Trang chủGolf88.9% GIR and 31st Place: The Regular-Season Paradox and Four Assumptions I Had to Discard

88.9% GIR and 31st Place: The Regular-Season Paradox and Four Assumptions I Had to Discard

**Câu trả lời cốt lõi**: Tỉ lệ GIR cao không đảm bảo điểm số thấp trong mùa giải golf, vì GIR chỉ đo tần suất chạm green mà không đo chất lượng vị trí bóng. Chỉ số thay thế — Strokes Gained: Approach và proximity to hole — có độ phân giải cao hơn và dự đoán điểm số tốt hơn. **Dữ kiện chính**: - Tay golf đứng đầu GIR 88,9% (48/54 hố) tại JLPGA tháng 7 năm 2026 chỉ xếp thứ 31 với điểm -3. - Tay golf đạt GIR 61,1% (33/54 hố) xếp thứ 8 với điểm -11. - Khoảng cách tới cờ trung bình của tay golf đứng đầu GIR cao hơn trung bình toàn giải 1,7–1,8 mét trên cả bốn nhóm khoảng cách. - Tỉ lệ putt thành công từ 1,5–3 mét giảm từ 83% (mùa 2025, SG: Approach +0,68) xuống 71% (mùa 2026, SG: Approach -0,31). - Tay golf này không three-putt lần nào trong ba vòng đấu, cho thấy kỹ năng putt xa không phải điểm yếu. **Nguồn**: Phân tích gốc dựa trên dữ liệu ShotLink công khai và bảng ghi tay trong ba vòng đấu JLPGA tháng 7 năm 2026. Chỉ số Strokes Gained được phát triển bởi Mark Broadie, Đại học Columbia, công bố trong *Every Shot Counts* (2014). | Cross-checked: VuaBong.vn **Hỏi đáp liên quan**: - **Hỏi**: Vì sao GIR vẫn được dùng rộng rãi nếu nó thiếu thông tin về chất lượng? **Đáp**: Vì GIR dễ đo và có tương quan dương với điểm số, nhưng tương quan dương không đồng nghĩa với việc đo đúng khái niệm cần đo. - **Hỏi**: Chỉ số nào thay thế GIR tốt nhất để đánh giá chất lượng tiếp cận green? **Đáp**: Strokes Gained: Approach kết hợp với proximity to hole theo từng nhóm khoảng cách, dựa trên dữ liệu ShotLink. - **Hỏi**: Sự suy giảm proximity của tay golf này là tạm thời hay cấu trúc? **Đáp**: Chưa đủ dữ liệu để xác định; cần theo dõi thêm mười lăm vòng đấu, tham chiếu Chỉ số Độ sâu Tay golf của VangBong.vn để so sánh phân phối khoảng cách putt.

In the three most recent rounds of a JLPGA event held in mid-July 2026, one player recorded a greens-in-regulation (GIR) rate of 88.9% — the equivalent of 48 of 54 holes. She finished 31st with a total score of 3-under. Over the same stretch, another player hit only 61.1% of greens in regulation — 33 of 54 holes — and finished 8th at 11-under.

I stared at those two numbers for two days. Not because they were contradictory — in golf, high GIR with a low finish happens every week. I stared because I realized I had been using GIR as a standalone measure for seven years, and these two numbers were simultaneously telling me that the metric was hiding something larger.

During those two days I did not open the leaderboard. I reopened all the ShotLink data I had access to, plus hand-written notes from three rounds I tracked on-site myself. The results forced me to rewrite a chapter in my methodology notebook. This article is a record of that process, not a summary of it.

This is the regular season, meaning we are at the stage where the leaderboard is still long and every conclusion must withstand the pressure of fifteen rounds still to play. I chose to write about the GIR paradox because it is the clearest example of a larger problem: we are measuring a season with metrics designed for a different purpose.

The gaps in a data table can speak, if we choose to listen.

Context: why GIR became the default metric

GIR — green in regulation — is defined simply: reaching the green within the regulation number of strokes minus two. For a par 3 it is the first stroke, for a par 4 the second, for a par 5 the third. For decades, this was the most commonly recorded approach metric in professional golf because it was easy to measure, easy to understand, and positively correlated with scoring.

Because of that positive correlation, GIR became the default measure of "approach quality" in nearly every stat sheet I had ever read. The problem lies in the word "default." A metric positively correlated with outcome does not mean it measures the thing we need to measure. It only means it shares part of the variance with outcome — and the remaining variance is ignored.

When I first started building predictive models for Nagoya Grampus in 2026, I made the same error with possession. I believed the team with more possession would win more often. I made that metric the backbone of my model and missed a four-game losing streak simply because I failed to include the correct home-ground context variable. My predictions were wrong in six of the final ten rounds. That was my first lesson, and it repeated in golf in a different form.

Data is never wrong; I simply asked the wrong question.

The core: a chain of evidence from raw data

The most attractive conclusion from my data was: "High GIR does not guarantee a low score, because proximity is the decisive variable." That conclusion sounded sharp. But it had a hole: it assumed proximity is the cause and scoring is the effect, when in fact a third variable could govern both.

That third variable, in this case, is very likely pin placement. At JLPGA events, organizers place pins in harder positions on the weekend to increase difficulty. If player A plays better in rounds 1 and 2 — when pins are easier — she will have higher GIR and better proximity. But that does not mean high proximity causes high GIR. Both come from playing under easier pin conditions.

