The Paspalum Screen With a Hole in It: The Metric Mexico Open Selection Logic Names and Then Abandons

Six metrics. Two profiles. Five numbers. The sixth metric has only a name — and the name is Paspalum.

I did not take that count from a betting slip. I took it from a technical document: GOLF.com's pre-event screening for the VidantaWorld Mexico Open, with Brady Kannon as analyst of record. The document lists six metrics — driving distance, SG: Ball Striking, SG: Approach, hole proximity from 200+ yards, birdies-or-better percentage, and a Paspalum-specific putting split. Five of them carry numerical outputs for the two players profiled. The sixth carries a name and nothing after it.

My conclusion, in one sentence: the Mexico Open selection logic points in the technically correct direction and is still incomplete — because the one metric on which this golf course is strangest is the one it goes quiet on; and every other number in the piece is field-relative, which means it stops meaning anything at the next event.

Let me footnote my own assumptions, as I do. This is not betting advice and it forecasts nothing about who wins. I am reading the source's "picks" strictly as market-expectation and handicapping-logic signals. The thing under audit is the model, not the decision.

Context: 7,300 yards, almost no rough, and one defence called wind

Vidanta Vallarta, Nuevo Nayarit, on Mexico's Pacific coast. Greg Norman design. Par 71. North of 7,300 yards. Wide fairways, Paspalum turf, very little rough. Heavy bunkering, water in play, and very large greens — also Paspalum. The winning score has averaged around 20 under in recent editions.

Put those four ingredients together and the course character speaks without help. Wide fairways and minimal rough mean a ball hit the wrong way off the tee is barely punished. 7,300-plus yards mean the second shot arrives long. Large greens mean greens-in-regulation go up and approach differentials compress. And 20 under means the field finishes bunched — which moves the winning mechanism from bogey-avoidance to birdie-generation.

That is why four of the last five winners fit a power profile: Jon Rahm, Tony Finau, Jake Knapp — plus Brian Campbell, whom the source itself calls the outlier. On a course where distance carries no price, the long hitter is not taking risk. He is taking a free option.

One thing can flip the whole equation: wind. The forecast held calm, with isolated afternoon storms. The course was open in its most defenceless form.

Now the field, because this is where the argument turns. This is a FedExCup Fall Series event, down-tiered from its former place in the main-season schedule. The source document records no top-50 OWGR player in attendance. Of the two detailed profiles, one player sits 112th in the FedExCup and the other is a Korn Ferry Tour regular. This is a weak field, and saying so should not embarrass anyone — what should embarrass the document is that it never states the field size.

I learned that habit in India, on PGTI stringer work. There, knowing the field size, the cut number and the prize fund is ordinary practice. When you have grown up around 120-126 player fields, you know that "second in the field" is an extremely cheap claim unless you also say how large the field was and how many players sat within a stroke of average. India has roughly 200 courses and about 250,000 registered golfers, no franchise league, no mass television contract. In that ecosystem every number is relative and every number's limit is public, because there is nothing else to lean on.

The American fall swing needs the same discipline. The document left it half-built.

One more thing in the headline and not in the body

The headline promises five picks. The document contains substantive profiles for two. That gives me a 40 per cent sample of the portfolio. I will not hide it — it is a limit on every conclusion below, and I print limits beside samples. Whatever follows is an analysis of the two profiles the source actually provided.

Core: four places the screen leaks

Hole one: the metric named and abandoned

The six-metric screen includes a Paspalum-specific putting split. Neither player has a number for it. This does not read like an oversight; it reads like a structural decision, and in my view a wrong one.

Why? Because putting at this venue is not the job people assume. When greens are very large, GIR% rises and the between-player spread in SG: Approach compresses. The course blunts the approach advantage and increases putting volume — but volume of long putts. On a 7,300-yard course, after a long approach, your first putt is 30-40 feet. SG: Putting here is not a make-rate test. It is a lag-putting and three-putt-avoidance test.

