Golf Analytics

How Golfers Win

Predicting Putting Performance by Distance

Mark Broadie’s research of the Shot Link data established a clear relationship between putt distance and % of putts made. PGA Tour pros make a very high percentage of their close putts, but only about half of their putts around 10 feet and only around one in six around 20 feet. Pros hole very few (~5%) of their longest efforts from 25 feet and beyond. That data on % of putts made for each distance now forms the backbone of the PGA Tour’s Strokes Gained Putting statistic where players are credited and debited for making or missing every putt from every distance. Over a single season Strokes Gained Putting is often an unreliable indicator of putting performance, particularly at the extremes and also for players who have putted much worse or much better than in previous seasons.

Putting performance is polluted by randomness; Tour players just don’t attempt enough putts over the course of the season to get an accurate picture of their underlying putting ability. However, to make accurate projections of putting ability, you need to know whether Graeme McDowell’s 0.9 putts gained this season represents more talent or more luck. I’ve broken down putting performance into four different distance buckets from the PGA Tour data: putts inside 5 feet, 5-15 footers, 15-25 footers, and putts outside 25 feet. The results show that putting performance is far more predictable and consistent at the short distances. Long putting is so noisy that it’s difficult to say anyone gains much of an advantage from their long putting over the long-term.

Inside 5 Feet:

These putts are almost always converted (average 96%). The spread in performance between 2011-14 was 93% to 99%. The spread in expected performance derived from weighting the previous four seasons is 94.3% to 97.8%. This indicates that we should expect every regular Tour player’s true talent from inside 5 feet to fall somewhere inside that 3.5% range. Based on an average of over 900 putts attempted inside 5 feet over a season, we should expect every regular Tour player’s talent in terms of putts gained or lost to fall between +0.2/round and -0.3/round.

The graph below shows the correlation between a three year average (2011-13) and 2014 performance for all players with qualifying rounds in all four seasons. The correlation (R=0.56) between prior performance and 2014 performance is strongest in this distance range.

inside5feet

5-15 foot Putts:

This length is either short birdie putts or par putts after a scrambling shot that are converted approximately half the time. The spread in performance between 2011-14 was 36% to 54%. The spread in expected performance derived from weighting the previous four seasons is 40% to 52%. Based on around 450 putts attempted from 5-15 feet over a season, we should expect every regular Tour player’s talent in terms of putts gained or lost to fall between +0.4/round and and -0.5/round. Compare that to the best putters on Tour gaining about 0.75 putts/round.

The correlation between three year average and 2014 performance is below. The correlation (R=0.53) is similar to that for the short <5 foot putts.

5-15 footers

15-25 foot Putts:

These length are normally longer birdies putts and are converted about 16% of the time. The spread in performance between 2011-14 was 8% to 26%. The spread in expected performance derived from weighting the previous four seasons is 12% to 20%. Based on around 225 putts attempted from 15-25 feet over a season, we should expect every regular Tour player’s talent in terms of putts gained or lost to fall between +0.15/round and and -0.15/round. There’s much less at stake from this range than the previous two, just because so few putts are attempted from 15-25 feet.

The correlation between three year average and 2014 performance is below. There’s not much of a relationship (R=0.28), showing that putting performance from this range is much more affected by random chance over a full season than the shorter length putts.

15-25 footers

Putts outside 25 feet:

These length are the longest birdie putts, often really lag putts just to get it close for par. The spread in performance between 2011-14 was 2% to 13%. The spread in expected performance derived from weighting the previous four seasons is 4% to 9%. Based on around 300 putts attempted from beyond 25 feet over a season, we should expect every regular Tour player’s talent in terms of putts gained or lost to fall between +0.1/round and and -0.1/round. Again, there’s very little difference in expected performance from this distance. Even the very best long putter on Tour will gain little from these putts – over the long term.

