BMW Championship

Every number on this board runs through DataGolf's strokes-gained model — the skill metric that measures how a player performs relative to the field on every shot, not just whether they had a good week. Win odds, top-finish probabilities, group picks, DFS projections: it all starts from SG. And every pre-event call gets frozen before the first tee shot and graded publicly after the final round. The record is here if you want to check it.

What's on the board

Five tabs, each a different lens on the same field:

• Board — live scores with our win and top-finish probabilities beside every player. Expand any card for the full round-by-round scorecard and that player's past results at this track.
• Groups — round-by-round 2-ball and 3-ball picks: who we project to post the lowest score in each pairing. Every pick is graded after the group finishes.
• Value — spots where our projected finish probability beats the book's implied odds by at least 1.5 points. If nothing clears that threshold this week, the tab says so.
• DFS — the DraftKings lineup optimizer for the main classic slate. DataGolf's projected fantasy points, salary, and wave data for every priced player.
• Report card — the graded archive. Every past event, what we projected, where players actually finished. Nothing hidden.

The model in brief

DataGolf reconstructs strokes-gained skill ratings from round-level data going back years. A player's SG reflects how they're actually striking the ball and putting — not just their most recent finish. Once the field is set, the model runs a simulation: given these SG ratings, how likely is each player to win, make the cut, or finish in the top 5, 10, or 20? The simulation accounts for real variance in SG — skilled players have bad weeks — and returns a probability for each finish bucket.

Once play starts, those projections update round by round. The live model blends current leaderboard position with the player's underlying SG — so a leader running hot gets a slight haircut and a skilled player sitting two back picks up more win equity than the raw gap implies. The full methodology is on the how it works page.

Where the edge is — and where it isn't

The honest version most golf content skips: most markets are efficiently priced, and we won't pretend otherwise. We check the outright winner market and make-cut lines in every snapshot. Both almost always come back clean — the books calibrate those markets closely, so real gaps are rare once you account for the vig. You'll see them on the Value tab listed as efficient on weeks when nothing qualifies.

Derivative markets are softer. Books invest less pricing effort in top-5, top-10, and top-20 finish lines — and that's where DataGolf's model finds real discrepancies. When our projected probability exceeds the book's implied number by 1.5+ percentage points, the play shows up on the Value tab. That threshold isn't arbitrary: below it, the vig alone eats the apparent edge. Some weeks, nothing qualifies. The board says so.

The receipts

Across 168 graded events, when MatchWiz projected a player inside the top 10 before the tournament, 28% of those picks actually finished top 10. Make-cut accuracy — players modeled as more likely to make the cut than miss — has landed at 72% across the same sample. Every projection is frozen at the opening tee shot: the numbers don't move once the round starts, and the grades are posted publicly after the final round. Pull up any past event on the report card and see exactly what we said before it happened.

How to use it

For betting research: start on the Value tab. Qualifying plays are the spots our model found a real price gap — the card shows the book's line and the implied probability it's set at. Pull up a player's board card for their SG rank and projected finish probability to understand why they cleared our threshold.

For DFS: go to the DFS tab, check the wave column, and sort by value (projected points per $1,000 of salary). The optimizer builds cap-legal DraftKings lineups around your locks automatically. Wave matters more than most DFS tools acknowledge: on a links course on a windy day, the AM and PM draws can separate field scoring by more than a full stroke. Tilt toward the better draw before you finalise.

For round-by-round interest: check the Groups tab before each round starts. You get a pick for every pairing, graded after the group finishes. It's the most granular lens on the model — one clear signal per group, per round.

For the full picture: browse the report card. Every past event, what the model said before it started, and where everyone finished. That's the ledger — the same one that tells us which inputs are earning their weight.

The traps

• Playing the outright winner market on SG rank alone. SG tells you who's best. It doesn't tell you whether the posted odds are fair — a great player can be correctly priced or overpriced. Always check the implied probability against our model number first.
• Treating last week's winner as a hot hand. SG averages out single-tournament noise on purpose. A player who won last week may have outshot their underlying SG by two strokes — that usually reverts. The model already accounts for it.
• Ignoring wave split in DFS. On calm weeks at parkland courses, the AM and PM draws matter less. On a links or mountain course on a windy day, the tee-time split can be worth more than a full stroke of projected SG. Check the wave column before you build.
• Building DFS lineups too early. The model's projections sharpen as the official entry list firms up — usually two to three days before the first round. An early look can miss late withdrawals. Check the DFS tab close to first tee.
• Pressing when the Value tab is empty. Some weeks the book is sharp across every qualifying market. When the tab posts nothing, that's a signal, not a glitch. We'd rather sit out a week than stretch a thin edge into a play.