The weekend at Colonial Country Club is shaping up exactly how our Charles Schwab Challenge 2026 Predictions & Picks suggested it would - precision iron play and putting are separating contenders. Jordan Smith sits atop the leaderboard at -10 after rounds of 65-65, backed by elite approach numbers that I’ll break down in detail below. What makes this weekend particularly interesting from a betting perspective is the DataGolf model’s updated probabilities after R1 and R2 diverge significantly from sportsbook pricing in several spots.
The cut line fell at -2, eliminating roughly half the field and concentrating value into a tighter group of weekend contenders. I’m seeing multiple spots where DraftKings, FanDuel, and BetMGM haven’t properly adjusted their numbers to reflect how live tournament strokes gained data is reshaping win probabilities.
This Week’s Betting Landscape
Jordan Smith enters the weekend at +760 on DraftKings with the DataGolf model giving him a 9.4% win probability. That’s solid value considering the outright leader’s historical conversion rates at this course. Right behind him, we’ve got a four-way tie at -9 featuring Hideki Matsuyama (+730), Michael Thorbjornsen (+950), Ryan Gerard (+1075), and Brian Harman (+1050).
Here’s what jumps out to me: Russell Henley sits T6 at -8 but the DataGolf model gives him a 9.8% win probability while DraftKings has him at +880. That’s one of the tightest model-to-odds alignments I’ve seen this week, suggesting the books respect his course fit and recent ball-striking. On the flip side, J.J. Spaun at +1200 with a 6.7% model probability represents solid longshot territory if his approach work from R1/R2 holds.
The real betting intrigue lies in how the model has shifted from pre-tournament expectations. Ludvig Åberg opened the week as the DataGolf favorite at 10.6% but now sits at just 4.8% after posting -6 through 36 holes. Meanwhile, Matsuyama’s model probability jumped from 2.0% pre-tournament to 10.0% after two rounds of elite iron striking.
This preview from Golf on CBS breaks down the course setup and precision requirements at Colonial that explain why approach play is dominating the leaderboard.
Value Plays: Where the Model Disagrees
Let me walk through the most compelling model-versus-market discrepancies I’m tracking heading into the weekend.
Hideki Matsuyama (+730 DraftKings, 10.0% DG model): The model now gives Matsuyama the second-highest win probability in the field despite him being fourth in the betting markets. His live SG:APP of +1.80 ranks third among players inside the top 25, and Colonial has historically rewarded this profile. The delta between his 10.0% model probability and the implied probability from +730 odds (12.0%) is narrower than you’d like, but I think there’s a case that his recent form justifies backing the model’s confidence.
Russell Henley (+880 DraftKings, 9.8% DG model): Henley’s 9.8% model probability translates to roughly +920 fair odds, making the DraftKings number almost perfectly efficient. What’s notable here is how consistent his strokes gained profile has been. His +2.35 SG:APP through two rounds leads players at -8 or better, and his career baseline of +0.524 SG:APP suggests this performance is sustainable rather than variance-driven.
Ryan Gerard (+1075 DraftKings, 7.3% DG model): This is where I start seeing real model edges. Gerard’s 7.3% win probability implies fair odds around +1270, meaning the current DraftKings line undervalues him by nearly 200 points. His live SG:APP of +2.96 leads the entire field, though the -0.86 SG:OTT raises questions about whether he can maintain greens-in-regulation without elite driving. The putter has been scorching at +2.17 SG:PUTT, which typically regresses over 72 holes.
Michael Thorbjornsen (+950 DraftKings, 7.0% DG model): At 7.0% model probability, fair odds sit closer to +1330. The market is pricing Thorbjornsen as if he’s carrying an 8-9% win probability, which suggests the books are respecting his T2 position more than the underlying strokes gained data warrants. His balanced profile (+1.02 OTT, +0.82 APP, +1.23 PUTT) doesn’t show a single elite category, making this more of a “everything clicking” performance than a course-fit dominance story.
J.J. Spaun (+1200 DraftKings, 6.7% DG model): Spaun’s model probability translates to approximately +1390 fair value, making the current DraftKings line short by almost 200 points. His +1.54 SG:APP and +0.54 SG:OTT through 36 holes match his season-long baseline strengths. The only concerning number is -0.156 SG:PUTT on the season, though he’s gained +0.71 on the greens this week.
The common thread across these discrepancies: the DataGolf model’s updated probabilities haven’t fully filtered into sportsbook pricing yet. Books are still leaning on pre-tournament power ratings while the model is reacting to live tournament data.
