New Golfweek's Best: 2014 Rankings

[Bart Bradley]

Pac Dunes average score is under 9.  That means lots of raters don't consider it among the top 15 modern courses.  In fact, its average score puts it (slightly) outside that ranking. 

Much more unanimity about which classic courses are the best.  Only one modern course averages above 9, while 6 classics do.  Also, the averages for PVGC and CPC mean most raters rank both courses in the top 5.   

I suppose that could be ... but even the rating for Sand Hills is only 9.22, which means less than half the panelists put it in the top 5 modern courses.  That seems unlikely to me.

I suspect that fewer 9's and 10's are awarded to the modern courses, because voters are comparing the numbers to the classics.  They are supposed to be separate polls, but the magazine has produced a bunch of derivative lists which combine the numbers, and the panelists vote on both at the same time, so it wouldn't be surprising if they think of them together.

Of course, there are lots of GOLFWEEK panelists here, so they can certainly correct me if my hunch is wrong.

Tom:

The rating of 9.22 does not mean that half the panelists voted it outside the top 5 modern....all of the panelists may have voted it the number one modern at it still could have scored a 9.22.  There is no 10 point must system for the number one course.

Bart

[BCowan]

Sean,

   You need to play Battle Creek CC (MI)

[Tom_Doak]

Tom:

The rating of 9.22 does not mean that half the panelists voted it outside the top 5 modern....all of the panelists may have voted it the number one modern at it still could have scored a 9.22.  There is no 10 point must system for the number one course.

Bart

Bart:  I get that.  But Jim Nugent was saying the score of Pacific Dunes meant that "quite a few panelists did not vote for it among the top 15," and I questioned that for the same reason. 

[Sean_A]

Sean,

   You need to play Battle Creek CC (MI)

B

"Need" implies desperation    I haven't yet hit that stage in life.  There are far too many wonderful courses to worry about being too particular.

Ciao    

[BCowan]

U are as particular as they come 

[Sean_A]

U are as particular as they come 

It depends on what it costs me in time and money 

Ciao

[John Crowley]

Congratulations to Dunlop White!

+1

Congrats!

+2

Let's all re-read Dunlop White and resist trying to apply science to an art form.

Having said that, rankings are with us for the duration and debates that follow here do tweak the mind.

[jim_lewis]

One more time!!!

Golfweek raters do NOT Rank courses. They assign ratings.  The rankings are done by averaging the ratings provided by all raters who submit a rating for a particular course. That is why I always object to being referred to as a "ranker: I don't rank. I rate. I may give two courses the exact same rating which means I would not necessarily rank one above the other. I leave the ranking to Brad and his large and capable staff of mathematicians

Jim

[Steve Burrows]

Good to see Jeff Brauer's work getting recognized on the public list, i.e. the top 3 courses in Kansas and the top 2 courses in Minnesota are each his.

[Joe Sponcia]

I'm confused how the worst Nicklaus got third in the entire state of TN when no one I know likes it better than Harrison bay or even Tim's Ford?

Also, how did Legends Moorland beat out Strantz True Blue, Tidewater, and Tpc of Myrtle in S Carolina? 

[Mac Plumart]

I simply must say it again,

The story behind this year's Golfweek Ratings is Old Town.

Kudos!!!

[Mark Steffey]

However, to your question, I'd put Winchester equal to or ahead of Worcester, Newport, Salem and Brae Burn.  Also, I haven't seen Vesper or Charles River, but I would be shocked if I thought either was better than Winchester.

Winchester is a better course than Vesper.  I've played all on that list except the River and Winchester compares very capably with those.  

I would rather though play Newport over Winchester any day.

[Scott Stambaugh]

Top 100 Classic #24- The Country Club composite course.

What is the criteria for compiling what I assume is a "best of" routing of an 18+ hole venue?  Seems to me there are other "composite" courses that would rank quite high if you could cherry pick the best holes on the property.

Scott

[Greg Holland]

also need a shout out to Kyle Franz and Mid Pines making an appearance at 77 on the Classic list.  I love the restoration and the course.  It is definitely a course on which I want to immediately return to the first tee after finishing and play again.  Well done.

[Nigel Islam]

Good to see Jeff Brauer's work getting recognized on the public list, i.e. the top 3 courses in Kansas and the top 2 courses in Minnesota are each his.

Also Gaylord Springs in Nashville

[Emile Bonfiglio]

My thoughts on Oregon's Ratings:

Oregon

1. Pacific Dunes, Bandon (No. 2, m) (Yes)
2. Old Macdonald, Bandon (No. 6, m)(Yes)
3. Bandon Dunes, Bandon (No. 8, m)(Yes)
4. Bandon Trails, Bandon (No. 22, m)(Yes)
4a Bandon Preserve (not sure how this could have been included)
5. Tetherow, Bend (m) (not even the best in Central Oregon)
6. Pronghorn (Nicklaus), Bend (m) (worthy)
7. Pumpkin Ridge (Ghost Creek), Cornelius (m)
8. Sunriver (Crosswater), Sunriver (m) (Worthy)
9. Running Y Ranch, Klamath Falls (m) (No comment)
10. Black Butte Ranch (Big Meadow) (m) (2nd best course at the resort)

Courses I feel should be in contention for this list
-Stone Creek
-Aspen Lakes
-Juniper

[Jim Nugent]

A course ranked 50th in the nation gets 8 points.  A course ranked 51st gets 7 points.  So minuscule changes in ranking can sometimes lead to big moves in the ranking. 

