We Called Our Shot. Now We Wait for South Bend.
The Week in Review: Issues #107–112
- #107Jul 13 — The NBA’s Last Year Is Worth a +65. The Pacific Northwest’s La Niña Is Worth a Minus 19. — Lag-one persistence rebuilt with each domain’s native predictor: last-year record for sports, ENSO state for snow. Basketball is foreseeable; Cascades snowfall is nearly random. A reader pushback prompted a cleaner answer.
- #108Jul 14 — For One Year, the Trophy Was Always Somebody’s. That Year Was 1986. — The greatest single year in the modern sports calendar, built game by game: the Mets, the Bears, the Celtics, Bird and Magic and Ferdinand’s screwball. Every title changed hands. No team repeated. A reader’s anniversary prompt became the week’s best retrospective.
- #109Jul 15 — Notre Dame Came Within a Game of the Title, Then Returned More of Its Roster Than Anyone in the Country. We’re Calling the Dynasty. — The explicit, dated prediction: Notre Dame wins the 2026 CFB national championship. CJ Carr returns with a 24-to-6 TD/INT ratio. Top-ranked recruiting class. Marcus Freeman on every preseason watch list. We read the checklist. Then we went further.
- #110Jul 16 — At the Break, the Standings Lie Politely. The Second Derivative Doesn’t. — Applied from finance: velocity and acceleration in baseball. The Rays’ first-place lead is a Pythagorean credit line (six wins over their engine). The Dodgers are cooling on a ten-game sample that may be noise. The Mets’ honest signal was never the May bounce — it was always the −64 run differential underneath.
- #111Jul 17 — The Math Said She Was Oscar Robertson. Then We Handed Him a Three-Point Line. — Part II of the Clark series. 47.6% of her 2024 rookie points came from distance; Robertson never took the shot. The only player at the top of both charts — Robertson’s assist fingerprint, Curry’s scoring geography — is Clark. She is not a throwback and not a gunner. She is the first.
- #112Jul 18 — Forty Summers Ago, Fernando Valenzuela Struck Out Five Straight All-Stars. He Was Only the Second Man Ever to Do It. — Published on the night of the 2026 All-Star Game. Hubbell in 1934 and Valenzuela in 1986 remain the only two men to strike out five consecutive batters in an All-Star Game. Forty years, and the record still belongs to a vanishing pitch.
The Prediction Scorecard
Five of six issues this week were retrospective or analytical — no forward claims to grade. One explicit prediction was logged. Two diagnostic observations carry implicit forward weight. All three are accounted for below.
| # | Prediction | Outcome | Grade |
|---|---|---|---|
| #109 | Notre Dame wins the 2026 CFB national championship. Made explicitly, dated July 15, before a snap is played. Our most open call of the year. | Season opens late August. Notre Dame is among the betting favorites; QB CJ Carr (24-to-6 TD/INT) returns; top-ranked recruiting class by consensus five-star count; Marcus Freeman on Dodd Trophy watch. No outcome available until January 2027. | PENDING — grade due Jan 2027 |
| #110 | Rays’ six-win Pythagorean surplus won’t survive the second half. Their engine (+38 run differential) supports ~50 wins; they had 56 at the break. | Second half begins today. Rays led AL East at 56–37. No second-half games played at time of publication. | PENDING — revisit mid-August |
| #110 | Dodgers “genuinely cooling, worth watching, nowhere near a crisis.” July run-differential velocity dropped to −1.20; smoothed 15-game window showed only −0.27. | Dodgers led NL West at 61–35 (+151 run diff) at the break. The smoothed signal was mild; the monthly boundary produced the alarming number. Three post-break games is insufficient to grade. | PENDING — revisit mid-August |
| #110 | Mets’ honest signal is the deeply negative level, not the May acceleration. The bounce was noise; the −64 run differential was the floor. | Mets closed the first half at 40–56, dead last in NL East by 16 games. Run differential confirms the diagnosis. No counterevidence in the data. | HIT — diagnostic confirmed |
| #107, #108, #111, #112 | No forward predictions. All four were retrospective or analytical: persistence methodology, the 1986 sports calendar, Clark’s statistical fingerprint across eras, Valenzuela’s 1986 All-Star record. | — | N/A |
What We Got Right
The Wins
The Mets diagnosis held. Issue #110 said the May acceleration was noise — that the team’s true story was the deeply negative level underneath any surface bounce. The first half ended at 40–56, last in the NL East, with a run differential of −64. That is not a team that was close to bouncing. The framework read it correctly.
