Yesterday's picks: how the model did (September 18)

Verdexed's graded line for Friday, September 18, is zero picks, zero correct, zero incorrect. Baseball was the only league in session, its board was played and settled, and not one game was scored by this desk, because the model's per-game probabilities were not published. That is the forty-third consecutive slate opening with the same number, and it is a bad result rather than a neutral one.

The scorecard

Picks made: zero. Correct: zero. Incorrect: zero. Accuracy is undefined, because the denominator is zero.

**MLB.** Verdexed's own best bets column for September 18 stated at the top that per-game win probabilities, confidence tiers and edge figures were not retrievable at publication, that there was no biggest edge to name, and that no probability it printed was the model's. It published a matchup and market read instead, and there is nothing in it to grade.

**NFL, NBA and NHL.** No games. Week 2 of the NFL played its Thursday opener on September 17 and did not resume until Sunday, the NBA remains in its offseason, and the NHL's 2026-27 preseason opens Saturday, September 19, the day this column publishes. Friday was a baseball-only slate.

No accuracy percentage and no Brier score is cited here. The verdexed.com prediction, edge and accuracy pages remain blocked by the network egress proxy this column is written behind, and a public search returns the published methodology, which says every prediction is logged before first pitch and scored afterward on win accuracy and on Brier score, rather than the graded ledger underneath it. That describes a process, not a number. There is no high-confidence tier to report either: the 65 percent and 70 percent bands went unpopulated for a forty-third straight day.

The hand-kept ledger of stated sides holds at 13-10 across 23 decisions, unchanged since Sunday. At that sample the distance between 13-10 and 10-13 is a few coin flips.

What can be graded is the market half the column printed. It quoted prices on three of Friday's games. Two held, one lost, and the one that lost is the game it led with.

Best call / worst miss

Nothing was picked, so there is no sharpest call and no worst miss. What follows grades the quoted prices.

**The price that lost: Toronto.** The column opened on Dylan Cease in Arlington, called the Cease against Kumar Rocker matchup the widest projected pitching gap on the board, and quoted Toronto near minus 130 while flagging that figure as unconfirmed. Other books had the Blue Jays at minus 137 and minus 145, so the favorite was real even if the exact price was not. It went the other way completely. Result: **Rangers 7, Blue Jays 1**. Cease was gone with two outs in the third, charged with four runs on three hits in 2.2 innings, one of them earned, with a season-high six walks and a season-low three strikeouts. Corey Seager's RBI double keyed a four-run second that a throwing error had opened up. Texas pulled even with Houston in the American League West. The one game the column built its lead around is the one the favorite lost, and that belongs at the top of this section rather than below the winners.

**The reads that held.** Seattle was quoted at minus 134 on the road, and the column's own caveat was that Coors Field cuts against confidence in either direction rather than toward it. Result: **Mariners 5, Rockies 4**, decided by a Lazaro Montes leadoff home run in the ninth off Jordan Romano, a 426-foot shot, after Colorado put fifteen hits on the board. The favorite survived, and the caution around the park was the part that aged well. San Diego, the only game on the slate with a clean two-sided price at roughly 64 percent once the sportsbook's margin was removed, won **8-2** over Miami and moved into a three-way tie with the Phillies and Cubs for the top National League wild card.

**The two the column did not price.** Milwaukee won **6-5 in ten innings** at Camden Yards on a Sal Frelick RBI single, after surviving a blown save. Boston beat Tampa Bay **4-2**. Neither carried a confirmable moneyline, and neither is scored here.

What it means

Two of three quoted prices holding is noise. Favorites near minus 130 are built to lose roughly two times in five, so a single loss in three sits comfortably inside the ordinary distribution, and nobody should read a trend into three games the model did not rate.

The pattern worth reporting is the other one, and it has not improved. Forty-three consecutive slates without a graded pick is no longer a gap in the record; it is the record. The honest read of Friday is not that the model ran hot or cold. It is that readers were handed a matchup essay where a probability was supposed to be, and the essay's headline game lost. Against the longer run, the only figure this desk can stand behind is 13-10 across 23 stated decisions, last moved on Sunday. Until per-game probabilities and the graded accuracy summary are readable again, this column's answer every morning will keep being zero, and readers are entitled to see the zero.

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 of legal betting 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.