Length Realllllly Matters

[Jason Topp]

I'd bet there is less "bomb and gauge" on the courses with tougher greens...

It would be interesting to look at that issue.  My course is 6600 yards with very tough greens and pretty tight fairways.  Two guys I play with hit it around 290 off the tee but have 0-3 handicap short games rather than pro or even good state level amateur short games.  They lay up a lot.

We host the state match play this year.  It will be interesting to see how guys attack the course.

Can you think of Tour courses that would be decent tests of your theory?  Pebble Beach? 


We are holding the state match play this year. 

[George Pazin]

Jason, I believe the real reason bomb and gouge works is simply that occasionally the bombers will have a week where they hit 70& of the fairways and make a bunch of putts. One great week and a few good weeks will keep you on Tour for awhile, as long as you can stand it mentally (the off weeks, that is).

George:

Woudln't that have been true in the 70's as well? 

In the past, you paid dearly for a mishit with a wooden wood. Not so with today's clubs, so there is no reason to play a controlled game.

[Pat Burke]

If you want shorter, accurate hitters to have a better chance of competing, all you have to do is play SHORTER courses.  Anybody can see those courses produce a different set of winners, and that Tiger is not as dominant in such conditions.

On the other hand, since Tiger is clearly the best player, why would you want to prevent him from winning?

The USGA seems more interested in protecting par than in finding a worthy winner.  They've altogether forgotten about keeping the game fun.  They seem to agree with the golf writers who are shills for the equipment companies on that score ... that better equipment makes the game more fun for average golfers ... forgetting that the same equipment is also promoting difficult architecture, and the USGA's role as the most visible champion of making courses tougher.

I agree Tom.  It always seemed to me that when you have a good golf course and let the guys play, you'll get great champions.  Ben Hogan at Riviera, gave way to Nicklaus then to Tiger in the US Open scoring records
The low 72 hole totals in The Open Championship show names like Tiger, Faldo, Norman, Watson.
Most times, a tricked up "par saving" effort gives a winner that is respected, but not The Man at that time

[Brent Hutto]

The made-putts-by-distance numbers from ShotLink in that range are quite significantly higher than the same numbers gathered by Dave Pelz 15-20 years ago. I trust the ShotLink number completely and the Pelz numbers unless someone has evidence his methods were flawed. I don't have the comparison at hand but they have come close to doubling their percentage made in the 6, 8, 10 foot range over just a couple decades.

Brett - when I looked at that, I do not agree.  Shotlink gives percentages in ranges, say less than 3 feet, between  5' and 10' etc. 

It is impossible, however, to convert that to the Pelz measurements because you do not know how putt lengths are distibuted within that range.  For example the stats showing a player makes 99% of putts of 3 feet or less is misleading because the vast majority of those putts will be tap ins.  If 99% of the putts are tap ins and 1% are three footers, the statistic would show the guy was a lousy putter, not a good putter.

Shotlink stats generally show players making more than 50% of putts between 5 and 10 feet.  If players have a lot more 5 footers than 10 footers, Pelz 6 foot 50/50 range may still be accurate. 

I suspect the Pelz number has changed but do not have enough information to be sure by how much.  12-18" was my best guess, but I cannot recall how I came up with it.

Jason,

I wasn't thinking of the crapola "stats" that the media publish off the ShotLink data. Someone wrote an article or maybe it was on a blog or somewhere about 2-3 years ago lining up the actual percentages made in 1-foot bins (1-2 feet, 2-3 feet, 3-4 feet, etc.) comparing then-current ShotLink percentages of putts made to Dave Pelz's similar numbers from back when he was researching his books. The difference was quite striking in the 6, 7, 8 foot range. Anyway, the gist of it was that when you compare now to then the players convert a lot more often from that more-or-less 50/50 range than they did in the 80's.

P.S. I get super annoyed every time a topic like this comes up because the Tour has all of that data but there's no way for me or you to access it on a subscription basis or anything like that. They apparently pre-digest it for media hype and never let people do real analysis (at least not for public consumption).

