OWGR - Week #18

First topic message reminder :

The future's bright, the future's... orange?
Whatever we may feel about his clothes, his choice of colors, his moustache or his god, there's no doubt that this has been a great first win for Rickie Fowler! We'll see what this will do now for his confidence and carreer moving forward. It certainly does pretty good things for his world ranking, with a nice jump up to #24 (not a carreer best though). As for his play-off rivals, we know by now that Rory is returning to #1, after another fine tournament that he will no doubt feel he could've won, while DA Points enjoys a sizable jump up of 34 places to #73, which I believe is a carreer best for him.

Congratulazioni to Francesco Molinari for a very fine win at the Open de Espana, and that is lifting him back inside the top 30 in the world (#29) and is also pushing him up close to the relevant top places in the Ryder Cup tables. Otherwise, Webb Simpson is returning to the top 10 after a fine tournament at home in Charlotte, while Ben Curtis continues his impressive climb up the rankings with another very good finish which should bring him somewhere inside top 120. But from those who are battling these weeks for a place in the top 60 to qualify for the US Open, nobody has been able to make a significant move upwards this week.

Finally, well done to the young Georgia graduate Hudson Swafford who won his first tournament on the NW tour after a brilliant 62 on Sunday. Not quite sure if he is still an amateur or not, but in any case he will be propelled around 750 places (!) as a result of this, somewhere around #430.

The OWGR table after week #18 should look as follows:

1 Rory McIlroy
2 Luke Donald
3 Lee Westwood
4 Bubba Watson
5 Hunter Mahan
6 Steve Stricker
7 Tiger Woods
8 Martin Kaymer
9 Webb Simpson
10 Phil Mickelson
------------------------
11 Justin Rose
12 Adam Scott
13 Louis Oosthuizen
14 Charl Schwartzel
15 Jason Day
16 Matt Kuchar
17 Dustin Johnson
18 Graeme McDowell
19 Bill Haas
20 Jason Dufner
21 Sergio Garcia
22 Keegan Bradley
23 Nick Watney
24 Rickie Fowler
25 Brandt Snedeker
26 Peter Hanson
27 Ian Poulter
28 K.J. Choi
29 Francesco Molinari
30 Mark Wilson
31 Bo Van Pelt
32 Zach Johnson
33 John Senden
34 Thomas Bjorn
35 Bae Sang-moon
36 David Toms
37 Carl Pettersson
38 Simon Dyson
39 Martin Laird
40 Jim Furyk
41 Ernie Els
42 Alvaro Quiros
43 Robert Karlsson
44 Paul Lawrie
45 Aaron Baddeley
46 Anders Hansen
47 Fredrik Jacobson
48 Ben Crane
49 Geoff Ogilvy
50 Paul Casey
-----------------------
51 Kyle Stanley
52 K.T. Kim
53 Nicolas Colsaerts
54 Gonzalo Fdez-Castano
55 Jonathan Byrd
56 Kevin Na
57 Ryo Ishikawa
58 Miguel A Jimenez
59 Robert Rock
60 Y.E. Yang
60 Matteo Manassero
----------------------------
62 Rafael Cabrera Bello
63 Retief Goosen
64 Greg Chalmers
65 Johnson Wagner
66 Branden Grace
67 Charles Howell-III
68 Gary Woodland
69 Ryan Moore
70 George Coetzee
71 Darren Clarke
72 Robert Garrigus
73 D.A. Points
74 Pablo Larrazabal
75 Hiroyuki Fujita
76 Robert Allenby
77 Spencer Levin
78 Sean O'Hair
79 Michael Hoey
80 Alexander Noren
81 Jeff Overton
82 Chez Reavie
83 Joost Luiten
84 Ryan Palmer
85 Vijay Singh
86 Marcus Fraser
87 Richie Ramsay
88 Jaco Van Zyl
89 Padraig Harrington
90 Brian Davis
91 Brendan Jones
92 Stephen Gallacher
93 Rory Sabbatini
94 John Huh
95 John Rollins
96 Michael Thompson
97 Bernd Wiesberger
98 Toru Taniguchi
99 George McNeill
100 Thomas Aiken
------------------------
101 Harrison Frazar
102 Thorbjorn Olesen
103 Jamie Donaldson
...

super_realist wrote:I think you also have to take Woods scoring average with a pinch of salt too as he's hardly played compared to most. Its a bit like saying someone who has a 100% sand save record but has only been in the sand once all year is the best at bunkers. When all it really says is that the one time they went in they were successful.

Understood, but there is a minimum divisor for OWGR (40 or whatever). I don't have a problem with the notion of rewarding those players who consistently play well but less frequently. They are running the risk of volatility (Woods again is possibly the best example in recent times) if they haven't actually played the minimum number of events over the 2 year period.

GPB... Check your IM's. RE: Rendleman Brodie study.

