Wimbledon Predictions, Picks, Odds for Fritz vs Sonego, Tiafoe vs Bublik, Khachanov vs Cobolli — 7/4

The third round of Wimbledon continues on Saturday, and our expert — Chris Gimino — has targeted three matches with great betting value.

Today, we'll focus on Taylor Fritz vs. Lorenzo Sonego, Karen Khachanov vs. Flavio Cobolli and Frances Tiafoe vs. Alexander Bublik.

Continue reading below to find our men's Wimbledon predictions and picks for Saturday, July 4.

Wimbledon Predictions, Picks, Odds — Saturday, July 4

Fritz vs Sonego Odds

Chris Gimino's model has two great projections for this matchup: a 5.9% edge on the total and a 3.9% edge on the spread.

Taylor Fritz has an 81% win probability, but given the market odds, with 99% of bets placed on the American's moneyline, the value is elsewhere.

The best edge (5.9) is on the under, which the model tags with a B+ grade. The second pick is on Fritz to cover the spread, which has a C grade.

Be sure to also check out box score projections for this matchup.

You can play them separately or go for a +118 same-game parlay on DraftKings.

Pick: Under 38.5 (-125), Fritz -5.5 (-115)

Playbook

Tail it instantly at DraftKings

Khachanov vs Cobolli Odds

Another great pick for Saturday's Wimbledon slate is on Flavio Cobolli to upset Karen Khachanov.

The Italian comes in as a +119 underdog, but the model marks his moneyline with a B grade, thanks to a 5.1% edge.

This is practically a pick 'em between the two players, and the head-to-head offers little, as they have only met once, with Khachanov winning on clay in Madrid (2024).

On grass, in this particular matchup, the model gives Cobolli a 52% win probability.

Pick: Cobolli Moneyline (+119)

Playbook

Tail it instantly at DraftKings

Tiafoe vs Bublik Odds

Finally, a 2.9% edge offers the best possible pick for this match, backing Alexander Bublik to cover a 4.5 alternate spread against Frances Tiafoe.

Tiafoe is a short favorite, and the model gives him a 54% win probability. However, as you can see in the matchup chart, both players are rated fairly close.

Pick: Bublik +4.5 (-330)

Playbook

Tail it instantly at DraftKings

How the Model Works

For every match, three independent signals get blended into a final probability. First, a walk-forward Elo rating for each player (separate for overall and by surface, regressed hard toward a below-average baseline for anyone with a thin sample or a history of playing qualifiers, so a hot streak against weak competition can’t fake elite form).

Second, a point-by-point Monte Carlo simulation built from each player’s real serve/return stats, which gets pulled toward what Elo says whenever the two disagree — Elo acts as a strong check on the simulation, not the other way around.

Third, the devigged market consensus price itself, blended in at the end as a sanity anchor. The result is a single win probability per match that reflects skill history, current form, and what the market already believes, not any one of those alone.

From there, the same win probability is used to grade three markets — Moneyline, Spread, and Total Games — by comparing the model's estimate against what the market is actually offering. The size of that gap, scaled consistently across all three markets on one shared scale, is the “edge” you see, and the letter grade is just that edge bucketed for a quick read.