[Jim_Kennedy]

Garland,
It took me a while to find the article, but here it is. Make of it what you will, but I'll take the word of these three guys, thank you very much.  

....."two years later while at a conference in Northern Ireland, I attended a session titled “Can a Smooth Clubface Increase Spin?  The leader of the session was Art Chou of Chou Golf Design Labs.  He and Deshou Liang of Drexel University had done a study where they found a smooth faced club at low loft angles produces more back spin than a rough surface.
         To explain this I am going to quote Art(Chou), “For lofts under 20 degrees, the period of contact between the ball and the club face can be divided into two stages.  During the first stage of contact the ball is compressed on the club face, making the ball stick to the club surface.  As it compresses, the ball experiences some deformation as the loft of the club causes the ball’s center of gravity to move up the club face.  As the ball tends to move upwards, the club exerts a downward frictional force on the ball surface, creating the ball rotation that becomes backspin.  Due to the large normal force at impact resulting from the relatively low loft angle, this frictional force is independent of the surface roughness of the club face.  This backspin increases until, at the end of the period, the ball experiences such a large backspin that its surface in contact with the club actually starts to move downward.”
         Art went on to explain, “During the second contact period, while the ball surface is moving downward, the club exerts an upward frictional force on the ball.  This upward force acts against the direction of the ball’s backspin, reducing the overall spin with which the ball leaves the club face.  Where the downward force experienced during the first stage was independent of the surface properties of the club face, the upward force during the second stage is entirely related to the roughness of the face material.  This is because the forces acting on the ball at this time cause it to slide along the club face instead of sticking.”
         This means that the face roughness, or coefficient of friction, does not contribute to the backspin produced during stage 1 but contributes significantly to the forces acting against backspin in stage 2.  So, a rough surface, such as rough steel, retards the overall spin of the ball more than a lower friction material such as Teflon.  Therefore the smoother surface generates greater backspin. 
         At higher lofts, such as over 40 degrees, the nature of the contact period changes.  The ball experiences only the first stage of  contact, except this time the increased loft means that the ball never sticks to the club face.  Instead, it only experiences a sliding up the club surface.  This means that the downward frictional force exerted on the ball is proportional to the surface roughness; a rougher club face creates a larger downward force, resulting in increased backspin.
           Many individuals attending the seminar questioned the study stating the results were questionable at best and the study needed to be replicated.  That is when Frank Thomas, Technical Director for the USGA stated it had been replicated and was absolutely true.  After his comments the attitude in the room changed from disbelief to trying to understand the research better.
           Understand, this is only true of clubs lofted at 20 degrees or under.  So, what does this mean to the average golfer?  The application would be this, if you want more back spin with your driver, fairway driver, or low lofted irons, buff the faces smooth.  There is no USGA rule against a smooth face.  The USGA is only concerned with how rough the face may be.  If you do not want as much backspin created with your low lofted clubs be sure the face does not become worn smooth. If they are worn have them sand blasted so they are cosmetically rough again.  I say cosmetically rough, because the USGA states that the roughness of a face can only be for cosmetic purposes and not to impart spin.
        Today with the thin Beta-Titanium faces if grooves were cut into the face it would create a weak spot and break.  You will rarely see grooves cut in the middle of titanium heads now and if they are there they are not cut deep due to the thin face."

[Matt_Ward]

Jason:

The greens at ANGC have a very large role to play. Mike Weir when he won had an incredible short game week. Zach Johnson benefited from tough conditions that negated many birdies and kept the scores higher than normal.

Frankly, the PGA Tour had a great site with Westchester CC with the old Buick Classic there. It was a course that routinely confounded Tiger when he played there -- sparingly I might add.

Total length was barely above 6,700 yards but the classic layout had great land which heaves and hoes in all different types of manners and provided a fantastic array of champions over the years. The silly idea that couses have to be lengthened to some inane distance is not really what's needed.

The course was not "tricked up" and it had holes where power and accuracy could be tied together quite nicely.

