All podcasts / People I (Mostly) Admire / Summary

23. Greg Norman & Mark Broadie: Why Golf Beats an Orgasm and Why Data Beats Everything

2021-04-09 - source

Steve Levitt (host)Greg NormanMark Broadie

Key insights

Counting statistics (putts per round, fairways hit) are bad performance measures because they treat wildly different outcomes as identical.

Broadie argues a 3-foot putt made and a 60-foot putt made both count as "one putt," and a ball just off the fairway counts the same as a ball out of bounds, even though the underlying performances are nothing alike. This was the core flaw strokes gained was built to fix.

data-driven-measurement

Strokes gained works by converting every shot type into the same unit: strokes relative to a scoring benchmark.

For any distance and lie, there's an average number of strokes it takes tour pros to hole out. A shot that leaves you closer to the hole than that benchmark predicts is worth positive strokes gained; a shot that leaves you further away is negative. This lets a drive, an approach, and a putt all be compared and summed on one scale.

data-driven-measurement

Ball-striking, not putting, is what separates elite pros from average tour pros.

Strokes gained analysis overturned the folk wisdom "you drive for show and putt for dough." Broadie found that driving and approach shots (ball-striking) account for roughly two-thirds of the scoring gap between top-10 players and the average PGA Tour pro, while short game and putting account for only about one-third, even though the two categories involve a similar number of shots per round.

data-driven-measurement

Extra driving distance is a bigger net win than conventional wisdom suggests, even accounting for lost accuracy.

A player who is 20 yards longer than tour average gains about 0.1 strokes per drive, or 1.4 strokes across 14 drives in a round. That costs roughly 0.3 strokes from hitting about one fewer fairway per round, for a net gain of about 1.1 strokes per round (over 4 strokes across a 4-round tournament). Distance is also more reliable than putting because it doesn't fluctuate day to day.

data-driven-measurement

The biggest score-killer for amateur golfers is under-respecting hazards, not poor swing mechanics.

Broadie's data shows amateurs hit 10-30% of certain tee shots out of bounds or into hazards when a safer line was available, and out-of-bounds carries a stroke-and-distance penalty (2 strokes for one swing), making it one of the largest avoidable score contributors.

risk-and-strategy-under-uncertainty

A randomized field experiment found that amateurs playing the "safe" recovery shot often did worse than amateurs told to attempt the heroic shot.

Levitt, Luke Donald, and Pat Goss randomized amateur golfers into playing safe vs. attempting an near-impossible shot from the woods. About a third of amateurs failed to execute the "safe" chip-out cleanly, and the heroic-shot group ended up roughly half a stroke better on average, because their failed heroic attempts still advanced the ball further than a botched layup. The intended book on optimal on-course strategy for amateurs was abandoned because strategy barely mattered relative to execution ability.

risk-and-strategy-under-uncertainty

For amateurs, chasing memorable heroic shots may be a more rational objective than minimizing average score.

Levitt argues most amateurs don't actually care about the difference between shooting 93 and 95; what they remember and talk about later is the one spectacular shot. He proposes amateurs are (unconsciously) optimizing for memorable moments, not stroke totals, and that this isn't irrational so much as a different, legitimate objective function than the one golf's scoring system rewards.

risk-and-strategy-under-uncertainty

A points-based scoring system (15/8/3/1/0 for eagle/birdie/par/bogey/worse) would better align golf's incentives with what actually makes it fun.

Under Levitt's proposed system, a bogey plus a birdie (9 points) beats two pars (6 points), rewarding aggressive play. It also lets one great hole erase the memory of many bad ones (e.g., 8 triple-bogeys plus 1 birdie still nets a respectable 8 points), and lets players pick up on disaster holes to speed play, without requiring any change to courses or equipment.

risk-and-strategy-under-uncertainty

Norman attributes elite skill acquisition to compartmentalized, targeted practice rather than generic repetition.

He isolated specific flaws (e.g., a poor 5-iron shot from a hanging lie during a tournament) and would go find that exact lie on the range and hit 50-150 balls until it was fixed, splitting practice roughly evenly across scoring irons, long game, and short game so no area was neglected.

deliberate-practice

Norman built power first and control second, rather than the more common approach of building accuracy and then adding power.

His coach told him to "hit it as hard as you possibly can, and then we'll figure out how to hit it straight," so his power was preloaded and he typically swung at only about 85% effort, scaling back from a higher ceiling rather than trying to add power to an already-maxed swing.

deliberate-practice

Handling defeat gracefully, for Norman, came from attributing losses to opponents' merit or his own execution rather than from suppressing competitiveness.

Norman explicitly separates the sport from his identity ("golf is just a game, you can't live by it and you can't die by it") and credits his father's advice to always tell the truth when asked why a shot failed, rather than deflecting or making excuses, as the root of his post-loss composure, including his famous 1996 Masters collapse against Nick Faldo.

grace-under-failure

Strokes gained succeeded at rapid adoption largely through a blind-test buy-in strategy rather than persuasion.