On the other hand, ranking a course 16th gives it the same score as ranking it 50th (8 points).  i.e. GW's system does not lend itself to much precision. 

A few other interesting points:

Pac Dunes average score is under 9.  That means lots of raters don't consider it among the top 15 modern courses.  In fact, its average score puts it (slightly) outside that ranking. 

Much more unanimity about which classic courses are the best.  Only one modern course averages above 9, while 6 classics do.  Also, the averages for PVGC and CPC mean most raters rank both courses in the top 5.   

Jim, I'm not a GW rater, but I'm fairly certain that they have the ability to enter in gradients and not just whole numbers.

As I understand it, you guys aren't quite right. 

There are two issues, or numbers, at stake.  How the raters score each course, and how they rank them.  They are different altogether.  The scoring is a tool, to help the raters analyze the various courses.  But the rankings are what really count.  In theory you could score a course at 9.90, but still rank it 50th or 100th -- so long as you gave 50 or 100 courses higher scores. 

Those average scores we see next to each course give the average of the RANKINGS.  I'm pretty sure for that there are no decimal points.  i.e you, the rater, give the course a 10 if you think it's in the top 5 courses.  You give it a 9 if you think the course is among 6 to 15. 

So if I'm right about that, then it's almost a sure thing that lots of raters rank Pac Dunes outside the top 15.  The only other alternative is that some of them give it a ranking score of 2 or 3.  Seems real unlikely to me -- the statistical models would probably throw them out as outliers -- and would still require a fair number of them, so long as Pac Dunes gets lots of ballots. 

btw, the average scores also tell us the majority of raters (though not all) ranked both Pine Valley and CPC as 10s -- they consider both courses among the top 5 classic in the country. 

[John Crowley]

One more time!!!

Golfweek raters do NOT Rank courses. They assign ratings.  The rankings are done by averaging the ratings provided by all raters who submit a rating for a particular course. That is why I always object to being referred to as a "ranker: I don't rank. I rate. I may give two courses the exact same rating which means I would not necessarily rank one above the other. I leave the ranking to Brad and his large and capable staff of mathematicians

Jim

Jim,
Yes, I also rate. Ranking is done by Brad etal. Never meant to imply anything else.
John

[Brad Klein]

John, I don't do the ranking. The numbers do that for us - obviously sequential, based on average rating. I know you knew that.

[Jim Nugent]

Guys, I'm wrong, confused or both.  Brad, perhaps you can clear up some of my questions. 

What exactly does the overall vote (score) each rater gives a course represent?  Is it how that rater ranks the course, given the scale in the GW Rater Handbook?  In that scale courses ranked in the top 5 get 10 points each.  Courses ranked from 6 to 15 get 9 points.

If it's not based on that scale, what exactly does the vote (score) mean?  i.e. what does giving, say, an 8 to a course mean to the rater?  And also, what then is the purpose of the scale in the handbook -- the one that gives points according to rankings?   

[Steve Lapper]

Although incremental, the move of Friars Head ahead of Whistling Straits into the #3 spot finally confirms the ascension of natural vs. manufactured architecture as well as the possibility that GW raters may even know what they are looking at. 

[jim_lewis]

Jim Nugent:

Brad cam correct or confirm my explanation, but I am going to take a stab are responding to your last post regarding the guidelines in the GW rater handbook.

All raters necessarily evaluate courses against some standard. Also, since raters are human, some are naturally inclined to assign liberal ratings while others are more conservative. The guidelines were developed in order to allow raters to have a common standard. All raters are well traveled and have played many of the great courses. A very few have played a high percentage of the great courses. The presumption is that in order to rate a top 100 course, the rater should have played several. Otherwise, a novice might overrate a course simply because it is the best he has ever played.

The guideline is just an aid to help raters have a similar, if not common, standard. A 10 should be assigned to a course only if you think it is deserves to be among the 5 best courses in the category. Naturally, that means that you should not give that rating to more than 5 courses. You are not necessarily comparing it to the other 4 best courses. In fact, it may be the only top 5 course you have ever played. The guidelines are just that, guidelines. They are not hard and fast. It just means that you should be very careful about assigning 8's, 9's, and 10's, because there are not many courses that deserve them. The rater is not being asked to RANK courses.  Personally, I resist the temptation to even try to rank courses. I know that I consider my 9's to be better than my 8's, etc., but that is as far as I go.

[Kyle Casella]

Top 100 Classic #24- The Country Club composite course.

What is the criteria for compiling what I assume is a "best of" routing of an 18+ hole venue?  Seems to me there are other "composite" courses that would rank quite high if you could cherry pick the best holes on the property.

Scott

This is an interesting point. Where would Royal Melbourne fall, for instance, if they played the composite there? I wonder how many raters actually played the composite course at TCC in sequence as opposed to playing the main course and the Primrose (I guess that doesn't really work though because of the combination of 1 & 2 on Primrose). I don't think they play the composite very often over there- once a month or so.

[Greg Tallman]

Just curious, how many GolfWeek panelists have actually played the composite course at TCC Brookline?

[K. Krahenbuhl]

Just curious, how many GolfWeek panelists have actually played the composite course at TCC Brookline?

I played it in their member/guest.  I know of others who have as well.