January 28, 2014

How We Ranked The 100 Best Modern Players

The ranking's creator, Brett Avery, talks through how he crafted our in-depth list

Nicklaus, Duval, Woods, Els, Singh, Love III, Floyd, McIlroy, Mickelson, Faldo (from left to right)

Lets start at the obvious: why did we decide rank the 100 best modern players?

When Jaime Diaz, the magazine's editor, proposed this project in early 2013, he said he was intrigued by the fact that no one had ever tried to rank PGA Tour players of this era. Everyone concentrated on career and all-time rankings, but not the best of the moment.

Ultimately, like college basketball and football polls, a major part of it is trying to place order on an unwieldy universe. Everyone following the PGA Tour knows the season money list, the playoff points list and the Official World Golf Ranking. They tell us something, but the first two reflect partial seasons and the World Ranking goes back 104 weeks. We were looking for something between those spans.

Throughout the list you mention the formula that you employed. In layman's terms, what is this formula?

The cornerstone of the formula rests on something Jaime kept describing during that first meeting: "How good was your good?" In a way the tour's all-time victory list does that, but it's a single metric and it doesn't account for how players reached those numbers. Take two guys with 10 victories. They're equals in wins, but if one does it in 100 starts he's a star and if the other does it in 750 he's had a tidy career.

"How good was your good?" is determined by who you beat and who beat you. Because most players never played in the same tournaments, it's tough to equitably compare a guy in the 1980s to today's players. What you can do, though, is measure performance in a season and then translate those grades into a number quantifying a career. All those 47,000-plus individual tournament starts and more than 110,000 rounds were divided into 34 yearly pools and then crunched from there.

The formula itself is fairly simple. I looked at the 1,945 individual seasons two ways. The first is called "performance." It's how many times you played in a year and how you finished in those starts. The fewer starts, the more weight each carries. The more starts, the less weight. There was a sliding scale of points for everything from wins down to 25th and a small deduction for missing a cut. Everything else -- making the cut but placing outside the top 25, the background noise -- counted for zero. We gave a 50-percent bonus for the majors and 20 percent for the Players after it moved to TPC Sawgrass. That part was a matter of adding up every player's points for a season.

The second part is called "versus peers," because it's a comparison of every eligible player in a season. There's a par every day on tour, the field scoring average. We calculated a player's cumulative field average for every round and a separate calculation for final rounds, including fifth rounds in 90-hole events. Those were our proxies for strength of schedule, because tournaments with the highest scoring averages, like the majors, tend to have the strongest fields and the tournaments with low scoring averages tend to have the weakest. We then calculated how much cumulatively more or less the player's scores stood to their all-round and final-round field averages. It's like an over par-under par score, expressed to three decimals.

We then indexed those four versus peer numbers. The top guy in a category received a score of 100, the lowest guy a 0. That's about as unflinching a way to separate these guys as you can imagine. And when we combined the versus peers number with the base score from the performance finishes, we had a season score. From there we counted the number of seasons a guy was eligible, totaled up the points and divided for his average.

I tried to wind up with a nice, round number for the average season score and thought 4 would be good, because that's par on most holes. All those seasons wound up with an average of 3.942, which isn't bad. I wanted to throw as wide a net as possible, so the bar was fairly low: three official wins since 1980 or two wins if at least one was a major. That brought 178 guys, which was a large enough pool to feel we could identify 100 players with solid records.

And how did you decide who you would enter into this formula?

I wanted to throw as wide a net as possible, so the bar was fairly low: three official wins since 1980 or two wins if at least one was a major. That brought 178 guys, which was a large enough pool to feel we could identify 100 players with solid records.

Why did we only go back as far as 1980?

Golf used to have three lists: money, wins and majors. Period. In 1980 the tour created a staff to track basic stats like driving distance, greens in regulation and putting that they hand-wrote onto a form. Now we have ShotLink data, where every shot is laser-measured to the inch, and all kinds of intricate stats. But at the center of the tour database is a record of every round from 1980 onward, which seemed like a good place to start this modern era.

Are there any surprises, in your mind, about how the list finished?

I really didn't concern myself with value judgments on any particular player while crunching the numbers. The blessing was that the No. 1 player was clearly identified, because it meant that there wouldn't be a situation where one weighting of the metrics would mean Player A was the best, and another weighting meant Player B, and perhaps a third weighting formula would mean Player C. This way I could concentrate on the numbers, especially in the middle of the pool of candidates and those guys who didn't make it.

Now, there are guys who I was curious about where they would finish. Payne Stewart, fabulous 1999, wins the U.S. Open at Pinehurst and then dies in a plane crash. Seve Ballesteros, such a heavy concentration of majors in so few tournaments on tour. Wayne Levi, a journeyman pro who kept a low profile yet won 12 times. Nicklaus, for obvious reasons, and then the young guys: Rory McIlroy and Sergio Garcia.

My curiosity now is how others will see the list. And I'm especially curious about whether any players will get in touch, either the ones in the Modern 100 or the ones who didn't make it.

With Greg Norman finishing at No. 2, despite ending his career with less majors than some of the guys further down, is it fair to say the list weighs heavily other factors aside from major wins?

Yes. Wins are like toggle switches, one guy a week gets an "on" and 143 guys get an "off." It's varying degrees of losing. But Norman was in contention so many times, we forget that because we've seen those clips of him in '96 getting that hug from Nick Faldo at the Masters or Bob Tway holing out the bunker shot at the '86 PGA or Larry Mize in the '87 Masters. Melodramatic footage is eyeball candy, but the fact is he was in the picture. Again and again and again. And remember, most of this happened in majors, which carry a 50-percent bonus. He had the highest point total in our formula six times from '86 to '95, with only two majors.

A lot of players who won before 1980 seem lower on the list than I expected. Is that unfair?

Depends on how you define unfair. Nothing before 1980 counted, which explains why a player of Johnny Miller's caliber is 63rd, or Hale Irwin is 35th, or Hubert Green didn't even qualify. The Modern 100 still reflects enough of their play that Irwin and Miller, and Nicklaus and Watson, and a bunch of other players, make it. But if you think about it, there's an unfairness, if you will, at the other end of the ranking because the players eligible today haven't finished their careers. You could also say they're at a disadvantage of sorts to the players who have their entire careers in consideration, like Norman or Singh or Els. We have no idea where the still-active players will end up once they stop winning. Someone in the second 50 could wind up in the top 10. And someone in the first 50 could slide out of the 100. That's what will make this interesting to watch.

Did you go back and figure out how the Modern 100 would have looked in earlier years?

It will be exciting to see how the Modern 100 develops from here. This moment is a lot like the late 1980s, which had a lot of volatility. Nicklaus and Watson had stopped winning, Norman was emerging but he wasn't winning in bunches. The two big questions were who would become the next Nicklaus and who was the best player without a major. A lot of guys had big single years, such as Bob Tway's four wins in '86, but the game went about a decade with a lot of players earning a lot of wins. And then there was Woods, and he sucked up a lot of oxygen.

Same scenario today as in the '80s: Els, Singh, Mickelson, Woods, all coming to the end of their victories. Who wins from here? If you enjoy watching the game this next stretch could be highly entertaining. If this is indeed a repeat of the late '80s and early '90s it could get wild, and this will have been a prescient moment to establish a Modern 100.