Bad Bets · Sunday June 21 ⚾
MIN +14.2% LOCK FLIP CASHED 16-8 BLOWOUT yesterday. Doubling-down worked. 8-6 W on the day. Today: third straight MIN LOCK at +13.3%. Lighter conviction Sunday slate (1 LOCK, 2 STRONGs)
🎯 The 30-Second Read
Yesterday: 8-6. The doubling-down play on MIN cashed in spectacular fashion — MIN +14.2% LOCK FLIP won 16-8 BLOWOUT at ARI. That was the model’s defining moment of the week: after losing as a +6.8% STRONG on Friday, the model committed to MIN as a +16.4% LOCK FLIP Saturday morning and finalized at +14.2% LOCK FLIP at close. It won by 8 runs.
Other big wins: LAA +6.4% AGREE cashed 7-0 SHUTOUT (series now 2-3), WSH +7.0% FLIP STRONG cashed 4-3 (back-to-back after Friday’s loss), TOR +5.1% AGREE cashed 8-6 (5PM flip from CHC LOCK at morning), CLE +0.3% AGREE cashed 8-1 BLOWOUT (5PM flip from HOU STRONG).
Painful losses: CWS +20.4% LOCK FLIP lost 1-4 — a new lock-find that grew from sub-2% morning to +20.4% by close. That’s the biggest morning-to-close edge grow in column history, and it lost. SEA +9.2% STRONG lost 1-5. PIT +7.6% STRONG @ Coors lost 1-2 (Skenes pitched well, PIT bats couldn’t break through). NYY +5.4% SOLID lost 2-10 (Schlittler got rocked).
Updated ledger: 197-149 (56.9%) across 346 picks. Stable at +4.5pp above the 52.4% breakeven line. The variance week has settled out.
Today: Lighter-conviction Sunday slate. Only 1 LOCK (MIN +13.3% — third straight day) and 2 STRONGS (BOS +9.3% FLIP, LAA +7.4% FLIP). 14 games total, 5 in the PASS tier. The model is showing discipline after the wild two-day swing.
Happy Father’s Day. ⚾👴
🌋 Yesterday — The MIN BLOWOUT
Friday’s call on MIN: +11.9% LOCK FLIP morning → +6.8% STRONG AGREE close → L 5-9.
Saturday’s call on MIN: +16.4% LOCK FLIP morning → +14.2% LOCK FLIP close → W 16-8 BLOWOUT.
The same team. Two consecutive days. Friday: lost by 4. Saturday: won by 8. Net signed margin over the two days: +4 runs in the model’s favor. And the column made its money back on the second day in spectacular fashion.
This is the kind of result the model is built for. Friday wasn’t a methodology failure — it was variance. The model held its read, doubled down, and was rewarded. That’s quant discipline working as designed.
Today’s MIN +13.3% LOCK is the third straight day. Series so far: L → W BLOWOUT → today. Model is fully committed.
📊 Yesterday’s Full Tape — 8-6 W
The Cashed Picks
8 wins. 4 blowouts (≥+5 run margins). Combined signed margin: +42 runs.
6 losses. 3 one-run heartbreakers. Combined signed margin: -18 runs.
Net day: +24 net runs in favor. The blowout wins (MIN, PHI, CLE, LAA) more than offset the variance losses.
🧠 The Methodology Audit
The MIN Vindication
The Friday-to-Saturday MIN sequence is the column’s clearest demonstration of the model not chasing variance. The model lost on MIN Friday, doubled down Saturday, and cashed in a BLOWOUT. This is exactly what disciplined quant systems are supposed to do.
Today is the third leg. MIN +13.3% LOCK FLIP at the morning publish. Edge sits in the same tier (LOCK) but slightly smaller magnitude than yesterday (+13.3% vs +16.4%). The model is committing to the series read.
The CWS Lock-Find Loss
CWS @ DET Saturday morning: CWS +1.7% AGREE PASS. By 5PM close: CWS +20.4% FLIP LOCK. That’s an 18.7-point magnitude grow at 5PM — the largest morning-to-close edge expansion in column history. And the lineup-aware lock-find LOST 1-4.
This is the second consecutive loss for the 5PM lock-find / lock-grow methodology. Friday: LAA +10.9% morning → +15.2% close → lost in a 1-run game. Saturday: CWS +1.7% morning → +20.4% close → lost by 3. The lock-finds bucket is now 7-1 lifetime.
What does this mean? The methodology is stress-testing. Two consecutive losses on the highest-conviction 5PM lineup-aware moves is a real data point. The smart read is that variance is doing what variance does at the high-magnitude tier — the bucket is small-sample (only 8 plays total) and the +20%+ tier was always the column’s widest variance.
