Visualizing Field Strength – Wells Fargo Championship 2016

May 3, 2016 at 12:53 pm

The following are Box plots of the most recent normalized adjusted scores for the top half of the field at the Wells Fargo Championship. The golfers are plotted in order of their expected score from the previous post. Clicking on each image will take you to the plot – click the “Full Screen” button in the top right for a full screen version. Box plots are a quick way to show the quartiles of each player’s playing history.

Each dot above a player’s name represents an adjusted normalized round score. Normalized adjusted scores are agnostic to field strength and course difficulty.  The average normalized score is 0. A score of 1 indicates a round that was one standard deviation worse than a standardized field. A score of -1 indicates a round that was one standard deviation better than a standardized field. Negative is good – this is golf.

The shaded box for each player shows the middle 50% (middle 2 quartiles) of that player’s normalized scores over the past 50 rounds that he has played. The horizontal line in the middle of each box represents that player’s median normalized adjusted score over the past 50 rounds. You could think of the top of each box as the player’s floor (it’s backwards I know, negative is good), i.e. 75% of the time the player will do better than the floor. For example, consider the leftmost box in the first plot, Henrik Stenson. The top of the box (his ‘floor’) is 0. Hover over Stenson’s box with your cursor to see. That means that 75% of the time Stenson will do better than the field average, which is 0. You can think of the bottom of each box as the player’s ceiling (again backwards). So, the lower the bottom of the box, the more upside. Consider Jamie Lovemark, the large purple box near the center of the first plot. The box is large, therefore the range of his scores tends to be large. But, he also has a lot of upside, the lower edge of the box is low.

In summary, the lower the score the better. So, the lower the box the better. The bottom of the box can be thought of as upside, the top as downside.

The thin lines that extend from the top and bottom of each box, the whiskers, represent the range of the worst and best 25% (top and bottom quartile) of each player’s rounds, respectively. Again, negative is good. The whiskers extending below the box are good rounds, the whiskers above are bad rounds. Rounds (dots) outside of the range of the whiskers are outliers.

If a box and its whiskers are wider, they indicate that a player has a wider range of scores. Narrower boxes and whiskers indicate a more consistent golfer.

These box plots are strictly empirical and not parameterized like the projections in the previous post. In English, the box plots show the actual data only. The projections in the previous post are obtained by fitting the data to parameters, mean and standard deviation, of a normal distribution. That parameterized distribution is then used to predict scores.



Two golfers really stand out to me: J.B. Holmes and Charles Howell III, 9100 and 8100 on DraftKings (pink and grey boxes on the 1st graph). Their play is consistently good and in a narrow range. Roughly 75% of their rounds are better than average (i.e. below 0). They lack the truly outstanding (very negative) rounds that players like Justin Rose and Rory McIlroy have, so they are less likely to win, but they also have fewer bad (positive) rounds. These two look like good values for cash game on DraftKings.

Some players with good upside that stand out for tournaments are: Kevin Chappell, Jamie Lovemark, Justin Kokrak, and Jonas Blixt.

What do you see? Let me know in the comments.