The Sunday Edition · Week 9 CFB / Week 8 NFL · All Sports
The Sports Page
Making the numbers mean something since the first pitch
Sunday Edition No. 030 · Issue No. 216 November 1, 2026 Free and Family-Friendly

The Option Made It Close. Notre Dame Won Anyway. We Had Both in Print.

Six issues this week, two gradeable predictions — both about the same game. Navy, a 30.5-point underdog, led Notre Dame 13–7 with under two minutes left in the first half. Then the Irish scored 28 unanswered. The option did what we said it would. The Brewers won the World Series. The family pool landed exactly on coin flip. Let’s go through it.
By The Sports Page · November 1, 2026 · Sunday Edition No. 030
216
Issues Published
2 HIT
Predictions Graded This Week
284
Issues Remaining to 500

The question we ask every Sunday: what did we claim this week, and did reality agree?

The scoreboard. Six issues this week, 2 gradeable predictions: 2 hit, 0 missed, 0 still pending from this week. Four pieces made no forward forecast. Two long-running picks — Notre Dame’s championship odds and the Bills’ Super Bowl odds — remain active from Issue #202.

Here is the honest note: no misses this week. The Rays lost the World Series, but Issue #208 described their overperformance — it did not predict their October. The Peralta story ended with an ALCS MVP and a World Series loss; that is new information, not a grading failure. A newsletter that invents misses is no more useful than one that buries them.

Why we bother: a prediction nobody grades is just a story.

The Week in Review: Issues #210–215

  1. #210Oct 26 — One in Seven Coaches Goes Home to Coach. Home Gives Them No Edge. — Since 1970, 181 major-college hires were former players at that school. They won 47% of their games. Other coaches at the same schools won 46%. The real number to look up is how the previous five years went.
  2. #211Oct 27 — The Mets Let Him Go. Then He Got Worse. Then He Won the Pennant. — Sal’s Column: Freddy Peralta posted a 7.2 ERA in 22 Mets starts, then allowed 5 earned runs in his first 10 appearances for Tampa Bay, then won the ALCS MVP. The devil did not move him. August did.
  3. #212Oct 28 — Notre Dame Got a Week Off. It Bought About One Point. — Across 13,000-plus college games, a bye added about one point. For a team coming off a loss, about a third of a point — same as nothing. The Irish came off a loss at BYU. This held.
  4. #213Oct 29 — The Best Hitter in Baseball Finished Fifth in Runs Batted In. — OPS alone explains only 43% of the variation in RBI across 145 hitters in 2025. Add plate appearances with runners on and the number jumps to 88%. The RBI is a window with your teammates’ fingerprints all over it.
  5. #214Oct 30 — Navy Beats the Spread Two Times in Three. It Rarely Beats the Team. — As big underdogs, academies have covered 31 times in 47 games. They have won 9 times where their lines predicted 4. Against spreads of 24 or more: zero wins in ten tries. Notre Dame was a 30.5-point favorite on Saturday.
  6. #215Oct 31 — Two in Three Home Run Hitters Do It Again. Most Doubles Hitters Don’t. — 67% of top-third home run hitters repeated the next season. Only 45% of doubles hitters did. Chance alone would keep 33%. Most fans assume home runs and doubles both measure power. They barely move together.

The Prediction Scorecard

#PredictionOutcomeGrade
1 Issue #214: Navy wins outright: roughly 1 chance in 21 (~5%) Notre Dame won 35–13 at Gillette Stadium. Navy did not come close to winning. HIT — low probability for Navy did not happen
2 Issue #214: Academy teams as 24-point-plus underdogs cover the spread historically (7 of 10 recent games) Notre Dame won by 22. The spread was 30.5. Navy covered. The option kept it from becoming a blowout. HIT — historical pattern held exactly
3 Issue #212: Bye worth about 1/3 point for a team off a loss — same as nothing Notre Dame won convincingly. The extra week did not transform the outcome. The Irish came off a loss and won by 22 — Notre Dame quality, not rest. HIT — confirmed by outcome; the bye was not the margin
4 Issue #202 (standing): Notre Dame wins CFB National Championship (~16% at issue time; updated to <5% after BYU loss) Notre Dame beat Navy 35–13. Now 7–1. As an independent it has no conference title game, so its way in is an at-large spot. Path narrow but open. PENDING — one ranked opponent left: No. 1 Miami, Nov. 8
5 Issue #202 (standing): Buffalo Bills win Super Bowl (~15%) Bills entered Week 8 at 5–1, playing today at home as −3 favorites. PENDING — season ongoing

What Yesterday’s College Football Showed

Notre Dame 35, Navy 13 at Gillette Stadium, Foxborough. Navy trailed 7–3 after one quarter and led 13–7 with 1:52 left in the half. Notre Dame answered 57 seconds later and led 14–13 at halftime, then scored three more touchdowns after the break: 28 unanswered, and a 22-point win. The spread was 30.5. Navy covered.