The correlation between three year average and 2014 performance is below. There’s almost no relationship (R=0.10), which means it’s almost impossible to predict how well a player will putt on these long putts. The top ten long putters from 2011-13 average hitting 7.6% of their putts (versus 5.5% average). They only hit 6.7% of their putts in 2014 – a regression of almost 50% to the mean.

outside25ft

The Big Picture:

This graph shows performance in all four ranges. The longer putts show little relationship to future performance, while the shorter putts do show a more consistent relationship. This means that players who gained a lot of putts last season based off their longer putts will start making putts at a lower rate, while those who gained a lot of putts based on shorter putts are better bets to retain that putting ability.

bigpicture

Most Improved Putters from 5-15 feet in 2014:

1. Graeme McDowell

2. Charley Hoffman

3. Billy Horschel

4. Justin Leonard

5. Michael Thompson

These guys have a better chance of retaining their putting performance into 2015.

Most Improved Putters from > 25 feet in 2014:

1. Rory McIlroy

2. Y.E. Yang

3. David Toms

4. Brendan Steele

5. Brian Gay

These guys look likely to regress in terms of putting performance, especially McIlroy who performed to career average on all other putts, but hit 8% more of his long putts – gaining almost a third of a putt per round over his career average.

Advertisements

13 responses to “Predicting Putting Performance by Distance

  1. Jeremy Miller October 9, 2014 at 12:10 PM

    Is two-putting from say 25+ feet a repeatable skill after backing out the short putts? In other words even if making the long putt is pretty random, I wonder if good lag putts are measurable.

    • jalnichols October 9, 2014 at 1:04 PM

      Yeah it’s tough to say without digging into the shot by shot data, but in general the guys who avoid three putts are the guys who are good inside 5 feet. In fact, if you’re trying to predict ‘3 putt % on putts outside 25 feet’ you’re better off using last few years average ‘putting inside 5 feet’ rather than last few years average ‘3 putt % outside 25 feet’. I don’t want to say lag putting doesn’t matter, but it probably doesn’t play a major role in putting performance.

    • Dan Freeman November 17, 2014 at 2:47 AM

      Good lag putts are definitely measurable. That’s exactly what strokes gained stats do: they don’t just measure whether you make a putt or not, they measure exactly where a putt started and where a putt finished, and each of those locations is associated with a strokes-to-hole benchmark based upon 10’s of thousands of putts from the same distance, down to a thousandth of a stroke. For example, a putt from 49 feet takes 2.127 strokes to hole out. If you putt it to 4 feet, that 4 foot putt has a 1.147 strokes-to-hole benchmark. So knocking a 49 foot putt to 4 feet, a PGA Tour golfer actually loses .020 strokes to the field. The formula is (2,127) – (1.147) – 1 = -.020. That’s why strokes gained putting over a season is an extremely good measure of putting performance: it is not just keeping track of makes and misses.

  2. Dan Freeman November 13, 2014 at 1:01 AM

    “That data on % of putts made for each distance now forms the backbone of the PGA Tour’s Strokes Gained Putting statistic where players are credited and debited for making or missing every putt from every distance. Over a single season Strokes Gained Putting is often an unreliable indicator of putting performance, particularly at the extremes and also for players who have putted much worse or much better than in previous seasons.”

    This statement very clearly shows that you don’t understand how strokes gained works. Strokes gained putting DEFINITELY does not simply determine whether you made or missed a 25 foot putt. The strokes gained for that shot is determined by taking into account the distance-to-hole before the shot and the distance-to-hole after the shot. Each distance to hole on the green has a strokes-to-hole baseline number determined by millions of PGA Tour pro shots. So for a a 25 foot putt the strokes-to-hole # is 1.934. Lets say you leave the putt 8 feet from the hole. The strokes to hole # for 8 ft. is 1.515. Formula to determine strokes gained fore the putt is (1.934) – (1.515) -1= -.581. Strokes gained for each shot is determined by where the shot ends up, NOT WHETHER IT WAS MADE OR NOT.

