Irish Open 2026 at Doonbeg: Rory McIlroy, Security Fences and the Atlantic Wind Variable

Doonbeg, an afternoon of westerly wind

Three layers of steel fencing run along the dirt path leading to the 1st hole. Behind the second layer sits a temporary checkpoint, a green canopy, two security staff standing with their backs to the sea. Further out, where the 4th fairway reaches toward Doughmore Bay, the Irish flag and the American flag are mounted on the same pole, and the wind bends both toward the south-east with a frequency that barely changes across Tuesday afternoon.

What strikes you first at Trump International Golf Links Doonbeg is not a golf course. It is a logistics system. Screening gates, restricted corridors, a separately fenced media area, spectator routes compressed into a single axis running alongside the 1st, 9th and 18th holes. A week of golf in County Clare, on Ireland's west coast, is being run to the standards of a state-level event.

And at the centre of it all, a 37-year-old man arrives as defending Irish Open champion, introduced on every tournament signboard with exactly one sentence: the greatest Irish golfer of his generation.

I logged the moment. Not because of the fences. But because of a data question I had carried with me through three years of working with golf data series: when a tournament is surrounded by so much that cannot be measured, which part of the outcome can still be measured?

Data is never in a hurry; it simply waits for someone who knows how to read it.

---

The course, and what we actually know

Doonbeg is a links course on the Atlantic, designed in the classic style, opened in the early 2000s, and later owned by Donald Trump. Its architecture carries the full signature of a true links: sand beneath, fine fescue, fairways that are firm and fast when dry, deep and punishing bunkers, heavily contoured greens, and most importantly, almost no wind breaks across the majority of the holes.

These features are not decoration. They shape the entire scoring calculus of the tournament.

On a dry links course, the ball runs after landing. That fundamentally changes the effective distance of every shot. A 300-yard drive on a parkland course might stop at 305 yards. The same shot at Doonbeg, with firm fairways and a downwind, can roll another 40 to 60 yards and finish in rough if the landing angle is wrong. Or it can settle 60 yards from the green with a favourable run. Same swing, two completely different outcomes, and the deciding variable sits in the turf, not in the player.

I mention this because it explains a paradox common to Irish Opens staged on links courses. The leaderboard shows scores. It does not show ball flight. And in links golf, the gap between those two things is wider than on any other course type.

In Nha Trang, where I work, we do not have links courses. We have sea wind, high humidity, mornings when the ball flies completely differently from the afternoon. But our courses are still tropical, the ball stops quickly, and everything happens within a narrow band of variables. Doonbeg is the opposite. That is why I track the Irish Open more closely than any other event on the DP World Tour calendar, despite the absence of any Vietnamese player in the field.

---

The data problem: what we are missing

One thing should be stated plainly from the outset, and I want to say it as someone who audits numbers rather than as a commentator: this week, almost every data point cited by media is outcome data, not process data.

Number of titles. Number of wins. Past Irish Open results. Rankings. Those numbers are real, verifiable, and entirely useless for predicting what happens over the next 72 holes.

What we need is Strokes Gained. Specifically four segments: Off the Tee, Approach, Around the Green, and Putting. On the PGA Tour, ShotLink provides these at shot level, allowing skill to be separated from luck with high reliability.

The DP World Tour does not operate an equivalent system at every hole of every event. At the Irish Open we get a thinner dataset: average driving distance, fairways hit, greens in regulation, putts per round, sand save percentage. These are macro-level metrics, and they carry a fatal weakness: they cannot distinguish a player performing well on an easy course from one performing well on a hard course.

At Doonbeg this week, the course is hard. That means the entire macro dataset will be compressed, and the differences between players will live in details the stat sheet does not capture.

Audiences applaud to emotion, but data hears a different rhythm.

I have spent most of my preparation for this week doing something I consider more useful than any ranking table: reconstructing Doonbeg's terrain profile by wind direction, and cross-referencing it against the technical profiles of the leading contenders.

---

The wind profile: golf's most underrated variable

Across three years of tracking golf, I have learned something that official stat tables almost never express. Wind is not a playing condition. Wind is an independent variable capable of inverting an entire leaderboard.

At Doonbeg, the prevailing wind blows from the west and south-west into shore. That creates an asymmetric structure: the opening nine is mostly into the wind, the closing nine mostly downwind, or the reverse, depending on how the organisers set the routing in each round. This asymmetry matters more than absolute wind speed.

A player hitting into a 30 km/h wind for six consecutive holes will expend far more strategically than one hitting downwind at the same speed. A shot into the wind forces a lower trajectory, more spin, less roll. A shot downwind does the opposite, and carries the risk of flying too far, landing in a bunker or running through a green.

This creates a form of tension that macro data cannot capture: the player who controls trajectory better gains an advantage that compounds hole by hole. Over 72 holes, that advantage can reach four to six shots. But it never appears in any statistical line. It only appears in ball position on the fairway.

And here is the part I want to emphasise: in links golf, the most important metric is not driving distance but the standard deviation of driving distance.