Issue #214 on Friday laid out exactly this shape: the option shortens the game; a short game is closer than the line; the better team still usually wins it. Navy had the ball for long stretches of the first half. Then the Irish scored four straight touchdowns. Notre Dame wins the game. Navy wins the spread. Both things were in print Friday morning, before kickoff.

The series record now stands 85–13–1. Notre Dame has won nine straight. As a 24-or-more-point underdog, Navy’s record since 2013 is now zero wins in eleven. What the option does is make the game shorter. What it does not do is make the gap disappear. You can beat the number for years and still never beat the team.

Ohio State fell to USC. Ohio State was a ranked program. The result matters for the playoff picture. Both teams were ranked; this was not an upset in the usual sense. Notre Dame remains in the conversation, depending on what happens in the standings elsewhere.

Miami (FL) remained undefeated at 8–0.


The Family Pool Scores Exactly a Coin

Every week since August the family has played a confidence pool: fifteen college football games, ranked by how certain you are, 15 points for your strongest pick down to 1 for your weakest. A perfect week is 120 points. A coin — pick every game by flipping a coin, confidence assigned randomly — averages 60 points a week.

Through eight weeks of play, the whole pool averaged 60.1 points per week. Sixty-one entries. Every entry from every week. The market cannot be beaten by the family, in aggregate, and this is exactly what should happen.

The Math Behind the Coin

A fair coin picks each game correctly 50% of the time. Points scored on a correct pick: 1 + 2 + ... + 15 = 120 total available. Expected weekly score for a coin: 0.50 × 120 = 60 points. Weekly standard deviation ≈ √(0.25 × (1² + 2² + ... + 15²)) ≈ √310 ≈ 18 points. One good Saturday moves you ±18 points. One great Saturday: ±36. Eight weeks of coin: expected 480, standard deviation ≈ 50 points. Our pool leaders: ~511 points through Week 8 — 31 points ahead of a coin. 31 points divided by 50 points of noise: about 0.6 standard deviations. That is not skill. That is one above-average Saturday.

The entries are not independent — everyone picks the same games from the same lines — so the error bars are even wider than the naive calculation suggests. The leaders might be ahead by more than luck, or they might not. Eight weeks is not enough games to know which.

Our own entry is 9th of 10 on total points (251 through the weeks we played). Here is why: we scored zero in weeks 1, 2, 4, and 8. Not because we picked poorly — because we never entered those weeks. In the five weeks we played, we averaged 63 points per week: a little above coin flip, consistent with knowing the math but not much more than that. Week 9 was sealed before kickoff and entered. You can check the digest; the text did not change after Saturday’s results were known.

The lesson has two parts. First: the spread is the market’s best estimate of the gap between the teams. When you pick with the spread, you are essentially confirming the market’s opinion and scoring 50% as a result. Beating the spread consistently — even by a few points per week — means finding games where the market is systematically wrong. Nobody in the family has found those games yet. The market is hard. That is not a complaint; it is a fact worth knowing before you argue with it at dinner.

Second: the short leaderboard through Week 8 measures attendance and luck before it measures anything about football knowledge. The leaders have played every week. We have not. The calibration table, the real test, needs 25 or 30 resolved games per confidence band before it says anything useful. We will run it in November and again at season’s end.

“The market is right until proven otherwise. Eight weeks is not enough to prove it.”

— The Sports Page, on the family pool, Week 8

What Readers Read · October 24 – October 30

RankIssue
#1The WNBA and MLS Were Born the Same Year. One Pays $584K. The Other Pays...
#2Bill James Says the Mets Should Be 22–24. The Bigger Story Is What That ...

Ranked by unique pageviews (GoatCounter), author visits excluded. Pulled 2026-10-31.


What to Watch Today — and Why Those Games

Week 8 of the NFL season is in progress. The Bills are off bye and playing at home today; the Seahawks are on Monday night tomorrow. Two games worth knowing about before you turn on the television:

MatchupSpreadWin prob.Why watch
Jets vs. Raiders (1pm ET) NYJ −1.5 54–55% Jets The board’s closest thing to a coin flip this week. The Jets are 2–5; the Raiders are 3–4. The market can barely separate them. One play decides both teams’ direction heading into the second half.
Bills at home BUF −3 ~59% Bills Buffalo enters at 5–1, first in the AFC East, off a bye. Three points at home is the market’s way of saying they are the better team but not by much. Our standing Super Bowl pick rides here.
Seahawks vs. Bears (MNF, Nov 2) Tomorrow night Line pending Seattle on Monday Night Football. Not today — listed so you know where the family game is.

