Facts and Distance

[Paul_Turner]

The PDF file states that the COR drops faster for solid balls than wound.

[DMoriarty]

The PDF file states that the COR drops faster for solid balls than wound.

Yes . . .  

Since the .pdf doesnt give any indication of the actual COR of the wound ball, this tidbit sounds interesting but is not at all illuminating.

Had the USGA chosen to include a wound ball in their tests, my guess is that we would have seen a slightly straighter line but also a lower and flatter line.  

[JohnV]

Still Geoff's suggestion of using 1995 would clearly be comparing all wound vs. the all solid 2005.

Ok, I did it.  I took the 60 players who were in the PGA Tour stats in both 1995 and 2005.

I put them in groups of 6 to get 10 groups.  Below are the average gains from 1995 to 2005 for the 10 groups with shortest 1995 hitters listed first.

Shortest 6  - 18.75 yards
2nd Shortest  - 18.97
3rd Shortest - 18.95
4th Shortest - 19.52
5th Shortest - 21.92
6th Shortest - 19.4
7th Shortest - 18.45
8th Shortest - 19.2
9th Shortest - 13.2
Longest       - 17.55

It seems like the shorter hitters have gained more than the longer hitters from the new balls and clubs, with the bottom 50% gaining 2 more yards than the top 50%.

The 6 shortest hitters in 1995 of this group were: Omar Uresti, Billy Andrade, Corey Pavin, Loren Roberts, Larry Mize and Jim Furyk.

The 6 longest hitters were Fred Couples, Woody Austin, Vijay Singh, Dennis Paulson, Davis Love III and John Daly.

If it wasn't for Corey Pavin's measly 3.8 yard gain, the lowest 6 would have been even better.

[DMoriarty]

John,

Perhaps you can explain to me why you and USGA chose to use the older of the statistics (2000 as opposed to 2005, 1995 as opposed to 2005) as your baseline?    If we are really interested in comparing who benefited from technology with those who didnt, shouldn't we use the later of the comparison years as the baseline?  Given the difference in the game compared to even 5 years ago wouldn't the 2005 distance numbers be a more accurate indicator of relative swing speed?

I tried it to see if the results differ.  And they do, significantly.  

I took the 80 players active in 2000 and 2005 and broke them down into groups of 8 (10%.)   Not suprisingly (not suprising to me at least) the longer hitters (presumably the faster swingers) gained the most.  By far.

With the longest drivers in 2005 listed first . . .

Top 10%   18.55 yards
11-20%   14.36
20-30%   14.09
30-40%   12.36
40-50%   14.68
50-60%   11.08
60-70%   8.63
70-80%   8.79
80-90%   5.04
 10%   7.31  yards.

So the longest 10% in 2005 had gained over 11 yards more than the shortest hitters.

 

[JohnV]

I have also done 1995-2000 comparison.  There were 106 players that were in the list for both years.  I put 11 players in the shortest and longest groups and 12 in all the ones in the middle and got 9 groups.  Here are the numbers:

Shortest - 10.39 yard average gain
2nd Shortest - 11.14
3rd Shortest - 7.88
4th Shortest - 7.71
5th Shortest - 6.77
6th Shortest - 6.73
7th Shortest - 6.51
2nd longest - 4.81
Longest - 2.69 yard average gain

So the shortest players in 1995 gained 4 times as many yards as the longest ones did by 2000.  Seven of the short 11 gained over 10 yards while only one of the 11 (John Daly) gained that much.

11 players lost yardage betwen 1995 and 2000 including 3 of the longest 11 (Robert Gamez, Woody Austin and Vijay Singh).

Nobody lost yardage between 1995 and 2005.

[Jim Nugent]

Still Geoff's suggestion of using 1995 would clearly be comparing all wound vs. the all solid 2005.

Ok, I did it.  I took the 60 players who were in the PGA Tour stats in both 1995 and 2005.

I put them in groups of 6 to get 10 groups.  Below are the average gains from 1995 to 2005 for the 10 groups with shortest 1995 hitters listed first.

