Yesterday's picks: how the model did (August 6)

Thursday's seven-game board produced no graded record for this desk, because the August 6 column published no stated sides. That is the second straight morning with a blank ledger, and the second straight morning on which Verdexed's per-game probabilities were not retrievable at publication. The honest entry for August 6 is zero picks, zero correct, zero incorrect.

What the column printed instead was a set of structural notes built on listed starters and surface numbers. Those notes had a bad day. Five of them carried a direction. Three were wrong on the outcome, and both that landed were built on a starting pitcher who did not pitch for the team the column named.

The scorecard

Zero stated directions, zero correct, zero incorrect. That is the entire model line for August 6.

Two things stay separate here. The model logs and scores its own predictions whether or not a column quotes them. This desk can only grade what was published. No slate-wide accuracy figure, no Brier value and no confidence-tier breakdown for August 6 was retrievable from Verdexed's public accuracy surface at the time of writing, which makes five consecutive mornings without one. That gap now matters more than any single day's outcome.

The confirmed finals on all seven games:

**Cubs 3, Blue Jays 2** in 11 innings at Wrigley Field. **Mets 13, Guardians 6** at Progressive Field. **Angels 4, Orioles 1** at Camden Yards. **Brewers 5, Pirates 2** at American Family Field. **Tigers 11, Mariners 0** at T-Mobile Park. **Reds 6, Athletics 5** in Cincinnati. **Phillies 7, Nationals 3** in Philadelphia.

Measured against the column's five directional notes rather than any model output, Baltimore lost, Cleveland lost, Seattle lost, Milwaukee won and Chicago won. Two of five on outcome, and neither winner for the reason given.

One correction belongs at the top of that list. Thursday's column listed Dylan Cease as the Chicago starter. Cease started for Toronto, and he threw seven shutout innings of two-hit ball with 10 strikeouts against the Cubs. The column had the pitcher on the wrong side of the game it called the widest pitching gap on the board.

Best call / worst miss

The nearest thing to a correct call was Milwaukee. The column pointed at the Brewers' rotation and Milwaukee beat Pittsburgh 5-2, with Jackson Chourio and Gary Sanchez homering. The pitcher the column named, Shane Drohan, was not the pitcher who threw. Dustin May made his Milwaukee debut after arriving from St. Louis at the deadline and gave up two runs on three hits in six innings. Braxton Ashcraft, the one starter the column got right in that game, took the loss. A correct side on a projection that missed the starter is not a call worth much credit.

The worst miss was Seattle. The column ranked Bryce Miller and his 2.80 ERA ahead of Framber Valdez and his 4.41, and Detroit won 11-0. Valdez held the Mariners to five hits over seven scoreless innings. Dillon Dingler homered off Miller in the first, his 26th of the season and the most among American League catchers, and Kevin McGonigle added a two-run shot in the seventh. Season ERA is a poor proxy for one afternoon, and this is what that looks like when it fails badly.

Two others went the same way. Cleveland's Foster Griffin, the arm the column called better by the surface numbers, allowed five runs in four innings in his first start for the Guardians as the Mets took the game 13-6 and swept the series. Baltimore, the only side on the page with a price attached at -163, was described as the closest thing to a straightforward mismatch and lost 4-1 to a 43-71 Angels club, with Zach Neto hitting a three-run homer.

The Cubs game turned on neither pitching nor the pregame read. Alex Bregman hit a two-out, two-strike, two-run homer in the ninth to force extras, and Pete Crow-Armstrong scored on a throwing error while stealing third in the 11th. Toronto went 1 for 19 with runners in scoring position and stranded 13.

What it means

The stated-side ledger does not move. Five graded cards read 2-0, 1-1, 2-2, 3-2 and 4-2, or 12-7 across 19 decisions, and two consecutive blanks add nothing to that in either direction. Nineteen decisions carry almost no signal to begin with.

The finding from Thursday is not about the model, which said nothing here that can be scored. It is about what fills the space when model output cannot be reached. Five straight mornings have passed without a retrievable accuracy or Brier figure, and on the second of the two blank days, the substitute reasoning went 2 for 5 on outcomes while misidentifying two of seven starting pitchers. Declining to invent a probability remains the right call. Printing surface stats in its place is not a lesser version of a model, and Thursday is the evidence for that rather than an argument about it.

Nothing above says whether the model would have been right on August 6. That counterfactual is unknowable and does not belong in a record.

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.