Here's a recent article (it's long) on new statistical research and the game of golf.
THE SLATE ARTICLE - MONEYGOLF
Much of what is being written about here in the article relates to the book "Moneyball" by Michael Lewis where they discuss how the baseball team the Oakland A's used a different approach to make their team a perennial winner.
For the international readers who don't know a thing about baseball, baseball has always been a very statistically driven game. This is because while baseball is a team sport, it has the unique dynamic of being very individually performance driven because the it's really a series of one-on-one matchups.
Unlike a sport such as Rugby or basketball or football, if you take a great baseball player on one team (and you can see how good they are by their statistics) and trade them to another team, they will still likely perform at a high level. In those other more team-oriented sports, a good player on one team may not perform well on another team because they might not fit into the coach's scheme, work well with other players, etc.
In baseball, there's no salary cap. Meaning that a team can spend as much as they want to on player's salaries. So, the best players in baseball usually command the most money, so the teams that can afford to spend the most money usually get the best players and thus usually win the most games. However, 'Moneyball' is about how the Oakland A's, who cannot afford to spend that much money on players, yet are a perennial winning team.
That's due to the A's using different type of statistics to judge players and not get so bogged down on traditional statistics and traditional scouting reports.
For instance, in baseball the common traditional statistics for hitting are usually batting average (number of hits / at bats), home runs and RBI's. However, in 'Moneyball' they prefer to look at On Base Percentage (number of times the batter gets on base, which includes hits AND walks / at bats) and Slugging Percentage (a percentage that adds points for doubles, triples and home runs hit).
So, let's say we have Player A and Player B.
.320 batting average
30 home runs
.280 batting average
20 home runs
In *traditional* statistics, people would take player A over player B. But, in the 'Moneyball' stats (often referred to as 'Sabremetrics'), their stats may look like this:
.320 batting average
.340 on base percentage
30 home runs
.550 slugging percentage
.280 batting average
.410 on base percentage
20 home runs
.575 slugging percentage
In 'Moneyball' stats, they would actually take player B every time because the on base percentage is higher (.410 vs. .340) and the slugging percentage is higher (.575 vs. .550). Essentially you could draw the conclusion that Player A is better at hitting home runs and hitting singles. But Player B isn't too far behind and hits a lot more doubles, triples and takes a lot more walks...all of which put the team in a better position to win the game.
And that's a big basis of what the Slate's article is about...disregarding the 'traditional' statistics for newer, more accurate statistics.
Are their flaws?
As the Slate describes, the Shotlink technology had some glitches early on which created inaccurate data. And when you're playing for that much money on the PGA Tour, you don't want inaccurate data influencing your decisions.
Also, Moneyball has some holes in it as well. As evidenced by the 2004 Boston Red Sox who hired Bill James, considered *the* Moneyball guru of baseball. James' Moneyball approach favored those high On Base Percentage and Slugging Percentage guys and worried about good defense last. It also didn't care for a 'closer' (a relief pitcher that usually pitches the last inning of the game to 'close out' the victory.
The Red Sox struggled heavily midway throughout the year and were in doubt to make the playoffs. Their defense was killing them in games and they couldn't sure up a win with such a shaky relief pitching staff.
They eventually decided to change course to a more 'traditional baseball approach.' They traded away high OBP and Slugging Percentage Shortstop, Nomar Garciaparra, for the good fielding, poor offensive player in Orlando Cabrera. They also got a formidable closer, traded for good defensive firstbaseman and got a good pinch runner in Dave Roberts.
They basically had all of the offense they needed, so they just needed to get a defense that wouldn't give up runs so easily.
That's part of the problem with these type of statistics, they can lead to erroneous conclusions...or conclusions that need to be tweaked for optimal effect.
But, here's what 'MoneyGolf' people have come up with so far:
It's not that putting doesn't count. It does. But a golfer without a world-class long game simply can't be world class. The importance of power is confirmed by Mark Broadie in a forthcoming paper. Thanks to his shot-value analysis, Broadie is able to isolate particular skills. The areas that have the most influence on a golfer's score, Broadie found, are long-distance tee shots, shots from 200-250 yards, and shots from 150-200 yards. It's these locations on the course—not the greens—where golfers are most able to distinguish themselves from the pack.As I mentioned, sometimes these theories need to be tweaked, but the interesting thing about good statistical data is how it can detect more of the 'truth' about a subject. How many times have we heard 'the key to success on the PGA Tour is being able to hit your fairway woods and long irons?' I would say not very often, if ever.
Of course, it should be known that statistics are about probability, not certainty. Meaning, that there's no statistic (that's not at 100%) that will guarantee an outcome, but it will give you an odds of that certain outcome.
That's one of things that kills me when you say something like in the NFL that 'the team that passes more efficiently will win the game about 80% of the time.' Then you hear somebody point to a game where one team passed less efficiently and still won. It's not that the statistic is 'wrong', it's just that the team happened to defy the odds in that game.
The big problem for the average golfer like myself is that we don't have access to Shotlink data. However, we have to remember that this data is in its infancy stages and as time goes along, there's a good chance that it will get better and there may be something that develops to allow the average golfer to figure out their own strengths and weaknesses.
In a world of probability, you just never know