R1/R2 Recap and Weekend Storylines
Jordan Smith’s opening 65-65 came with elite work across multiple categories - +1.31 SG:OTT, +1.03 SG:APP, and +1.35 SG:PUTT. That’s not a hot putter masking poor ball-striking. That’s a complete performance. His 75.0% greens-in-regulation rate and 85.7% driving accuracy suggest he’s found something with his swing at the exact right time. As Jake Humphry noted in our analysis of how to find value after round 1 in live golf betting situations, identifying players who gain strokes in multiple categories early in a tournament tends to produce better value than chasing low scorers propped up by hot putting alone.
The four-way tie at -9 tells distinctly different stories when you examine the strokes gained splits. Matsuyama’s +1.80 SG:APP towers over the field among contenders, while Harman posted an absurd +2.98 in that category with 86.1% GIR through two rounds. Gerard’s +2.96 SG:APP matches Harman’s precision, but his -0.86 SG:OTT creates sustainability concerns. Thorbjornsen’s balanced +1.02/+0.82/+1.23 profile suggests he’s hitting his ceiling rather than operating below it.
Get the Edge on Every Tournament
AI-powered predictions, data-driven analysis, and expert picks delivered before every PGA Tour event.
Try Golf Agent ProRussell Henley at -8 represents the most interesting chase position from a model perspective. His 9.8% win probability (second in the field) despite being two shots off the lead reflects how the DataGolf model weights his skill set for Colonial’s finishing holes. The cut at -2 eliminated some pre-tournament favorites who never found their footing. Ludvig Åberg made the weekend at -6, but his model probability collapsed from 10.6% pre-tournament to 4.8% after 36 holes despite posting a respectable score.
Notable misses include several players who carried top-20 pre-tournament win probabilities but couldn’t navigate Colonial’s precision requirements. The course is playing exactly to type - rewarding accurate approach work and penalizing spray drivers who can’t hold these small bentgrass greens.
Strokes Gained Breakdown
Colonial Country Club historically separates the field via approach play more than any other skill category, and this year’s live tournament data confirms that pattern. Among the top 25 players on the leaderboard, SG:APP dominance is universal and aligns with what we know about strokes gained as the stat that predicts PGA Tour winners.
Brian Harman’s +2.98 SG:APP leads the field among contenders with legitimate win equity. His 86.1% greens-in-regulation rate through two rounds is elite-tier execution, though the flat 0.00 SG:OTT suggests he’s hitting fairways without gaining distance. That profile works at Colonial where the course sits at just 7,209 yards and favors positioning over power.
Ryan Gerard’s +2.96 SG:APP nearly matches Harman’s number, but the context differs significantly. Gerard is losing -0.86 strokes off the tee, meaning he’s finding greens despite poor drives through scrambling and wedge precision from the rough. That’s typically unsustainable over four rounds as fatigue compounds and approach distances from non-fairway lies increase. His scorching +2.17 SG:PUTT is masking some ball-striking variance that could bite him on the weekend.
Hideki Matsuyama’s +1.80 SG:APP at -9 represents the most projectable weekend profile among leaders. His season-long baseline of +0.896 SG:APP (elite) confirms this isn’t a two-round hot streak. The +1.52 SG:PUTT through 36 holes is well above his +0.238 seasonal baseline, which introduces regression risk, but his 75.0% GIR rate built on solid iron work should hold.
Russell Henley’s +2.35 SG:APP from the T6 position tells you everything about why the DataGolf model likes his weekend chances at 9.8%. His 83.3% GIR rate trails only Harman, Gerard, and Castillo among players at -6 or better. The concerning number is just +0.34 SG:PUTT - if the flatstick heats up even moderately toward his +0.226 season baseline, he’s got the ball-striking foundation to chase down leaders.
What’s notable across the entire leaderboard: SG:OTT correlations to scoring position are weaker than usual. Gary Woodland’s +1.89 SG:OTT leads all players at -6 or better, yet he sits T15 rather than competing for the outright lead because his -0.17 SG:APP has been mediocre. Jordan Smith’s +1.31 SG:OTT pairs with +1.03 SG:APP for the most complete tee-to-green profile among leaders.
The putting surface matters more than usual this week. Colonial’s bentgrass greens are running firm and fast, rewarding players who can control distance and read subtle breaks. The live putting stats show Rasmus Neergaard-Petersen gaining +2.28 strokes on the greens despite sitting just T15 at -6, while Ricky Castillo has lost -0.56 strokes putting despite ranking third in SG:APP at +2.40.
