Determinants of Scoring at the St Jude Championship

This analysis is based on scores and stats from individual rounds in the last 10 Tour events at TPC Southwind: 3,765 rounds in total.

Section 1: Absolute Correlation Coefficients with Score

Correlation Coefficients Analysis (2014-2023)

Absolute Correlation between Score and SG Metrics
Figure 1: Absolute Correlation between Score and SG Metrics

SGTee: Throughout the analysed period, the correlation between Score and SGTee has remained moderately consistent, hovering around the 0.3 to 0.4 range. Notably, there was a subtle uptick in 2017-2019, with the highest correlation recorded in 2023. This steadiness suggests that the ability off the tee contributes moderately to a player's success at TPC Southwind. However, given that the course places a premium on precise approach shots and challenging greens, the impact of driving proficiency, while important, is not the sole determinant of scoring excellence.

SGApp: SGApp exhibits a robust and consistent correlation with Score, with peaks observed in 2017 and 2023. The correlation coefficients consistently surpass 0.6, underscoring the critical role of approach shots in influencing player performance. TPC Southwind's design, characterised by its strategically placed hazards and undulating greens, demands precision in approach play. Thus, players who excel in this metric tend to have a competitive edge, reinforcing its significance in handicapping assessments.

SGATG: The correlation of SGATG with Score shows fluctuations across the years, with a noticeable dip in 2021. Despite these variations, the metric maintains a moderate correlation, generally within the 0.3 to 0.4 range. This trend indicates that proficiency around the green - encompassing skills like chipping and bunker play - can vary in its impact on overall scoring. Factors such as course setup, weather conditions, and green firmness could influence the relative importance of around-the-green skills in different years.

SGP: The correlation between Score and SGP presents notable variability, with higher correlations in years like 2017 and 2019. The coefficients often exceed 0.5, highlighting the significance of putting performance at TPC Southwind. The course's complex green structures mean that players with superior putting skills can capitalise on scoring opportunities, making SGP a pivotal factor in distinguishing top performers.

Absolute Correlation between Score and Traditional Metrics
Figure 2: Absolute Correlation between Score and Traditional Metrics

Driving Distance: The correlation between Score and Driving Distance is consistently low, often below 0.1, indicating a minimal impact on overall scoring at TPC Southwind. This suggests that the course's design does not disproportionately reward long hitters. Instead, it favours strategic play and precision, aligning with its tight fairways and the presence of hazards that penalise errant long shots.

Driving Accuracy: Exhibiting a more substantial correlation, particularly peaking in 2022, Driving Accuracy underscores its importance at TPC Southwind. With correlation coefficients reaching up to 0.36, maintaining positions on the fairway is evidently crucial. The course's layout, featuring narrow fairways bordered by water hazards and dense rough, penalises inaccuracy, thereby elevating the significance of this metric in influencing scores.

Greens In Regulation (GIR): GIR consistently displays a strong correlation with Score, often exceeding 0.5. This trend highlights the imperative of reaching the green in the regulation number of strokes. Given the challenging nature of TPC Southwind's greens, players who secure more GIR opportunities mitigate the risk of high scores, reflecting the metric's critical role in successful play.

Scrambling: With correlation coefficients typically around 0.5, Scrambling emerges as a vital skill, indicating a player's ability to recover and save par after missing the green. The consistent importance of this metric suggests that even proficient approach players will inevitably face challenging situations, where adeptness in scrambling can salvage a round.

Putts Per GIR (PPGIR): PPGIR shows a strong correlation with Score, particularly notable in years like 2017. This metric's significance underscores the necessity of capitalising on GIR opportunities through effective putting. At TPC Southwind, where the greens can be treacherous, efficient putting after reaching the green is essential for low scoring.

Absolute Correlation between Score and Par Metrics
Figure 3: Absolute Correlation between Score and Par Metrics

Par 3: The correlation between Score and Par 3 performance remains moderate, generally around the 0.4 to 0.5 range. This consistency indicates that while scoring on Par 3s contributes meaningfully to overall performance, it is not the predominant factor. Given that Par 3s at TPC Southwind often feature water hazards and require precise club selection, proficiency on these holes aids in maintaining competitive scores.

