top of page

The Analytics of Shinnecock

Updated: 20 hours ago

Explaining the Changes between 2018 and 2026, & What These Changes Can Tell Us About the Future of US Opens & USGA Setup Philosophy


Photo Credit: Andy Lack
Photo Credit: Andy Lack

Editor's Note: This article is being provided for free! Get this and so much more content when you subscribe today.





Let’s start big picture. Shinnecock played nearly two strokes easier than it did in 2018, but it still remained on the harder end of U.S. Opens over the past decade.


  1. Shinnecock 2018: +4.39

  2. Winged Foot 2020: +4.07

  3. Oakmont 2025: +3.82

  4. Pinehurst 2024: +2.78

  5. Shinnecock 2026: +2.63

  6. The Country Club 2022: +2.60

  7. Torrey Pines 2021: +2.27

  8. Los Angeles Country Club 2023: +1.76

  9. Pebble Beach 2019: +1.22

  10. Erin Hills 2018: +1.1


Now let’s look at the standard deviation of scores, which shows the variance of scoring. Shinnecock went from displaying the highest scoring variance of the last decade in 2018, to featuring around the same scoring variance of a PGA Tour event. This is a clear result of a more conservative setup, and we certainly saw much fewer “car crashes” (big numbers) this year, something that was previously a staple of prior U.S. Opens at Shinnecock.


  1. Shinnecock 2018: 3.18

  2. Oakmont 2025: 3.12

  3. Erin Hills 2017: 3.05

  4. Winged Foot 2020: 2.96

  5. Pinehurst 2024: 2.91

  6. Los Angeles Country Club 2023: 2.82

  7. Torrey Pines 2021: 2.80

  8. Shinnecock 2026: 2.77

  9. The Country Club 2023: 2.77

  10. Pebble Beach 2019: 2.69


Less scoring variance is not always bad, as there are plenty of golf courses that feature bunched up scores but still allow the best players to separate. Did Shinnecock do that in 2026? Let’s take a look at the skill separation slope (1.0 is Tour average. Anything above 1 indicates that the tournament separated skill at a higher level than Tour average.)


  1. Oakmont 2025: 1.52

  2. Winged Foot 2020: 1.52

  3. Los Angeles Country Club 2019: 1.19

  4. Pinehurst No. 2 2018: 1.18

  5. Shinnecock 2018: 1.15

  6. Pebble Beach 2019: 1.11

  7. Torrey Pines 2021: 1.07

  8. Erin Hills 2017: 1.03

  9. The Country Club 2022: 1.02

  10. Shinnecock 2026: 0.93


Interestingly enough, this U.S. Open featured the lowest skill separation slope of any of the last 10 U.S. Opens, and it’s the only U.S. Open in the last decade to separate skill at a below PGA Tour average level. My guess is that this had to do with a high variance of putting (40.6% of the variance decomposition was attributed to putting at this U.S. Open, compared to 34.1% at Shinnecock in 2018). Keep in mind, while a higher number generally indicates a better test of skill, there is still a lot of putting noise in a four-day tournament. For example, if many of the best players putt poorly in that given week, or simply don’t play well, the golf course is going to feature a low slope. 


Now let’s look at this wonderful table from DataGolf, which dives more into the differences between 2018 and 2026.



My Takeaways


  • Every skill at Shinnecock was easier in 2026 than it was in 2018, although putting and approach play largely stayed the same. The real change was that ARG play and OTT play became a lot easier in 2026. This can be attributed to two key differences from 2018: Softer greens and a lower missed fairway penalty. 

    • Softer greens made chipping easier. It allowed players to get up and down at a higher rate, and balls were not bounded 30 yards into the fescue the way they were in 2018. Also, players that were short-sided were less “dead" ARG, and actually had the ability to get their ball to stop closer to the hole. 

    • The missed fairway penalty penalty in 2018 was 0.50 -- which is extremely high. It went down to 0.41 this year, which is still very high, but it falls more in line with courses like Bay Hill and Sedgefield, and well behind courses like Muirfield Village (0.47), Royal Troon (0.48), LACC (0.51), Erin Hills (0.54) and Oakmont (0.60). 


Here’s how the missed fairway penalties have stacked up at the last 10 U.S. Opens, along with field adjusted driving accuracy, to really start to paint the picture of what were the toughest driving tests.


