Yesterday's picks: the model's cleanest favorite falls as the Brewers lose, but the Dodgers' road chalk holds

Yesterday graded to a roughly even day for the Verdexed model on a 15-game Sunday MLB slate, the only league still on the board with the NBA, NHL, and NFL all in their offseasons. The result that defines June 28 is an awkward one for a model that ranks its own confidence: the side it tabbed as its cleanest favorite, the Milwaukee Brewers, went down 4-3 in extra innings, while the heavy road chalk it flagged as the day's sharpest question, the Los Angeles Dodgers, came through 4-2. When the spot you trusted most loses and the spot you doubted most holds, the day lands near even, and that is where this one settled.

This column grades the model in public, win or lose. June 28 was a mixed day with a blemish at the top of the card, so it gets reported as it happened, not spun into something cleaner.

The scorecard

The model went roughly even across its named leans on the Sunday slate, a split rather than a sweep in either direction. The headline number is where the trouble sits. The Brewers were the model's clearest high-confidence side, posted around -210 at home behind Brandon Woodruff, and they lost. That is the kind of result that dents the high-confidence tier even on an otherwise balanced day, because the tier is supposed to be carried by exactly these spots. Set against it, the model's heaviest road favorite, the Dodgers at roughly -200, did hold, and a thin lean on the Phillies at the Mets cashed as well. Two of the three named favorites came in; the one that did not happened to be the most confident of the bunch.

The near-coin-flip games broke in mixed directions, which is what near-coin-flip games do. The Yankees, framed in the preview as closer to even than their record suggested against a hot Sonny Gray, lost 5-4 in 10 innings as Boston completed a four-game sweep. The Phillies, projected a hair above even in Queens, won. That is calibration working roughly as advertised on the toss-up tier: the model did not claim an edge it did not have, and the results split.

Verdexed publishes its graded record, Brier score, and per-confidence-tier breakdown on the public accuracy pages, and those season-long figures were not cleanly retrievable at the time of writing, so the day is described here qualitatively rather than with a precise hit-rate. One Sunday does not move a season much. High-confidence MLB sides tend to land somewhere around three in five across a month, and a single losing favorite, even the cleanest one on a given card, is well inside that expectation. A price near -210 still implies the favorite loses close to one time in three.

Best call / worst miss

The sharpest call was the Dodgers, and it is sharper for having been the spot the model itself was most nervous about. The preview flagged Los Angeles as heavy road chalk handing the ball to Emmet Sheehan and his 5.32 ERA against San Diego ace Michael King, who had thrown seven scoreless against this same lineup back in May. It read as the loudest warning on the board. Instead, the Dodgers won 4-2 at Petco Park, with Freddie Freeman drawing a bases-loaded tiebreaking walk and Mookie Betts following with a two-run single in the fifth. When the favorite you doubted most delivers anyway, the roster edge the model priced in earned its keep.

The worst miss is unavoidable: the Brewers, the cleanest favorite on the card. Milwaukee fell 4-3 to the Cubs in 10 innings in the series finale, and the loss was not Woodruff's fault. He threw 5 2/3 shutout innings and struck out six, allowing one hit into the sixth. The game turned in the 10th, when Seiya Suzuki delivered a two-run single and Gary Sanchez, who had homered early, grounded into a game-ending double play with the bases loaded. A strong start wasted by a late bullpen sequence is precisely how a heavy favorite loses, and the model does not get to discount it. The cleanest side lost; the column says so plainly.

What it means

The honest read is noise, not a pattern. A day where the model's most confident side falls and its least confident heavy side holds looks bad on the marquee but lands near even once the full slate is counted, and that is closer to expectation than to alarm. The high-confidence tier takes a small hit on the Brewers, but a 5 2/3-inning shutout start losing in extras is a bullpen-variance outcome, not a signal that the projection was wrong. Across the longer run the model is still tracking near its usual clip, roughly three in five on the sides it leans hardest, and a single Sunday does not move that. The caution worth keeping is the one the preview already named: heavy chalk, home or road, is a probability and not a promise, and June 28 paid that lesson out in both directions on the same card.

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.