Par 4: Demonstrating the strongest and most consistent correlation, often exceeding 0.8, Par 4 performance is evidently the most critical determinant of overall scoring. TPC Southwind's Par 4s are known for their length and strategic complexity, demanding a combination of accurate driving, precise approach play, and adept putting. Mastery of these holes significantly differentiates top performers from the rest of the field.

Par 5: With lower correlation coefficients, typically ranging from 0.3 to 0.4, Par 5 performance appears less influential on total scores. However, these holes still offer birdie opportunities that can be pivotal in tightly contested tournaments. Efficient scoring on Par 5s can provide the necessary cushion or advantage, especially when coupled with solid performance on the more challenging Par 4s and Par 3s.

Section 2: Importance of Each Metric in Determining Score

Random Forest Regressor and Feature Importance

Random Forest Regressor is an ensemble learning method that constructs multiple decision trees during training and outputs the average prediction. It combines the predictions of several models to improve accuracy and robustness.

Feature importance is a technique used to interpret a machine learning model. It refers to the score that quantifies the contribution of each feature to the prediction made by the model.

In a Random Forest, the importance of a feature is computed by looking at how much the feature decreases the impurity (e.g., variance for regression tasks) across all the trees in the forest. The more a feature decreases the impurity, the more important it is considered.

The calculated importance scores for all features are then normalized to give relative importance as a percentage. This shows the relative contribution of each feature to the prediction task.

Interpreting Feature Importance

Features with high relative importance percentages have a strong impact on the model's predictions. They are crucial for accurate predictions and indicate key areas where performance matters most.

Features with low relative importance have a minimal impact on the model's predictions. While they can still contribute, they are less critical.

Importance of Each Metric (2014-2023)

Relative Importance of SG Metrics on Score
Figure 1: Relative Importance of SG Metrics on Score

SGTee: 15.91% - Lower than the PGA Tour average, indicating that tee shots are less decisive at TPC Southwind.

SGApp: 39.53% - The most significant factor, highlighting the importance of approach shots on this course.

SGATG: 13.66% - Less influential, reflecting the course's lesser emphasis on short game skills.

SGP: 30.90% - Critical for success, given the complex greens at TPC Southwind.

Summary: Approach play and putting are paramount, with notable deviations from PGA Tour averages. Focus on these areas when predicting player performance.

Relative Importance of Traditional Metrics on Score
Figure 2: Relative Importance of Traditional Metrics on Score

DrivingDistance: 4.99% - Less impactful, aligning with the course's focus on precision.

DrivingAccuracy: 2.96% - Also less influential, reflecting the need for strategic play rather than distance.

GreensInRegulation (GIR): 24.99% - Important, though slightly lower than the PGA Tour average.

Scrambling: 35.65% - The most critical factor, underscoring the importance of recovery skills.

PPGIR: 31.41% - Vital for scoring well, especially on TPC Southwind's challenging greens.

Summary: Scrambling and putting are key at TPC Southwind, with traditional metrics showing notable differences from PGA Tour averages. Prioritise players with strong short games and putting abilities.

Relative Importance of Par Metrics on Score
Figure 3: Relative Importance of Par Metrics on Score

Par3: 20.56% - Important, especially given the challenging Par 3s at TPC Southwind.

Par4: 68.66% - The most critical factor, reflecting the significance of Par 4 performance at TPC Southwind.

Par5: 10.78% - Less influential, indicating that while Par 5s offer scoring opportunities, they are not as decisive as Par 4s.

Summary: Par 4 performance is paramount at TPC Southwind, contributing nearly 70% to the overall importance. Par 3s also play a significant role, while Par 5s are less influential compared to PGA Tour averages. Focus on players who excel on Par 4s to predict success at this course.

Top 5 Ranked Players - 2024 St Jude Championship

The table below shows the top-5 ranked players and their average estimated scores from the three different Random Forest models above.

Player Score
Scottie Scheffler 66.92
Xander Schauffele 67.44
Ludvig Aberg 67.57
Davis Thompson 68.29
Erik Van Rooyen 68.55

Estimated scores for all players can be found here.