Golf Analytics

How Golfers Win

Monthly Archives: March 2015

What’s Wrong With Justin Rose?

I’ve rated Justin Rose as a top five golfer in the world since 2013, but his awful start to the season has dropped him to 11th in my ratings – the lowest level since his re-emergence as a star in 2012. Rose has missed the cut in three of his four PGA Tour starts in 2015, only staying to play the weekend at the no-cut WGC event at Doral where he was a non-factor and finished 55th. He’s always relied on a combination of great iron play and one of the pre-eminent short games in the world to contend, hitting it close to generate birdies and scrambling really well when he missed the green. Putting has always been circumspect (normally at or below PGA Tour average in Strokes Gained Putting), but the rest of his game has been so good it hasn’t mattered.

Rose PGA Tour stats

Well, this season it hasn’t worked. Rose would rank 178th on the Tour in scoring average if he had enough rounds to qualify, and instead of the top five Tee to Green game he’s had the past three years, he’s fallen to PGA Tour average. Looking at his stats, it’s easy to identify the main culprit in his collapse this year. His Strokes Gained Putting is just completely awful; he’s losing over a stroke per round with the putter – similar to what you would expect from Boo Weekley or Lucas Glover. His misses have mainly been focused on the short putts (4 to 12 feet), so that’s where he should be making adjustments. I’ve made the case the putting is subject to extreme fluctuations in small samples (Rose has only 9 rounds tracked this season), so it’s likely that his putting will rebound to his career levels in the coming tournaments. That’s not a significant concern in my mind.

What is concerning is the state of the rest of his game. I’ve charted his results across a wide range of shot types between 2014 and 2015 below. Adj. Dist measures his driving distance on all holes relative to the field. Rel. Accuracy measures how much more accurate he’s hitting his drives than the field; +20% is Jim Furyk level elite and -20% is Andrew Loupe level wild. Other Drives measures the number of par 4/5 tee shots that end up in water/out-of-bounds/trees/desert/etc. Hit Greens measures how many more greens he’s hitting relative to the average Tour pro on his par 4 2nd shots. The rest are measuring strokes gained relative to the field on each shot type.

Rose SG Stats

Rose’s decline has been spread across a number of shots. His SG on drives has declined, his SG on short game shots has fallen from elite to awful, and he’s lost a ton of ground on par 4 approach shots. In other words, his strengths last year (short game and par 4 approach shots) have turned into weaknesses. Now is the time to state again that this is only 9 rounds. In fact, right before this stretch of misery, Rose played back-to-back tournaments in the Middle East and finished T12 and T13. However, this is a guy who has missed six cuts in three years; this is as close to panic as it’s likely to get for Rose.

One interesting thing that pops out of his numbers is that Rose is hitting more greens with his 2nd shots than last year, but he’s not hitting anything close to the hole to generate birdie looks. That’s not good with the Masters looming; at Augusta simply bailing out and hitting the greens won’t be enough. If Rose isn’t able to hit his spots with his irons, he’ll be facing a bunch of long, slippery chances and likely racking up the three putts. That’s just the way it is on those huge, undulating greens. I’m very interested to see how he’s playing in Houston this week after two weeks off to reassess his game. Ignore his putting results and how many greens he’s hitting; focus instead on how often he’s able to get it inside 20 feet and set-up birdie.

Valero Texas Open Preview – 2015

The Tour shifted to TPC San Antonio in 2010, hoping a more modern track would be a more competitive venue than the previous birdie-fest at the La Cantera resort course. Unfortunately, the course is widely panned among Tour pros – who consider it too long and difficult. They are certainly correct that it’s difficult (73.5 stroke average since 2011 on a par of 72); pros have only hit 55% of fairways and 56% of greens here since 2011. TPC San Antonio also has a lot of what the Tour calls “native area” (brush, desert, etc. – pretty much where Kevin Na made his 16 from in 2011) in play. Last year, it was among the leaders in terms of courses where players ended up in the native area off the tee.

