The greens at the 2004 U.S. Open did not die because the grass gave up. They died because a golf course superintendent named Mark Michaud was managing a surface that was never built to survive what the USGA decided to do to it. The surface was Poa annua — annual bluegrass — a grass most superintendents in the country spend their careers trying to kill. And the one growing on the putting surfaces at Shinnecock Hills when the national championship arrived on Long Island. By Sunday afternoon, play stopped so crews could hand-water the seventh green to keep it alive. Not to improve it. To keep it from dying on television.
You watched that if you were watching golf that year. Retief Goosen putting on a surface the color of a dirt road. Ball struck cleanly, landing on the green, and rolling off into the bunkers because there was nothing left to hold it. Twenty-eight of the 66 players who finished shot 80 or worse in the final round. Nobody broke par. The story everyone told afterward was simple: the greens baked out, the grass died, the USGA let a great golf course get away from them. That story is half true. The half nobody tells you is the half that matters — because it explains why it took 14 years and three separate rebuilds before anyone trusted Shinnecock Hills with an Open again.
The grass was not the villain. The grass was the evidence. Almost every problem you have ever watched a golf course have on television is a maintenance decision wearing the costume of an accident. The dust clouds, the dead green, the playoff that should not have happened. None of it is weather. All of it is a choice made about turf weeks earlier, surfacing on camera at the worst possible moment. Here is what you were actually looking at in 2004.
Why Poa Annua Makes a Beautiful Surface Right Up Until It Doesn’t
Poa annua is not a grass anyone plants on purpose for a championship. It arrives. It invades. It is a shallow-rooted opportunist that germinates in the thin, stressed margins of a putting surface and then takes over. And at a links course exposed to ocean wind on three sides, it had taken over completely at Shinnecock. The reason superintendents fear it is exactly the reason it failed in 2004. Its roots are short, so its access to water is short. It wilts faster than bentgrass, it dies faster under heat and drought stress, and it cannot survive long stretches without frequent shallow watering to keep the surface cool and alive. It makes a beautiful, dense, fast putting surface right up until the moment you stop watering it. And then it collapses faster than almost anything else you could have under your feet.
Shinnecock’s greens in 2004 were a meadow-type stand of Poa and rye — broadleaf, poorly adapted to drought — and held alive through tournament week by hand. Michaud’s crew was out there with hoses because the irrigation system could not be trusted to do it. That system was roughly 20 years old. It threw water in broad arcs from a long distance, and on a calm day that was merely inefficient. Shinnecock does not have calm days. The wind carried a large share of every cycle off the greens and into the rough before it ever landed. So you had a grass that needed precise, frequent water to survive stress, sitting on top of an irrigation system that could not deliver precise, frequent water. That is the machine that was already running when the USGA made its decision.
The Decision That Made the Seventh Green Inevitable
Two weeks before the 2004 Open, USGA officials decided to firm the golf course by withholding fairway irrigation. The field staff running the setup effectively took control of all water from the club. The intent was defensible. They wanted the firm, fast, demanding test a U.S. Open is supposed to be. But they were making that decision about a course whose greens were a drought-intolerant weed sitting on an irrigation system that could not rescue them, in a wind that stole water faster than anyone could replace it. After round one, the watering stopped across much of the course. Overnight wind did the rest. By the final round, the greens were not firm. They were dying.
The seventh green went first, exactly as its shape predicted. The seventh is a Redan — a green that slopes away from the player, running downhill from a high front-right to a low back-left, set at roughly a 45-degree angle to the tee, and it usually plays straight into the wind. A green like that survives only on a thin margin of moisture holding a ball on the slope. Take the water away on a drought-tolerant Poa surface, on a green built to shed anything that lands above the hole, and the margin vanishes there first. The seventh did not get unlucky. It was the green whose very shape punishes a dry, fast surface harder than any other. It was always going to be first.
This is the gap between what you saw and what was happening. You saw fast greens that got too fast. What was happening was a collision between a biology that needed help, a decision that removed it, and infrastructure that was never capable of bridging the difference. Speed was the symptom. The grass was the symptom. The cause was that three different systems had to fail at once — and in 2004, all three did.
