11.12 Applying park factors
Let’s retrieve Andrés Galarraga’s 1996 plate appearances ending with a batted ball, and # defining a column was_hr is defined indicating whether the event was a home run.
# query <- '
# SELECT "away_team_id", LEFT("game_id", 3) AS "home_team_id", "event_cd"
# FROM "events"
# WHERE "year" = 1996
# AND "event_cd" IN (2, 18, 19, 20, 21, 22, 23)
# AND "bat_id" = \'galaa001\';
# '
#
# andres <- dbGetQuery(con, query) |>
# mutate(was_hr = ifelse(event_cd == 23, 1, 0))
andres <- read_rds('./data/andres.rds')
head(andres, 20)## away_team_id home_team_id event_cd was_hr
## 1 COL ATL 2 0
## 2 COL ATL 2 0
## 3 COL ATL 2 0
## 4 COL ATL 20 0
## 5 COL ATL 2 0
## 6 COL ATL 2 0
## 7 COL ATL 2 0
## 8 COL ATL 20 0
## 9 COL ATL 18 0
## 10 COL ATL 2 0
## 11 COL ATL 20 0
## 12 COL ATL 23 1
## 13 COL ATL 2 0
## 14 COL ATL 2 0
## 15 COL ATL 2 0
## 16 COL ATL 23 1
## 17 COL ATL 21 0
## 18 COL ATL 2 0
## 19 COL ATL 2 0
## 20 COL CHN 20 0Now, let’s calculate the pakr factor for Galarraga.
andres_pf <- andres |>
inner_join(ev_compare, by = c("home_team_id" = "team_id")) |>
summarize(mean_pf = mean(pf))
andres_pf## mean_pf
## 1 129The mean_pf indicates that Galarraga ad his home run frequency increased by an estimated 29% relative to a neutral environment.
## mean_pf
## 1 36.4According to our estimates, Galarraga’s benefit from the ballparks he played in (particularly his home Coors Field) amounted to roughly 47 -36 = 11 home runs in the 1996 season.