Course fit:
I wrote in last year’s preview that the course favored the longer/inaccurate hitters over the more accurate/shorter hitters. Concentrating on everyone who had played the course from 2011-2013, golfers who hit for more distance during the tournament played better than those who hit more fairways. Looking at 2014, Steven Bowditch and Andrew Loupe both had their best finishes on the season – they’re the archetype of the very long and very inaccurate hitter on Tour. That’s not to say every long/inaccurate hitter will play well, but when I broke the field up into seven different groups based on their accuracy and distance off the tee and compared just their tee to green play from 2014, the longest/least accurate group played the best last season – even better than the group of similarly long, but more accurate pros.

Performance at Texas 2014

The important question is why the course doesn’t punish inaccuracy to the level of a normal PGA Tour course. Most importantly, the rough here just isn’t very difficult to play out of. I only have shot-by-shot data for the final round in 2014, but the rough played the easiest of any course on Tour in that round (relative to the difficulty of fairway shots). What this means is that when I compare the results of shots hit from the rough to shots hit from the fairway on the same hole, shots hit from the rough only resulted in scoring 0.15 strokes worse than shots from the fairway. Normally that number is around 0.30 strokes. Sample size issues are a concern with only one round of data, but this measure tends to be consistent across the four rounds of an event. This is certainly an advantage for guys like Bowditch or Loupe who play from the rough more often.

Distance is critical also because the par 5s are so long at TPC San Antonio, only the longest hitters have a chance to go at them in two shots. In last year’s final round, pros who normally hit their drives over 295 yards went for the green in two on 46% of their opportunities; pros who normally hit their drives under 280 yards went for the green in two on only 2% of their opportunities. Now, long hitters typically have a large advantage in going for the green chances, but normally more like 65% to 35%. Here, short hitters are basically forced into lay-ups by the length – regardless of how they would like to play the hole. That turns the par 5s (already extremely difficult) into par holes for anyone who’s not long off the tee.

Masters Invite Watch:
The focus this week has to be on the bubble boys for Masters qualification. The top 50 in the OWGR after this weekend’s events earn invites to the Masters. Right now, Paul Casey is in the most precarious position – unqualified and not in this week’s event, but sitting only 49th in the world right now. Of those out of the field Marc Warren chose to enter this week’s PGA event rather than the European Tour event in Morocco. I have him projected for between 2 to 2.5 OWGR points in either location, which is basically too close to call on his chances to get in the field.

Harris English has a good argument to be the best guy not invited to the Masters; he’s got an outside shot with a top ten in San Antonio. My rooting interests are the young guys playing well so far this year. Augusta will be better with one (or both) of Justin Thomas or Daniel Berger in the field.

Below is the full break-down of what the guys out of the field need to do to get in this week. If you’re at all interested in tracking this/anything about the Official World Golf Rankings, follow @VC606.

Best Course History:
These are the guys who have played best here relative to their typical performances. In other words, for each year they’ve played I’m comparing their TPC San Antonio performance to their average performance for the year (minimum 2 starts here since 2010).

1. Charley Hoffman
2. Fredrik Jacobson
3. Martin Flores
4. Martin Laird
5. Ben Curtis
6. Pat Perez
7. Brendan Steele
8. Brian Harman
9. Daniel Summerhays
10. Cameron Tringale

View the full field course history at this Google Doc.

Why Rickie Fowler Doesn’t Win More

Amid Rickie Fowler’s torrid summer last year – which included top five finishes in each major championship – the collective golf media was focused on Fowler’s need to win tournaments to validate his status as one of the big stars on Tour. Bring up Rickie’s name among golf fans now, and the question immediately turns to why he doesn’t win more often. These sorts of questions are nothing new. Phil Mickelson was hounded about his inability to win a major for years before he broke through at the Masters. Up to a week before McIlroy’s Open Championship win last summer there were questions about his ability to close out tournaments when he jumped into the lead. And just last week, Jordan Spieth captured his second PGA Tour win – hopefully distracting the critics who think he needs to win more often for a few months.

