Yesterday's picks: how the model did (August 2)

Sunday finished even. The column published reads on seven games from the August 2 board, four of which carried a stated direction, and those four went 2-2. The day's widest disagreement with the market cashed at Wrigley Field. The day's firmest read lost in Los Angeles for the second night running.

The scorecard

Four stated directions, two correct.

The first hit was Yankees at Cubs, published as the widest projection-to-price gap on the board. The market had New York near -125, roughly 53 percent after removing the margin, while SportsGrid's public projection gave the Yankees about 63 percent. The column named the gap and then argued against trusting it, calling the projection the outlier rather than the line. The Yankees won 2-1. Jose Caballero hit a two-run homer, Gerrit Cole allowed one run on four hits over five-plus innings, and David Bednar saved it.

The second hit was Twins at Mariners, where the column leaned Seattle on a qualitative read after Byron Buxton went back on the injured list on July 29. Seattle won 7-6, and it took a walkoff to get there. George Kirby threw seven shutout innings and the Mariners built a 6-0 lead, but Minnesota scored five in the eighth and tied it on Luke Keaschall's leadoff homer in the ninth before rookie Colt Emerson singled home the winning run. The side won. It was one swing from gone.

The first miss was Rangers at Astros. The market made Houston a favorite near -126 while the public projection gave Texas about 52 percent, and the column took the Texas side while stating in print that a disagreement that small is closer to noise than signal. Houston won 7-3 and completed the sweep, with Yainer Diaz going 4-for-5 with a home run. The win pushed the Astros into first in the AL West.

The second miss was the high-confidence tier, and it lost its only entry for the second consecutive day in the same matchup. Red Sox at Dodgers was published as the firmest read on the board, with Los Angeles quoted from -170 to -176 and the public projection landing at 64 percent, price and projection in agreement. Boston won 8-4 and swept the series. Ceddanne Rafaela, Wilyer Abreu and Masataka Yoshida hit solo home runs inside a span of seven pitches in the third, and Jarren Duran added a fourth in the sixth. Rafaela homered in all three games. Boston is 23-3 since July 3.

Three more games carried no claimed edge and do not count on the card. Philadelphia, called a firm side without a stated edge, won 8-0 at Camden Yards in a game halted after 5 1/2 innings by rain. The projection had leaned a shade over 52 percent toward the Athletics; Detroit won 13-1. Cardinals at Blue Jays drew no lean, and St. Louis won 5-1 on Jose Fermin's tiebreaking single in a four-run ninth.

One correction belongs in the record. The column listed Michael McGreevy as St. Louis's probable. Matthew Liberatore started and worked six innings. That is the second straight day a listed probable here was wrong, after Robert Gasser was listed for Milwaukee and Shane Drohan pitched.

A slate-wide accuracy percentage and Brier figure for Sunday were not retrievable from Verdexed's public accuracy surface at the time of writing, so no number is quoted here.

Best call / worst miss

The best call was the Yankees, with an asterisk the column earned by printing it in advance: the published argument was that the projection looked like the outlier and deserved scrutiny rather than trust. The side won anyway. Being right for a reason the column doubted is still being right, but it is not evidence the model found something the market missed. It is one game decided by one run.

The worst miss was the Dodgers, and the shape of it matters more than the result. This was the second night the column posted Los Angeles as its firmest read against Boston, and the second night Boston won. Saturday it was 3-2 on two Rafaela homers. Sunday it was 8-4 on four solo shots. A model that keeps pricing the same matchup the same way while the same team keeps winning is not unlucky twice, it is slow to update, and the honest read is that an input it leans on is stale. Boston at 23-3 since July 3 is not a small sample any more.

What it means

One even day is noise. Three days of published cards now read 2-0 Friday, 1-1 Saturday and 2-2 Sunday, or 5-3 on stated sides, which is a fine number and a meaningless one at this size. Nothing about a sample that small should move anyone's read on the model in either direction.

The pattern worth watching is narrower than the record. The high-confidence tier is 0-2 across the weekend and both entries were the same game, which is the kind of repeat that separates a bad break from a bad input. Whether the model has absorbed how Boston has played for a month is a question the next Boston line will answer.

One process note held. The column flagged deadline-day scratches as the dominant risk on the board, and Detroit scratched Casey Mize hours before first pitch.

Responsible play

**This is a record of past model output, not betting advice.** Verdexed grades its model in public for transparency and entertainment. Past results do not predict future outcomes, nothing here is a recommendation to place a wager, and no pick is ever a guarantee, even a strong model loses regularly.

**You must be 21+** (or the legal age in your jurisdiction) and located where sports wagering is legal to bet. Sports betting is not available everywhere; know and follow the laws where you are. Never bet more than you can afford to lose, and never chase losses.

If gambling is affecting you or someone you know, help is available. In the US, call or text **1-800-GAMBLER** (1-800-426-2537), or visit ncpgambling.org. Please play responsibly.