Shortest 6  - 18.75 yards
2nd Shortest  - 18.97
3rd Shortest - 18.95
4th Shortest - 19.52
5th Shortest - 21.92
6th Shortest - 19.4
7th Shortest - 18.45
8th Shortest - 19.2
9th Shortest - 13.2
Longest       - 17.55

It seems like the shorter hitters have gained more than the longer hitters from the new balls and clubs, with the bottom 50% gaining 2 more yards than the top 50%.

The 6 shortest hitters in 1995 of this group were: Omar Uresti, Billy Andrade, Corey Pavin, Loren Roberts, Larry Mize and Jim Furyk.

The 6 longest hitters were Fred Couples, Woody Austin, Vijay Singh, Dennis Paulson, Davis Love III and John Daly.

If it wasn't for Corey Pavin's measly 3.8 yard gain, the lowest 6 would have been even better.

Great job, John.  With one exception (9th shortest, i.e. 2nd longest), the differences don't seem significant.  My interpretation is, everyone gained about the same.  

Two questions.  Suppose equipment had not changed.  The players are now ten years older.  Do you think they, on average, would have lost yardage?

Second, these guys all learned their swings with the old balls, the old clubs.  As I understand it, the new tech lets you swing harder.  So could it help young players learning the game more than older ones who already have built their swings?  

[JohnV]

David, I chose it to be consistent with the USGA's report.

I believe the proper way to measure change is to measure from a baseline and go forward in time, not backwards.

Somehow it doesn't surprise me that you would like to go back in time.  

[DMoriarty]

John, which year--- 2000 or 2005--- do you think is a more accurate indicator of relative swing speed between the players?

After all, isnt that the point-- to try to best capture swing speed?  

[JohnV]

David,

I had 98 players who were active in 2000 and 2005 and on the driving distance list.  I then added Ernie Els and Nick Price since they didn't play enough rounds to be counted and it made it an even 100.

When done the same way as the USGA, I get almost identical numbers (except for the Ernie/Nick factor)  Ernie was a big gainer and modified his group pretty heavily.

When I do things the same way you did, I do see that the longer hitters would show the most gain, .  I don't believe your numbers are as valid though.  Since you are sorting based on the end of the series rather than the start, your numbers get biased towards showing the changes.  I took a few statistics classes in college over 30 years ago and I can't remember much, but it seems that when you plot trends, you should go forward rather than backwards.

[JohnV]

David,

The relative swing speed of two players can be based on many things.  As to which is a more accurate representation, I don't see how you can say either is, both are what they are at that point in time.  Given that we are talking about 2006, then 2005 is a more accurate measurement of today.  That doesn't mean anything when deciding how you sort the data to me.

How much of 2005's gain is swing speed (and which of the following affects that), how much is the ball, how much is the club head going from 200 CC to 460 CC, the COR going up, the MOI going up, the shaft getting better and longer, the players working out more.

Some players swing faster in 05 that in 00, others might swing slower if you gave them the same club as 2000.  But, they all have gained because the ball and the clubs have gotten better.  No doubt about that.  If someone has gone to a longer shaft, they are getting more swing speed.  If they have worked out more, they probably got more swing speed.  If they got a more forgiving club, they might be swinging harder.  If they are trying to keep up with Tiger, they might be doing all of the above.  

Tiger switched from steel to graphite, he went to a longer shaft, he is stronger and fitter than he was in 2000, he modified his swing to make better contact, he switched clubs and balls.  And out of all that, he gained 18.1 yards.  I can't tell you which of those added the most or the least.

[DMoriarty]

When I do things the same way you did, I do see that the longer hitters would show the most gain, .  I don't believe your numbers are as valid though.  Since you are sorting based on the end of the series rather than the start, your numbers get biased towards showing the changes.  I took a few statistics classes in college over 30 years ago and I can't remember much, but it seems that when you plot trends, you should go forward rather than backwards.