Sad to see it go because it demonstrated that a control game can still play a winning hand. You can see a similar type of situation with the likes of Harbour Town -- even Colonial has fared well on this front.

Westchester CC will be missed by me -- although I believe they have one year left to host the event -- likely in 2012.

[Kyle Harris]

Does anyone else on this thread think that one of the main reasons players have adopted the "Bomb and Gauge" approach is that 6 - 8 foot putts are just not that difficult for them anymore?

I'd bet there is less "bomb and gauge" on the courses with tougher greens...

Jim:

Interesting question.  I've been reading a Bill James baseball statistics book recently so forgive this analysis which I will break into three posts:

History of Bomb and Gouge -

Statistics and results show that top players changed strategies in the the early 2000's - increasing their focus on length and worrying less about accuracy.

Vijay Singh and Tiger Woods show this well.

2000

Tiger hit the ball 20 yards farther than Singh off the tee (298 to 279) and hit more fairways (71.2 (ranked 54th) to 67.9 (ranked 112th)  Woods won $9 million and Singh $2.5.  Woods won 9 out of 20 tournaments.  Singh won 1 out of 26.

2004

The driving distance difference shrunk to nothing and Singh was slightly more accurate than Woods (60% to 56%) but 4% less accurate than the year before and 12% less accurate than in 2000.  In other words, Singh decided to sacrifice accuracy to make up his distance gap 8behind Woods.

Singh won 9 tournaments and $10.5 million and Woods won 1 tournament and $5.3 million. 

2005 

Woods cranked up his driving again - adding 15 yards and taking a 15 yard advantage over Singh (316 - 301) and a little hit in accuracy, now behind Singh by 6% (60-54).  Woods won $10.5 and Singh won $8.5 million that year.  Woods won 6 out of 19 tournaments.  Singh won 4 out of 30.


Since 2005, Woods has won 20 more tournaments.  Singh has won 6. 

The plural of anecdote is not data.

[Jason Topp]

The plural of anecdote is not data.

  I have no idea what you mean.  My entire discussion is based on data.  I then pose some theories from that data that may or may not be accurate. 

My theories can more broadly be tested both looking forward and looking backwards.

Looking Forward:


Because courses have gotten longer, I predict that absent another increase in driving distance (which seems unlikely) we will see:

1.  It is more difficult for individual players to dominate (9 and 10 win seasons will not happen)
2.  Driving accuracy will become more relevant to winning money on the PGA Tour

This theory can also be tested on a backwards basis.  I also suspect that if one went back to Tour results over the last few years, you would find that shorter hitters do better on two types of courses (1) courses that take driver out of long hitters hands (Harbour Town, Colonial) and courses that are long enough that bombers are hurt by poor driving accuracy (I have not analyzed which courses fit this description).   

[Niall C]

Jason,

I've nothing really to add to the discussion but would like to say that I have found it great to read. I think the next stage for you might be to have a go at analysing previous tournaments to see if your theory holds up. I for one would be interested to find out if it does.

Niall

[Kyle Harris]

The plural of anecdote is not data.

  I have no idea what you mean.  My entire discussion is based on data.  I then pose some theories from that data that may or may not be accurate. 

My theories can more broadly be tested both looking forward and looking backwards.

Looking Forward:


Because courses have gotten longer, I predict that absent another increase in driving distance (which seems unlikely) we will see:

1.  It is more difficult for individual players to dominate (9 and 10 win seasons will not happen)
2.  Driving accuracy will become more relevant to winning money on the PGA Tour

This theory can also be tested on a backwards basis.  I also suspect that if one went back to Tour results over the last few years, you would find that shorter hitters do better on two types of courses (1) courses that take driver out of long hitters hands (Harbour Town, Colonial) and courses that are long enough that bombers are hurt by poor driving accuracy (I have not analyzed which courses fit this description).   