  1. Oakmont 2025: 0.60 (54.4%)

  2. Erin Hills 2017: 0.54 (74.5%)

  3. Los Angeles Country Club 2023: 0.51 (66.2%)

  4. Shinnecock 2018: 0.50 (71.5%)

  5. Pebble Beach 2019: 0.42 (64.9%)

  6. Shinnecock 2026: 0.41 (71.2%)

  7. Pinehurst 2024: 0.39 (66.9%)

  8. Winged Foot 2020: 0.30 (46.3%)

  9. Torrey Pines 2022: 0.30 (48.7%)

  10. The Country Club 2022: N/A (N/A)


Oakmont feels the toughest test of driving of all of these, combining the third-lowest driving accuracy with the hardest missed fairway penalty. Winged Foot and Torrey Pines both featured nearly impossible fairways to hit, but despite their reputation with thick rough, they featured much lower missed fairway penalty than the sandy soil, fescue-based courses like LACC, Erin Hills, and Shinnecock. This combination of wider fairways and an extremely high missed fairway penalty is certainly my preferred choice when testing driving skill. Let’s see if the numbers agree when we rank the hardest OTT U.S. Open courses (lower number indicates the harder that a shot plays)


  1. Winged Foot 2020: -0.069

  2. Oakmont 2025: -0.050

  3. Torrey Pines 2022: -0.003

  4. Shinnecock 2018: -0.009

  5. Pebble Beach 2019: -0.048

  6. Erin Hills 2017: +0.016

  7. Los Angeles Country Club 2023: +0.033

  8. Pinehurst 2024: +0.037

  9. Shinnecock 2026: +0.072

  10. The Country Club 2022: N/A


So in 2026, Shinnecock went from being the toughest version of “Wide Fairways, High Missed Fairway penalty” to the easiest. That is a pretty large departure from the driving test that we saw in 2018. Yet the even larger drop was with the rough penalty, which went from 0.56 in 2018 (again, historically high), all the way down to 0.27, which isn’t even in the top 10 hardest rough penalties this season. Not to victory lap, but we clocked this pretty early on site. Everyone was talking about the softness of the greens, and rightfully so, as that played a major impact on Shinnecock losing its bite ARG, but the bigger story all along may have been the significant decrease in rough penalty this year, which decreased by a whopping 51.8%.



So what are the trickle-down effects from the rough penalty decreasing by over 50%? You guessed it. Inaccurate players performed better. In 2018, Inaccurate players performed 0.20 strokes worse than their baseline. In 2026, they performed 0.13 strokes worse than their baseline. Honestly, that’s not a huge difference, yet perhaps another downstream effect was the advantage that accurate players lost with a lower missed fairway penalty. In 2018, accurate players performed 0.23 strokes better than their baseline, in 2026, they performed only 0.11 strokes better than their baseline. 


Another downstream effect of the lower missed fairway penalty and softer greens was that the field strength adjusted greens in regulation percentage saw a relative increase of 11%. This higher greens in regulation percentage seemed to allow putting variance to play a larger role in this tournament. The relative increase of putting variance was 19.1%, which again, plays a role in this being the only US Open in the last decade that separated skill at a rate lower than your average PGA Tour course. 


Adding all of this together, I’m frankly a little surprised this U.S. Open wasn’t criticized more in comparison to Aronimink, given that a lot of the frustration (high putting variance, low penalty for extremely inaccurate players) levied at Aronimink was also true of the 2026 version of Shinnecock. Part of that is probably the empathy fans may have had towards the USGA given the pickle they faced with setup this year based on prior mishaps at this venue. This is not an opinion piece, but I think most can understand why the USGA watered the golf course as much as they did. However, I’m not sure I have a good explanation for why the rough was so patchy this year and less penal. I had the pleasure of walking the golf course on Wednesday with the DataGolf brothers, and the patchiness of the rough was the first thing that immediately stood out to us. I was keen to see the data back up what our eyes were telling us. 


While it’s important to note that Shinnecock was still a very penal golf course off the tee compared to other PGA Tour courses, a higher rough penalty is not only a staple of US Opens, it’s a staple of Shinnecock, which has historically done an excellent job at neutralizing distance and significantly penalizing inaccurate players. That was less so the case in 2026. Not to give the USGA another pass, but there could have been myriad reasons why the rough was patchier and less penal this year, including lead-in weather patterns that could have possibly stunted fescue growth. I’m not an agronomist, but that would be my most rational hypothesis.



Despite the softness of the greens and lower missed fairway penalty being a rather stark departure from prior U.S. Opens at Shinnecock, Mike Whan and others at the USGA viewed this U.S. Open as a rousing success, which is in stark contrast with how they have spoken about 2018, a tournament they viewed as a massive failure. The success of a U.S. Open is up for them to judge, not me, but the data would strongly suggest that 2018 was statistically a better test of skill than 2026. Yet, there is a lot of noise in that small of a sample size. While I do believe that if the U.S. Open was played 100 times at the 2018 version vs. the 2026 version of Shinnecock, 2018 would be the premier test of skill, a major talking point of Mike Whan this week was that they wanted to “let Shinnecock be Shinnecock,” so perhaps the 2026 version of this golf course was a more honest portrait of the golf course. 