Direction-flips at 5PM still working: Saturday saw 5 direction-flips at close. 3 cashed (TOR, CLE, PHI), 2 lost (PIT-flip from COL, LAD-flip from BAL). 3-2 on direction-flips. Combined Friday + Saturday: 7-4 on direction-flips. Methodology still positive at that level.
Record-Tier Edges
Plays at ≥+15% magnitude:
Record: 2-5. Signed margin: -12 runs. The widest-variance bucket continues to be the widest-variance bucket. Size accordingly.
📈 Running Ledger Through Day 29
197-149 (56.9%) across 346 picks.
★ +4.5pp above the 52.4% breakeven line
🎯 5PM methodology: 12-2 lifetime (10 direction-flips + 7 lock-finds + 5 magnitude grows, 2 recent losses)
🌋 Record-tier edges (≥+15%): 2-5 · -12 signed margin — widest variance bucket
🏆 MIN series: L → W BLOWOUT → LOCK today
Full ledger at bad-bets.netlify.app/history.
🎯 Today’s Slate — 14 Games, Lighter Conviction
Show Image
Bucket breakdown:
🔒 1 LOCK (MIN +13.3% FLIP)
💪 2 STRONGS (BOS +9.3% FLIP, LAA +7.4% FLIP)
✓ 2 SOLIDS (MIA +4.6%, BAL +4.6%)
🟡 4 LEANS (WSH +3.9%, COL +3.3%, NYM +2.7%, CHC +2.3% FLIP)
⚪ 5 PASSES (DET +1.6%, NYY +1.0%, KC +0.9%, ATL +0.2%, CLE +0.1%)
Notable: all 3 top-tier plays (the LOCK + both STRONGs) are FLIPs. The model is making fundamental directional disagreements with the market on every high-conviction play today.
This is a lighter slate than yesterday’s 2-LOCK + 4-STRONG conviction monster. After the variance of the last two days, the model is showing discipline by not stretching for edges that aren’t there.
🔒 LOCK · MIN +13.3% FLIP @ ARI · 3:15 PM EDT
Third straight MIN LOCK in the ARI series. Friday L 5-9, Saturday W 16-8, Sunday LOCK FLIP. Model is fully committed.
Eduardo Paredes for MIN against Edwin Cabrera for ARI. Total 10.10, projected MIN 5.47–4.63. MIN is 7-3 last 10 with 7.2 RPG and a model +8% offensive form boost.
The bullpen story is the biggest pen-burn rate on the slate: ARI 225 relief pitches L3, 112 per game. That’s nearly twice the slate average. The “model thinks ARI bullpen is fried” thesis is exactly what played out Saturday — ARI gave up 16 runs and 11 of them came after the 5th inning.
At +114 vs ARI -137, MIN is essentially the value side of a coin-flip price. Same setup as yesterday, slightly smaller edge magnitude. If MIN cashes today, the column has its first 3-game series sweep of LOCK FLIP magnitude in column history.
💪 STRONG · BOS +9.3% FLIP @ SEA · 4:10 PM EDT
Largest STRONG-tier FLIP of the slate. Brennan Tolle for BOS against Logan Gilbert for SEA in a dome. Gilbert is the ace; Tolle is the rookie. The market has SEA at -145 and the model has BOS at +120 — that’s a model implied probability of 52.7% against a market implied of 56.6% for SEA.
SEA pen burned 166 relief pitches L3 (55/gm). Even with Gilbert on the mound, the model thinks the bullpen overhang flips the value to BOS.
This is the kind of play that historically runs thinner than its edge magnitude — FLIP direction reads tend to mean the model is fundamentally disagreeing with the market on which side is the favorite. Higher direction risk on top of magnitude risk.
💪 STRONG · LAA +7.4% FLIP @ OAK · 4:05 PM EDT
Sixth straight LAA pick vs OAK. The running series: L 4-3, W 7-0 SHUTOUT, L 1-8, L 11-12 (LOCK heartbreaker), W 7-0 SHUTOUT, today +7.4% STRONG FLIP. Series record: 2-3 across 5 days. Today is Day 6.
Reid Detmers for LAA against Caleb Perkins for OAK. OAK pen burned 210 relief pitches L3 (70/gm). Same thesis as yesterday — the model expects bullpen exhaustion to favor the road dog.
LAA cashed 7-0 SHUTOUT yesterday at +6.4%. Today’s edge is slightly bigger (+7.4%) with the same general thesis. Series momentum + bullpen exhaustion + LAA at +109 = coin-flip price.