Jason,

I don't think your data is nearly significant enough to merit any sort of conclusions. Comparing two golfers over three years is hardly representative of the affect of equipment, any equipment changes and golf strategy. Especially two golfers in the upper echelon of the game using a measure such as driving distance (done on one or two holes per round), prize money (changes per tournament and management of the tournament), and victories (based on the strength of field and frequency of play).

[Jason Topp]

The plural of anecdote is not data.

  I have no idea what you mean.  My entire discussion is based on data.  I then pose some theories from that data that may or may not be accurate. 

My theories can more broadly be tested both looking forward and looking backwards.

Looking Forward:


Because courses have gotten longer, I predict that absent another increase in driving distance (which seems unlikely) we will see:

1.  It is more difficult for individual players to dominate (9 and 10 win seasons will not happen)
2.  Driving accuracy will become more relevant to winning money on the PGA Tour

This theory can also be tested on a backwards basis.  I also suspect that if one went back to Tour results over the last few years, you would find that shorter hitters do better on two types of courses (1) courses that take driver out of long hitters hands (Harbour Town, Colonial) and courses that are long enough that bombers are hurt by poor driving accuracy (I have not analyzed which courses fit this description).   

Jason,

I don't think your data is nearly significant enough to merit any sort of conclusions. Comparing two golfers over three years is hardly representative of the affect of equipment, any equipment changes and golf strategy. Especially two golfers in the upper echelon of the game using a measure such as driving distance (done on one or two holes per round), prize money (changes per tournament and management of the tournament), and victories (based on the strength of field and frequency of play).

Kyle:

I agree the data is not significant enough to merit a conclusion.  I would suggest that it is significant enough to try out some tentative theories, which I have done.  if you have a better one, I would love to hear it.

Here is why I think the data is a reasonable basis upon which to try and come up with theories:

(1) The two top players represent precisely what it took to be the best player in the world during that time period.  If you throw in 2-4 more players you have the entire population of players with any shot at being considered the best.

(2) the statistics are matched by descriptions by the players themselves regarding what they were doing at the time;

(3) The examples illustrate more generalized data demonstrating that driving accuracy became irrelevant to tour success and that the driving accuracy of top players went down significantly during the last decade despite improved equipment that everyone thinks makes the ball go straighter

(4) prize money is probably the best indicator of relative performance within a year and did not change enough in that three year period to render comparisons between years meaningless

(5)  While driving distance measured over two holes is imperfect, it is probably the best measuring stick available for driver distance given that the tour attempts to choose driver holes and counteract wind.   It is not a very good indicator for how often players used driver.  The best measure would be to figure out if the players hit driver more often during the years where I have suggested a strategy change occurred.  I don't have that data.


 

[George Pazin]

Jason, I think you could greatly simplify things if you looked first at driving accuracy in a specific win. I think trying to find a correlation between season long stats is inherently misleading and inaccurate.

Brent Hutto is a stat guru, you should reread his posts.

[Jason Topp]

Quail Hollow is a course that should be a perfect test case for my theory.  It is 7500 yards long.  Under conventional theory this long course should favor long hitters.

Under my theory, this is a course where length would eliminate the double advantage and give short accurate players have a better chance than they do on a typical tour course. 


Although I do not know about wind and elevation factors, based on yardage alone 10 par fours are 450 or longer, 2 are about 350 and 2 fall within the "Bomb and Gouge Sweet Spot."  Under my theory, shorter hitters should have a better chance at Quail Hollow than they do in general.



General statistics for 2008 demonstrate that bomb and gouge players dominate the PGA Tour.  Of the top 10 money winners, 8 were bomb and gouge players ranking in the top 50 for driving distance and out of the top 100 for accuracy:(Driving Distance/Driving Accuracy rank)

   Singh (25/150)
   Woods (44/169)
   Mickelson (35/181)
   Garcia (43/153)
   Kim (11/159)
   Villegas (38/102)
   Cink (28/181)

Kenny Perry was balanced (31/87) and Justin Leonard finished 10th on the money list as a short and accurate player (132/21).




Results from Quail Hollow that I could easily access indicate that short accurate players have a better chance on the course.

