unexpected_points.Rmd

---
title: "Unexpected Points"
output: html_notebook
---

This notebook reads in the data from the Unexpected Points Substack.

```{r libraries, include=TRUE, warning=FALSE, message=FALSE}
library(tidyverse)
library(googlesheets4)
library(nflverse)
library(gt)
library(gtExtras)
library(ggtext)
library(ggrepel)
library(patchwork)
library(ggimage)
```

```{r nfl_analytics_theme, include=TRUE}
nfl_analytics_theme <- function(..., base_size = 12) {
  
  theme(
    text = element_text(family = "Lexend",
                        size = base_size,
                        color = "black"),
    axis.ticks = element_blank(),
    axis.title = element_text(face = "bold"),
    axis.text = element_text(face = "bold"),
    plot.title.position = "plot",
    plot.title = element_markdown(size = 16,
                                  vjust = .02,
                                  hjust = 0.5),
    plot.subtitle = element_markdown(hjust = 0.5),
    plot.caption = element_markdown(size = 8),
    panel.grid.minor = element_blank(),
    panel.grid.major =  element_line(color = "#d0d0d0"),
    panel.background = element_rect(fill = "#f7f7f7"),
    plot.background = element_rect(fill = "#f7f7f7"),
    panel.border = element_blank(),
    legend.background = element_rect(color = "#F7F7F7"),
    legend.key = element_rect(color = "#F7F7F7"),
    legend.title = element_text(face = "bold"),
    legend.title.align = 0.5,
    strip.text = element_text(face = "bold"))
}
```

```{r teams, include=TRUE}
teams <- nflreadr::load_teams(current = TRUE)
```

```{r df_aqe, include=TRUE}
df_aqe <- read_sheet('https://docs.google.com/spreadsheets/d/1ktlf_ekms7aI6r0tF_HeX0zaxps-bHWYsgglUReC558/edit?gid=1212845642#gid=1212845642',
           sheet = 'Adjusted Quarterback Efficiency')

df_aqe %>% 
  filter(plays > 100,
         season == 2024)
```

```{r}
df_aqe %>% 
  filter(season == 2024) %>% 
  arrange(epa_rank)
```

```{r play_by_play_2024, include=TRUE}
play_by_play_2024 <- nflreadr::load_pbp(seasons = most_recent_season())
```

From Congelio:

> In the case of Roethlisberger, the three pass attempt difference was the result of the passer_player_name grouping including three QB spikes in the data. **Aside from when attempting to replicate the official statistics, it is better to use just passer as it removes those instances where a QB spike, scramble, or sack my skew the results of your data.**

```{r}
play_by_play_2024 %>% 
  filter(!is.na(passer)) %>% 
  group_by(passer) %>% 
  summarise(qb_epa = sum(qb_epa, na.rm = TRUE),
            plays = n(),
            .groups = 'drop') %>% 
  mutate(qb_epa_per = qb_epa / plays,
         epa_rank = rank(-qb_epa)) %>% 
  arrange(-qb_epa)
```



```{r}
df_aqe %>% 
  filter(name == 'R.Wilson',
         season == 2024)
```

```{r}
play_by_play_2024 %>% 
  # select(contains('epa'))
  filter(game_id == '2024_07_NYJ_PIT',
         play_type == 'pass') %>% 
  group_by(passer) %>% 
  summarise(sum_epa = sum(epa, na.rm = TRUE),
            sum_epa_qb = sum(qb_epa, na.rm = TRUE),
            sum_epa_yac = sum(yac_epa, na.rm = TRUE),
            sum_epa_air = sum(air_epa, na.rm = TRUE),
            sum_epa_comp_yac = sum(comp_yac_epa, na.rm = TRUE),
            sum_epa_xyac = sum(xyac_epa, na.rm = TRUE),
            plays = n(),
            .groups = 'drop') %>% 
  mutate(test = sum_epa_air + sum_epa_yac) %>% 
  select(passer, test, everything())
```

The AQE methodology says that it uses the charting data from FTN.

```{r ftn_charting, include=TRUE}
ftn_charting <- nflreadr::load_ftn_charting(seasons = most_recent_season())
```

```{r ftn_charting_head, include=TRUE}
ftn_charting %>% head()
```