Three Tracks to the Fix — and Why None of Them Was the Seed
That is also why it could not be fixed quickly. You cannot reseed your way out of that. A panicked over-reseed the following spring fixes nothing, because none of the three real problems is the seed. The first track was the grass itself. Over the years that followed, under superintendent Jonathan Jennings — who took over in 2010 and built on tree and vista clearing started even earlier — the greens and fairways were converted toward far more bentgrass. Bentgrass is the deeper-rooted, more drought-resilient cool-season surface that a championship test wants underneath it. But here is the precise version, because the loose version is wrong and the loose version is what you will hear everywhere else: they did not switch to bentgrass. They converted toward it. Shinnecock’s greens to this day carry a heavy Poa component. What changed was the balance and the resilience, not the species list. The weed did not get evicted. It got outnumbered.
The second track was the water. The old broadcast irrigation system was rebuilt to concentrate water where it was needed and to stop throwing it into the wind. Spot control instead of long-distance arcs. This is the part that never makes television because a new irrigation grid is invisible the moment the sod goes back over it — but it is the single change that most directly addresses what killed the seventh green. The 2004 crew was hand-watering because the system could not. The rebuilt system meant the next crew would not have to fight that fight by hand.
The USGA Rewrote Its Own Setup Philosophy
The third track did not belong to the golf course at all. It belonged to the USGA. After 2004, the organization rewrote its setup philosophy. New guidelines on how rough is grown in, how water is managed through a championship, how close to the edge the course is pushed and how fast it gets there. The president of the USGA said plainly that they let the course get away from them in the final round — that it was an error in judgment in water management. That admission became the seed of the protocol the organization still uses.
By the time the Open returned to Shinnecock, the setup team was working with a firmness meter called TruFirm, with soil moisture sensors and evaporation monitors, with forecasting that let them manage moisture by measurement instead of by feel. The 2004 decision was made by instinct. The replacement was made by instrument.
Augusta as the Control Group
Augusta National is the control group for everything that went wrong at Shinnecock. Augusta grows A-1 and A-4 creeping bentgrass on its greens through brutal Georgia heat and humidity — a climate that should make those greens impossible. It survives because Augusta engineered the weather out of the equation. Every green at Augusta sits on a sub-air system, the subsurface air technology developed by the club’s long-time grounds director Marsh Benson: a network of pipes that can vacuum moisture out of a green after rain or push cooled air up into the root zone in heat. On the 12th green, subsurface pipes hold the soil temperature in a tight band — roughly 72 to 82 degrees Fahrenheit — no matter what the Georgia afternoon is doing above. Augusta never has to gamble on the weather because Augusta built a machine underneath the grass that makes the weather irrelevant.
Shinnecock had none of that in 2004. No buffer underground, no vacuum, no cooling, no margin. Just exposed Poa, an old sprinkler grid, and the Atlantic wind. So when the USGA made the same kind of aggressive setup call that Augusta’s infrastructure quietly absorbs every April, there was nothing beneath the surface to catch it. Same biology, same demand, wildly different scale and wildly different cost. Augusta solved the problem by spending whatever it took to build a controlled environment. Shinnecock had to solve the same problem the slow way — over 14 years — by rebuilding the turf and the water and the rules around it until the course had enough margin to survive a hard setup on its own.
The Thing the Fix Could Not Reach
The fix worked — and it still was not enough to make the problem disappear. When the Open came back, the agronomy held. The greens did not die. But the course still drew complaints on the weekend because the one variable nobody can rebuild is the architecture. These are old greens designed in an era of slower surfaces, and when you run a modern green speed across a green William Flynn shaped for the 1931 season, you get trouble that has nothing to do with the grass or the water. As one former Open champion put it, the line you cross is old-school architecture meeting modern agronomy.
You can convert the turf. You can rebuild the irrigation. You can rewrite the setup manual. You cannot unbuild the slope of a green that has been there since before the Second World War. So the weed did not wreck a U.S. Open by itself. That is the headline, and the headline is the trap. The Poa was a surface that could not survive a decision on infrastructure that could not rescue it, set up by an organization working on instinct. Fourteen years and three rebuilds later, two of those three things are fixed for good. The grass is deeper, the water is precise, the setup is measured. The third thing — the ground itself — is exactly as old and exactly as unforgiving as it was the day the seventh green died on camera. That is the part still being managed every single time the championship comes back.