I did my initial analysis sometime last year and had ended up with 80 players..  I dont recall but I might have cut my numbers off at 150 or something.  I'll look again to pick up the other 20, or if you want you can send me your spreadsheet and I can work off that as I do trust it.

But I disagree with you about the validity of using the old baseline vs. the more recent baseline.   Our goal is to determine how the faster swingers benefited from the technology vs. the slower swingers from 2000 to 2005 (or for whatever years.)   So our baseline should be whatever statistic most closely correlates to swing speed for these given players.   Given the changes in the game and the transitory state of the technology in 2000 (ex. 40% already using solid balls by the end of the 2000) I'd say that 2005 is a much better bellwether for swing speed.  

Let me ask you again,  which year do you think is a better indicator of relative swing speed?  

Would you split the difference?  

Here are the numbers given my 80 players and averaging their distances for both years . . .

Longest 10%   14.4
Next 10%   14.6
Next 10%   12.8
Next 10%   12.0
Next 10%   12.4
Next 10%   13.2
Next 10%   10.9
Next 10%   9.8
Next 10%   7.6
Shortest 10%   7.3
 

[Garland Bayley]

How it is unfair for a player who is able to swing faster than me to hit it farther than me escapes me.  That's just the way golf is.
...

It is fair for a player who swings faster to hit it farther. It is unfair if his advantage is increased by artificial technological tricks.

[Garland Bayley]

Still Geoff's suggestion of using 1995 would clearly be comparing all wound vs. the all solid 2005.

Ok, I did it.  I took the 60 players who were in the PGA Tour stats in both 1995 and 2005.

I put them in groups of 6 to get 10 groups.  Below are the average gains from 1995 to 2005 for the 10 groups with shortest 1995 hitters listed first.

Shortest 6  - 18.75 yards
2nd Shortest  - 18.97
3rd Shortest - 18.95
4th Shortest - 19.52
5th Shortest - 21.92
6th Shortest - 19.4
7th Shortest - 18.45
8th Shortest - 19.2
9th Shortest - 13.2
Longest       - 17.55

It seems like the shorter hitters have gained more than the longer hitters from the new balls and clubs, with the bottom 50% gaining 2 more yards than the top 50%.

The 6 shortest hitters in 1995 of this group were: Omar Uresti, Billy Andrade, Corey Pavin, Loren Roberts, Larry Mize and Jim Furyk.

The 6 longest hitters were Fred Couples, Woody Austin, Vijay Singh, Dennis Paulson, Davis Love III and John Daly.

If it wasn't for Corey Pavin's measly 3.8 yard gain, the lowest 6 would have been even better.

Thanks for doing this John. It helps the USGAs argument some. However, as I wondered before, why isn't the mathematical model used to compare the wound balata ball? The data such as yours and that given in the USGA report suffers from mixing in several factors among which are the change in club head size and COR, and the Tiger conditioning factor (players working on their conditioning to try to keep up with Tiger). My college technical writing teacher taught me to not change methods in the middle of a paper. If the paper is going to present the formulae describing the physics, then it should complete the argument that way instead of switching to questionably applicable tour driving stats.

[Bryan Izatt]

How it is unfair for a player who is able to swing faster than me to hit it farther than me escapes me.  That's just the way golf is.
...

It is fair for a player who swings faster to hit it farther. It is unfair if his advantage is increased by artificial technological tricks.

Are you suggesting that there are still some "artificial technological tricks" that lead to disproportionate gains?  If so, what are they?

[Bryan Izatt]

For those of you who want a comparison with a wound balata ball, where would you propose to get some to do the test with?  Does anybody still manufacture them?  Any that have been lying around for 10 years would certainly have lost some of their distance due to loosening of the windings and deterioration of the rubber.  I doubt that it is possible to effectively construct the test you're suggesting.

[Bryan Izatt]

But I disagree with you about the validity of using the old baseline vs. the more recent baseline.   Our goal is to determine how the faster swingers benefited from the technology vs. the slower swingers from 2000 to 2005 (or for whatever years.)   So our baseline should be whatever statistic most closely correlates to swing speed for these given players.   Given the changes in the game and the transitory state of the technology in 2000 (ex. 40% already using solid balls by the end of the 2000) I'd say that 2005 is a much better bellwether for swing speed.  