    I am so sick of people who claim to know a lot about stats not understand the simple math behind strokes gained. It is WAY MORE RELIABLE as a measure of putting performance than you realize. Educate yourself before you start spouting nonsense.

  3. Dan Freeman November 13, 2014 at 1:12 AM

    Also, you don’t think 300 putts from 25+ feet is statistically significant? You don’t think you can get an accurate measure of a player’s putting skill after 300 putts from a given distance range? How many putts do you consider enough to be a large enough sampling to determine a player’s skill? Seriously, 300 putts from 25+ feet is not enough to indicate skill? Using strokes gained, every single one of those 300 putts is precisely compared to the PGA Tour average based on hundreds of thousands of shots in that range. NOT ON WHETHER THE SHOT WAS MADE OR missed, but by the precise distance the shot was left from the hole and the corresponding shots to hole down to a thousandth of a stroke.

    Wake up, dude. You need to look into a different line of work.

    • jalnichols November 13, 2014 at 11:46 AM

      Man, I know how it’s calculated; in fact, I calculate it myself for every tournament round from the raw data. The Strokes Gained stats are awesome. I’m using the official PGA Tour stats here because 1) everyone can access them from the PGA Tour’s site & re-do these studies easily and 2) because using strokes gained from a certain range doesn’t produce materially different results than using % of putts made from a certain range. They’re very highly correlated over the full-season of data that I’m using for these studies.

      Unfortunately, when you break putting results – as measured by strokes gained – down into single round, single tournament, 3-5 tournaments, half a season, etc., they tend to be unreliable predictors of how well a player will putt in the future – as measured by strokes gained. I’ve compared all the above samples of putting in previous posts and that is the one consistent result my research has found. What that means is that when predicting future putting performance (ie, saying this guy is the best putter on Tour or this player is a better putter now than a year ago) you have to put those smaller samples of putting performance in the context of what they’ve done previously. Feel free to show otherwise; I’m confident in the results I’ve found.

      You can run the numbers and instead focus on Strokes Gained on Putts outside 25 feet rather than % of putts made outside 25 feet, but the results aren’t materially different. The guys who tend to miss more >25 foot putts also tend to leave more >25 foot putts outside the gimme range. In short, making more >25 footers is strongly correlated with avoiding 3 putts from that distance. I used the numbers posted by PGA Tour in case anyone was interested in replicating the research because the raw strokes gained stats aren’t publicly available.

  4. Dan Freeman November 13, 2014 at 7:11 PM

    That was cool of you to get back to me so quickly. I wasn’t exactly nice in my first comment.

    I still have major issues with your analysis:

    1) “Putting performance is polluted by randomness; Tour players just don’t attempt enough putts over the course of the season to get an accurate picture of their underlying putting ability.”

    This just makes no sense. You go on to say a player takes an average of 900 putts <5 feet, 450 putts between 5 and 15 feet, 225 putts between 15 and 25 feet, and 300 putts over 25 feet. That's 1875 putts over the course of a season. Your saying that's not enough to determine putting skill or "talent" for the year? I'd say you are wrong. Strokes gained takes every one of those 1875 putts and compares it to the field average, taking into account the specific starting distance and finishing distance. Making a few more or less putts over 25 feet than the field over the course of a season is not very significant in the context of 1875 putts. Strokes gained results after 1875 putts contain close to zero "luck".

    2) "Over a single season Strokes Gained Putting is often an unreliable indicator of putting performance,…"

    This is simply not true. Strokes gained putting over the course of a season measures putting performance almost perfectly. In fact, it measures performance during a round perfectly as well: you can't get much better than measuring your performance in relation to the field down to the thousandth of a stroke. There is no better way to measure putting performance, and we will never have a better way to measure putting performance.

    You seem to get "measuring putting performance" confused with "predicting future performance": two completely different tasks The problem isn't with the accuracy of strokes gained stats, the problem is that HUMAN BEINGS DO NOT PERFORM CONSISTENTLY. You are looking for a consistent measure of underlying skill that will consistently predict the future. Strokes gained measures skill perfectly for a single putt, a round, a tournament, 3-5 tournaments, a season, etc. The problem is that golfers are not that predictable. Putting is a little bit predictable, but there is a lot of variation. But that doesn't mean we can't accurately measure putting skill!!!