I’m not a member of Shinnecock, so it’s hard for me to say what version displayed more architectural integrity. I’d certainly be curious to speak to members about how their course plays on an extremely windy day in June, and whether the golf course receives this much water. I had the privilege of playing golf at another private club about an hour from Shinnecock on Thursday afternoon that is known for their pristine conditioning and extremely fast and challenging greens. In fact, it’s another Jimmy Dunne club! I can certainly report that the grounds crew was not watering and syringing the greens throughout the morning and before the afternoon tee times, and we all got our asses kicked! That golf course was insanely challenging under those conditions and felt borderline unfair (at least for a bad golfer like myself), but we all had a good laugh about it after. That’s just the rub of the green. Yet the biggest difference between my experience on a golf course in extreme winds and fast greens on Thursday is that I am not playing for millions of dollars. It’s fairly understandable (and predictable) why the USGA veered on the side of caution given that forecast and what was at stake with their championship. 


I would note however, that where the USGA ran into problems in 2018 was not actually just the firmness of the greens, it was the relationship between that firmness and the pin positions that they chose. Not only were the greens extremely firm, but the pin positions were on slopes that made it nearly impossible for the ball to lay to rest. This forced the USGA to water the golf course in between groups and compromise the integrity of the championship. Interestingly enough, however, this blunder didn’t have a major impact on how the golf course separated skill. The 2018 U.S. Open at Shinnecock was won by the No. 1 DataGolf player in the world and featured a higher skill separation slope than what we saw in 2026. The USGA took a much more conservative game plan into 2026 and spoke candidly about their pursuit of “fairness of tee times" and still barely avoided a huge wave edge on Thursday and Friday. Yet this remains the extreme impossibility of setting up Shinnecock. Damned if you do and damned if you don’t. I’d probably quibble with the concept that the USGA “let Shinnecock be Shinnecock” more in 2026 than in 2018, but there are pros and cons to each approach, and it would likely take a much larger dataset to understand the true impact of a highly conservative approach vs. letting the golf course approach the edge. 


The USGA made a massive effort to make this a more equitable test, but in doing so, they unequivocally diminished one of the key tenets of Shinnecock in short game skill. It seems like “fair” is a word that we will hear a lot going forward from the USGA, but the data doesn’t entirely support that a conservative approach will yield a more “fair" golf tournament (especially when we still haven’t found a way to accurately penalize large misses on golf courses that feature tons of foot traffic and spectators). Furthermore, it’s worth noting that the emphasis on “fair” is a clear departure from previous identities of this tournament, which embraced carnage and has been marketed as the most vexing, mental examination of the season. The point of this essay isn’t really to generate nostalgia from a bygone era and wax poetically about how we may be losing the spirit that this inherently unfair game was built upon, but rather examine the efficacy of the USGA’s decisions through data. 


I’m not really one to complain about the challenge of the golf course so long as good shots are being rewarded and poor shots are being penalized, but it’s challenging to make a statistically sound case that golf courses with thicker rough 5-10 yards off the fairway than 10-30 yards off the tee breed equanimity. The rough penalty decreasing by 50% is still unexplained to me, as a higher missed fairway penalty in 2018 had nothing to do with the problems the USGA encountered on the greens.  Shinnecock has been historically known as a supreme test of accuracy, and what was so wonderful about 2018 is that players were forced to play an old school style of strategic golf. It was entirely antithetical to the growing trend of players simply bombing away aimlessly on most PGA Tour courses. We barely saw Brooks hit driver all tournament in 2018, while Wyndham Clark took a far more driver-heavy approach and ranked 51st out of the 72 players who made the cut in driving accuracy. Penalizing misses has been a core tenet of US Opens, and fans were apoplectic about this lack of penalization at LACC, yet there was seldom a peep about it this year at Shinnecock, despite LACC statistically penalizing misses more than 2026 Shinnecock. 


As far as the greens go, while they certainly could allowed firmer surfaces with the pin positions that they chose (which would have presumably restored some of the challenge ARG we have grown accustomed to at Shinnecock), it’s hard to really fault their logic given the prism through which they’ve viewed prior Championships at Shinnecock. There are so many other setup concerns that the USGA has to worry about to ensure a successful championship, including pace of play, although I hope their goal is not to simply appease the players, as we see enough of that week to week on the PGA Tour. In conclusion, while you can probably see that I have some questions about the efficacy of a more conservative approach from the USGA, the 2026 version of Shinnecock still beats every other U.S. Open venue in the rotation for me. I enjoyed 2018 more as a viewer, chaos and all, but Shinnecock remains the premier Championship test in America. Even with softer greens and a diminished missed fairway penalty this year, the combination of a golf course with reliable wind, undulation, and strategic architecture that continues to do a solid job of neutralizing distance is hard to come by in modern professional golf, and I eagerly await its return in 2036. 

bottom of page