Read more at No Laying Up

Tampa Bay Preview – 2015

Innisbrook’s Copperhead course is a tree-lined track just inland from the Gulf. It offers a distribution of 4 par 5/5 par 3/9 par 4s for a par of 71. It had played close to par until 2013; the scoring average has ballooned over 72 for the past two tournaments. The course offers the potential to play out to almost 7350 yards from the tips – long for a par 71 with 5 par 3s. This is another three wood heavy track – despite a number of downhill tee-shots, pros only hit it 272 yards off the tee in last year’s final round. Most notably, about 7% of tee shots last year ended up in what the Tour defines as Tree Outline – in the trees off the fairway. This course is regularly near the top of the list in terms of courses with drives ending up somewhere other the the fairway, rough, or bunkers.

This is often talked about as a ball-strikers haven, and it is. There just aren’t many wedges into these greens. Pros who are comfortable hitting the mid to long irons will be most successful here – not only because of some of the brutally long par 4s, but because four of the five par 3s can play over 200 yards.

What I’m Watching:
I touched on a number of the most improved players on Tour for 2015 yesterday. One I didn’t discuss in Ryan Palmer. Palmer’s really upped his game at the end of last year and beginning of this year by dramatically improving his results on approach shots (+1.1 strokes gained/round better in 2015). He’s seen serious regression in his performance off the tee however (-0.5 strokes/round worse) – driven by a loss of distance and a major decrease in accuracy. I wrote in my most improved piece that my research has shown that tee to green improvements tend to be retained much more than putting improvements. So for Palmer, his tee to green play has still improved considerably (+0.6 strokes/round better).

However, as I’ve expanded my shot-by-shot database it turns out that when I sub-divide the tee to green game into performance on drives, approach shot performance, and short game performance it turns out that performance on drives is the most stable indicator of performance. In other words, golfers who improve or decline on their drives tend to retain almost all of that improvement or decline, while golfers who improve or decline on approach shot performance or short game performance retain less of those improvements or declines.If that’s true, Ryan Palmer may be in for a decline soon as his approach shot performance erodes and his struggles off the tee continue.

When I was looking through the data for yesterday’s piece, Brendon Todd’s name came up as a guy who has dramatically improved his tee to green game in the first two months of the season. However, he wasn’t anywhere close to the top of the list of most improved. It turns out the culprit there has been his normally outstanding putting. Todd was 6th best on Tour last year at +0.66 strokes/round, following up a great season in 2012 and what looks like an outstanding putting season on the Tour in 2013.

In 19 rounds to start 2015 he’s been basically average. Todd has historically putted about a third of a stroke better on bermuda greens, so he could be in for a natural rebound in Florida. If he can go back to putting as well as in the past three seasons, he could emerge as a very good all-around player, instead of someone who relies on their putting/short game to carry them.

Best Course History:
These are the guys who have played best here relative to their typical performances. In other words, for each year they’ve played I’m comparing their Innisbrook performance to their average performance for the year (minimum 3 starts here since 2008).

1. Sang-moon Bae
2. Justin Leonard
3. Gary Woodland
4. Luke Donald
5. Retief Goosen
6. Chez Reavie
7. Jason Dufner
8. John Senden
9. Jim Furyk
10. Jonathan Byrd

The value of my approach is best illustrated by Justin Leonard. Leonard has been a mediocre player for years now, but has consistently raised his game at Innisbrook. A good finish for him normally is just making the cut, but since 2010 he’s 5/5 in making the cut with a T4 and two T20s. Innisbrook has provided his best, 4th best, 5th best, and 4th best finishes in 2013, 2012, 2011, and 2010.