I found an article where they had the B-R rankings for the end of 2010: was it just a one-off or have they published further rankings?

LastDamnation.... Actually it was a "one off". It really wasn't intended so much to be a ranking, but more a study to determine if there was bias in the OWGR. It pretty well determined that there was significant bias against the full time PGA Tour player (regardless of nationality)...

The key sentence from their conclusions: "We find a persistent, large and statistically signicant bias in the OWGR rankings against PGA Tour golfers; a golfer of a given estimated SBSE skill level, or a given Sagarin rank, is likely to be penalized in the OWGR rankings for playing events on the PGA Tour and rewarded for playing elsewhere"

Other portions of their study indicate the average PGA Tour player is being penalized some 37 places in the OWGR rankings at the beginning of the year when the OWGR bias is at it's worst, and 26 places at mid-year when the bias is near it's least.

NOTE: From one of the authors of the study: "... we are not proposing a new OWGR system, at least not yet. Instead, what was reported in Golf World as our system is actually just a performance benchmark against which we are able to estimate potential bias in the OWGR."

For those of you that are interested... Richard Rendleman, one of the authors of the April Golf World article on OWGR bias (and ranking) has kindly provided a paper he and Brodie prepared that contains the underlying basis for their methods and conclusions. It explains in general terms how this mathematical study was conducted, (a good read for everybody, and understandable to anyone with even a basic grasp of the current OWGR). And for those of you with a a more complete understanding of statistical analysis and methods... there's plenty in there for you too.

I also received word this evening that I am free to share the paper with anyone who would like to have a copy. Below find the link to a copy I uploaded on my Google Docs account. Further distribution IS allowed, but only the complete paper in its entirety with NO edits or changes from the original. Also feel free to the share or discuss the information in any forum or communication, but please do so "in context" and with proper credit if the entire document is not to be provided.

EDIT: Below link to the research paper has been changed to the version published directly on Mark Broadie's website.

http://www.columbia.edu/~mnb2/broadie/Assets/owgr_20120507_broadie_rendleman.pdf

Again... thanks to

Mark Broadie, Graduate School of Business, Columbia University

and

Richard J. Rendleman, Jr., Tuck School of Business, Dartmouth College

Thanks for providing the link to the original article Robo, it will be an interesting read.

I know what I will be doing instead of working today, I will just claim I am carrying out some sort of learning exercise.

robo, Broadie's paper is openly available on his website...
Just had a quick browse through the abstract and conclusions and can't say I'm too convinced, but will try to read through details later when I'll have some time. The general 'tone' of those paragraphs sounds to me like coming from someone who knew already what he would like his conclusion to look like before he started the study...

I don't want to prejudge anything, but I would be also a bit concerned (to say the least) when I read that Broadie used to be Vice President at Lehman Brothers and one of their 'experts' in risk management, etc. We know how much these people were aware or connected to the real world and how correct their theories and predictions were. Also, if I remember correctly, these days the Lehman Brothers crowd (or what's left of it) are owned by Barkleys, one of the big PGAT sponsors...

princedracula wrote:robo, Broadie's paper is openly available on his website...

Thanks PD... I wasn't aware they had put the paper up as it was only finalized on May 7. I changed the link directly to Broadie's website above.

princedracula wrote:Just had a quick browse through the abstract and conclusions and can't say I'm too convinced, but will try to read through details later when I'll have some time. The general 'tone' of those paragraphs sounds to me like coming from someone who knew already what he would like his conclusion to look like before he started the study...

PD... I've yet to run simulations of their formulas on my own as I struggle through some of the methods employed (Guess I should have paid more attention to those 3 or 4 statistics classes in college... sigh). But that said... what I DO like about the analysis is the methodology is OPEN and made freely available to peer (and our) review (unlike Sagarin) . That in itself mostly removes any questions that I may have about "methods" or possible "agenda" or "past employment" of the researchers. Now I will fully admit that if PGA Tour executives had promulgated this study on it's own, then I would be MUCH more suspect. That would be almost as suspect as if Euro Tour executives ran the OWGR... :-)

And yes... I will agree they already knew, at least to some of the extent, the outcome of their analysis would show bias in the OWGR against the PGA Tour player. Well duh... IMO, anyone who has taken the time to fully understand and ponder the OWGR methodology would already understand the ratings are biased to favor players who are successful on the "lesser tours", or even those that are successful in "lesser" events on the "major" tours. I mean how many times have we seen some "high OWGR ranked" player from some "minor" tour, come out and fail to perform against the big boys and prove the point. For instance, nobody is EVER going to convince me that Yuta Ikeda was ever a top-40 player in the world, at any point in his career (not that he may not be at some point). So I can't fault Broadie and Rendleman for having a "pre determined understanding" that OWGR bias does exist. What this study does (and Sagarin does to some extent as well) is give us at least some idea just how significant the bias is.

