Introduction
What influences driving distance in the game of golf? Throughout the many years, the sport of golf has seen many changes in the way that players approach the game. It is no question that Tiger Woods is the greatest golfer and one of the greatest athletes of all time. Woods has always been must see since he burst onto the scene as an amateur. He has been a vital figure in changing the way that the game is played. Early on in his career, Tiger’s power was matched by very few. He could hit distances that others couldn’t even think about. Tiger said that at that time, his daily routine consisted of a four-mile jog, going to the gym to lift weights, two to three hours of range practice, playing some holes, returning to work on his short game, running another four miles, and then playing basketball or tennis. Ever since then there has been a shift in the game to longer hitters and especially off the tee with the driver. Today, one of the faces of long hitters is Bryson DeChambeau. He has received the nickname “The Scientist” because of his unique approach to breaking down the swing to maximize distance. He looks at every detail and every measure to improve his distance.
Bryson DeChambeau and his all-out desire to increase his distance has led many people to wonder what contributes to longer distances with the driver? Learning about what in the golf swing contributes to longer distances is key in the evolution of golf. I wanted to test this so I created a regression model with average driving distance as my Y variable. Average driving distance helps analyze longer hitters because hitting it longer means further distances. Setting average driving distance as the Y variable will allow me to see what parts of the swing (predictor variables) are correlated with higher average distances.
Model
To run this regression, this was the model that I created:
Average Driving Distance = α +β1AverageClubHeadSpeed + β2AverageLaunchAngle + β3AverageSpinRate + β4AverageHangTime + β5Age + β6Age2 + ε
In my model, I wanted to see what impacts average driving distance. The attributes that I wanted to incorporate into my model were mostly all about the swing and the main components that makeup the swing. I wanted to choose variables that I thought best encapsulated that.
Figure 1
I got my data from the PGA Tour website and from ESPN. All the averages about the swing and distances came from the PGA Tour’s website and all the age data came from ESPN. This data is from the 2021, 2019, 2018, 2017, and 2016 seasons. I decided to skip the 2020 season because that season was affected by Covid. The variables that I chose to represent in my model were average clubhead speed, average launch angle, average spin rate, average hang time, and age. Figure 1 shows the summary statistics of the variables that I included in my model. Other than age, all the variables were pretty normally distributed. Most of the data was centralized near the averages. Age was skewed right a little but that makes sense because golf is a sport that doesn’t have as much wear and tear on your body compared to other sports. This allows people to play into their later ages as can be seen by Figure 1 which shows the maximum age of a player on tour was 52. I chose to represent Age as a quadratic because I wanted to see if there was an age that maximized average driving distance.
Average driving distance is the average distance that each player hits their driver throughout a season. Drive distance is measured in yards from where the ball is on the tee to where the ball comes to rest after a player hits it. Average clubhead speed is a player’s average clubhead speed on their drives throughout a season. Clubhead speed is the speed in miles per hour when the club impacts the ball. Average launch angle is the average launch angle on a player’s drives during a season. Launch Angle is the vertical launch angle of the ball right after it has been hit. Average spin rate is the average spin rate that a player had on his drives during a season. Spin rate is the revolutions per minute of the ball right after it has been hit. Average hang time is the average hang time that a player had on his drives during a season. Hang time is the time in seconds of the ball from when it leaves the tee to when it hits the ground. Age is a player’s age during a particular season.
Results
Figure 2
The results of the regression that I ran are in Figure 2. The adjusted R-squared value for my model was 0.8403 meaning that 84.03% of the variation in average driving distance can be explained by the variables in the model. Although it is not always appropriate to base the effectiveness of a model solely on R-squared, having an R-squared close to 1 is very good. In my model, 4 of my predictor variables were statistically significant at the 0.01% level. Those variables were average clubhead speed, average launch angle, average spin rate, and average hang time. In my model I also checked for multicollinearity to see how correlated by predictor variables were with each other. To do this I used the vif function in R and all my non quadratic variables had a vif between 0 and 3. This is good when testing for multicollinearity, so I knew that multicollinearity wasn’t going to have a big impact on my results.
Average clubhead speed had a coefficient of about 1.884. This means that for every 1 mph increase in average clubhead speed, average driving distance increases by 1.884 yards. A faster clubhead speed leads to further distances. This was statistically significant at the 0.01% level. Average launch angle had a coefficient of about 0.537. This means that for every 1 degree increase in average launch angle, average driving distance increases by 0.537 yards. This was statistically significant at the 0.01% level. Average spin rate had a coefficient of about -0.010. This means that for every 1 rpm increase in average spin rate, average driving distance decreases by 0.010 yards. This was statistically significant at the 0.01% level. Average hang time had a coefficient of about 4.505. This means that for every 1 second increase in hang time, average driving distance increased by 4.505 yards. This was statistically significant at the 0.01% level.
The coefficients for Age and Age2 were positive and negative respectfully. This means that there is an age where average driving distance is maximized. Players tended to increase their average driving distance as they get more experience and older, but it peaks at a certain point and goes back down as they get older and older.
The results of my model showed what impacts the average driving distance of a PGA Tour golfer. Average clubhead speed, average launch angle, average spin rate, and average hang time all had an impact on average driving distance. The quadratic term for age shows that the age where average driving distance is maximized is around 24 years old. Although this is what it predicts, this number can be impacted by some bias. Golf is a sport that players can play for many years. As can be seen in figure 1, the age range of players was from 20 to 52. Typically when younger players in the low 20’s join the tour, they were superstars at the collegiate level. To join the tour at that young of an age means that they must be a very good player. These players are also younger which gives them more opportunities to swing out of their shoes without risk of injury. These could all have an impact on why the age where average driving distance is maximized is closer towards the lower end of the range.
Golf has seen many changes throughout its history since the PGA Tour was founded in 1929. Today it has seen players try to alter their games to increase their distances and in particular their driving distances. Players have many statistics on their swings and have broken them down piece by piece. Average clubhead speed, average launch angle, average spin rate, average hang time, and age are all factors that have an impact on a player’s golf swing and their average driving distances.