The Model · Full Transparency
HOW THE MODEL WORKS
No black boxes. Every win%, every fair price, every fit adjustment on this site traces back to one spine — strokes gained — projected forward, adjusted for the course, and simulated thousands of times. Here is exactly how, and just as importantly, where our edge really comes from.
Strokes Gained — the spine
OTT · APP · ARG · PUTT → T2G → TotalOff the Tee · OTT
Driving
Distance and accuracy off the tee, valued against the field's expected score from each position. The stickiest skill on Tour — elite drivers stay elite.
Approach · APP
Iron Play
The single biggest separator in golf. Proximity to the hole from every yardage band, scored against tour baselines. Nearly as sticky as OTT — and where most tournaments are won.
Around the Green · ARG
Short Game
Chips, pitches, bunker play inside ~30 yards. A real, repeatable skill, but with a smaller spread across the field than OTT or APP.
Putting · PUTT
On the Greens
Strokes gained on the greens. The noisiest category — week to week it swings hard, so the model trusts it the least and regresses it the most.
Tee to Green · T2G
OTT + APP + ARG
Everything except putting. Because it pools three relatively sticky skills, T2G is the most predictive single number for who contends — a player who is T2G-positive every week is genuinely good, not hot.
Total · SG TOTAL
T2G + PUTT
The full picture: how many strokes better (or worse) than a baseline Tour pro a player is per round. This is the foundation the Golf King rating is built on — and the headline number on the rankings page.
The Golf King Rating
recency-weighted · field-adjusted · regressedStep 1 · Recency
Recency-Weighted
Recent rounds count more. We apply an exponential decay so last month's form outweighs last year's, but without overreacting to a single hot or cold week. Form is real; it just fades.
Step 2 · Field
Field-Adjusted
A −2.0 SG round in a major against the world's best is worth far more than the same number in a weak opposite-field event. Every round is adjusted for the strength of the field it was played in, so cross-tour numbers are comparable.
Step 3 · Regression
Regressed to Baseline
Small samples lie. Each player's projected SG is pulled toward a tour-average baseline in proportion to how little we've seen — heavily for a rookie with 12 rounds, barely for a veteran with hundreds.
Why the categories aren't treated equally
Each strokes-gained category is regressed by its own reliability. Off-the-tee and approach are the stickiest — they carry the most signal and get regressed the least, so they drive the rating. Putting is the noisiest — a player who putts lights-out for three weeks is mostly getting lucky, so we regress it hard toward the mean. The result is a rating that rewards durable skill and refuses to crown a player for a hot putter.
Course Fit
attribute importance × player skill → strokes/roundAttribute Importance
What This Week Rewards
Every course rewards different skills. A 7,800-yard bomber's track with wide fairways leans on distance; a tight, firm, fast-greened classic leans on approach and putting. We estimate how much each course historically rewards distance, accuracy, approach, short game, and putting from the scoring record of who has played well there.
Per-Player Adjustment
± ~1 Stroke / Round
We then match the course's demands to each player's skill profile. A long, accurate iron player at an approach-heavy venue gets a positive fit adjustment; a short bomber at the same course gets dinged. The swing is bounded — typically within about one stroke per round — because fit is an edge, not a different sport. Course history is folded in as a complementary, regressed signal.
Monte-Carlo Finish Simulation
skill → distribution → win / top-N / cutFrom a number to a probability
A skill rating tells you how good a player is. It doesn't tell you their chance to win — for that, golf has to be played. So we play it, thousands of times.
Simulate the Field
Thousands of Tournaments
Every player's fit-adjusted skill rating becomes a scoring distribution — a mean and a spread that reflects how streaky they are. We deal out full tournaments hole by hole, over and over, letting variance do what variance does.
Count the Outcomes
Win · Top-N · Make Cut
Across all those simulated tournaments we tally how often each player wins, finishes top-5 / 10 / 20 / 40, leads after round one, and survives the cut. Those frequencies are the probabilities you see on the model pages.
Convert to Fair Odds
Probability → Price
A 16% win probability is fair odds of about +525. We publish that fair price next to the market so a mispriced book stands out instantly — the raw material of every +EV play on the site.
The Live Model
Bayesian hole-difficulty update · wave-aware · 5-min cadenceDuring the tournament · ~5-minute refresh
Once play starts, the pre-tournament model hands off to a live model that re-simulates the rest of the event roughly every five minutes. It blends each player's pre-tournament skill with what's actually happening on the course — current score, holes remaining, and how hard each remaining hole is playing — including the AM/PM wave split when weather divides the draw.
It also distinguishes sustainable from lucky. A player who is +3 on the day because of crisp approach play is treated very differently from one who got there by holing a 40-footer and chipping in — the live strokes-gained read strips the noise so the projection reflects how someone is actually playing, not just their position on the leaderboard.
Where the Edge Really Comes From
analytics · line-shopping · CLV — not “beat Pinnacle”The part most sites won't tell you
We are not going to claim we beat the closing line. Golf betting markets — especially the sharp ones — are efficient, and public models rarely beat the closing price at a book like Pinnacle. Anyone who tells you their golf model prints money against the close is selling something.
So where is the real edge? Three places, and they're all things you can verify:
Analytics
A Sharper Read
A transparent, field-adjusted, fit-aware model that surfaces the strongest angles faster than scanning raw tables — so you spend time deciding, not digging.
Line Shopping
Best Of 10+ Books
The same bet pays differently across books. We show every price side by side and flag the best number, which is free EV on every ticket you place.
Closing Line Value
The Honest Scoreboard
Our tracker measures the price you got versus the closing line. Consistently beating the close — CLV — is the most reliable indicator that you're actually finding value, long before results stop being noisy.
That's the deal: best-in-class analytics, ruthless line-shopping, and an honest CLV scoreboard. Not a magic number that beats the market — a toolkit that helps you beat your own bad bets.
Model outputs are probabilistic estimates, not guarantees, and are refreshed as new data lands. The Golf King rating, course-fit adjustments, simulation, and prices are our own, built on comprehensive professional tournament data. Questions about a specific number? Start with the
glossary or the
FAQ.