This is where the source's own screen turns on itself. If large greens make the putting split the least reliable discriminator at this venue, the correct move was to print it with that caveat, not delete it. A metric you name and then abandon stops being a metric and becomes an unstated claim.

I have an old example of what that costs. July 2026, a junior data-writer contract at a Los Angeles digital sports desk — I parsed three seasons of ShotLink shot-level data, 412,000 shots, into my first strokes-gained model and pointed it at the Open Championship. Jordan Spieth won at Royal Birkdale at 12 under, by three. My chart showed his approach play and putting worth more than seven strokes combined, against a driving week that lost him ground. The piece ran 1,100 words, drew 40,000 reads, and produced one correction request. Since that day my rule has been sample size and a confidence band attached to every number. The note above my monitor still reads: no claim without 30 rounds or 1,000 shots behind it.

Every window in this document is 20 rounds. That is below the rule. I am not dismissing it — 20 rounds point somewhere — but a signal is not proof, and if you are going to load this much weight onto signals, print all of them.

Hole two: "second in this field" is not "second on tour"

Every rank in the source is field-relative. Brown is second in the field in SG: Approach, fourth in SG: Ball Striking. Penge is fourteenth in Ball Striking, eighth in 200+ proximity, sixteenth in birdies-or-better. No absolute strokes-gained figures. No ShotLink or Data Golf citation. No field size.

This is not a small omission. "Second in SG: Approach" in a depleted FedExCup Fall field and "second" across a full PGA Tour season are different worlds, and the gap between them is the gap in the entire selection argument. A field-relative rank is not a measurement; it is a comparison — and a comparison is valid only inside the room where it was made.

I learned this one in Russia. June 2026, World Cup data desk, an xG and PPDA model across all 64 matches, back-tested against 2026. Spain's round-of-16 exit became my test case: 1,029 passes completed, 79 per cent possession, 24 shots worth 1.8 expected goals, 1-1 after extra time, beaten 4-3 on penalties by a Russia side that took seven shots worth 1.1. I wrote that possession without penalty-box entry is a rounding error. The piece was syndicated to three outlets and my editor handed me a standing analytics column.

The direct transfer of that lesson here: proximity from 200+ without the number is as hollow as pass counts without box entries — and so is distance praise without proximity. The source does join the two, and that join is its best feature. It still never says how large the field was in which those ranks were earned.

Hole three: Penge — strength in one segment, the rest unaccounted for

The strongest single claim in Penge's profile is that he is the longest player in this field over the last 20 rounds. Alongside it: eighth in 200+ proximity, fourteenth in Ball Striking, sixteenth in birdies-or-better, and five top-20s in his last eight worldwide starts.

Now the number the source reports and does not resolve: 112th in the FedExCup. He needs top 100 to keep his card.

Sit with that contradiction. A player who leads the entire field in the defining skill of this golf course sits 112th in the season standings. The most economical explanation — and the source never states it — is week-to-week dispersion. Some weeks very good, some very bad, and the average sinks.

That explanation matters because Vidanta Vallarta has water and heavy bunkering despite the wide fairways. "Distance is free" is not an absolute. The longest hitter takes the largest share of the risk that a lateral miss on the wrong hole finds a penalty area and becomes a double bogey. The source's metric set does not model that tail risk at all. I regard that as a bigger technical omission than the missing putting split.

And one detail my archive habit caught: Penge's "five top-20s in eight worldwide starts" and Brown's "five top-20s in six PGA Tour starts" are presented in parallel syntax. They are not equivalent achievements. Penge's top-20s come from fields of heterogeneous strength across tours; Brown's come from a single, more uniform standard. Parallel syntax flatters Penge. That is not carelessness — that is what happens when two datasets are folded into one table. In 2026 I spent eleven weeks assembling a database of 1,240 matches played behind closed doors — Bundesliga from 16 May, Premier League from 17 June — and testing home advantage. In my sample it fell from 0.36 goals per game to 0.19. I published the method, the code and the caveats. The same discipline applies here: two numbers built on different field bases cannot share a sentence.

