Who Shines on Sunday? Evaluating Professional Golf Final Round Performance from 2016 to 2020
It is the first week of 2021 and the PGA Tour is back, whether are ready for it or not. But before we get into the weekly rhythm of the Tour and professional golf in a new year, I wanted to look back at how the best players in the game have performed when they have a reasonable chance of winning over the last five seasons. Who has outperformed expectations on Sundays and who has underperformed? Who has been most unlucky? Can we finally answer the long-running DJ Pie vs. Soly debate on Tony Finau?
In order to evaluate this, I’ve relied on Data Golf’s fantastic Pressure Tool. If you are an avid golf fan and do not subscribe to Data Golf already, I highly recommend it; it is the Kenpom.com of golf analytics. I’ve selected a group of 22 players who were already playing professionally in 2016. In my opinion, these are the best players in golf over that span (players like Morikawa, Wolff and Hovland don’t have enough data yet). I only looked at tournaments where a player entered the final round in the Top 10. If you are outside the Top 10, you typically have less than a 3% chance of winning.
The Pressure Tool allows us to analyze final round performance in three ways, as illustrated by these examples:
Expected Wins entering Round 4: This is a player’s expected wins before a shot is hit, based on their power rating, the power rating of others on the leaderboard, and where they stand. For example, going into Sunday of the 2020 Masters, Dustin Johnson had 0.79 expected wins; he was expected to win 79% of the time. DJ was one of the top 3 players by power ranking going into the week and had a 4-shot lead.
Expected Wins given Actual Play in Round 4: This reflects the player’s expected wins given how he played in Round 4, without accounting for what other players shot. This one is a little more complicated, but bear with this example: in the 2016 Open Championship, Phil Mickelson famously lost to Henrik Stenson by 3 shots, but beat third place by 11. Phil’s Expected Wins entering Round 4 were 0.33 (he should win 33% of the time). His Expected Wins given Actual Play — he shot 66 — without accounting for the fact that Stenson shot 63, were 0.98 (he should win 98% of the time)! In other words, Phil played very well, but was extremely unlucky to be beat by the historic performance of Stenson.
Actual Wins: Call it the Vince Lombardi stat. Forget whether a player was unlucky or lucky, played well on Sunday or not. Ultimately, how many times did a player win?
The combination of these metrics allows us to examine how players have performed with a chance to win while acknowledging the nuance that luck and the performance of others, outside of a player’s control, can determine actual wins.
No one has outperformed his Sunday morning expectations by more over the last four years than Bryson DeChambeau, and it isn’t particularly close. Bryson has 4.4 more wins than expectation based on his starting position, and 2.9 more wins than expected given his actual play, first and second in the group.
The others at the top of the list are likely not a surprise to golf fans over the past five years. Justin Thomas and Jon Rahm have been consistent winners and Dustin Johnson has been the most consistent winner of the last decade. Johnson’s numbers here are staggering: he’s 4.3 wins ahead of JT for most expected wins, and despite a reputation for losing Sunday leads, he’s won more than expectations. This is a good reminder that winning a golf tournament is really hard, even if you are one of the best players in the world and enter the final round in the top 10.
The names at the bottom end of list are not a surprise, either. Rickie Fowler and Tony Finau’s lack of wins have become memes on golf twitter and likely a source of consternation to both players. Finau and Fowler haven’t just been unlucky, either; they both have underperformed their starting positions on Sundays.
This can be illustrated through a stat I call Win Equity, which is defined as Expected Wins given Actual Play minus Expected Wins entering Round 4. Generating positive Win Equity means that a player outperformed his expectations, ignoring what others shot. This stat best removes luck from the equation of final round play.
Finau’s performance in Win Equity especially noteworthy: in the 36 final rounds where Finau has entered in the top 10, he has added win equity through his play in only 3 of them (2017 Valero Texas Open, 2019 HSBC Champions and 2018 Safeway Open). That is by far the lowest rate amongst these players.
On the other end of the spectrum, Brooks Koepka, has both added the most Win Equity and been one of the most consistent positive Win Equity players over the last five years, which is in line with his reputation as a Sunday closer. He, Bryson and Justin Thomas had added win probability in over 35% of the final rounds where they enter in the top 10.
It is worth noting that Data Golf have found that leaders consistently underperform their expected Strokes Gained (and thus expected score) in final rounds. This is likely due to a combination of reversion to the mean (i.e., if you are leading, you likely outperformed your mean skill level for the first three rounds), tougher playing conditions on Sunday afternoons, and, of course, pressure. Thus outperforming 35% of the time as the top players do is really quite impressive.
While Finau is unlucky to have not won since his 2016 Puerto Rico Open victory, he clearly has not made his own luck through his Sunday play. Neither has Fowler, although his numbers are somewhat skewed by losing 0.6 Win Equity in his 2019 Phoenix Open win when the weather turned nasty in the desert and everyone played poorly. While the reasons for Finau and Fowler’s Sunday struggles are mysteries to us, their existence cannot be denied. It will be interesting to watch whether that changes in 2021 and beyond.