• Analytics Blog
July 6th, 2018
How big is the ability gap between the professionals vying to one day compete on the PGA Tour, and the ones who have already made it?

In this blog post we analyze data from 2015-2017 on the PGA Tour and three developmental tours - the Web.com, Mackenzie (Canadian), and Latinoamerica tours- to answer that question. The three latter tours all provide a pathway to the PGA Tour: the top 25 players on the Web.com money list earn PGA Tour status, while the top 5 on both the Mackenzie and Latinoamerica tours earn Web.com status.

To make direct comparisons across tours, we use a statistical technique we’ve used several times before in other blog posts. By leveraging the fact that there is overlap between the sets of players that play on each of these tours (e.g. there are players who play in both PGA Tour and Web.com events in the same season), we are able to calculate a strokes-gained measure that is comparable across any tour. Intuitively, we can compare the quality of a PGA Tour field to that of a Web.com field by looking at how the same player performed relative to each of these fields. That is, if Joel Dahmen beats a PGA Tour field by 0.5 strokes on average, but beats a Web.com field by 1.5 strokes on average, we conclude that the Web.com field is 1 stroke weaker (ignoring statistical noise). Of course this is a simplified example; in practice we are aggregating many such comparisons simultaneously. For the full statistical details, see the endnote of this earlier article of ours.

After running the numbers, we obtain a strokes-gained measure that allows for direct comparisons of each round of golf from 2015-2017 on our 4 tours of interest. The interpretation of our strokes-gained measure is relative to the average PGA Tour professional. For example: Austin Connelly shot 71 in round 4 of the 2016 GolfBC Championship on the Mackenzie Tour. The field average that day was 70.3, but our strokes-gained measure for Connelly’s round was -3.01. Therefore, while Connelly only lost 0.7 strokes to the field that day, we estimate that he would have lost about 3 strokes to an average PGA Tour professional playing the course on that day.

On to the results: first let’s look at some summary measures of the quality of golfers on each tour. Plotted below are the strokes-gained averages of high-quality (95th percentile), average, and low-quality (5th percentile) players on each tour. For those a little stale on their first-year statistics, the golfer that occupies the 95th percentile has a better strokes-gained average than 95% of golfers on that tour.

Notes: Plotted are the average strokes-gained of the 5th percentile, average, and 95th percentile golfers on each indicated tour. Data includes all official stroke-play events from 2015-2017 on the PGA, Web.Com, Mackenzie, and Latinoamerica tours. A player had to play at least 25 rounds to be included. Each golfer is assigned to a tour based on where they played the majority of their rounds (but all of their rounds, irrespective of tour, are included to construct their strokes-gained averages).
On the Web.com tour, there are very few players better than the average PGA Tour professional. In our sample of data from 2015-2017, there were just 22 players who had seasons where they averaged positive strokes-gained (i.e. were better than an average PGA Tour pro). On the Mackenzie tour and Latinoamerica tours, it is an exceedingly rare occurrence to find a player consistently playing at the level of an average PGA Tour professional. It happened just 3 times from 2015-2017: Aaron Wise (+0.45) on the 2016 Mackenzie Tour, Ryan Ruffels (+0.09) on the 2016 Latinoamerica Tour, and Dan McCarthy (+0.13) on the 2016 Mackenzie Tour. It’s also interesting to note that the skill gap between the Web.com and the PGA Tour appears to be at least as large as the skill gap between the Mackenzie Tour and the Web.com.

Next we do a simple analysis that is inspired by the concept of expected wins introduced by Jake Nichols of 15th Club. The idea is that a given strokes-gained performance at a tournament is expected to result in a finish position in a certain range. For example, focusing on the PGA Tour for now, if a player gains 4 strokes per round (remember that, using our strokes-gained measure, this is 4 strokes better than an average PGA Tour pro), his expected finish position in a PGA Tour event is 1.9. The expected finish position will vary by the quality of the field at the tournament of interest. Playing 4 rounds at a level that is 4 strokes better than an average PGA Tour professional is more likely to result in a win against the relatively weak field at the Greenbrier than it is against the relatively strong field at the U.S. Open. But, without identifying the specific tournament, a golfer averaging 4 strokes-gained per round in a PGA Tour event has an expected finish position of 1.9.

Let's extend this logic to each of the tours in our data. If a golfer performs at a level of 1 stroke per round better than an average PGA professional at a Web.com event, what is his expected finish position? As it turns out, it’s about 9.4. Intuitively, we are calculating the average finish position on each tour by strokes-gained performance at the tournament. Another example for clarity: playing at a level of 1 stroke per round worse than an average PGA professional is actually quite good at a Latinoamerica event, as the expected finish position associated with this performance is 16.7.

The following plot visualizes the relationship between strokes-gained over an average PGA Tour professional and expected finish position on each of the 4 tours we analyze (the data has been smoothed, details in graph note). Hover over the graph for details!

Notes: Data is from 2015-2017. All official stroke-play events are included on the Web.com, Mackenzie, and Latinoamerica tours. For the PGA Tour, the World Golf Championship events, and the TOUR Championship and Tournament of Champions, are not included due to the small field size and absence of a cut. Only players who make the cut are included in the analysis: this is to more clearly understand the relationship between strokes-gained per round and finish position. For each tour, a cubic spline is fit to smooth the relationship between finish position and strokes-gained per round. The relationship looks very similar if you simply average finish position by strokes-gained bins (e.g. 0.2-0.4, 0.4-0.6, etc.).
A performance equal to that of an average PGA Tour professional over 4 rounds is expected to result in a finish of 51st on the PGA Tour, 25th on the Web.com Tour, 7th on the Mackenzie Tour, and 6th on the Latinoamerica Tour. Depending on your prior, this may be surprising for different reasons. To win on any of these tours, a performance that is above-average by PGA Tour standards is required. However, the quality of performances falls off quickly as you move down the leaderboard.

This paints a clear picture of what sort of consistent performances are required on each of the developmental circuits to be ready to compete with guys on the main Tour. On the Mackenzie and Lationamerica tours, consistently finishing in the mid-pack of those golfers who make the cut is simply not good enough: these performances are roughly equivalent to losing 1.5 to 2.5 strokes per round to an average PGA Tour pro.

We will end with a specific prediction. A name to watch out for on the Web.com Tour is Joseph Bramlett. Bramlett has played 12 events in 2018, making 11 cuts and notching 7 top 25s. Despite currently ranking outside the top 25 on the Web.com money list, we have Bramlett rated as the best player on the Web at +0.4 strokes-gained per round. With his performances so far in 2018 on the Web.com Bramlett has shown he can consistently perform at a higher level than PGA Tour calibre golfers; if he keeps it up, next year he'll get to do it for real.