Four Golf Stats We'd Like to See

Earlier this week, while the rest of the golf world was at the Westchester airport watching Tour execs throw rose petals at Phil Mickelson’s feet, I was sitting at my desk thinking, why not take a break from the FedEx hubub and go off-topic for a few paragraphs?

The unhelpfulness of golf statistics has bugged me all year. My irritation began in January, during the Sony, with the Golf Channel’s WinZone stat telling me what a brilliant Sunday closer Charles Howell was, while my memory screamed otherwise. Then came the week when I noticed that the Tour's Total Driving stat was compiled by surprisingly simple process of adding together a player's ranks in Driving Distance and Driving Accuracy. Then there was the spell when the accuracy stats said Tiger was driving the ball wildly, but he himself insisted he was driving it well.

Funny thing is, this was supposed to be the golden age of golf statistics. ShotLink, now about five years old, was supposed to be revolutionary, giving us more numbers than we knew what to do with. Sadly, that prediction was right. We have all the data we could hope for, but we still can't put it to good use.

The problem is that we're not asking the data the right questions. Golf stats, in other words, still await their Einstein (or at least their Bill James), the kind of person who can not only crunch the numbers, but figure out how to make them tell us instructive things about the game.

In the hopes of encouraging golf's first Sabermetrician to step forward, I hereby offer four new golf statistics I’d like to see.

1. Everybody knows that the Tour's putting stats are unreliable. Putts Per Round is compromised by guys who miss a lot of greens but have great short games, and never have anything longer than five feet. Putts Per GIR is greatly affected by how close a player tends to hit it. ShotLink has given us percentages from certain distances, which has been great. Better would be something Peter Kostis (and perhaps others) proposed a long time ago, but only now, with ShotLink, is practical. Let's call it Average Length of Putt Holed (or "ALPH"-- cute, no?). The premise is that better putters tend to hole out from greater distances; poorer putters wind up most often holing out from a foot or two after missing the putt that really mattered. How the stat would work: take the total distance of putts holed, and divide by the number of holes played.

2. ShotLink also ought to make possible a better Total Driving Stat by incorporating its Fairway Proximity numbers (which, through GPS data, measure the distance by which a player misses the fairway when he misses it). Tiger, of course, loves the Fairway Proximity stat-- it allows him to insist that his accuracy off the tee is better than it looks. And he has a point, at least insofar as relative wildness should be part of any measure of who drives the ball best.

How to improve the current stat, however, would make for a long conversation. To begin with, one might start by using ShotLink to compute an average miss from an imaginary fairway centerline (instead of just using hit fairways and missed fairways). It would be crucial, however, to build in a way to reward distance, so that the guy who usually missed the centerline by 2 yards at 305 would be ranked higher than the guy who missed by 2 yards at 285.

(It turns out, by the way, that I'm not the only person unhappy with current Tour Total Driving stat. Three Northeastern University professors recently published their own overhaul. I won't go into it too deeply, but two things trouble me about their alternative stat: (a) its best driver, Charles Warren, is also the current Tour stat's best driver, indicating, say, that the Tour's stat may not be so bad after all; (b) when there are radical differences in rank, they're easily attributable to a dimunition of importance in driving accuracy (Bubba Watson, who hits 55.34% of his fairways, goes from 82nd to 2nd; J.B. Holmes, at 56.37%, goes from 61st to 3rd).)

3. My Sunday Pressure Index (another clever acronym: "SPI"!) would provide the greatest mathematical and creative challenge, but perhaps the greatest reward. If you look closely at the Golf Channel's WinZone statistic (which I'm told is still cited on their Thursday-Friday broadcasts), the first thing you notice (at least if you're me) is that it doesn't make any distinction between performance on different days. To me that makes no sense, if you want an indication of the likelihood someone will win. If you watch any golf at all you know that some of the best players on Tour on Thursday are some of the worst players on Tour on Sunday. How the new stat would work, generally speaking, is this. First, isolate the way a player's leaderboard standing changes over every Sunday he's played on Tour, but only when he's in contention. (This stat shouldn't care if you go from 70th to 50th). Then develop a point system by which you're awarded for upward moves not just on the basis of positions climbed, but also, geometrically, for how high you eventually get. (In this way, a move from 3rd into 1st is rewarded more than a move from 4th into 2nd.) Conversely, you would be punished for moving backward on Sunday, the more so if you're in the last couple of groups. You wouldn't need ShotLink for this, but you would need to dump an awful lot of data from the Tour's record books into your computer.

4. Same thing with the kind of player consistency index I've mentioned here before. A few months ago I tried to illustrate Tiger's consistency at majors with a preliminary "Majors Consistency Index," using standard deviation to show that, at majors, there was less variance in Tiger's day-to-day scores than in Jack Nicklaus's. Now, I'm not sure that stat's worth keeping for every player. (In fact, I'm pretty sure it's not.) Much better would be stats that tracked every player's consistency (a) from round to round, and (b) from week to week. Naturally such stats would judge consistency and what we call streakiness. Also, by indicating especially steady periods, they would also be pointing out for us the periods most representative of a player's true talents (since a thing's nature is best judged when it's at equilibrium).

Comments

[al]

Chris
these are great. Anyone at Golflink or TGC picking up these and running with them?

Maybe for all the tourneys after the fedex cup they could test them out to see how interested viewers are. I know I would be.

[Jucad Golf Trolley]

Mini golf games are usually developed for mobile phones. PC versions of these games are also available. These games have the enthusiasm, skill and pleasure of playing real time golf to mobile phones. Mini golf games enable players to develop new and requires different skills in the game of golf.