The Clark analysis held in real time. Issue #111 placed her at the intersection of Robertson’s playmaking fingerprint and Curry’s three-point geography — the only player who lives at the top of both charts simultaneously. The day Issue #111 published, she scored 45 points and distributed 10 assists against Seattle. We did not predict that game. But a 45-and-10 line from the player who combines Robertson’s vision and Curry’s distance scoring is precisely what the framework said should occasionally happen. The model was right about what she is.
The cross-sport persistence scale from Issue #107 directionally matched everything the data should show: NBA at +65, NHL at +62, MLB at +49, NFL at +34. The order is exactly right by roster geometry — basketball locks in superstars for years; football cycles rosters aggressively. Utah snowfall at +22 and Colorado at −48 are the two ends of the ENSO story for ski mountains. Every directional call was correct.
Factual Updates — Not Forecasting Misses
The 2026 All-Star Game Added a Name to a Different List
Issue #112 published on the night of the 2026 All-Star Game with the claim that Hubbell (1934) and Valenzuela (1986) are the only two pitchers ever to strike out five consecutive batters in an All-Star Game. The game itself — a 4–0 AL shutout at Citizens Bank Park, started by Dylan Cease, finished with 15 strikeouts total across 11 pitchers — produced a relevant update.
Justin Wrobleski struck out five or more batters in the game. That places him on a short list alongside Valenzuela (1986) and Pedro Martínez (1999) of pitchers with five or more total All-Star strikeouts since the mound was lowered in 1969. That list and the Hubbell-Valenzuela list are not the same thing.
The record Issue #112 described is for five consecutive strikeouts — five in a row, no baserunner between them. Wrobleski’s five Ks were distributed across a multi-pitcher game in which the NL recorded three total hits. They were not five in a row. The consecutive record, forty years old, appears intact as of Tuesday night. If box-score review finds otherwise, the correction runs here, promptly, without hedge.
The AL’s Cody Bellinger was named MVP after a three-RBI first inning. Juan Soto got one of the NL’s three hits. The Mets outfielder playing in Philadelphia, playing for a team 16 games out of first, collected one of three. There is something poetic about that, and none of it is the newsletter’s fault.
“A prediction made in the open and dated is not arrogance. It is what makes accountability possible. Notre Dame wins it all. Check back in January.”
— The Sports Page, Issue #109, July 15, 2026Over-Reactions and Under-Reactions
Over-Reactions
The Dodgers cooling call in Issue #110 was framed correctly — “genuinely cooling, worth watching, nowhere near a crisis” — but the piece still opened with the alarming −2.46 monthly acceleration before delivering the corrective. Lead with the smoothed number. Bury the raw one. If the Dodgers hold their run differential through August, the opening paragraph will have been more dramatic than the evidence warranted.
Under-Reactions
Issue #109 spent three paragraphs hedging before planting the flag. The hedge was honest, and the doubt was real. But the evidence was strong enough to lead with the call: CJ Carr’s 24-to-6 TD/INT ratio is not a soft number. The returning production rate is not a soft number. The blue-chip count is not a soft number. We earned the confidence the prose declined to show.
What Readers Read
Readership data pending — analytics will populate here when the GoatCounter connection is available from the publishing environment. Check back on the live site for the final top-5 list from the prior week.
The Road Ahead
The trade deadline is July 31 — twelve days away — and the next three issues build toward it with intention.
Issue #113 publishes Monday: “The Diamond in the Rough Is Mostly Rough.” A Beta-Binomial model applied to the Mets’ prospect pipeline. The thesis is uncomfortable. At 40–56 and 16 games out in July, the Mets are almost certainly sellers. The question is what they are selling toward, and whether the system underneath can sustain what comes next. The model has an opinion. Monday we publish it.
Issue #114 shifts to college football for variety — “A Champion Needs 50% Blue-Chip Talent. Indiana Won With 8%.” The blue-chip ratio is the methodological companion to the Notre Dame dynasty call. That piece made the flag-planting; this one explains why the checklist is real and what it says about who belongs in the conversation. The exception — Indiana in 2024 — is what makes the rule worth examining.
Issue #115 returns to deadline economics: “Eight Teams Have Quit. The Other Twenty-Two Are About to Overpay.” Winner’s curse at the trade deadline, log5 estimates of what each rental is actually worth versus what the market will charge. This is the piece that reads best on the morning of August 1, when the receipts are in.
The Notre Dame call sits open. It will sit there until January. Five months of patience is the cost of the flag. We are comfortable with that cost.
“The second derivative on the Mets was never an alarm. It was a confirmation. Minus sixty-four is not a trend. It’s the answer.”
— The Sports Page, on reading the level before the slope
