optimize backend when data and annotations are big

[seankross]

library(dplyr)
library(Lahman)

small_batting <- Batting %>% 
    filter(yearID >= 2010) %>% 
    select(playerID, yearID, HR)

small_batting %>% 
  group_by(yearID) %>% 
  arrange(desc(HR), .by_group = TRUE) %>% 
  slice(1)