind_rows(df,
                    asa_client$get_team_xpass("mls", start_date = date, end_date = date) %>% mutate(game_date = date))
}