The Paper Ledger #2 — a just-for-fun PAPER sports-betting experiment, run in public. Paper money only. No real bets, no sportsbook links, ever. StratIQ's research process pointed at football: this week the card is three picks, not five — the full $8,103.47 simulated bankroll split across them by conviction — and every pick gets graded in public on Thursday, October 29.
The scoreboard so far: 11 settled (5-6-0), down $1,896.53 on $10,575.91 staked (-17.93% ROI), average closing line value -0.10 points. No edge yet — finding out is the point of the series. Week one: 2-3-0, -$1,872.44 (-18.77%) — Carolina and Cleveland won as thesis-driven picks; Jacksonville, San Francisco and Notre Dame lost as bad theses.
Chapters:
0:00 Cold open — week one's honest shape: 2-3, -$1,872.44
0:14 The frame + the full scoreboard (5-6, -$1,896.53, -17.93% ROI, CLV -0.10)
1:07 Week-one receipts, graded by thesis
3:31 Why the card is three picks, not five
4:43 Pick 1 — Steelers/Saints Under 44.5: the Paris wind, 8/10
6:08 Pick 2 — Raiders +4.5: the Rams injury cluster, 7/10
7:34 Pick 3 — Dolphins +4.5: the below-bar fill, 5/10
8:58 The rule: if the line already moved with the news, the news is priced
9:09 The honest math — if all three hit
Stat table — this week's three (all logged before kickoff, Thu Oct 22, 2026; bet-017 to bet-019). All prices are the market price at log time; odds shown as -110.
1. Steelers at Saints, Under 44.5 (-110), Sun Oct 25 (Paris) — stake $3,240.00, stands to win $2,945.45 of paper money. Implied 52.4%. Conviction 8/10. Why: 14 mph sustained wind, gusts to 36 — situational math takes 4+ off a total; total opened 44.5 and has not moved since the forecast (news unabsorbed). Intent: closes 42.5-43.5. Counter: our own model leans over — and the model is a falsified edge source (27-season backtest: 51.1% ATS, -2.5% ROI; its ~75% claimed bucket covered 52.6%), so its lean is a reason to watch, not a reason to believe.
2. Raiders +4.5 vs Rams (-110), Sun Oct 25 — stake $2,836.21, stands to win $2,578.37 of paper money. Implied 52.4%. Conviction 7/10. Why: Rams injury cluster — Adams (hamstring surgery, out), Turner (groin, IR), Ford (triceps, season), Garrett (IR); market anchored to its preseason Rams rating (2-4, lost at home as -12.5 favorites in Week 6). Line moved LV +5.5 -> +4.5 since Monday — ticks toward us, wrong-way disqualifier not triggered. +4.5 crosses keys 3 and 4. Intent: closes +4/+4.5.
3. Dolphins +4.5 at Jets (-110), Sun Oct 25 — stake $2,027.26, stands to win $1,843.87 of paper money. Implied 52.4%. Conviction 5/10. MARKED BELOW-BAR FILL, disclosed: Miami off its bye (+1 situational), Breece Hall DNP Wed (quad), model +2.0 edge at the threshold (screen only — never quoted as a chance). Counters: Miami starting a backup QB on the road, Achane on IR, and this is the Jets' biggest favorite line of the season. It clears the bar; it does not inspire. Intent: closes -3.5/-4.
Disqualified: New England +3 at Chicago — five Patriots starters out, but the line moved with the news (-2.5 to -3 against us) → killed by the wrong-way-move rule. Cincinnati at Baltimore — the Lamar Jackson news was fully absorbed (line flipped to CIN -1.5). Giants +10 and Cleveland -1.5 sat below the bar.
If all three hit: +$7,367.69, bankroll to $15,471.16 (+90.9%) — about a 14.4% parlay at the market's own implied odds.
The model note (from episode 1's backtest segment): before trusting the ratings model with a public bankroll, we backtested that exact model walk-forward over 27 seasons — 6,549 graded bets. Cold-start variant: 51.1% against the spread, -2.5% ROI — a coin flip that loses to the juice. Its probabilities are not credible: spots where the model claimed ~75% covered 52.6% of the time, and across edge buckets win rates sat flat at ~51-52% while claimed probabilities climbed 57% to 69%. So in this series the model is a screen, not an oracle: its fair lines are quoted as "what the model said," never as real chances. The market…