In 2008, 12 players finished in the top 10 (with ties).
Four were Bomb and Gougers (Kim, Cink, Couples (16/169), Scott (8/185)), five were Short and Accurate (Curtis (121/47), Bohn (111/72), Slocum 164/3), Furyk (162/28); two balancedl (Allenby (61/70), Pampling (79/100) and one was poor in both categories (Hart 179/129).

I also pulled winner and runner up results from 2003-2008 - five were Short Accurate (Toms (72/2); Furyk; Furyk; Sindelar (143/66); Curtis), six were Bomb and Gouge (Singh; Geiberger (39/170), Singh; Garcia (71/129); Woods; Kim) two were balanced – Gamez (85/99); Immelman (58/115) and two were bad in both categorieas  – Oberholser (129/123); Stricker (135/178)

Eliminating the doubles; good finishers at Quail Hollow can be categorized as follows:

Short Accurate - 10
Bomb and Gouge - 9
Balanced - 4
Bad - 3


While I have no idea at what point these results would be considered statistically siginficant, there can be no doubt that the Quail Hollow results are very different from general tour results. 

We will see how the trend holds up this week.

[Jason Topp]

Jason, I think you could greatly simplify things if you looked first at driving accuracy in a specific win. I think trying to find a correlation between season long stats is inherently misleading and inaccurate.

Brent Hutto is a stat guru, you should reread his posts.

Thanks for the suggestions George.  It does make sense to look at a specific tournament set of stats.  I'm just not sure where they are.  We can look at Quail Hollow this week as an example but of course it will be a small sample.

I generally agree with Brent on everything he has said here except:

1.  I have not seen data to back up his recollection of a significant increase in length to get to the point where tour pros make 50% or their putts.  Even if there is an increase, I am not sure it impacts my analysis.

2.  I disagree with his viewpoint that fairways hit is not a meaningful measure of accuracy.  While there will no doubt be ocassions where a player is trying to hit a tee shot in the rough (or on the green on a short par four) I think those situations are exceedingly rare.  A better measurement would be to know where the player is aiming and measure the distance away from that spot but I do not have access to that data and I doubt anyone has it.

[Jerry Kluger]

Here's a couple of factors which are hard to put into any equation:

1. Tiger is a freak - that is, he is so freakin' good that it doesn't really matter.

2. Top players can choose what tournaments they want to play and certainly the course is part of his consideration.

BTW: Getting back to one of my earlier posts - I would say that the shorter putts are more difficult on fast greens - to me, speed equals break, which means trouble.

[Brent Hutto]

2.  I disagree with his viewpoint that fairways hit is not a meaningful measure of accuracy.  While there will no doubt be ocassions where a player is trying to hit a tee shot in the rough (or on the green on a short par four) I think those situations are exceedingly rare.  A better measurement would be to know where the player is aiming and measure the distance away from that spot but I do not have access to that data and I doubt anyone has it.

I'm less concerned about not knowing when the players might be aiming somewhere other than down the middle (moot point, as you point out nobody will ever have that data) than with the reifying the characterisation of Fairway vs. Not Fairway which is in many cases totally unrealistic in capturing what matters about where the ball ends up. If they want a binary measure (and I most certainly do not) it would have to be Limits the Next Shot vs. Does Not Limit the Next Shot.

Imagine the difference between a ball sitting on an uber-tight Tour fairway versus one sitting 10 yards away in an inch of fescue rough. For any given situation estimate the expected loss of scoring opportunity presented by the former vs. the latter. Maybe zero, maybe a tenth of a stroke?

Now imagine the difference between that ball in an inch of rough vs. one 10 yards away in thick knee-high fescue, the kind that wraps around the hosel of your club, and with a bush a few yards in front of the ball. What's the expected loss of scoring? Not zero, not a tenth of a stroke. More like a half a stroke if you catch a good lie and maybe closer to a full stroke if it's a wedge out sideways because of the bush.

The "Fairways Hit" percentage treats those two balls as an identical outcome of the tee shot. Doing so is idiotic and no, it does not average out across players.