Matchup Analysis
The weekend pairings create several exploitable matchup edges when you cross-reference the DataGolf model’s pairing probabilities with actual sportsbook lines.
Justin Thomas (+158) vs Robert MacIntyre (+213): The model gives Thomas a -119 edge in this pairing despite both players sitting at -4 through 36 holes. Thomas posted -0.04 course fit in the pre-tournament projections while MacIntyre clocked -0.02, so this isn’t a venue-specific advantage. What separates them is Thomas’s baseline skill edge - his 1.417 SG:Total season baseline dwarfs MacIntyre’s 1.264 mark. Through two rounds, MacIntyre has gained +0.495 strokes putting (well above his +0.495 baseline that I’m seeing repeated), while Thomas has leaked strokes on the greens. If Thomas’s putter wakes up even marginally, this matchup swings hard in his favor.
Ben Griffin (+143) vs Tom Kim (+218): Griffin carries a -127 model edge despite both sitting at -4. The DataGolf model loved Griffin pre-tournament at a 2.8% win probability (7th overall), and his +0.187 SG:APP baseline fits Colonial’s precision requirements. Kim posted a strong opening 64 but followed with 68, and his live SG:Total of +1.65 through two rounds suggests he’s riding a hot start rather than gaining separation via skill edges. Griffin’s 2.8% pre-tournament win probability vs Kim’s absence from the top-20 model projections tells you everything about their relative course fits.
Max Homa (+180) vs Max McGreevy (+170): This one’s nearly dead even in the model at -106 vs +106, making it a pure toss-up matchup from a skill perspective. Both sit at -3 through 36 holes. Homa’s name recognition is keeping his line slightly shorter despite the model seeing no meaningful edge. What’s interesting here is Homa’s +0.66 baseline SG:APP (solid) vs McGreevy’s weaker iron game creating a theoretical Homa advantage, yet the live tournament data hasn’t reflected that split through two rounds. I’d lean Homa on skill but not at this price.
Sahith Theegala (+182) vs Ben Griffin (+143): The model gives Griffin a substantial -127 edge in this pairing. Theegala posted 67-68 to make the cut at -3, but his live SG:APP through two rounds has been mediocre compared to his season baseline. Griffin’s pre-tournament course fit and baseline strokes gained advantage create a clear model edge that the pairing odds don’t fully capture.
Davis Thompson (+150) vs Christiaan Bezuidenhout (+145): The model sees Bezuidenhout with a -117 edge despite both sitting at -4. Bezuidenhout’s +0.17 course fit adjustment (strong positive) from the pre-tournament model explains the confidence. His 1.5% pre-tournament win probability vs Thompson’s 1.5% suggests the model viewed them as near-equals entering the week, but the live tournament data is shifting toward Bezuidenhout’s iron play holding up better.
Key Stats to Watch
Hole 11 carnage (Par 5, 4.832 scoring average): This hole is absolutely destroying the field in R2 with just 1 eagle, 39 birdies, and 17 bogeys or worse among 132 attempts. As the hardest hole on the course this week, it’s becoming a weekend momentum killer. Players who can navigate this par 5 without giving back strokes gain massive separation. Watch for GIR percentages on this specific hole - the players finding the green in regulation are the ones cashing birdies rather than scrambling for par.
Short game scrambling on Par 4s: Colonial’s Par 4 scoring averages show Hole 1 (4.470) and Hole 9 (4.303) as the two toughest non-par-5 holes. Both require precision approaches to avoid difficult up-and-downs. Players with elite scrambling percentages (80%+) who miss greens on these holes maintain scoring pace, while weaker scramblers leak multiple shots across the weekend. Akshay Bhatia’s 92.3% scrambling rate through two rounds is sustaining his -8 position despite just 63.9% GIR.
Putting on Holes 13, 16, 8 (birdie holes): These three holes rank as the easiest on the course with sub-3.0 scoring averages in R2. The field is collectively making birdies, which means weekend leaders MUST capitalize on these opportunities to stay ahead of the pack. Hole 13 (Par 3, 2.954 average) gave up 23 birdies in R2 alone. If you’re not gaining strokes on the field’s easiest holes, you’re effectively going backward in a bunched leaderboard, particularly at historically tight venues where course history in golf betting often reveals horses-for-courses patterns.
**Driving accuracy on Holes