    3) Predicting performance off the green is not very easy either, but performance can be very accurately measured. Again, the problem isn't with the statistics but with the variability in golfer performance. I just went back into the PGA Tour stats and looked at strokes gained putting and strokes gained tee to green, following 20 players back to 2004. In both putting and tee to green, there is significant consistency from year to year. Each player generally falls within a range of performance. But there are years that are outliers, there are mini-trends, and there is a lot of randomness. Again, this is not due to strokes gained stats being unable to accurately measure performance, but rather the result of the players themselves being somewhat inconsistent.

    4) The challenge you are taking on is admittedly difficult: predicting performance based on past performance and "underlying skill", while separating out factors like "luck". I just think you are making a mistake every time you move away from strokes gained statistics. As you know, they will dominate the future of golf stats and make most other statistics obsolete. Total putts is a complete waste of everyone's time. But so is "greens in regulation". Strokes gained "greens in regulation" will overtake the original by taking into account WHERE YOUR APPROACH SHOT ENDS UP. By analyzing strokes gained after the second shot on a par 4, the first shot on a par 3, or the third shot on a par 5, the stat is greatly improved. Strokes gained scrambling, strokes gained sand, strokes gained rough, strokes gained for specific distances/lies will eventually take over all other golf stats. It will be able to quantify factors like the difficulty of the rough on specific courses and holes, temperature, wind, precipitation. It will even be able to help people get fitted for clubs.

    5) If I'm trying to predict future putting performance, I'm looking strictly at strokes gained putting. This is just as available as any other PGA Tour stat. You can even get strokes gained results for each tournament for every player, by mathematically separating the most current week/rounds from what was posted before. Yes, if a player has gained 4.63 strokes over a 4 day tournament, then their performance will surely "come back to the mean" the following week. But if I were you, I would determine what the "mean" is FOR EACH PLAYER. You could look at their last year's performance or their career performance, but there is no point in saying that Luke Donald or Greg Chalmers are going to come back to the field average in putting: they will come back to their own mean which is generally over half a stroke per round better than the field average no matter how you decide to determine it. Likewise, Boo Weekly is unlikely to come back to average either: he's going to lose at least a stroke per round.

    Cool. I feel better now. Thanks for listening.

  5. jalnichols November 13, 2014 at 10:28 PM

    Yeah I think strokes gained does a damned good job of measuring what happens on the course. In that sentence you quoted I should’ve said it’s an unreliable indicator of future performance over samples smaller than a year. That’s what a half-dozen studies have shown across the board.

    For #1, a season’s worth of putts is not enough to measure their underlying talent. Performance =/= talent. Again, it’s the best way to measure their observed performance, but a season of putting doesn’t tell me very much about how “good” a player is at putting or how well they will putt the next season (https://golfanalytics.wordpress.com/2014/03/27/repeatability-of-golf-performance-by-shot-type/). With one season, you would have to regress everyone almost halfway to PGA Tour average to predict their following season performance.

    For #5, totally agree. When I need an estimate for putting true-talent for anything, I use as much historic SGP data as I can for each player (up to ~4 seasons if available).

  6. Dan Freeman November 17, 2014 at 2:33 AM

    Again, thanks for your response. I respect what you are trying to do, and it’s fun to have a chance to discuss strokes gained with someone (there aren’t that many people who understand how it works at all!).

    I think I have honed in on my issue, and I say this respectfully: I think when you are trying to find the “underlying talent” of a player to predict future results, you are clearly chasing a ghost. This “underlying talent” is simply not as stable as you are trying to make it. I think you are looking for a statistical foundation that is not there. Ask the players: they will tell you that round to round, tournament to tournament, week to week, season to season, etc. that their “underlying talent” is constantly changing/evolving.