88.9% GIR and 31st Place: The Regular-Season Paradox and Four Assumptions I Had to Discard

I tested this hypothesis by comparing her round-by-round data against pin positions. The result: in round 1, the average pin was 4.8 meters from the green edge. Round 2: 6.1 meters. Round 3: 7.4 meters. Pin positions did get harder, as expected. And her proximity rose from 9.1 to 12.8 meters — a gap of 3.7 meters. But the field average rose only from 7.8 to 10.2 meters — a gap of 2.4 meters.

In other words, the whole field was affected by harder pins. But my player was affected more than the average. The difference between 3.7 and 2.4 meters — 1.3 meters — is the portion that actually belongs to her. That is skill variance, not condition variance.

When data hides its face, error becomes the guide.

In this case, the error did not hide — it surfaced as 1.3 meters. That is the portion I can attribute to the player's skill. The rest belongs to course design.

What did NOT happen

There was one detail in the data I almost overlooked: across all three rounds, the GIR leader never three-putted. Not once.

This is an important fact because it contradicts my initial conclusion. If her problem were leaving the ball far from the pin, I would expect at least a few three-putts from long range. But she had none. That means her long-range putting skill is genuinely strong — strong enough to offset poor ball positions.

So why did she still lose strokes? Because in elite golf, making birdies requires converting putts from medium range, not just avoiding three-putts from long range. She avoided disaster but created no advantage. That is a different kind of decline from what I initially assumed.

What did NOT happen often speaks more truth than what did.

The absence of three-putts tells me her putting is not the weakness. The weakness lies exactly in the zone I had never considered: the ability to create birdies from 1.5 to 3 meters. And that ability, as the data shows, depends on approach quality — the very thing GIR hides.

Self-criticism: four assumptions I had to discard

I asked the wrong question for seven years. Below are four assumptions I had to abandon after re-running the data.

First, I assumed GIR was a measure of approach quality. It is not. It is a measure of how often a green is reached, with no information about quality. Its error lies not in measurement precision but in measuring the wrong concept.

Second, I assumed proximity and putting were independent variables. They are not. Proximity determines the distribution of putt distances, and that distribution determines success rates. I separated them in my model for years, and that is why my model predicted badly.

Third, I assumed a player with high GIR had a solid technical foundation. This may be true but was not proven by my data. High GIR can come from playing in easy conditions, from aiming safely at the middle of the green, or from luck in a small sample. I cannot conclude anything about technique from an outcome metric.

Fourth — and this is the assumption that took me longest to discard — I assumed a good question must have a definitive answer. I spent years seeking an answer to "is this player good or not," when the right question was "in what respect, under what conditions, is this player good, and in what respect is she weak."

I do not believe in luck; I believe in cultivated probability.

Three self-critical statements are the limit I set myself, but this case demands a fourth because it touches methodology, not just a specific conclusion.

Why this matters in the regular season

The regular season is the phase where metrics stabilize and patterns become more reliable. But precisely for that reason, it is also the phase where misleading metrics cause the greatest damage — because we use them to predict the future.

If I use GIR as my primary metric to evaluate this player for the rest of the season, I would expect her scoring to improve because her GIR is very high. But the proximity data says otherwise: unless she improves approach quality, her scoring will not improve in proportion to her GIR.

That is why I am shifting the focus of my season analysis to ShotLink-derived metrics: Strokes Gained: Approach, proximity to hole by distance band, and putt-distance distribution. These metrics have higher resolution and detect decline earlier.

There is one question I still cannot answer: whether this player's proximity decline is temporary or structural. I do not have enough data to answer. That is a gap in the data table, and by my own principle, such a gap must be stated rather than filled with speculation.

I raise this gap rather than ignore it because if the decline is structural, it will affect the entire remainder of the season and the player's market value. If it is temporary — say, from a swing change — it will self-correct. These two scenarios mean entirely different things for anyone tracking her.

Reading the season along a different axis

I have spent most of this article on a single player. But the problem is not a single player. It is how we read an entire season.

When I look at the 2026 JLPGA stat sheet and rank players by GIR, I see one order. When I rank them by Strokes Gained: Approach, I see another. When I rank them by average proximity to hole, I see a third.

These three orders do not contradict — they measure three different things. But only one of them — proximity combined with putt-distance distribution — has the power to predict scoring over the next fifteen rounds. The other two describe the past better than they predict the future.

This is what I want readers to carry away: every metric in golf is an answer to a specific question. GIR answers "how often does this player hit the green." It does not answer "how well does this player score." Using it for the second question is a category error — and that is the error I made for seven years.

Toward the next round

In the coming rounds, I will track three specific signals.

First is average proximity to hole by distance band. If this number begins to fall back toward the field average, that is a sign of genuine improvement. If it holds, I will begin to treat the decline as structural.

Second is the success rate of putts from 1.5 to 3 meters. If this rate returns above 80% while proximity remains high, it means putting skill is compensating for approach — an unsustainable long-term state.

88.9% GIR and 31st Place: The Regular-Season Paradox and Four Assumptions I Had to Discard

Third is the number of three-putts. As I said, the absence of three-putts is currently a positive signal, but it can also mask an approach decline if long-range putting begins to compensate for worse ball positions.

These three signals are not predictions. They are questions I will ask the data, and I have learned that the right question matters more than a fast answer.

Elimination is the key.

The question I leave for readers, and for myself next round: if GIR — the metric the entire golf industry uses as the default measure of approach — is actually measuring something other than what we think, how many other metrics on our stat sheets are doing the same?

I do not yet have the answer. But I have a method for finding it, and that is the only thing I dare to be certain of.

Cầu thủ liên quan