But I guess the bottom line is... To this point all this study was intended to do was attempt to quantify the extent of OWGR bias. Even they freely admit that it is not necessarily a ranking system. IMO the problem is that ANY ranking system becomes Subjective the moment it tries to answer the question: "What constitutes the best player".

IMO this study (and Sagarin) do the close same thing... they identify, in rank order, which player is likely to perform better in a group of "random" head-to-head encounters. But that in itself does not necessarily identify which one is the better player. Take for example...

Player A regularly finishes in the top-5 or 10 in tournaments... and wins a regular PGAT or ET event once or twice a year...

Player B may finish top-10 far less often, but a lot of top-25's, or even 60th places and even a MC from time to time. But when he plays well, he wins 3 times a year including at least one, maybe even two big events like a Major, TPC, or WGC.

In Sagarin and Broadie... Player A will almost certainly be rated the best player, as he will likely have finished ahead of Player B way more often in their common tournaments. And week in and week out I would probably pick player A to finish higher than player B too. BUT... I think most people would choose Player B as the better player, and select player B as the most likely player to WIN any given tournament.

So as soon as we say "it feels like" player B should be rated better than Player A... we start subjectively choosing which "modifiers" we will apply to the data to obtain the "desired" result. And after that... when we form federations and want to gain a "consensus of ranking acceptance" then we further "monkey" with the methodology in an attempt to satisfy the sensibilities (and politics) of all. Out of such efforts... things like event minimum values, equal event minimums between un-equal tours, and home tour SOF values are born.

Obvious that there'll be plenty of movement in a big event, but there are a lot of missed cuts by players in high positions and borderline 40's, 50's, 60's who haven't done much recently and whose pace of fall will just accelerate:
Day
Stricker
Vijay
Yang
Crane
Goosen
Casey
Woodland
Snedeker
Badds
Stanley
McDowell
Simpson
Freddie J

= Turnover!

Looking at Casey's situation (at #50) and his prospects of being in the top 60 at the deadline.

He has 0.3 lead on #60 (Yang and Manassero) so he is not really in danger of 'slipping' behind through bad play (or no play). I think he going to be in the top 60

Probably only two players will pass him this week (Na and Byrd have the best chance right now).

BTW, Ryo Ishikawa missed the cut in Japan this week

some quick OWGR calculations (maybe PD can confirm)

Depending on how others perform of course.

Na moves into the top 30 with a win?
Kuchar moves to the #5 spot with a win?
Fowler can move up to #7 with a win?

As it stands now, Na moves to #24, Kooch to #6, Fowler to #18...

The #50-60 area looks as follows...

.....
45 Aaron Baddeley
46 Anders Hansen
47 Robert Karlsson
48 Fredrik Jacobson
49 Jonathan Byrd
50 Geoff Ogilvy
--------------------------
51 Ben Crane
52 Nicolas Colsaerts
53 Kyle Stanley
54 Paul Casey
55 Kim Kyung-Tae
56 Gonzalo Fdez-C.
57 Johnson Wagner
58 Miguel Jimenez
59 Ryo Ishikawa
60 Robert Rock
------------------------------------
61 Matteo Manassero
62 Rafael Cabrerra-Bello
63 Y.E. Yang
64 Greg Chalmers
65 Branden Grace
66 Retief Goosen
67 Charles Howell III
68 Ryan Moore
69 Gary Woodland
70 Darren Clarke
...

Contrasting Swedish fortunes:
Jacobson misses the cut and goes up two places.
Karlsson makes it and may drop five! Get on yer bike Robert!!

How d'you like 20 putts from Del Moral? Always used to follow him via goldwolf's game.

Thanks for that comment kwini. It's not that bad for the two Swedes, forgot to count Jacobson as playing this week, so there's the ranks updated above...

Del Moral putting was great this week, but it's been pretty poor until now this season, so maybe he's found the problem. One of his main issues remains his accuracy from the tee... Unfortunately he seems to have found some bigger problems on top of that the last couple of holes.....

Maybe jp could tell us who is Harto....

A breakthrough for Lagergren or Santos this week?

Very solid round from Santos and you can't see him losing it now, nice home win for him! First Portugal-born player to win on the ET I think.

Nice earnings at CT level for Harto though, as I recall he won both as an amateur and a pro on that tour, would expect him to make the breakthrough sooner or later. Shame for Del Moral.

There it is for Santos. If he turns out to be a top player I would have thought that would be a big positive for European golf.

Disappointing for Wilson, and very disappointing for Dougherty.

Meanwhile, looks like a second place for Peter Baker in Brittany.

Great maiden/home win for Santos, hopefully it will turn out a top player...

Casey out of Volvo MP next week, replaced by Karlsson.
Well, it's Volvo...