    But don’t take my word for it or even the word of the players. Instead, go take a good hard look at the strokes gained statistics back to 2004. Over the course of ten years, MOST players change from year to year fairly significantly, and yet there is definitely a loose order there as well. The stats clearly show that the “underlying talent” of players changes over time. Taking a ten year average is not going to show you a solid statistical foundation. THIS IS NOT BECAUSE THEY DON”T TAKE ENOUGH PUTTS IN A YEAR OR BECAUSE THE STATISTICS FAIL TO ACCURATELY MEASURE PERFORMANCE. IT IS BECAUSE THE SKILL OF PLAYERS CHANGES. THIS IS ALSO TRUE FROM TEE TO GREEN.

    Take a look at Adam Scott’s strokes gained putting numbers from 2004 to present. His progression/regression is dramatic. In 2004, Scott’s strokes gained for the season was .880. THE NEXT YEAR, his putting changed to -.016. HE then stayed negative for 7 of the next 8 years, dropping as low as -.888 and -.746 in 2009-2010. The last two years, his numbers have gone positive again. How are we supposed to go about finding the “underlying talent” of Adam Scott for the last 10 years? He has been the best putter some years and nearly the worst putter some years, and mostly he has been relatively average. But there is some “order” to the stats as well: clearly he had it, lost it, and seems to be regaining it. So it’s not random, but it is CONSTANTLY CHANGING.

    His results are dramatic. Some other players are much more consistent.

    I did a quick study manually: I tracked about 20 players back to 2004, and asked myself this simple question: what is the best predictor of next season’s strokes gained putting results? Of course, this is not looking round to round, but season to season. I decided to look at 2013 (already calculated), and tried to determine what was the best thing to look at if you were trying to predict that season before it started. I considered 1)results from 2012 2)average results from previous 2 years 3)average results from previous 3 years and 4) average results since 2004 and 5)the PGA Tour average 6) previous season of the player taken halfway to the field mean.

    The results were quite variable, with no method proving to be a great predictor. But what was the BEST predictor? It turns out the previous season overall was the best predictor. Interestingly, the WORST predictor was the field average, and the second worst was the player’s average for all recorded seasons. In other words, looking at a players all-time average as an indicator of “underlying talent” gives very poor results. You are better off looking at the previous season alone.

    I did this all manually, and it was a pain in the ass. I have a challenge for you since you seem to have the resources: looking at strokes gained stats only, what is the best predictor of the following year’s strokes gained putting results? You can use each of the last five years as your tests: going into each of those years, what is the best way to predict the following year’s results? I think the answer might be something like this: average the figure from the previous year with the player’s all-time average.

    I think if we take all judgement out and put it into the hands of science, we’ll find that nothing is a great predictor, and furthermore that the idea of “underlying talent” is not helpful.

    Then, I challenge you to do the same thing from tournament to tournament: for a given tournament, what is the best predictor of strokes gained putting results for that tournament? All of this is testable, just a little tedious. What’s cool about it though is you can take subjectivity out of it. In other words, we may not find the perfect predictor but something will clearly show itself as the best.

    Just remember: if you don’t have the numbers to back up “underlying talent”, then the idea is as thin as air.

  7. Pingback: Don’t Trust a Hot Putter | Golf Analytics

  8. Pingback: Quick Masters Thoughts | Golf Analytics

  9. Frank November 6, 2015 at 4:01 PM

    Hello. Do you have specific stats of % made from every feet for the PGA Tour?
    thanks

    • jalnichols November 7, 2015 at 11:16 AM

      0 100%
      1 100%
      2 99%
      3 96%
      4 87%
      5 75%
      6 66%
      7 57%
      8 50%
      9 44%
      10 38%
      11 34%
      12 31%
      13 28%
      14 25%
      15 23%
      16 21%
      17 19%
      18 17%
      19 16%
      20 14%
      21 13%
      22 13%
      23 12%
      24 10%
      25 10%

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s

%d bloggers like this: