---
title: "UFC Fight"
author: "Luke Keirn"
output:
flexdashboard::flex_dashboard:
source_code: embed
orientation: columns
vertical_layout: fill
---
<style>
.chart-title
{
/* chart_title */
font-size: 18px;
font-family: Arial;
}
body
{
/* Normal */
font-size: 18px;
}
</style>
```{r setup, include=FALSE}
library(flexdashboard)
library(tidyverse)
library(DT)
library(plotly)
raw_fighter_details <- read_csv("raw_fighter_details.csv")
data <- read.csv("data.csv")
```
UFC
===
Column {data-width=1000}
---
### Introduction
<h1 style="font-size:3vw"><b>Influences on a UFC Fight</b></h1>
<br>
**What was I looking to find? **
- What is better a striker or a wrestler?
- What external factors can influence the result of a UFC fight?
**How will this be looked at?**
- Strike Attempts
- Take Down Average
- Overall Records
- Reach
Strikers/Grapplers:Strikes/Min
===
Column {data-width=450}
---
### Strike Attempts Per Minute For Top Five Strikers
- More than two strikes per minute
- Five minutes per Round
- Three to five rounds
- 50 or more per fight
<br>
#### Result
```{r plot1, fig.align= 'center', out.width="80%"}
library(ggplot2)
selected_fighters <- c("Conor McGregor", "Anderson Silva", "Israel Adesanya", "Max Holloway", "Jose Aldo")
striker_data <- raw_fighter_details[raw_fighter_details$fighter_name %in% selected_fighters, ]
ggplot(striker_data, aes(x = fighter_name, y = SApM, fill = SApM)) +
geom_bar(stat = "identity", position = "dodge", width = 0.7) +
scale_fill_gradient(low = "white", high = "blue", limits = c(1, NA)) +
labs(x = "Fighter Names", y = "Strike Attempts Per Minute (SApM)", title = "SApM Distribution for Top 5 Strikers") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
```
Column {data-width=450}
---
### Strike Attempts Per Minute For Top Five Grapplers
- Everyone, but Daniel Cormier
- Less strikes thrown
- Less exhausting
<br>
#### Result
```{r plot2, fig.align = 'center', out.width="80%"}
library(ggplot2)
selected_wrestlers <- c("George St. Pierre", "Matt Hughes", "Randy Couture", "Daniel Cormier", "Chael Sonnen")
wrestler_data <- raw_fighter_details[raw_fighter_details$fighter_name %in% selected_wrestlers, ]
ggplot(wrestler_data, aes(x = fighter_name, y = SApM, fill = SApM)) +
geom_bar(stat = "identity", position = "dodge", width = 0.7) +
scale_fill_gradient(low = "white", high = "blue", limits = c(1, NA)) +
labs(x = "Fighter Names", y = "SApM", title = "SApM Distribution for Top 5 Grapplers") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
```
Strikers/Grapplers:Takedown Avg.
===
Column {data-width=450}
---
### Take Down Average For Top Five Strikers
- Not even 1 per match
- Don't look for the ground
<br>
#### Result
```{r plot3, fig.align = 'center', out.width="80%"}
library(ggplot2)
selected_fighters <- c("Conor McGregor", "Anderson Silva", "Israel Adesanya", "Max Holloway", "Jose Aldo")
striker_data <- raw_fighter_details[raw_fighter_details$fighter_name %in% selected_fighters, ]
ggplot(striker_data, aes(x = fighter_name, y = TD_Avg)) +
geom_tile(aes(fill = TD_Avg)) +
scale_fill_gradient(low = "white", high = "blue", limits = c(1, NA)) +
labs(x = "Fighter Names", y = "TD_Avg", title = "TD_Avg Heatmap for Top 5 Strikers") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
```
Column {data-width=450}
---
### Take Down Average For Top Five Grapplers
- Higher average per match
- Looking for the ground
<br>
#### Result
```{r plot4, fig.align = 'center', out.width="80%"}
library(ggplot2)
selected_fighters <- c("George St. Pierre", "Matt Hughes", "Randy Couture", "Daniel Cormier", "Chael Sonnen")
striker_data <- raw_fighter_details[raw_fighter_details$fighter_name %in% selected_fighters, ]
ggplot(striker_data, aes(x = fighter_name, y = TD_Avg)) +
geom_tile(aes(fill = TD_Avg)) +
scale_fill_gradient(low = "white", high = "blue", limits = c(1, NA)) +
labs(x = "Fighter Names", y = "TD_Avg", title = "TD_Avg Heatmap for Top 5 Grapplers") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
```
Strikes Absorbed:Top Fighters
===
Column {data-width=450}
---
### Strikes Absorbed For Top Grapplers and Strikers
- Fairly even
- Different styles but similar results
Column {data-wdith=550}
---
### Result
```{r plot5, fig.align = 'center', out.width="80%"}
library(reshape2)
raw_fighter_details <- data.frame(
fighter_name = c("George St. Pierre", "Matt Hughes", "Randy Couture", "Daniel Cormier",
"Chael Sonnen", "Conor McGregor", "Anderson Silva", "Israel Adesanya",
"Max Holloway", "Jose Aldo"),
SApM = c(1.5, 2.0, 1.2, 1.8, 2.2, 2.5, 1.7, 1.6, 1.9, 1.4)
)
hc <- hclust(dist(raw_fighter_details$SApM), method = "average")
plot(hc, main = "Hierarchical Clustering for Strikes Absorbed",
xlab = "Fighters", ylab = "SApM", labels = raw_fighter_details$fighter_name)
```
Record: Strikers/Grapplers
===
Column {data-width=450}
---
### Record For Top Five Strikers
- Blue = Blue Fighter
- Red = Red Fighter
- More of a mixture
<br>
#### Result
```{r plot6, fig.align = 'center', out.width="80%"}
library(ggplot2)
selected_fighters <- c("Anderson Silva", "Max Holloway", "Conor McGregor", "Israel Adesanya", "Jose Aldo")
selected_dates <- as.Date(c("2010-08-07", "2011-02-05", "2011-08-27", "2012-07-07", "2012-10-13",
"2013-07-06", "2013-12-28", "2015-01-31", "2016-02-27",
"2016-07-09", "2017-02-11", "2019-02-09", "2019-05-11", "2020-10-03",
"2012-02-04", "2013-05-25", "2013-08-17", "2019-04-13", "2019-12-14",
"2020-07-11", "2016-03-05", "2018-10-06", "2021-01-23",
"2021-03-06", "2015-12-12", "2017-06-03", "2017-12-02",
"2019-05-11", "2019-12-14", "2020-07-11"))
filtered_data <- data[(data$B_fighter %in% selected_fighters | data$R_fighter %in% selected_fighters) &
data$date %in% selected_dates, ]
filtered_data$color <- ifelse(filtered_data$date %in% c("2013-07-06", "2013-12-28", "2016-02-27", "2016-07-09",
"2019-02-09", "2019-05-11", "2020-10-03", "2012-02-04",
"2013-05-25", "2013-08-17", "2019-04-13", "2019-12-14",
"2020-07-11", "2016-03-05", "2018-10-06", "2021-01-23",
"2021-03-06", "2015-12-12", "2017-06-03", "2017-12-02",
"2019-05-11", "2019-12-14", "2020-07-11"),
"Red", "Blue")
ggplot(filtered_data, aes(x = date, y = Winner, color = color)) +
geom_point() +
scale_color_manual(values = c("Red" = "red", "Blue" = "blue")) +
labs(x = "Date", y = "Winner", title = "Top Five Strikers Overall Record") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
```
Column {data-width=450}
---
### Record For Top Five Grapplers
- Blue = Blue Fighter
- Red = Red fighter
- Less mixture of colors
<br>
#### Result
```{r plot7, fig.align = 'center', out.width="80%"}
library(ggplot2)
selected_fighters <- c("George St. Pierre", "Matt Hughes", "Randy Couture", "Daniel Cormier", "Chael Sonnen")
selected_dates <- as.Date(c("2004-10-22", "2007-04-07", "2000-12-16", "2004-01-31", "2006-11-18",
"2007-12-29", "2008-06-07", "2010-11-20", "2002-03-22", "2002-09-27",
"2004-01-31", "2005-04-16", "2006-02-04", "2008-11-15", "2009-08-29",
"2011-04-30", "2015-01-03", "2019-08-17", "2005-10-07", "2006-05-27",
"2009-02-21", "2010-08-07", "2012-07-07", "2013-04-27", "2013-11-16"))
filtered_data <- data[(data$B_fighter %in% selected_fighters | data$R_fighter %in% selected_fighters) &
data$date %in% selected_dates, ]
filtered_data$color <- ifelse(filtered_data$date %in% c("2004-10-22", "2007-04-07", "2000-12-16", "2004-01-31",
"2006-11-18", "2007-12-29", "2008-06-07", "2010-11-20",
"2002-03-22", "2002-09-27", "2004-01-31", "2005-04-16",
"2006-02-04", "2008-11-15", "2009-08-29", "2011-04-30",
"2015-01-03", "2019-08-17", "2005-10-07", "2006-05-27",
"2009-02-21", "2010-08-07", "2012-07-07", "2013-04-27",
"2013-11-16"),
"Red", "Blue")
ggplot(filtered_data, aes(x = date, y = Winner, color = color)) +
geom_point() +
scale_color_manual(values = c("Red" = "red", "Blue" = "blue")) +
labs(x = "Date", y = "Winner", title = "Top Five Grapplers Overall Record") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
```
Reach:Red/Blue Fighter
===
Column {data-width=450}
---
### Reach in Fights (Advantage For Red Fighter)
- Across all weights classes
- Red had reach advantage
- Resulted in wins
<br>
#### Result
```{r boxplot2, fig.align = 'center', out.width="80%"}
library(ggplot2)
filtered_data <- data[data$R_Reach_cms > data$B_Reach_cms, ]
win_counts <- table(filtered_data$Winner)
win_data <- data.frame(Winner = names(win_counts), Wins = as.numeric(win_counts))
ggplot(win_data, aes(x = reorder(Winner, -Wins), y = Wins, fill = Winner)) +
geom_bar(stat = "identity") +
labs(x = "Winner", y = "Number of Wins", title = "Red Wins With Reach Advantage") +
theme_minimal()
```
Column {data-width=450}
---
### Reach in Fights (Advantage for Blue Fighter)
- Across weight all classes
- Blue had reach advantage
- Did not result in wins
<br>
#### Result
```{r boxplot3, fig.align = 'center', out.width="80%"}
library(ggplot2)
filtered_data <- data[data$B_Reach_cms > data$R_Reach_cms, ]
win_counts <- table(filtered_data$Winner)
win_data <- data.frame(Winner = names(win_counts), Wins = as.numeric(win_counts))
ggplot(win_data, aes(x = reorder(Winner, -Wins), y = Wins, fill = Winner)) +
geom_bar(stat = "identity") +
labs(x = "Winner", y = "Number of Wins", title = "Blue Wins With Reach Advantage") +
theme_minimal()
```