P.S. So my complaint to the USGA is this. If you're going to spend years and millions of dollars on a campaign to bring tee shot "accuracy" back into the game of elite players as the most important determinant of scoring, save putting, then start by defining "accuracy" in some meaningful way and then actually measuring it. Do not claim some useless convenience garbage stat as justification for what you claim is the most important issue facing the game.

[Jason Topp]

2.  I disagree with his viewpoint that fairways hit is not a meaningful measure of accuracy.  While there will no doubt be ocassions where a player is trying to hit a tee shot in the rough (or on the green on a short par four) I think those situations are exceedingly rare.  A better measurement would be to know where the player is aiming and measure the distance away from that spot but I do not have access to that data and I doubt anyone has it.

I'm less concerned about not knowing when the players might be aiming somewhere other than down the middle (moot point, as you point out nobody will ever have that data) than with the reifying the characterisation of Fairway vs. Not Fairway which is in many cases totally unrealistic in capturing what matters about where the ball ends up. If they want a binary measure (and I most certainly do not) it would have to be Limits the Next Shot vs. Does Not Limit the Next Shot.

Imagine the difference between a ball sitting on an uber-tight Tour fairway versus one sitting 10 yards away in an inch of fescue rough. For any given situation estimate the expected loss of scoring opportunity presented by the former vs. the latter. Maybe zero, maybe a tenth of a stroke?

. . .

P.S. So my complaint to the USGA is this. If you're going to spend years and millions of dollars on a campaign to bring tee shot "accuracy" back into the game of elite players as the most important determinant of scoring, save putting, then start by defining "accuracy" in some meaningful way and then actually measuring it. Do not claim some useless convenience garbage stat as justification for what you claim is the most important issue facing the game.

Thanks for the clarification Brett.  I certainly agree that there are degrees of misses and that a ball in the bushes is worse than a ball in short rough.  I'm not sure I agree with your viewpoint that there is virtually no difference in impact on the next shot between a ball in short rough vs. a ball in the fairway.  There is no difference for you or me.  I do not know about a tour pro.  If it did not matter, "second cut" at the Masters would be utterly meaningless.

Furthermore, I do not think one needs to be real precise with this measurement.  I would be surprised if anyone seriously thinks, Woods, Mickelson, Singh, Garcia, Cink etc are accurate off the tee compared with tour norms.  I would be surprised if anyone labeled Furyk, Slocum or Leonard as anything other than short and accurate compared with tour norms.

[Brent Hutto]

Yeah, but the origin of this discussion is the USGA's (in my opinion specious) claim that lack of correlation between "Percentage of Fairways Hit" and winning on Tour is casus belli for their war on modern wedge grooves. That's a lot of claim to make for such a poorly thought-out "statistic".

[JMEvensky]

P.S. So my complaint to the USGA is this. If you're going to spend years and millions of dollars on a campaign to bring tee shot "accuracy" back into the game of elite players as the most important determinant of scoring, save putting, then start by defining "accuracy" in some meaningful way and then actually measuring it. Do not claim some useless convenience garbage stat as justification for what you claim is the most important issue facing the game.

Could the "graduated rough" used at the US Open be a result of the USGA coming to your same conclusions through their own findings regarding accuracy?

[Jason Topp]

Interesting Golf World column on the topic which compares statistical evaluation with players acecdotal beliefs on the groove rule:

http://www.golfdigest.com/golfworld/columnists/2009/05/golf_grooves_rule_johnson_0501


One excerpt:

I don't completely buy the USGA's premise there is little-to-no correlation between accuracy and success on the PGA Tour, which is important because it is pretty much the premise for the rule in the first place.

In compiling its correlation, the USGA used year-end accuracy stats and year-end money stats and if you went solely by that, you can see their point. A quick look at the top 10 in earnings last year would be nothing short of appalling in terms of accuracy as nine of the 10 finished 120th or worse in hitting fairways. Only two players in this group (Kenny Perry and Justin Leonard) managed to hit more than 60 percent of fairways for the year.

