Fourteen Issues, Five Predictions, and the Truth: Our First Report Card
This Week’s Issues: April 5–11
- #8Apr 5 — The $467M Infield — Lindor .143, Bichette .111, and why 15 PA is 10% of the evidence needed for batting average to stabilize.
- #9Apr 6 — Mets Alumni All-Stars — −3.7 WAR lost in departures. The PCA catastrophe. Nolan Ryan redux.
- #10Apr 7 — The Relocation Myth — A’s to Vegas, Raiders as cautionary tale. 60% of moves improve a franchise, but only by +.050.
- #11Apr 8 — Three NHL Franchises, Three Trajectories — Rangers, Sabres, and Vegas: what patience, spending, and draft capital actually predict.
- #12Apr 9 — The $20M Mentor Myth — Kirk Cousins, the Raiders, and why “veteran mentor” is a story teams tell themselves.
- #13Apr 10 — The Portal Freeze — EO Series Part 1: What happens when the transfer portal closes? Winners, losers, and a 35–40% reduction in movement.
- #14Apr 11 — The NIL Gini Coefficient — EO Series Part 2: College sports inequality measured — Gini = 0.37, a 14x gap between top and bottom.
The Prediction Scorecard
Every testable prediction the newsletter has made, graded against reality as of this morning. Data pulled fresh April 12, 2026.
| Issue | Prediction | Current Outcome | Grade |
|---|---|---|---|
| #1 | Skenes’ 67.5 ERA would dilute quickly; recovery to career 1.97 ERA in ~4–5 starts | After 3 starts: 2-1, 5.25 ERA. Starts 2–3 were excellent (5 IP/1 ER, 6.1 IP/1 ER). Recovery underway but pace slower than modeled. | PARTIALLY HIT — direction correct, pace slower than our Gamma-Poisson model projected |
| #2 | 60.2% chance Alonso finishes below .250 in 2026 | Alonso hitting .176 through 13 games with Orioles. OPS .538. 2 hits in last 28 AB. | PENDING — tracking strongly toward HIT but only 13 games in |
| #3 | Rangers’ −11.8% decline matches portfolio’s −14% | Rangers 33-36-9. Eliminated from playoffs. Fire sale completed. Panarin traded. | HIT |
| #4 | .500 after 6 games has no predictive power; median projection 88 wins | Mets now 7-7 (.500 exactly) after 14 games. Still within the wide confidence interval. | HIT — the model’s point was that early record is noise. 14 games in, they’re at exactly .500. |
| #7 | 54.4% chance Soto misses 10+ games from calf tightness | Placed on IL April 3. Expected back April 20–27 (17–24 days missed, ~15–22 games). All baseball activity shut down. | HIT — our most precise prediction. Pre-injury: .355 BA, 1 HR, 5 RBI in 31 AB. |
| #8 | Bichette’s 53% K-rate is “not just noise” (1-in-1,430 event) | K-rate now 25% (down from 53%, but still well above career ~18%). BA .200. OBP .229. | PARTIALLY HIT — the 53% magnitude was noise, but the elevated signal is real. Something changed in his approach. |
| #13 | 35–40% reduction in portal transfers under EO | EO doesn’t take effect until August 1, 2026. | PENDING — evaluation date: Aug 2026 |
What We Got Right
The Wins
The Soto injury model was our cleanest hit. We gave a 54.4% probability of 10+ games missed based on a Bayesian calf-strain prior, and he’s now expected to miss 15–22 games. Manager Mendoza has shut him down from all baseball activity. The model correctly weighted the severity distribution for right calf strains — the most likely outcome was always longer than the initial “day-to-day” framing suggested. Pre-injury, Soto was hitting .355. The Mets are paying $765 million for a player they can’t use in April.
The Rangers collapse. Issue #3 paired a portfolio drop (−14%) with the Rangers’ decline (−11.8%) as a misery-index experiment. The Rangers went on to be eliminated from playoff contention entirely, finishing at 33-36-9, fire-selling Panarin and other assets. The Misery Index earned its name.