Most Improved in 2015

This is about the time of the season when small-sample issues start to wash away and genuine performances changes can be detected. Last year around this time I wrote these pieces which attempted to tease out 1. how much weight to place on performance in the first two months of the season, 2. whether age was a meaningful factor to answering #1, and 3. whether it mattered if the change in performance was occurring in the tee to green game, with the putter, or both.

I found that past performance should be weighted about 3.5 times more than performance in the first two months. Also, younger golfers who show a lot of improvement tend to retain that improvement more than middle-aged or older golfers. It also turns out that golfers who have improved their putting tend to play worse going forward than those who improved their tee to green play.

What about 2015:
The ten most improved PGA Tour players in 2015 [1] are listed below:

most improved 2015

Coming into 2015, I had Daniel Berger rated as a below-average Tour player – largely because he hadn’t particularly stood-out in his 2014 season on the Tour (he performed slightly below what you’d expect an average PGA Tour player to do on that tour). Berger clearly had potential – he had finished as Golfweek/Sagarin’s 7th best NCAA golfer in his final season in 2012-13 – but he hadn’t clearly emerged as a future star like the similarly aged Justin Thomas or Jordan Spieth. Well, all that is moot after Berger’s start on Tour; 5/6 made cuts, performance similar to what you’d expect from a top 10 player in the world, and a playoff defeat in his twelfth PGA Tour event.

Applying the criteria from above – young players and those with improved tee to green play retain more of their improvement – Berger grades out well. Not only is Berger only turning 22 in April, but he’s dramatically improved his long game play in his 2015 Shot Link rounds compared to his 2014 Fall Series Shot Link rounds. He was carried a bit in the fall by a hot putter (+0.9 strokes gained/round), but this season’s success has been entirely driven by his play with his irons/driver (+2.5 strokes gained/round). I doubt Berger will continue to play at a top ten in the world level, but he already looks like a clear future star.

James Hahn has already won in his break-out 2015 season. Coming into 2015 he didn’t project as particularly likely to remain on Tour. In fact, I had Hahn rated as 205th out of all players with any PGA Tour status entering the season. Hahn has improved across the board in all key stats: he’s improved his iron shots by +0.5 strokes/round, his drives by +0.2 strokes/round, his putting by +0.5 strokes/round, and his short game by +0.5 strokes/round. I’m still not totally sold on Hahn as he’s has seasons worth of play at below Tour average, but he’s certainly moving in the right direction.

It seems like a long time since Brendan Steele’s rookie year win at the Texas Open in 2011. That win earned Steele notoriety and regular place in the Mickelson practice round matches. Designed in part to prepare guys for the rigors of Ryder Cup play, Steele watched Keegan Bradley and Rickie Fowler earn spots on those teams, while he suffered through a string of mediocre (at least in terms of performance) seasons. However, last year was his best yet on Tour and he’s kept the momentum going so far in 2015 – 5/5 in cuts and a 2nd at the Humana. Steele isn’t that young, but has improved across the board, especially in the long game (+0.5 strokes/round over last year and he’s hitting his drives 5 yards further).

Two other notables are Lucas Glover and Boo Weekley. Both have long track records of being awful at putting, but both have improved their putting numbers a ton in the past few months (Glover is now merely one of the worst rather than the absolute worst of all time). Neither have made many strides in the rest of their games, so I seriously doubt whether Boo Weekley especially can continue to play at this level. He has a decade of awful putting in his past and it strains credulity that he has all of a sudden become average.

Quickly running through the others, Wheatcroft and Laird are hitting their irons much better this year, Knost is relying on putting+short game improvements, and Collins is also riding a great short game. Zac Blair stands out as a young guy who should have no trouble keeping his card as a rookie.

Improvements among the stars:
Of golfers who entered the year in my top 50, the ten most improved are below:

MIP 2015 stars

Of these Shane Lowry sticks out. Lowry has played great in three US stops so far in 2015, earning enough FexEx points where he would be inside the top 100 in only four events. In my ratings he’s climbed all the way to 33rd after spending last year hovering around 100th. Lowry is only 28 and his success this year has been fueled by great long game (approach shots+drives), so I like his chances to remain around his current ranking. Lowry also has a very good shot at earning his Tour card for next year. He’s already at 222 FedEx Cup points – only 216 short of last year’s 125th finisher – and is pretty much guaranteed entry into at least seven more events plus any regular events he qualifies for or earns sponsors invites into.