Hole four: Brown — three results, one surface, one confound

Brown's profile is the best thing in the document and I want to give credit where it is due. Korn Ferry base, but: third at the Puerto Rico Open in March (Paspalum, coastal), eleventh at Corales Puntacana (Paspalum, coastal), runner-up at the Tulum Championship in April ("wall-to-wall Paspalum"), five top-20s in his last six PGA Tour starts, second in the field in SG: Approach, fourth in Ball Striking, eleventh in 200+ proximity.

That is not one independent signal. It is three, and all three point the same way. This is a pattern, not a hot week. In my language it is the most trustworthy signal anywhere in the document.

But I want to stop here, because this is where correlation and causation part company, and it is the boring part of my job.

Corales, Puerto Rico and Tulum are among the weakest-field events on their respective schedules. A Korn Ferry-level riser would be expected to contend there as a matter of course. So the question is whether Brown's record reflects Paspalum mastery or field-quality effect. Both are possible, and the document separates neither.

My own rule says: print the base rate next to the outlier. The base rate here would be how many Paspalum starts Brown has made in his career and in how many of them he finished top-20. The source does not have that number. Three good results are three good results; they do not become hard evidence until we know the denominator.

Hole five, and the easiest to miss: one weak link in the comparable-course set

The source offers four comps: Corales, TPC Craig Ranch, Memorial Park and Torrey Pines.

The first three hold. Corales — coastal, Paspalum, wide, exposed — is a genuinely strong analogue, and the players' actual results there corroborate it. Craig Ranch and Memorial Park are reasonable wide-fairway comps.

Torrey Pines is not. Thick rough, tighter driving corridors, a U.S. Open history — a fundamentally different examination. Placing it in the same list as a no-rough resort course thins the screening logic. One weak link casts doubt on the whole list, because the reader no longer knows on what basis the other three were selected.

I have a recent example of why this matters. August 2026, Tokyo. Kasumigaseki had no spectators and no on-site LPGA shot data. I hand-charted all 72 of Aditi Ashok's holes from the world feed and rebuilt her strokes gained from zero. She finished fourth at 15 under, one stroke out of bronze. Her putting week, in my numbers, was plus 1.9 strokes per round — best in the field. I filed 900 words within two hours of her last putt.

I mention it not to display credentials but because the lesson transfers directly: when official data is absent you must publish the charting method, and when official data is present you must publish what you left out. Both ask the same thing of the reader's trust.

Contrarian: card pressure is not fuel, and it is not risk-free

The document presents Penge's 112th standing and his top-100 requirement as a positive signal. Hunger, motivation, aggression. I read it the other way, and I think it is the largest undiscussed risk in the piece.

Motivation is not monotonically performance-positive. A player competing to keep a card will be aggressive at exactly the holes where water and bunkers sit, because that is where the birdie temptation is largest. The course punishes that temptation hardest. And on Sunday's final nine, with next season's earnings hanging on each stroke, decision load rises. This is not psychology, it is economics — risk preference does not stay linear when the stake rises.

I also note that the claim "he needs it, therefore he delivers" has no base rate behind it. How often do players sitting on the bubble actually reach the top 100? I do not have that dataset, and I will not invent the number. When you lack a base rate, the honest line is "I lack a base rate" — not a story.

This is where I take the one human detail I allow myself per piece. The player sitting 112th walks onto the 18th green on Sunday knowing that a bogey carries a price. He knows it. His caddie knows it. And nobody standing in front of a scoreboard can tell whether what is in that man's hands is distance or nerves. The data does not say. The data only says his distance leads the field.

Now the second contrarian point, which I like best because it argues for the source while still cornering it.

If large greens inflate GIR and compress approach differentials, then at this venue the Paspalum putting split may in fact be the weakest discriminator available. So was dropping it correct?