This reminds me of that old saying - lies, damn lies and statistics.  You can use numbers to say anything you want.

I'm with John; my statistics training was always based on selecting your group and then moving forward - not backward.

In trying to address your new goal, why don't you use 1995 so that the transition of players from wound to solid balls is removed from the equation.  John's numbers for those years support the USGA's graph.  No disproportionate gains for high speed swingers.  If you use 1995 as your base year, is it not axiomatic that the longest hitters of that era were also the fastest swingers?  Does anybody dispute that the fastest swingers were also the longest in any era?  Or do you believe that Garland's "artificial technological tricks" were at play in that era as well.

[DMoriarty]

For those of you who want a comparison with a wound balata ball, where would you propose to get some to do the test with?  Does anybody still manufacture them?  Any that have been lying around for 10 years would certainly have lost some of their distance due to loosening of the windings and deterioration of the rubber.  I doubt that it is possible to effectively construct the test you're suggesting.

Well Bryan, the USGA made the claim that the COR drops less for wound balls as swing speed increases, so apparently they have already done the tests.  If you don't think the balls they used were up to snuff, then why don't you contact them and express your concerns.  I am sure TEPaul can put you in touch with them.  

And Bryan, I don't really think it is my goal at all.  But if you say so . . . then perhaps you can explain to me just what exactly the USGA was trying to prove by comparing the pro statistics?  Just what variables were they trying to test?  Did they hold all the other variables constant?   Do you consider their comparison "scientific?"  What have they proven?

[Jeff Fortson]

For those of you who want a comparison with a wound balata ball, where would you propose to get some to do the test with?  Does anybody still manufacture them?  Any that have been lying around for 10 years would certainly have lost some of their distance due to loosening of the windings and deterioration of the rubber.  I doubt that it is possible to effectively construct the test you're suggesting.

Are you suggesting that Titleist couldn't produce 100 brand new Tour Balatas by tomorrow if asked?  I'm sure new wound balls could be created with ease.


Jeff F.

[Steve Lang]

 

hey.. ol' Billy Andrade was 6 under par yesterday on his own ball in a Pro-AM at the SHO.. hit the ball beautifully on the new long Tourney Course at Redstone..

Isn't there the corrolary and perhaps stronger argument that to capture any benefits from swinging in the 120+ mph range, one better hit the ball square with less side spin imparted!  Otherwise you're long in the woods or the next county..

I vote for the "approach to a perfect human swing" theory explaining the largest portion of the variance behind distance gains (say > 40%), lagged by club & ball improvements (25% each), followed by course improvements (10%).

Everyone can be in violent agreement or disagreement, but there's certainly a mix to this issue, emblematic of the game, open to opinion.

As much as it hurts, from my experience with the SHO leaving the TPC at the Woodlands for Redstone,  I say let the pros move on to new venues..

[Brent Hutto]

Well Bryan, the USGA made the claim that the COR drops less for wound balls as swing speed increases, so apparently they have already done the tests.  If you don't think the balls they used were up to snuff, then why don't you contact them and express your concerns.  I am sure TEPaul can put you in touch with them.  

They said the COR decreases less over the range from 90mph to 125mph for a wound ball than a solid ball, which is true. The COR of a wound ball is lower and more constant throughout that range than for modern solid-core balls. They didn't do a test because that is known as surely as what phase the moon will be tonight, old news, settled fact, nothing to test.

Note that they did not claim that the curve of distance versus clubhead speed was flatter for a wound ball than a solid one because it is not. Due to the higher spin of wound balls (and even more so wound balata balls) the gain in distance per unit increase in clubhead speed drops off dramatically at high clubhhead speeds. So do not take anything in this test as a contradiction of that fact.

It is a pity they didn't use the time and effort that went into this study to quantify the more interesting question of what has changed between the normative "Tour" ball of ten years ago and the solid-core "Tour" balls today at high clubhead speed.

