Yesterday's picks: the body of the board holds while the Dodgers get routed again

Saturday, the last full slate before the All-Star break, graded as a genuinely split day for Verdexed's MLB board. The bulk of the model's favorites did their job, with the Yankees, Red Sox and Rays all holding as chalk. The problem was the top of the board again. The model's marquee lean of the night, the Los Angeles Dodgers with Yoshinobu Yamamoto on the mound, was routed 9-2 by the Arizona Diamondbacks, the second straight day Arizona buried Los Angeles by a wide margin. That is the honest headline: the body of the board held, and the single game the model liked most cracked for the second night running.

The scorecard

On a full Saturday slate, the shape of the day was the near-inverse of the marquee slot. Underneath the Dodgers, the shorter favorites the model leaned largely behaved like favorites. The Yankees held on the road, the Red Sox pitched a shutout, and the Rays kept rolling at home. That is a healthier day at the raw-count level than July 10, when the top favorite lost and the read tilted toward a washout. Here the cluster of favorites converting keeps the day from grading as a loss on volume.

The exact graded accuracy figure and Brier score for the day were not retrievable at publication time from the public accuracy surface, so this recap leans on the confirmed final scores rather than a precise hit-rate. Qualitatively, the day reads as roughly a split to a modest positive on raw count, dragged back toward even by where the loss landed. When the miss comes from the highest-confidence slot, the calibration cost is larger than a simple win-loss tally suggests, because a routed top favorite pushes the Brier the wrong way harder than a coin-flip lean ever could. A model can win more games than it loses on a slate and still leave the day feeling worse than the count, because the one it was surest about got buried by seven runs.

That is the tension worth grading plainly. On the tally, Saturday was fine. On calibration, the highest-conviction pick was on the wrong side for the second consecutive night, and that is not something the favorites underneath it fully erase.

Best call / worst miss

The sharpest correct call was the New York Yankees over the Washington Nationals. New York entered as the favorite and had to grind it out, with Trent Grisham's two-run homer in the eighth inning flipping a tight game into a 4-2 win and handing Aaron Boone his 750th victory as manager. It was not a comfortable hold, but it was the favorite converting a game the model leaned. The Boston Red Sox backed it up in the cleanest way possible, blanking the New York Mets 4-0 behind rookie Eduardo Rivera and finishing the American League's ninth shutout of the season. The Tampa Bay Rays added a third, beating the Seattle Mariners 6-1 to protect the best home record in the league. Those three are why the day did not grade as a loss.

The worst miss was the Dodgers, and it is not close. Los Angeles was the model's marquee side of the night, and Yamamoto labored in his final start before the break, allowing six runs over six innings as James McCann hit two home runs and drove in four for Arizona in a 9-2 rout. A seven-run defeat from the top confidence slot is the single most damaging outcome a probabilistic board can post on a given day, and there is no available spin on it. It also came one night after the Diamondbacks routed the same Dodgers 9-3, so this was not a one-off flukey blowout but a second straight game in which the model's favorite was not just wrong but overrun. This column graded that miss plainly on Friday, and it grades it plainly again.

What it means

One bad marquee slot is noise, and two in a row against the same opponent is a slightly louder noise, but it is still not evidence that the model is broken. Arizona has now handled Los Angeles twice by a combined 18-5, and the honest read is that the Diamondbacks are playing well and caught the Dodgers at the wrong time, not that the model has lost the plot on the best team in baseball. What the back-to-back routs do illustrate, again, is the downside of concentrating conviction at the top of a board: when the pick the model rates near the top of its range loses by seven, the day grades worse than the count no matter how many shorter favorites hold underneath it.

The longer-run picture does not move on two routed favorites in the same series. The model's tracked win accuracy has hovered in the low-to-mid 50s across the season and does not swing much on any one slate, and a split Saturday with a buried marquee pick is a normal, if unpleasant, entry in a long ledger. The Dodgers rout goes on the record in plain view, the Yankees, Red Sox and Rays holds go next to it, and the board grades a split, banks the favorites that converted, and moves into the All-Star break.

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.