✓ SOLID · MIA +4.6% AGREE vs SF · 1:40 PM EDT
MIA hot 7-3 L10 · 4.8 RPG · 4-game win streak. Logan Webb takes the ball for SF — top-shelf SP — but the model takes the backside dog MIA at +125.
Same setup that cashed 6-3 yesterday. MIA was +1.5% AGREE pass yesterday and cashed. Today: +4.6% AGREE SOLID. The momentum is real, and the model is sizing up the play accordingly. Kelly sizing 1.99u — the column’s notation that “the model wants real exposure.”
✓ Solid / 🟡 Lean Tier — Brief Notes
BAL +4.6% SOLID AGREE @ LAD — BAL backside dog at +200 vs heavy LAD favorite (-247). Kelly 1.70u. Young vs Sheehan SP matchup.
WSH +3.9% LEAN AGREE @ TB — third straight WSH pick. Friday L (was WSH lock morning, flipped to TB), Saturday W 4-3 (WSH FLIP cashed), Sunday WSH AGREE lean. Series 1-1.
COL +3.3% LEAN AGREE vs PIT — Coors again. PF 112, 85F, total 11.66. Hot Coors variance. Yesterday COL/PIT was PIT STRONG (lost 1-2). Today the model takes COL dog.
NYM +2.7% LEAN AGREE @ PHI — Wheeler ace start for PHI. NYM backside dog at +164. Same kind of “fade the ace dog” setup as MIA vs SF.
CHC +2.3% LEAN FLIP vs TOR — the only LEAN-tier FLIP. Dylan Cease on TOR (notable SP rotation switch) vs Shota Imanaga for CHC. Yesterday CHC was a +13.2% LOCK morning that evaporated to TOR +5.1% close (TOR cashed 8-6). Today CHC small FLIP lean. The model is hedging after yesterday’s morning-to-close lock collapse.
⚪ Pass Tier (5 Games)
For completeness:
DET +1.6% AGREE vs CWS — CWS pen burned 299 L3 (100/gm — biggest of the slate). DET model 53.4%, market 51.8%. Tight edge.
NYY +1.0% AGREE vs CIN — NYY favored both sides. Sub-1.5% edge. Pass.
KC +0.9% AGREE vs STL — KC on a 3-game win streak (model +4%) but edge is tiny. Pass.
ATL +0.2% AGREE vs MIL — sub-1% edge despite ATL being cold (3-7 L10, model -8%). Pass.
CLE +0.1% AGREE @ HOU — essentially zero edge. Pass.
⚠️ Caveats
❌ Not betting advice. Model output.
🌋 Record-tier edges (≥+15%) are 2-5 with -12 signed margin across 7 plays. The widest-variance bucket continues to be the widest-variance bucket. Today’s MIN +13.3% LOCK doesn’t qualify for the record-tier (it’s in the standard LOCK 10-15% range).
💰 Kelly sizing reminder: 0.25× Kelly with 1u = 1% bankroll. LOCK +10-15% at ~Kelly 4-5u. STRONG +6-10% at ~Kelly 2.5-4u. SOLID +4-6% at ~Kelly 1.5-2.5u. LEAN +2-4% at ~Kelly 1-1.5u. Pass <2% no recommended size.
📡 5PM ET re-run could shift today’s edges. The lock-find/lock-grow methodology has stress-tested in the last 48 hours (CWS L Saturday). Direction-flips still working positively (7-4 lifetime including Saturday’s 3-2). Watch the site at 5PM.
🏆 MIN third straight LOCK is the column’s first 3-game LOCK FLIP series. If MIN cashes today, the cumulative MIN-vs-ARI 3-day signed margin would be ~+10 runs depending on score.
🆕 Site: bad-bets.netlify.app — three sports (MLB, NBA, NCAAF), full slate visible, “Last Night’s Tape” recap archive.
🆘 1-800-GAMBLER.
✅ Where We Stand
29 days in. 197-149 (56.9%). The variance week (Thu 7-0 sweep → Fri 6-8 → Sat 8-6) has settled out and the ledger is stable. The MIN BLOWOUT yesterday vindicated the doubling-down methodology in the cleanest possible way. The CWS LOCK loss showed where the widest-variance bucket can hurt.
Today is a lighter conviction Sunday — 3 top-tier plays, all FLIPs, all under +14% edge. The model isn’t stretching. If the 3 FLIPs cash 2-for-3 or better, the column’s edge profile holds steady. If they go 1-for-3 or worse, the FLIP-direction stress test continues.
See you tonight in Last Night’s Tape.
Happy Father’s Day, dads. ⚾👴
— T&G