[Jonathan Cummings]

The Coefficient of Restitution is a scalar - it has no dimensions. It is best to view COR as an measure of efficiency - how well (how efficient) does the dynamic clubhead transfer mechanical energy to the stationary golf ball. COR is purely a material property and invariant to the clubhead speed. The COR for an electron hitting the clubhead is the same for a bowling ball hitting it.  If the COR was a function of impact velocity - if would have to have units of restitution/impact velocity.

Given the same ball and same club (same COR) by physics Daly hits it further than Pavin purely based on the square of the impact velocities, not due to the COR.

Now, if you let Pavin use a "springier" club head with higher COR (Ping to a Taylormade, for example) than does Daly, then you're comparing apples to oranges and must adjust for the "trampoline" (COR) effects to make a meaningful comparison.

I'm baffled by the USGA data.

JC

[Paul_Turner]

The COR for a given collision is not invariant with clubhead speed.  The more that ball is deformed, the less efficient it is at recovering its shape and releasing the elastic strain energy into kinetic.

[Brent Hutto]

COR is purely a material property and invariant to the clubhead speed. The COR for an electron hitting the clubhead is the same for a bowling ball hitting it.  If the COR was a function of impact velocity - if would have to have units of restitution/impact velocity.

Not true. For starters, the COR is demonstrably different at different clubhead speeds. Just because a quantity is dimensionless does not mean it is invariant. At full-swing speeds the collision between clubhead and ball is far from inelastic (which seems to be type of "physics" you're quoting) and restoration is a non-linear function of the both the material(s) in the golf ball and the impact velocity.

Given the same ball and same club (same COR) by physics Daly hits it further than Pavin purely based on the square of the impact velocities, not due to the COR.

Not true. You're confusing kinetic energy with the clubhead speed and ball speed. Initial ball speed scales (approximately) with the first power of clubhead speed, not the square.

Now, if you let Pavin use a "springier" club head with higher COR (Ping to a Taylormade, for example) than does Daly, then you're comparing apples to oranges and must adjust for the "trampoline" (COR) effects to make a meaningful comparison.

The study we're discussing used a single clubhead to do all their tests.

I'm baffled by the USGA data.

Obviously. Start by just reading the actual published results rather than trying to adapt it to your own incomplete understanding of the physics behind ball/club impact. The paper itself gives you all the relavant physics you need to understand the results and the physics he presents are correct.

[Paul_Turner]

Have any wound balls been manufactured in the last few years?  The COR of a wound ball decreases as it ages.  

I'd be interested to know why the COR is more constant for a wound ball over the range of impact velocities.

I've only scanned the paper.  But I think the USGA must have used a Ti club without the SLE (Spring Like Effect) to minimize variables.  i.e. the effect of the SLE will not be exactly constant through the impact velocities and will change COR of the collision.

(I believe the SLE is not constant, but don't have data.)

[Brent Hutto]

Paul,

A clubface has a characteristic time (the frequency at which it flexes) and when that time/frequency matches closely to the compression and rebound time constant (or frequency) of the golf ball you get a higher COR (more efficient rebound). According to my limited understanding, the rebound time constant of the ball varies with impact speed while that of the clubface is constant (or varies only negligably).

The so-called "spring-like effect" consisted mainly of building drivers with faces sufficiently large and thin that their frequency was lowered into the range that matches up with modern golf balls. A typical 250cc stainless steel driver would have a characteristic time much shorter than the time it takes for a golf ball to rebound off the face, making the collision dynamics almost completely a function of the restoration that takes place in the golf ball. With thin-faced titanium drivers and large clubheads you can slow down the flexing of the clubface just enough to match up with the ball and let the driver do some of the flexing instead of the ball. The rebound of a metal clubface from a very small deflection is orders of magnitude more efficient than the rebound of a elastomer golf ball from a deflection of nearly half its diameter. So to the extent you can substitue clubface flex from golf ball compression you get more efficient rebound, hence the advantage to big, thin-faced drivers.