Golf Analytics

How Golfers Win

US Open 2014 Statistical Preview


This year’s US Open returns to a renovated Pinehurst #2, site of Michael Campbell’s shock win in 2005 and Payne Stewart’s final major victory in 1999. Pinehurst underwent significant renovations in 2011 to return it largely to the original Donald Ross designed conditions. Most importantly, the thick, bermuda rough of 1999/2005 has been removed, replaced by waste areas of sand, pine straw,  and wire grass, and the fairways have been widened by 50%. The course has also been lengthened to over 7500 yards. The extremely difficult greens weren’t modified; they remain small and harshly sloped.

The Course:

The 1999/2005 course was a traditional US Open set-up – thick rough/narrow fairways off the tee, hard to hit greens, and fast, firm conditions on the green. Traditional US Open set-ups are believed to favor accurate drivers, but my analysis of the last 10 US Opens doesn’t indicate a bias towards either distance or accuracy off the tee [1]. This is very important. Much discussion is made of the need to avoid the rough at a US Open, however avoiding the rough seems to confer no greater advantage at a US Open set-up than week-to-week on the PGA Tour. This year’s US Open course will have wider than normal fairways compared to PGA Tour courses and the same green conditions as before. The change is certainly the replacement of the thick rough with those sandy waste areas. While before any shot into the rough was a difficult approach because of the thickness of the rough, this year a shot off the fairway will be a random draw – sometimes you’ll end up with a good lie on the sand and other times you’ll be stuck behind a patch of wire grass or pine straw. In addition, the fairways will be firm and fast, waiting to usher slightly off-line drives to roll into the waste areas.

All in all, I don’t think this set-up favors distance over accuracy or vice-versa; in many ways it introduces more randomness into the tee to green game with the variability of lie if you miss the fairway. There are courses on Tour where wayward drivers are heavily penalized (Harbour Town) and others where length is the most important (Kapalua). US Opens, especially ones on long courses, fall into the middle. Hit it short to avoid trouble and you’re hitting a long approach shot into a fast and firm green. Hit it long and you’re at greater risk of landing in trouble. As with any US Open set-up, golfers who hit their irons precisely out of trouble will do well this week. On most courses week-to-week it pays to fire at most pins and generate birdie opportunities; most of the time at US Open set-ups you’re just aiming to hit the greens. As always, the very best golfers are in the best shape to win.

Course History:

A lot will be made of how Phil Mickelson is well set to contend and possibly win this week, completing his career Grand Slam. This is mostly wishcasting by the media. Phil did finish 2nd here in 1999, but he finished tied for 33rd in 2005, below what we would expect considering he was a top 5 golfer in the world that year. I’ll write more about Phil later in the week, but it’s important to note he is not the golfer he was in 1999 or 2005. Age has eroded his talent from top 5 in the world to more like top 15 in the world.

As with any course that has been used competitively three times in 15 years (1999/2005 US Open, 2008 US Amateur), a huge grain of salt must accompany any recitation of past performance. And once you factor in the substantial renovations that completely change the course off the tee, I don’t think it’s important at all how a player finished in 1999 or 2005.

The numbers back that up. There was no correlation in performance for the 48 golfers who played in both the 1999 and 2005 tournaments. That fails to control for a bunch of different factors (age, change in ability, etc.), but there was genuinely no correlation at all. In fact, the fact that those 48 golfers had previously competed at Pinehurst didn’t confer any extra advantage – they played 0.75 strokes better than the field on average in 1999 and 0.76 strokes better in 2005. Add in that this is a totally different course and I wouldn’t put any stock in prior results.

Course Statistics:

Par: 70, with two normal par 5s converted to par 4s. #3 will also play as a driveable par 4 for at least some of the rounds.

Length: using True Distance it will play +377 yards, longer than any normal PGA Tour set-up except Congressional (+455). Much longer than Merion (-197), Olympic Club (+52), or Pebble Beach (-210) and comparable to Bethpage (+308) among recent US Open venues.

Course Average (2005): 296.5, with a winning score of even par 280.


[1] – My analysis compared a golfer’s driving distance and % of drives that ended in locations besides the rough/fairway with their performance in the US Open, controlling for overall ability. Those drives in non-rough/fairway locations are catastrophic drives (water/trees/out of bounds/etc.), typically costing the player ~0.67 strokes. Only golfers with qualifying PGA Tour performances that season were included, yielding a sample of 823 tournaments over the past ten years. Both distance and % of catastrophic drives were only significant at the 0.10 level, while overall ability was significant at the 0.0001 level. Both distance and accuracy were roughly equally predictive of US Open performance, though overall ability is vastly more important (~25 times more important). This analysis explains ~40% of US Open performance.

A second analysis which included scrambling ability again revealed that the overwhelming factor in US Open performance is overall ability. The ability to scramble successfully, driving distance, and avoiding catastrophic drives were all small positive factors – of which avoiding catastrophic drives was the most valuable. Scrambling successfully is important because US Open courses often feature low GIR rates, meaning a golfer must hit 1-2 additional scrambling shots/round. In short, golfers who are good scramblers could pick up a very small advantage relative to normal.

3 responses to “US Open 2014 Statistical Preview

  1. Clayton June 9, 2014 at 10:00 PM

    I’m curious as to why you say that the 3/4 shot advantage by the 48 golfers who competed at Pinehurst before the 99/05 Opens wasn’t significant. Was this because of a high p – value or is there some other explanation you have. It would help to clarify because I would think, as im sure many readers do, that 3/4 shot would be significant at a par 70 course, but I don’t currently have access to that data.

  2. jalnichols June 9, 2014 at 10:20 PM

    The same group played the course 0.75 strokes better than the field in 1999 and 0.76 strokes better than the field in 2005. That is, their performance relative to the field didn’t change at all (it’s likely those 48 golfers on average ~0.75 strokes better than the field overall). This indicates they derived no advantage from having competed at the course six years prior.

  3. Pingback: The Most Overrated Golfers at the US Open | Golf Analytics

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: