Daily Bet Research Report — Tuesday, September 8, 2026

Report generated at 2:40 PM ET on Tuesday, September 8, 2026

Bankroll: $23.91 | Stake (12.5%): $2.99 Record: 43-39-3 | Hit rate: 52.4%

All times ET. Earliest first pitch 6:35 PM ET — the full 15-game slate is bookable (no game starts within 30 minutes of report time).

Pending bets: None. Log's last entry is Bet #98 (2026-09-03), settled. Nothing Open to carry.

Bankroll provenance: read this run from bet_tracker.xlsx → Log sheet, row 111, column H (Closing BR) = $23.91.

Drawdown note: ATH bankroll was $54.33 (2026-08-30); current $23.91 is a −56% drawdown. Per the Strategy tab, this shifts preference among qualifiers toward higher-hit-rate plays — it does not raise the edge bar. Both staked picks below are the two highest-hit-rate qualifiers (51% and 47%), not the 35% longshot.

---

Section 2: Schedule & Summary Analysis

Decimal odds only. One row per game — the single best angle after pricing ML (both sides), run line, total, and the starter K prop.

Line-shopping note. Spreads were material today and every price above is the best found. Examples: Cubs ML ranged +175 (bet365) to +195 (RiversCasino) — 20 cents; AJ Smith-Shawver o3.5 K ranged −140 (bet365) to −192 (RiversCasino) — 52 cents; Royals ML ranged −105 (HardRock) to +105 (Caesars) — 10 cents. bet365 was consistently the softest K-prop book, Caesars the softest ML book.

---

Section 3: My Recommendations

Four picks clear positive edge today; a fifth was not found. Every row below has Edge % > 0 — no minimum threshold applied and no unstated buffer. Ordered by Kelly fraction, highest first. Straight single-leg plays only.

Staked Stake $ sums to $2.99 = $23.91 × 12.5%. ✓

Kelly fractions (f = (p×d − 1)/(d − 1)): Royals 4.33% · Peterson 2.83% · Rays 2.49% · Cubs 1.67%. Normalized across the top 2: 60.5% / 39.5%.

Why the split: September is a diagnostic month. Listing all qualifiers produces graded observations across the edge spectrum toward the 120-pick target so October can set a threshold empirically; staking only the top 2 keeps money exposure small while that sample accumulates.

Known limitation: normalizing Kelly into a fixed 12.5% budget preserves relative weighting but discards Kelly's answer on total exposure — this is edge-weighted allocation inside a fixed-fraction budget, not true Kelly. The Hit Rate feeding f is the model's own estimate, which historically runs high; sizing inherits that optimism.

Calibration note: Option 2 / Option 4 slot gaps are monitoring-only and were not used as a sizing input.

---

Section 3b: Prop Watchlist

Researched K-prop context for SGP construction. Watchlist, not recommendations.

1. Brandon Young (BAL) Over 3.5 K — 1.840 (DraftKings). Tightest amplifier line on the board; priced by all 8 books, spread only −119 to −139 (20c). De-vigged ~53.7%, so it is not a standalone play, but it is the cleanest low-line amplifier tonight.
2. Cam Schlittler (NYY) Over 3.5 / 7.5 K — o7.5 at 1.833 (bet365). 2.04 ERA facing Colorado's road offense; bet365 is 27 cents better than Fanatics (−147) on the same line. Standalone EV −1.6%, but the best amplifier price if you want Yankees exposure without touching Yankees ML (tilt-fade active).
3. Jacob Misiorowski (MIL) Over 7.5 K — 1.730 (PrizePicks) — FADE FLAG. Do not use as an amplifier. Season line says 1.97 ERA / elite K rate, but his last five starts produced 9, 6, 6, 6, 5 strikeouts — the over cleared once in five. His 9/2 start against tonight's opponent was his worst of the year (4.0 IP, 5 ER, 4 BB, 92 pitches). The market is still pricing the season aggregate.

---

Section 4: Reasoning

Pick 1 — Kansas City Royals ML, 2.05 (Caesars) — STAKED $1.81

1. The math. Best price 2.05 → implied 48.8%. VegasInsider 8-book consensus is AZ −120 / KC −101 → raw 54.55% / 50.25%, sum 104.80%, de-vigged KC 47.95%. My estimate is 51.0% (adjustment below). Edge = 51.0 − 48.8 = +2.2%. EV = 0.51 × 2.05 − 1 = +4.6%. Kelly f = (0.51 × 2.05 − 1)/1.05 = 4.33%.

2. The evidence — recency. The decisive fact: Corbin Burnes has made zero appearances in 2026. The MLB StatsAPI game log for player 669203 returns an empty stats array for the 2026 season, and ESPN's odds page lists him with no W-L record and no ERA while every other starter on the slate carries one. Tonight is his season debut. Two market tells confirm the books know it: his strikeout line is set at 3.5 — the lowest on the board among established starters, and lower than Gabriel Hughes (6.19 ERA) — and only 4 of 8 books will price him at all. A season-debut starter on a strict pitch count (typically ~65-75 pitches, 4 innings) hands the middle innings to Arizona's bullpen on the road.

Michael Wacha (KC) is 9-8, 3.36 with a normal workload, at home. Team quality: Arizona 77-68 (.531), Kansas City 64-81 (.441) — the records favor Arizona and are the reason the line sits where it does, but records do not price a bullpen game. Adjusting KC from the de-vigged 47.95% up to 51.0% is a ~3-point bump for the pitch-count asymmetry, which I consider conservative.

3. Risk factors. (a) Kansas City is genuinely the worse team — this is a bet on start-length asymmetry, not talent. (b) The total moved down 9.5 → 9, which is mild sharp support for run suppression and cuts slightly against a bet on Arizona's bullpen leaking. (c) Burnes could be stretched further than I assume if he built up in the minors. (d) The AZ ML price barely moved (−113 open → −120), so the market is not obviously mispricing this — my edge rests on the pitch-count read, not on line movement.

4. Named source. https://statsapi.mlb.com/api/v1/people/669203/stats?stats=gameLog&group=pitching&season=2026 (empty splits, confirming no 2026 appearances) and https://www.espn.com/mlb/odds (Burnes listed without record/ERA).

---

Pick 2 — David Peterson Over 4.5 Strikeouts, 2.20 — STAKED $1.18

1. The math. Best price 2.20 → implied 45.5%. Seven books price o4.5 between +108 and +120 (one outlier at −137); median +119 → 45.7% raw. With a typical two-way prop vig the de-vigged market sits near 43.7%. My estimate is 47.0%. Edge = +1.5%. EV = 0.47 × 2.20 − 1 = +3.4%. Kelly f = (0.47 × 2.20 − 1)/1.20 = 2.83%.

2. The evidence — recency. Peterson's 2026 starts (19 of them) produced ≥5 K in 9 of 19 = 47.4% — already above the de-vigged market. His recent form is much stronger: last six starts went 8, 3, 6, 8, 6, 4 K, clearing 4.5 in four of six (66.7%). Recency and season agree in direction here; they diverge only in magnitude.

Peterson's ERA tells the same story of a pitcher better than his headline: season 5.39, but his 8/1–8/28 stretch was 9 ER in 27.1 IP = 2.96 — a 2.4-run divergence from the season aggregate, flagged per the recency rule. His 9/2 clunker (3.2 IP, 6 ER) came against tonight's opponent.

Opponent adjustment (why I did not use the 66.7% figure). Milwaukee is the one team that has suppressed him: two starts against the Brewers produced 4 K (9/2) and 2 K (6/27) — both under 4.5. That is a small but directly relevant sample, and it is the reason I shade my estimate down to 47.0%, essentially his season rate, rather than anywhere near his recent rate. The edge survives that haircut but is not large.

3. Risk factors. (a) 0-for-2 on this line against this specific opponent. (b) Peterson has been pulled early repeatedly (3.2 IP on 9/2, 3.2 IP on 7/3) — outing length is the main failure mode. (c) Milwaukee is 89-56, the best record in baseball, and a good contact team. (d) My 47.0% is close enough to the de-vigged 43.7% that a modest error in the vig assumption erases the edge.

4. Named source. https://statsapi.mlb.com/api/v1/people/656849/stats?stats=gameLog&group=pitching&season=2026 (full 2026 game log) and https://www.vegasinsider.com/mlb/odds/player-props/ (8-book strikeout board).

---

Pick 3 — Tampa Bay Rays ML, 2.01 (Caesars) — GRADED ONLY

1. The math. Best price 2.01 → implied 49.8%. Consensus TB −103 / ATL −117 → 50.74% / 53.92%, sum 104.66%, de-vigged TB 48.48%. My estimate 51.0%. Edge = +1.2%. EV = 0.51 × 2.01 − 1 = +2.5%. Kelly f = 2.49%.

2. The evidence — recency. AJ Smith-Shawver has made only four starts in 2026 and every one has been short and leaky: 4.1 IP / 2 ER, 4.1 IP / 4 ER, 5.0 IP / 3 ER, 4.0 IP / 5 ER. That is 17.2 IP over four starts (4.4 IP average) and a 7.13 ERA; his most recent outing (9/1 vs Washington) was his worst. He has never completed five innings. Season and recent form agree — there is no divergence to flag because there is no longer baseline.

Freddy Peralta (4.96 ERA) is not an ace either, and that is precisely why this is a small edge rather than a large one. The lean rests on team quality plus start length: Tampa Bay is 85-58 (.594), Atlanta 85-59 (.590) — functionally identical records, with Atlanta holding home field. Getting the better-by-a-hair team at plus money while the opposing starter cannot reach the fifth inning is worth roughly 2.5 points over the de-vigged line.

3. Risk factors. (a) Atlanta is home and equally good — this is close to a coin flip and the edge is thin. (b) Peralta's own 4.96 ERA means Tampa's pitching offers no real advantage. (c) The total moved up 8 → 8.5, i.e. the market expects runs, which raises variance and helps the team with the better bullpen — not clearly Tampa. (d) Four starts is a very small book on Smith-Shawver; a normal outing from him flips this.

4. Named source. https://statsapi.mlb.com/api/v1/people/700363/stats?stats=gameLog&group=pitching&season=2026 (all four 2026 starts) and https://www.vegasinsider.com/mlb/odds/las-vegas/.

---

Pick 4 — Chicago Cubs ML, 2.95 (RiversCasino) — GRADED ONLY

1. The math. Best price 2.95 → implied 33.9%. Consensus CHC +179 / MIL −219 → 35.84% / 68.65%, sum 104.49%, de-vigged CHC 34.30%. My estimate 35.0%. Edge = +1.1%. EV = 0.35 × 2.95 − 1 = +3.3%. Kelly f = 1.67%. Note that most of this edge is pure line shopping: bet365 offers +175 (2.75) on the same side, where EV would be −3.7%. The play only exists at RiversCasino.

2. The evidence — recency. This is the clearest season-vs-recent divergence on the slate. Jacob Misiorowski's season line is 14-5, 1.97 — genuinely elite. But his last five starts produced 11 ER in 28.0 IP = 3.54 ERA, a 1.57-run divergence from the season figure, which clears the ~1.5 flag threshold. His strikeout rate has fallen with it: 9, 6, 6, 6, 5 over those five starts after a stretch that included a 15-K complete game. Most directly, his most recent start on 9/2 was against these Cubs, and it was his worst of the year: 4.0 IP, 5 ER, 4 BB, 92 pitches.

Chicago's David Peterson (see Pick 2) has a 5.39 season ERA but pitched to 2.96 across 8/1–8/28. Both starters are trending away from their season lines — Misiorowski down, Peterson up — and the moneyline still reflects the season aggregates. Team quality remains firmly with Milwaukee (89-56 vs 81-64) and they are home, which is why my estimate moves only from 34.3% to 35.0%.

3. Risk factors. (a) Milwaukee has the best record in baseball and is at home; a 35% estimate is still a heavy underdog and this loses roughly two times in three. (b) The total moved down 7.5 → 7, a sharp signal that the market expects Misiorowski to be sharp, cutting directly against the recency read. (c) Five starts is a thin sample to declare decline; Misiorowski's 9/2 outing may simply be one bad night. (d) The edge is entirely dependent on getting 2.95 — at any other book this is negative.

4. Named source. https://statsapi.mlb.com/api/v1/people/694819/stats?stats=gameLog&group=pitching&season=2026 (per-start ERA and K totals, including the 9/2 start vs Chicago) and https://www.vegasinsider.com/mlb/odds/las-vegas/ (8-book moneyline board).

---

Candidates evaluated and rejected

Documented across different market types, with the math that disqualified each:

• AJ Smith-Shawver Over 3.5 K, 1.714 (bet365) — K Prop. Looked like the single best line-shopping edge on the board: 52 cents of spread (−140 bet365 vs −192 RiversCasino) against a de-vigged consensus near 60.2%, which prices out to +3.2% EV. Recency killed it. His four starts produced 4, 3, 3, 5 K — the over cleared twice in four. At 15 K in 17.2 IP across 4.4-IP outings, Poisson λ ≈ 3.7 gives P(≥4) ≈ 50.6%, not 60%. EV = 0.506 × 1.714 − 1 = −13.3%. The under is the value side; it is not priced on this board.
• Drew Anderson Over 4.5 K, 1.870 (bet365) — K Prop. 40 cents of spread (−115 to −155) and a de-vigged consensus near 55% suggested +2.9%. But his 2026 starts run 7, 5, 0, 4, 5 K over outings of 4.2, 2.2, 3.2, 4.0 and 5.0 IP — a 4.0-IP average. At λ ≈ 4.2, P(≥5) ≈ 41%, giving EV ≈ −23%. A 4.5 line does not fit a pitcher who rarely finishes the fourth.
• Patrick Sandoval Over 5.5 K, 1.769 (bet365) — K Prop. Cross-book de-vig implied +1.0%. His actual distribution disqualifies it: ≥6 K in 4 of 10 starts (40%) on the season, 3 of 5 recently, and just 2 K on 80 pitches in his last start (9/2). Blending to ~48% gives EV = 0.48 × 1.769 − 1 = −15%.
• Philadelphia Phillies ML, 1.741 (Caesars) — ML. The closest miss on the board at −0.2% (de-vigged 57.34% × 1.741). Painter's 5.55 ERA against Wesneski's 3.18 is a real angle, but the Phillies drifted −156 → −135, a 21-cent move in the direction that would favour us. Per the line-drift rule (>5c better than model = sharp signal on the other side), that is a reason to pass, not to press. Houston at 2.25 prices out at −4.0%.
• Chicago Cubs / Milwaukee Over 7, 1.80 (ESPN book) — Total. De-vigged two-way (o7 −125 / u7 +104) gives the over 53.1%; the price needs 55.6%. My model liked the over on two shaky starters and Milwaukee's offense (~57% → +2.6%), but the total moved down 7.5 → 7, i.e. sharp money is on the under. Respecting the move, and with no multi-book totals available to shop (see limitation below), this was passed.

---

Research limitations this run

Stated plainly rather than papered over:

• Multi-book totals and run lines were not obtainable. VegasInsider's /mlb/odds/las-vegas/ fetch truncated partway through the first row of the totals table; its /totals/ and /run-line/ sub-pages returned empty; OddsShark redirected to a JavaScript-rendered covers.com page containing no game data. ESPN supplied two-sided totals and run lines plus opening-vs-current for all 15 games, which I used for de-vigging and line-movement reads — but with a single book there is no shopping edge to capture, so no total or run line was recommended. Moneylines and strikeout props were shopped across 8 books each and are unaffected.
• Strikeout props are priced over-side only on the VegasInsider board, so under-side value (notably Smith-Shawver u3.5) could not be verified or recommended.
• TEX @ SEA is a genuine starter-confirmation failure, not a pricing failure: ESPN lists Seattle "Undecided" while VegasInsider lists "Miller (R)". Two sources conflict, so the game is a no-play under the 2-source rule.

---

Sources

Fetched or opened directly this run:

• https://statsapi.mlb.com/api/v1/schedule?sportId=1&date=2026-09-08&hydrate=probablePitcher,team,linescore
• https://www.vegasinsider.com/mlb/odds/las-vegas/
• https://www.vegasinsider.com/mlb/odds/player-props/
• https://www.espn.com/mlb/odds
• https://www.oddsshark.com/mlb/odds (redirected to covers.com; no usable data)
• https://statsapi.mlb.com/api/v1/people/669203/stats?stats=gameLog&group=pitching&season=2026 (Corbin Burnes — empty)
• https://statsapi.mlb.com/api/v1/people/656849/stats?stats=gameLog&group=pitching&season=2026 (David Peterson)
• https://statsapi.mlb.com/api/v1/people/694819/stats?stats=gameLog&group=pitching&season=2026 (Jacob Misiorowski)
• https://statsapi.mlb.com/api/v1/people/700363/stats?stats=gameLog&group=pitching&season=2026 (AJ Smith-Shawver)
• https://statsapi.mlb.com/api/v1/people/663776/stats?stats=gameLog&group=pitching&season=2026 (Patrick Sandoval)
• https://statsapi.mlb.com/api/v1/people/623454/stats?stats=gameLog&group=pitching&season=2026 (Drew Anderson)
• C:\Users\mitch\Documents\BettingTracker\bet_tracker.xlsx — Log sheet (bankroll), Strategy sheet, Picks sheet