Perhaps. But even then the document erred, because it named it. If you put a metric on your list, you cannot drop it; you must either give its number or state the reason for removing it. A list with an item and nothing after it tells the reader the analyst probably never had that data — and at that point the credibility of the other five metrics falls too.

The third contrarian point is the least comfortable, so I will place it last. I called Brown's three Paspalum results a pattern. The other name for a pattern is "three times". And three times is a small sample against a career. Apply my own rule — 30 rounds or 1,000 shots — and Brown's Paspalum record fails the test. I do not want to use the rule against him, because that would be the same unfairness I argue against. But I will not hide it either. Three results are a signal; they are not proof; and much of the American golf media is currently printing signals as proof.

Fourth, and held with the least confidence: fatigue and travel. Brown was runner-up at the (Korn Ferry) Tour Championship a few weeks ago and is now playing a PGA Tour fall event in western Mexico. Post-peak emotional letdown and travel load are real and unquantifiable. I log them as a low-confidence flag and go no further.

A structural problem that bothers me

Neither player has a disclosed OWGR in the source. A reader might treat that as a formality. It is not.

Without world ranking you cannot demonstrate that a player is objectively strong; you can only show that he looks strong relatively in a thin fall field. That difference tilts the whole selection argument. Any document claiming a player "fits" a course has to establish that the player is good, not merely better than the people standing nearby.

And one more archival habit. The source gives surnames only — "Penge" and "Brown". Not adding first names is discipline, not laziness. Attach the wrong identity and the entire form narrative changes, and the reader would never know. So I write it plainly: identity verification is outstanding. The same applies to "Tour Championship". Given Brown's stated Korn Ferry affiliation, the most plausible reading is the Korn Ferry Tour Championship — but I will not hang a course-fit thesis on "most plausible".

In 2026 I counted empty fixtures for eleven weeks. The Hero Indian Open was cancelled in 2026 and again in 2026; the PGTI stringer work that funded travel vanished, and I spent six weeks assuming the golf desk would come back. What those eleven weeks taught me is that the story is what the archive does not contain. That is my angle here. What the source lacks — field size, absolute strokes-gained values, the putting split's number, three of five picks, OWGR, a clear publication reference — is not decoration missing. It is the text.

Takeaway: what to watch next round

I do not predict. I list the signals that will let us evaluate this model next time.

First, whether field size and absolute strokes-gained values appear. Next preview of this event, check whether "second in the field" arrives with "how many players were in the field." If it does, preview journalism has made real progress. If not, the same half-built frame gets reused.

Second, the return of the Paspalum putting split. If it comes back named, check whether it is presented as lag-putting and three-putt avoidance or as make rate. One word changes the meaning of the entire metric.

Third, Penge's dispersion. If he really is the longest and week-to-week variance is his true story, it will show on the water par-5s and the short par-4s, where distance becomes aggression and aggression becomes double bogey. Watch his scorecard, not his driving-distance rank.

Fourth, whether Brown's approach number survives a stronger field. The Paspalum test is not Corales or Tulum; it is the week when ten top-50 players are in the field and proximity differences are real.

Fifth, and most important in my reading: Torrey Pines leaving the comparable-course list — or, if it stays, one line explaining why, given that thick rough and tight corridors cut against the course's entire premise.

What this model cannot see

My habit, at the foot of every piece, and it has saved me from two public errors.

This model cannot see which afternoon the wind arrives. It cannot see whose formative golf was played on Paspalum and whose on Bermuda. It cannot see how the man in 112th slept on Sunday morning. It cannot see how much the greens quickened between Monday and Thursday. It cannot see whether a player changed caddies. Some of these are measurable and were not measured; some are not measurable at all. The difference matters: the first category is the document's failure, the second is the sport's. A good analysis names which is which.

This document is honest about the second category. It is silent about the first.

Last line

The spreadsheet kept the score, but the course kept the story — and this time the story was that a six-metric list left behind the name of one metric and never its number. Next time a preview tells you a player fits a course, ask one question: fits it compared with whom, and how many people were in that comparison?