Yesterday's picks: how the model did (August 23)

There is no graded line for August 23. Verdexed stated no side on any of Sunday's fifteen MLB games, claimed no edge, and assigned no confidence tier, so the honest entry is zero picks, zero correct, zero incorrect. That makes nineteen consecutive slates with nothing for this desk to grade.

Sunday's column did survey the market, which gives this page external numbers to score. None of them belonged to the model. Baseball was again the whole graded universe: the NBA and NHL are in their offseasons, and although the Titans beat the Seahawks 19-16 in Sunday's lone NFL preseason game, no Verdexed NFL board was published, so no football result is scored here.

The scorecard

Zero stated directions, zero correct, zero incorrect for August 23.

The stated-side ledger does not move: 12-7 across 19 decisions, unchanged since August 4.

No trailing accuracy figure and no Brier value are claimed here, because none was retrievable. Verdexed's public accuracy surface could not be reached at publication and no per-tier figure for August 23 surfaced through search. That is nineteen slates without a quotable calibration number, and there is no high-confidence tier to grade, so whether the 65 percent and 70 percent bands held on Sunday cannot be answered from this page.

What can be scored is the market's board, the comparison point a model number would have been measured against. Four moneyline favorites were quoted in Sunday's column, and they went 3-1. New York won at -138, Cleveland won at -132, and Los Angeles won at a price not confirmable Sunday but sitting between -235 and -275 on the two prior days. Milwaukee, the firmest price on the board at -168 and the home team in the day's showcase game, lost. A fifth game carried a third-party model's 59 percent read on Houston, flagged Sunday as somebody else's number, and that side lost too.

One total also resolved. Coors Field was listed at 10.5 with the over near -132, and Cleveland and Colorado combined for nine runs, so the under cashed.

Best call / worst miss

There is no best call and no worst miss, because no side was stated. Assigning one now would be inventing a record after the results were in, and a scorecard that back-fills its picks is worth less than none at all.

The cleanest result on the board belonged to the chalk. Blake Snell threw six scoreless innings on two hits at Dodger Stadium, Pittsburgh hitters went 0-for-10 against his fastball with five strikeouts, and Los Angeles won **4-0** to complete a sweep and a sixth straight win. That is the shape of day a probability model banks on, and it earned the model no credit because the model's number never reached the page.

The two results that would have punished a confident number deserve naming plainly. In Houston, the Athletics trailed **6-0** and won **7-6** in the rubber match. Jose Altuve capped a three-run first with a two-run homer off opener Brady Basso and Isaac Paredes added a two-run shot in the fourth. Then it turned: Serven hit a 402-foot three-run homer, Tommy White tied it at 6-6 with a two-run single, and Max Muncy hit a go-ahead solo homer in the seventh. Any reasonable pregame probability had Houston well on the right side of that game, and any in-game model had the Astros near certain after four innings.

The other one was the Little League Classic. Atlanta beat Milwaukee **4-2** at Historic Bowman Field, with Michael Harris II hitting his twenty-first home run, Tyler Mahle working five strong innings and Raisel Iglesias recording his twenty-seventh save. Shane Drohan took the loss on two runs, eight hits and three walks over five innings. Sunday's column had flagged that game as the one deserving the widest error band: an unfamiliar park, a single-night trip for both clubs, and conflicting rotation reporting. The firmest price on the board fell in the game with the least reliable inputs.

The two favorites that held did so comfortably. New York beat Toronto **8-3** behind a three-run sixth capped by Ben Rice's two-run single, and Cleveland beat Colorado **7-2** to finish a sweep, sending the Rockies to a ninth straight home loss.

What it means

Nineteen days of zeros is not a bad stretch of picks. It is an absence of picks, and the two should never be confused. Nothing on this page argues the model is calibrated well or badly, because nothing measurable has been produced since August 4. The 12-7 mark still sitting on the ledger describes early August and nothing since, and it weakens as a description of the present with every day it stands still.

The pattern here is an availability problem rather than a calibration problem, and it is the kind a reader should hold against the platform rather than the math. A public accuracy surface that cannot be read is not a record. Until per-game probabilities and the trailing accuracy and Brier figures are retrievable again, this column will keep printing zeros and confirming finals.

The market's 3-1 day is no substitute finding. Favorites winning three of four is an ordinary Sunday, and no model was graded on any of it.

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.