A player averaging 295 yards with a 12-yard standard deviation is far more valuable than one averaging 310 yards with a 25-yard standard deviation. On parkland, that difference gets absorbed by thick rough that can still be recovered from. On a links course in strong wind, a 25-yard miss can put the ball into tall fescue or a bunker, and from there saving par becomes a probability problem.

I have tested this rule against my own data. Across Irish Opens and Open Championships staged on links courses within the period I track, players finishing inside the top 10 generally show a lower standard deviation of driving distance than those outside the top 20, even where their average distance is shorter. This is an observation from my dataset, not a conclusion from published research, and I state it with that limitation.

---

Rory McIlroy and the home-course paradox

McIlroy enters this week as defending champion. In the file I hold, his Irish Open record includes one title, a runner-up finish two years ago, and six major championships across his career.

That is a handsome record. It is also a record with problems.

The first problem is methodological. The label "defending champion" is time-weighted toward the most recent result. It assumes recent form has higher predictive value than distant form. That holds in many sports. But in golf, where a tournament spans 72 holes and course variables change completely week to week, the correlation between last year's title and this year's result is far lower than public intuition suggests.

And here is a specific detail: last year's title was not won on this course. That win came at a different venue. Doonbeg is a coastal links with a completely different wind structure. The defending champion label carries emotion, but it does not carry transferable data.

This leads to a conclusion I regard as the core of the whole week: last year's title provides information about general winning ability, but provides no information about winning ability on a specific links course.

The second problem concerns the home-nation factor. This is where I frequently disagree with mainstream coverage.

The prevailing assumption is that playing at home confers a clear advantage. Commentators point to crowd support, familiarity with weather conditions, psychological comfort. All of it is intuitively reasonable.

But when you look at the data, home advantage in individual golf is far smaller than in team sports. In football, home advantage was once measured at close to a 46% win rate under conditions with crowds, falling to around 34% when stadiums were empty during the pandemic. That is a large, measurable margin, and I once devoted a three-thousand-word analysis to it.

In golf, that margin compresses. Golf is a sport in which every shot takes place in relative silence. Crowds do not physically affect the ball. Crowd pressure is cognitive pressure, not physical pressure. And cognitive pressure, in my data, tends to affect putting more than driving.

That is a very specific point. If you want to find traces of home pressure at Doonbeg this week, do not look at the drives. Look at putts from 1.5 to 2.5 metres in the final two rounds, once spectators have massed around the greens nearest the clubhouse.

---

What I am actually tracking

I do not have ShotLink for this event. I have a leaderboard, a course map, a wind forecast, and a spreadsheet. That is the entire toolkit. But it is enough to build three tracking layers.

Layer one: score distribution by hole.

On a links course, the gap between the easiest and hardest hole within a single round is far wider than on parkland. I will record average score per hole, per round. If a hole's average spikes between round two and round four, that signals a wind shift, not a skill shift. Distinguishing those two causes is the entire value of the analytical work.

Layer two: fairway retention into the wind.

This is a metric I have yet to see on any official DP World Tour stat sheet, and I regard that as a gap. Fairways hit downwind carry low informational value. Fairways hit into a 30 km/h wind carry very high informational value. I will separate those two groups.

Layer three: green performance during peak hours.

At the Irish Open, peak hours are not tee times but spectator hours. The final groups of the afternoon play in front of the largest crowds. I will compare their average putts over the opening nine against the closing nine, then cross-check against the early starters.

These three layers will not tell me who wins. They will tell me who is playing better than their score suggests.

I write the report, close the file, and the market reopens on its own.

---

The blind spot: security fences and the informational value of misplaced attention

I have to write this section carefully, because it sits outside my data field.

This week, a substantial share of coverage around the Irish Open will revolve around course ownership, security systems, and visits by political figures. That is the reality of the event. It belongs to the nature of a golf course bearing the name of a former United States president, located on Ireland's west coast, hosting a national championship.

But here is what I want to say as a data practitioner: media attention and tournament outcome are two almost independent data series. They may correlate this week, but correlation and causation are different categories.

A security fence can change spectator movement. It can shorten warm-up time. It can disrupt the atmosphere of a venue. Those are real effects, but they are small and hard to measure.

What it cannot do is change wind direction, fairway firmness, or green speed.

And in the four great variables of links golf, three belong to nature.

An empty stadium does not lack noise; it lacks one dimension of data.

I repeat this because it is the most common trap in weeks like this. When the political story becomes loud, people start hunting for its effect on the leaderboard. That is the wrong question. The right question is: across all measurable variables this week, what share of scoring variance can be explained by wind, by turf, by tee-time scheduling — and what share remains with things no human controls?

In my experience, on links courses, nature's share is usually half or more.

---

The counterintuitive angle: a defending champion is not a favourite, but a data point

If I had to choose one sentence to summarise my position before the ball is struck, it would be this.

In golf, the label "contender" is built from outcome data, while a title is decided by process data. Those two datasets overlap less than the public believes.

There are three reasons.