Win probabilities use a 13.5-point NFL standard deviation: a 1.5-point favorite wins about 54% of the time, a 3-point favorite about 59%. We hold no NFL model. Eight weeks of data is not a model. What you see is the market’s estimate, not ours. Spreads from DraftKings and Bovada as of Sunday morning.


The World Series: Closed

The Milwaukee Brewers beat the Tampa Bay Rays 4–1. Game 5 finished October 28 at Tropicana Field. Milwaukee won their first World Series in franchise history.

Issue #208 described the Rays as one of the rarer overperformers in recent history: they won 98 games against a model that said 79. Of 750 team-seasons since 1999, only 14 beat the baseline by more. Then they lost the World Series to a team that also came out of nowhere — the Brewers won 103 games against a forecast of 90. Both finalists were at the far right of the distribution. The one that overperformed more on paper (Rays, by 19 games) lost to the one that overperformed a little less (Brewers, by 13). When everything is extreme, the extremes stop distinguishing who wins.

Issue #208 was not a prediction. It was a description of what the Rays had done. The result does not grade it. The piece stands: teams that win 19 more games than their forecast are rare, and they tend to be in the playoffs. What happens when they get there is a different question, and we did not answer it.

Freddy Peralta, who Issue #211 was about, won the ALCS MVP pitching for Tampa Bay and then lost the World Series. The Mets traded him in July after a 7.2 ERA in 22 starts. The Rays bought the bounce. He did not win the title. The Mets did not win the title. Nobody from this year is in the market for a regret.


Over-Reactions and Under-Reactions

Over-Reactions

Ohio State losing to USC will be treated as a program-defining disaster this week. Ohio State was a ranked team. So was USC. The spread determined who was the underdog; the final score settled it. This is not a collapse. It is a ranked team losing to a ranked team in a sport where that happens about 30% of the time. The broadcast will spend four days building a narrative. The math already had a name for it: the score on a line that put one team in the 60% range.

Under-Reactions

The pool landing on coin flip is being received as a neutral result. It is the correct result. The spread is set by a market with billions of dollars and decades of data behind it. The people in the family pool are not wrong to pick with their gut — they are wrong to expect a systematic edge from it. A pool that scores coin flip after eight weeks is a pool that has not yet found a mistake the market made. That is a compliment to the market, and it is the most honest thing the pool has told us so far.


Correction — 2 November 2026

This edition first went out saying Navy led Notre Dame 13–7 at halftime (and, in the opening, that Navy trailed 13–7 at halftime, which contradicted it). Both were wrong. Notre Dame led 14–13 at halftime: Navy went ahead 13–7 with 1:52 left in the half, and Nolan James Jr. ran 52 yards for a touchdown 57 seconds later. The 28 unanswered points began with that run, not in the second half. After one quarter it was 7–3 Notre Dame, not 7–6. Checked against ESPN’s scoring plays and CollegeFootballData’s line score.

It also said Notre Dame was 8–1, had to win the ACC Championship, and faced three ranked opponents. Notre Dame is 7–1 (it had a bye in week 8), plays football as an independent with no conference title game, and has one ranked opponent left: No. 1 Miami on November 8. The corrected text is above. The original wording is quoted here so the change is visible rather than silent.

What to Take Home, and the Road Ahead

Two things worth holding from this week. First: a spread and a win probability answer different questions. Issue #214 said it plainly and the game confirmed it Saturday. Navy had a 5% chance to win the game and a 70% chance to cover the spread. Both numbers were correct. If you read only one of them, you will always be surprised by the right thing.

Second: Notre Dame is 7–1. The loss to BYU dropped the championship odds below 5%. As an independent, Notre Dame has no conference title game to win; its way into the playoff is an at-large spot. Saturday’s win over Navy was exactly what it needed to be. The schedule has one ranked opponent left, No. 1 Miami this Saturday, and that is the game where the number actually moves.

The queue this week covers the QB wins debate (do wins belong on the stat line of a single player?), the Every Program chart that went live on the site this week, what it means to have your back against the wall in the playoffs, blockbuster trades and what they actually buy, and why you can outlive the owner but not the market. The calendar is full through Saturday November 7.

The World Series is over. Baseball goes dark until spring training in February. The newsletter follows the calendar: NFL and CFB until the bowl season, then the offseason pieces that can live without a live game underneath them.

“A spread and a win probability answer different questions. So do eight weeks and a full season. Be patient with both.”

— The Sports Page, on what the Navy game taught us about reading numbers
Pass it on.
A few minutes to read. A few seconds to send.
Share on X Facebook LinkedIn Email
thesportspage.net  ·  Permalink