The PGA Tour showed golf influencers (players, writers, coaches) two anonymized putting rankings, A (putts per round) and B (strokes gained putting), and let them independently judge which one better reflected real performance. People consistently picked ranking B without knowing what it was, which let the Tour adopt strokes gained without ever having to explain or defend the underlying math.

data-driven-measurement

Media referenced

Freakonomics Radio - podcast - Sister show; Stephen Dubner appears in the cold open bantering with Levitt about his golf obsession.
No Stupid Questions - podcast - Mentioned as part of the Freakonomics Radio Network in the outro.
Sudhir Breaks the Internet - podcast - Mentioned as an upcoming show on the Freakonomics Radio Network in the outro.

Companies

PGA Tour - Broadie spent years trying to get access to PGA Tour shot data before the Tour adopted his strokes-gained metric as an official statistic starting around 2008.
Reebok - Norman references being a brand ambassador for Reebok as one stage in his evolution from athlete to "living brand."

Techniques and frameworks

Strokes gained - Broadie's core metric: measures every shot (drive, approach, putt) against a benchmark of average strokes-to-holeout from that distance/lie, converting all shot types into one comparable unit.
Compartmentalized deliberate practice - Norman's practice method: isolate a specific flaw (e.g., a bad lie or a bad club selection from a real round) and repeat it in practice, sometimes 50-150 balls at a time, until it's fixed, rather than generic range practice.
Randomized on-course field experiments - Levitt, with golfer Luke Donald and coach Pat Goss, rented golf courses and randomly assigned amateur golfers to "play it safe" vs. "go for the heroic shot" from identical bad lies to test golf-strategy hypotheses.
Points-based alternative scoring system - Levitt's proposal to replace stroke counting with points (15 for eagle, 8 for birdie, 3 for par, 1 for bogey, 0 for double-bogey-or-worse) to reward risk-taking and let bad holes fade from memory.

Summary

Steve Levitt, an admitted golf obsessive who spent years chasing a quixotic dream of qualifying for the Champions Tour, interviews two very different giants of the sport: Greg Norman, the former world number-one golfer known as "The Shark," and Mark Broadie, the Columbia Business School professor who invented the strokes-gained statistic that transformed how golf performance is measured. The episode braids two threads: what elite physical mastery and grace under repeated failure look like from the inside, and how a rigorous but intuitive metric can displace decades of bad conventional wisdom almost overnight.

Norman describes a punishing, self-directed practice regimen (an estimated 5 million balls hit over his career) built around compartmentalizing flaws and drilling them relentlessly, including deliberately bad lies, one-handed shots, and practicing in rain and wind. He credits a coach's early advice to build power first and control it later, rather than the more conventional path of building accuracy and adding power afterward. The conversation's most memorable passages concern Norman's repeated, heartbreaking near-misses at the Masters (three second-place and three third-place finishes, including blowing a six-shot final-round lead in 1996) and his explanation for why he handled those losses with famous grace: he separates his identity from the outcome of any single tournament and, per his father's advice, always tells the truth about why a shot failed rather than making excuses.

Broadie explains how he built strokes gained after realizing the PGA Tour's existing counting statistics (putts per round, fairways hit) were meaningless because they treated very different outcomes as identical. Strokes gained instead measures every shot, drive, approach, or putt, against a benchmark of how many strokes it takes an average pro to hole out from that position, converting all shot types into one comparable unit. The metric overturned the old saying "you drive for show and putt for dough": ball-striking (driving and approach shots) actually accounts for roughly two-thirds of the scoring gap between elite and average tour pros, with short game and putting responsible for only about a third. He also walks through a concrete, favorite finding on the distance-accuracy trade-off: 20 extra yards of driving distance nets a player about 1.1 strokes per round even after accounting for the accuracy lost, because distance is a repeatable weapon in a way that hot-and-cold putting isn't.

A particularly striking segment covers a field experiment Levitt ran with world number-one Luke Donald and his coach Pat Goss: randomizing amateur golfers into "play it safe" versus "attempt a nearly impossible heroic shot" from bad lies in the woods. Contrary to their hypothesis, the heroic-shot group scored about half a stroke better on average, because amateurs frequently botched even the "safe" recovery shot. The finding upended Levitt's planned book on optimal amateur strategy and led him to a broader, half-serious conclusion: since most amateurs don't actually care about the gap between a 93 and a 95, and mostly remember the one spectacular shot from a round, golf might be more fun, and more honestly aligned with what people want, under a points-based scoring system that rewards risk-taking (15 points for an eagle down to 0 for a double bogey or worse) rather than the standard stroke count that punishes bad holes disproportionately.

The episode also surfaces a smaller but sharp lesson in how good ideas actually get adopted: rather than trying to explain the math behind strokes gained to skeptical golf insiders, the PGA Tour ran a blind test, showing influencers two anonymized putting rankings and asking which one seemed more accurate. People consistently and independently picked the strokes-gained ranking without knowing what it was, which let the Tour roll it out as an official statistic without ever needing buy-in on the underlying methodology.

Notable Quotes

"Hitting the perfect golf shot is better than having an orgasm. Right? I mean, it really is a great feeling." - Greg Norman

"It's sport. Really it's as simple as that. Golf is just a game. You can't live by it and you can't die by it." - Greg Norman

"A putt that sunk from three feet counts the same as a putt that sunk from 60 feet. They're both one-putt, but they're very different performances." - Mark Broadie

"I think doing golf analytics has exceeded my wildest expectations as far as the impact that it's had." - Mark Broadie

"It was really shocking to me that no matter what setup we did, we basically couldn't find a lot of examples where strategy mattered. It was really discouraging." - Steve Levitt