First, sample size. A professional golfer plays roughly twenty to twenty-five events a year. In each event, only about twenty to thirty holes produce meaningful technical separation. That gives us a few hundred high-quality observations per season. That is a small sample for a sport with golf's level of variance.

Second, course conditions. Golf outcomes depend on the course to a degree most sports do not. A player can win on parkland with soft fairways and slow greens, then fail to crack the top 30 on a dry links in 35 km/h wind. That is not decline. That is structural mismatch.

Third, and this is the point I consider most undervalued: consistency is not a skill measurable by average score. It is a distributional property. Two players with the same average score can have entirely different outcome distributions. The first finishes top 10 in twenty events. The second wins four and misses sixteen cuts. On a season ranking they may sit side by side. But in a given week, their win probabilities differ markedly.

McIlroy belongs to the second group, with a high win frequency and a wide amplitude. That is a statistically attractive profile, but it does not mean he is this week's number one contender. It means his outcome distribution has a fatter right tail than most players.

A fat right tail is a good property. It is not a forecast.

---

A view on transfer markets and tournament structure

I work with player data, across both football and golf. In both fields I observe the same pattern: valuation models overrate unproven potential and underrate unmeasurable factors such as fit within a working environment.

In golf, this pattern does not show up in transfer fees but in sponsor exemptions and in priority order for tee groupings.

A young player with a few good results often receives more invitations than a veteran with steady output. The reason is that growth potential is easier to sell than stability. But potential is a forecasting variable, not a confirmed one. And in a four-round tournament on a links course, the confirmed variable carries more weight.

At the Irish Open, this means the group of experienced Irish players — men who have hit hundreds of rounds on coastal courses since they were teenagers — tends to be underrated in prediction tables. They have no flashy metrics. They only have experience of wind direction.

And at Doonbeg, wind direction is a metric.

---

What I will not do

I will not offer a winner prediction without conditions attached.

This is a personal discipline. When I was younger and working as a data assistant for a football blog, I was dismissed for an analysis that ran against consensus. The rebuttal I wrote afterwards, two thousand words long, was widely shared and forced the editor into silence. From that I drew a principle: if my data is right, I do not need to argue. If my data is wrong, I have no right to argue.

But that principle carries an obligation in the other direction: I must state my limits clearly.

This week, my limit is the absence of shot-level data. I do not know the current swing state of any player in the leading group. I do not know fitness. I do not know practice schedules. Any claim about those would be speculation, and I do not put speculation into a report.

What I know is terrain, forecast wind direction, course structure, and competitive history at outcome level. That is the basis for discussing probability distributions, not for naming a winner.

People watch the goal, I watch the run before the goal.

In golf, that run is ball position on the fairway at the 15th and 17th, the two holes I consider most likely to decide Doonbeg when the Atlantic wind blows hard in the afternoon.

---

What to watch in the first 36 holes

The opening two rounds will give me three decisive pieces of information.

One is green speed. If organisers run greens fast in strong wind, three-putts will rise across the field, and the gap between leaders and mid-pack will widen faster than usual. That condition favours players with stable putting technique over those with explosive putting.

Two is driving strategy. If most players choose long irons or hybrids on the short par-4s to keep the ball low, that confirms wind as the governing variable. If they still pull driver, the wind is weaker than forecast, and scores will be lower.

Three is sand save percentage. On links, saving par from bunkers and tall rough is the separating skill. A player converting above 65% of par saves across the first two rounds at Doonbeg will almost certainly be there at the weekend.

None of these requires ShotLink. They require a spreadsheet and patience.

Data is never in a hurry; it simply waits for someone who knows how to read it.

---

A note on how to read results

This weekend, when the leaderboard settles, there will be plenty of articles telling the winner's story. That is the instinct of journalism, and it is not wrong. But it tells only half the story.

The other half lies with those who did not win. Specifically, those with the best hole-by-hole scoring distribution across four rounds, who lost to two bad holes. On the leaderboard they are losers. In the data, they may be the best player of the week.

That distinction matters to people in my profession, because it shapes how we evaluate players for the events that follow. If we read only the leaderboard, we will keep buying tops and selling bottoms.

I do not need recognition in a press room. The numbers know how to tell their own story.

---

The signal for the next cycle

I will close the Doonbeg file after the 72nd hole, log all the data, and not reopen it until a new data cycle establishes itself.

That is how I work. The event ends, the data stays, and the next cycle begins from a different point.

But there is one prediction I am willing to place with a clear time limit, and I will check it on Sunday evening.

I expect the 2026 Irish Open champion to come from the group of three players with the highest fairway retention rate into the wind across the final two rounds — and that player may not be the longest hitter in the leading group.

If this prediction is wrong, I will log it and adjust the model. If it is right, it confirms a hypothesis I have pursued for three years: in links golf, stability outperforms distance.

And if that holds at Doonbeg this week, it means that on a tropical coastal course twelve thousand kilometres away, where the wind also blows and the ball also runs, there is a principle that transfers.

Golf is not a game of the biggest shots. It is a game of the shots you can repeat.

The ball rolls on Thursday. I will be here, with the spreadsheet open.