The small-sample message keeps holding. We said .500 after 6 games was meaningless noise, with a median projection of 88 wins. The Mets are now exactly 7-7 after 14 games — the regression to the mean is happening in real time. The signal is starting to emerge, but slowly, exactly as the model projected.
What We Got Wrong (or Aren’t Sure About)
The Misses and Uncertainties
Skenes’ recovery is slower than modeled. Our Gamma-Poisson model projected a return to his 1.97 career ERA within 4–5 starts. After 3 starts, his ERA is 5.25. The good news: starts 2 and 3 were genuinely excellent (5 IP/1 ER vs the Reds; 6.1 IP/1 ER with 6 K vs the Padres, allowing only a solo HR to Bogaerts). The bad news: even two strong starts barely dented the denominator damage from that 0.2 IP opening. The model correctly predicted the direction (rapid improvement) but overestimated the pace. The 67.5 ERA created a deeper statistical hole than the model’s simulations expected — because 5 ER in 0.2 IP is so extreme that even elite subsequent performance needs more volume to dilute it.
Bichette’s K-rate: right signal, noisy magnitude. We flagged 53% as a 1-in-1,430 event and said it wasn’t just noise. His K-rate has fallen to 25% — so the 53% WAS partly noise in magnitude. But 25% is still significantly above his career average of ~18%. His slash line is .200/.229/.244 through 45 AB. The signal we detected was real; the extreme number we highlighted was not sustainable. We called the direction correctly but let the headline run ahead of the evidence.
Lindor is the quiet concern we haven’t written about yet. He’s hitting .135 with a .577 OPS. The hamate bone surgery from February 2026 looms large. This is exactly the kind of slow-developing story that deserves its own piece when the sample gets large enough to mean something.
“The purpose of the Sunday Edition isn’t to celebrate being right. It’s to learn from being wrong. A Bayesian model updates when new evidence arrives. So should a newsletter.”
— The Sports Page, Sunday Edition No. 1Over-Reactions and Under-Reactions
Over-Reactions
The $467M infield headline (Issue #8) used contract dollar figures as the hook. It was effective — it got shared — but it framed a statistical noise event as a financial crisis. Lindor has risen from .143 to .135 (barely), and Bichette from .111 to .200 (meaningfully). We taught the right lesson (small samples lie) but dressed it in the wrong costume (panic). Going forward: lead with the math, not the money.
Under-Reactions
We may have under-reacted to Lindor’s .135 average. We wrote about Bichette’s K-rate (correctly!) but Lindor is the bigger concern: .135 BA, .577 OPS, coming off hamate bone surgery. The hamate bone is the hand bone most responsible for bat grip and wrist torque. If this isn’t just rust, it’s a mechanical problem that surgery may not have fully resolved. Worth a dedicated piece once we have 60+ PA of data.
The Road Ahead
This week brings three things to watch:
The EO series continues. Part 3 drops next, examining Notre Dame’s peculiar position — simultaneously protected and vulnerable under the executive order. The Irish may be the poster child for the “boring winner” theory: conservative NIL, high-school-first recruiting, and a model the EO actually rewards. Part 4 follows with Curt Cignetti and the question of whether portal-built Indiana can survive the freeze.
The NFL Draft is 11 days out (April 23–25 in Pittsburgh). Sixteen prospects confirmed as attendees. Ohio State is sending five. The Jets’ needs haven’t changed since Issue #6, but the mock draft class has. And the combinatorics from Issue #5 — 1057 possible draft sequences — remain as humbling as ever.
Soto watch. Earliest return date is April 20. The Mets are 7-7 without him for most of it. When he comes back, we’ll rerun the injury model and compare reality to our Bayesian projection. That’s the deal.
“Fourteen issues in. Four hundred eighty-five to go. We’ll be here every Sunday to check our work. That’s the deal.”
— The Sports Page, on the long gamePitch a Story
Noticed a weird stat? Saw something that doesn’t add up? Send it in. The best ideas become issues.