My numbers are also coming around slowly on Patrick Reed. Reed’s career to this point has been marked by turmoil – four wins but also a lot of MCs and really inconsistent overall play – but he hasn’t missed a cut since July and his play this year has been driven by increases in his driving distance (+5 yards) and better approach shot play (+0.4 strokes/round). I’m not sold on Reed as a top ten player, but he’s definitely better than my numbers thought two months ago.

[1] – For this, first I calculated their performance in terms of strokes better than the field per round and then I compared that to their projection from the first of week of January 2015. I realize this is a bit of a black-box, but basically I try to find who is playing much better than my system expected going into the season. I’ve included only golfers with at least eight rounds played in 2015.

Golfer Statistical Comparables

I often find it useful when I’m watching a round or previewing a tournament to look at which types of golfers are doing well at a course, or more generally which types of golfers are similar to each other in terms of the results they are achieving. Commentators will normally focus on similarities between different golfers’s swings or their ball-flights, but I’m more interested in the results of their shots. I’ve generated some simple statistical comparables to use in comparing which golfers are most similar or most unique.

Methodology was fairly basic; I took every player who had at least 20 Shot Link rounds in my database and I standardized all of their results in six categories – driving distance, driving accuracy, greens hit above/below expected [1], strokes gained on approach shots/tee shots, strokes gained on the greens, and strokes gained with the short game. All of these numbers are adjusted for the field and include the results for almost 200 pros from about April 2014-present. I then found the “distance” each golfer was away from the others in each category and ranked each golfer from most to least similar to each other.

I’ve linked a Google doc with a top twenty and bottom five for each golfer here.

Robert Streb and Bill Haas were the most similar to the largest number of golfers, appearing in the top five most similar for 11 of 192 pros. Jim Furyk was easily the most unique, appearing in the bottom five least similar for 78 of 192 pros! Furyk is obviously an extreme outlier in fairways hit, greens hit, and approach shot SG.

As for the most similar players, some were fairly obvious. Bubba Watson’s top comp was Rory McIlroy, which makes sense as they are both very long, average in fairways hit, and don’t really move the needle much in terms of putting/short game. Zach Johnson is matched up with good ball-strikers who are short/accurate off the tee (Ryan Moore, Tim Clark, etc.). Dustin Johnson matches up with Rory, Adam Scott, and Jimmy Walker as guys with elite long games (drives+approach shots).

The value in this exercise comes in the surprising comparables though. Daniel Berger, contender last week at the Honda, has Horschel, Keegan, Paul Casey, Watney, and DeLaet as his top five, which shows just how fantastic his debut on Tour has been. Another rookie Carlos Ortiz shows up as similar to Russell Henley, showing he’ll make his money by driving the ball well and holing putts. Brendan Steele’s top matches just show the kind of potential his game holds: Jimmy Walker & Ryan Palmer.

My favorite comp is easily Bill Haas being the #1 match with the guy who weaseled his way onto the Ryder Cup team in his place, Webb Simpson.

These can be taken in a bunch of different directions, but I’m hoping to present some expanded comparables in the future. These will take into account actual predictive factors that can be used to judge course fit: ability to play from the rough, putting ability on different green surfaces, likelihood to aim for the center of the green vs. hunt for pins, etc. Today is just to show some basic similiarities between golfers. Again, the list of comps is linked here.

[1] This basically asks: from 156 yards to the pin in the rough, 189 yards to the pin from the fairway, etc. how often does the average PGA Tour golfer hit the green. So, (each pros actual GIR) – (expected GIR based on all their approach shot distances/lies).