But is that a fair measure? You can argue about what qualifies as being a "success," but there is one metric that is not up for discussion: winning. If you win, you are a success that week. And the winners on the PGA Tour in 2008 did OK when it came to keeping the tee shot in the short grass. In 23 of the 46 events in which driving accuracy was measured, the winner managed to hit 66 percent or more of the fairways -- or basically two out of every three. The average of the 46 winners was 64.97 percent -- above the tour average of 63.16 percent. So the aim of the game's best isn't as bad as it appears at first blush.

[Rich Goodale]

I, too, was present at the Valley Club that evening and what I took away from Jim Vernon's talk was:

1.  The USGA has absolutely no interest in controlling distance, particularly by rolling back the ball.  Why?  Just because......
2.  They think that by controlling grooves they can mitigate the advantage given to the longer hitters.  Why?  Just because......

IMHO, Brent is right regarding their lack of convincing statistical proof and Pat is right that no matter what they do the best pros will find a way to manufacture shots that send them back to their drawing boards again.

If you are a betting person, put your money on the status quo for the forseeable future.

[Ken Moum]

The made-putts-by-distance numbers from ShotLink in that range are quite significantly higher than the same numbers gathered by Dave Pelz 15-20 years ago. I trust the ShotLink number completely and the Pelz numbers unless someone has evidence his methods were flawed. I don't have the comparison at hand but they have come close to doubling their percentage made in the 6, 8, 10 foot range over just a couple decades.

Brett - when I looked at that, I do not agree.  Shotlink gives percentages in ranges, say less than 3 feet, between  5' and 10' etc. 

It is impossible, however, to convert that to the Pelz measurements because you do not know how putt lengths are distibuted within that range.  For example the stats showing a player makes 99% of putts of 3 feet or less is misleading because the vast majority of those putts will be tap ins.  If 99% of the putts are tap ins and 1% are three footers, the statistic would show the guy was a lousy putter, not a good putter.

Shotlink stats generally show players making more than 50% of putts between 5 and 10 feet.  If players have a lot more 5 footers than 10 footers, Pelz 6 foot 50/50 range may still be accurate. 

I suspect the Pelz number has changed but do not have enough information to be sure by how much.  12-18" was my best guess, but I cannot recall how I came up with it.

Jason, the 2005 PGA Tour media guide included some putting info from 2004, and it's a lot more detailed:

Inside 3 feet 99.1%
3-4 feet 90.9%
4-5 feet 80.8%
5-6 feet 69.8%
6-7 feet 61.6%
7-8 feet 54%
8-9 feet 47%
9-10 feet 42.2%
10-15 feet 31%
15-20 feet 18.9%
20-25 feet 13.2%
Over 25 feet 5.8%

Given those numbers, it looks like the 50% line has moved about three feet in the 20 years since Pelz collected his data.

And there was also this re. proximity to the hole for approach shots:

All shots 32'2" feet
From Rough -- Left 42'4" feet, Right 42'5"
From fairway 24'11"
From inside 75 yards 16'5"
From 75-100 yards 18'1"
From 100-125 yards 20'8"
From 125-150 yards 23'11"
From 150-175 yards 28'10"
From 175-200 yards 34'4"
From outside 200 yards 42'7"

And:

Proximity to the hole from sand 9'10"
Sand saves 49.1%  (Which tells me the average proximity is skewed by a small # of bad shots. Because that's a better putting average than they should get from 10 feet.)
Scrambling from the fringe 85.3%
Scrambling from the rough 83.8%
Scrambling from 10-20 yards 61.3%
Scrambling from 20-30 yards 47.9%
Scrambling from more than 30 yards 26.8%

Finally, a driving distance quirk. The Tour has maintained its driving distance stat as two holes per round, but Shotlink keeps it for all drives.

Driving distance (two holes) 287.3 yards
Driving distance (all drives) 279.4 yards
They hit 64.2% of fairways

Ken