Daily Bet Research Report — Monday, September 14, 2026

Report generated at 1:52 PM ET on Monday, September 14, 2026

Bankroll: $48.26 | Stake (12.5%, one play): $6.03 Record: 0-0-0 (Main bets, main leg, since 9/14 reset) | Hit rate: n/a — first day of the new era Slate: 11 games across MLB (10) and NFL (1) · 12 shortlisted for deep analysis

Bankroll source note. The Log's last row is Bet #105 (2026-09-11) closing at $51.60 — that is the final line of the pre-reset book. The clean reset dated today re-bases the current era to $48.26, per the Strategy tab (CLEAN RESET 2026-09-14 - bankroll re-based to 48.26) and site/build.py (RESET_BANKROLL = 48.26, with a comment stating the $3.34 gap is deliberately unreconciled). Today is the one day the "last Log row" rule and the live bankroll diverge, because no post-reset row exists yet. I used $48.26 so the report matches the site. From tomorrow the Log's last row resumes as the sole source.

Pending bets: none. Every Log row is settled (Bet #101 is an Adjustment/promo row and carries no outcome by design).

Section 2: Shortlist Analysis

Games skipped without deep analysis: 2. SD @ COL (Coors Field — standing no-play) and MIA @ ARI (Alcantara vs Burnes; both prop markets projected at or below their lines with no bet365 price advantage, so nothing to work with). The other 9 games all contributed at least one shortlisted candidate.

Section 3: My Recommendations

Ordered best-first by Kelly fraction (12.78%, 3.41%, 3.12%, 2.09%, 0.21%).

Hit Rate is my own estimate and historically runs about 5 points high. Today it deserves an extra discount: see Research Limitations — I could not independently verify a single pitcher's last-5 form, so these estimates lean on a third-party projection blended with the market rather than on my own game-log work.

All five are strikeout props, which is not the spread across markets I'd prefer. It is not a stylistic choice. Every game line I could price at bet365 came back negative or failed a hard rule (odds floor, Coors, unverifiable NFL injuries, Yankees tilt-fade). Props were the only market where bet365's price was the best in the market often enough to generate a positive number — which is the expected pattern, since bet365 prices props well and moneylines badly.

Section 3b: Prop Watchlist

Watchlist only — not recommendations.

• Dean Kremer (MIN) Over 4.5 K @ 1.741 (bet365). Projection 5.08 sits above the line and bet365 has the best over price in the market, but at −135 the price has already eaten the gap: blended estimate 56.2% vs 57.45% implied. If this drifts to −125 or better it becomes playable.
• Jonah Tong (NYM) Over 5.5 K @ 1.909 (bet365). Rookie right-hander, projection 5.81. Essentially a coin flip against a −110 price. Worth watching for a line move to 4.5, which would change the shape entirely.
• Tarik Skubal (LAD) Under 7.5 K @ 1.952 (bet365, −105). bet365 holds the best under price on the slate's best arm, but the projection (7.67) is above the line, so the number is pointing the wrong way. Listed because it is the sharpest single market on the board and worth a look if the line moves to 8.5.

Section 4: Reasoning

#1 — Kade Anderson (SEA) Under 5.5 strikeouts @ 2.05

The math. bet365 +105 → decimal 2.05 → implied 48.78%; removing ~4% prop vig gives a fair market read of 46.90%. A Poisson model on the ScoresAndOdds projection of 4.77 strikeouts gives P(K ≤ 5) = 65.62%. I do not trust that number at face value, so the estimate is a 55/45 blend of market-fair and model: 55.33%. Edge = 55.33 − 48.78 = +6.55%. EV per unit = 0.5533 × 1.05 − 0.4467 = +13.4%. Kelly fraction 12.78%.

The evidence. MLB framework, starting-pitcher lens. Anderson is a rookie left-hander making a late-season start for a Seattle club already eliminated from contention at 70-80, against a 56-93 Angels team. The situational factor doing the work is the gap between the market's line and the projection: the board sets Anderson at 5.5 while the projection says 4.77, a −15.3% divergence, the largest on today's slate. Two plausible readings — either the model is under-rating a rookie on thin sample, or the market has not adjusted his workload for a September start on a non-contender. Los Angeles Angels @ Angel Stadium is a neutral park for strikeouts.

Risk factors. This is the pick I am least comfortable ranking first. The entire edge is one third-party projection disagreeing with seven sportsbooks, and the strategy's own rule is explicit: "If book's implied deviates significantly from your raw estimate, respect the market." A 12.8% Kelly fraction off a single unverified input is not a 12.8% Kelly fraction. I could not pull Anderson's game log to check his actual recent strikeout rate or pitch counts, and rookie innings limits in September are a real and unmodelled risk in both directions. It ranks #1 because the instructions say to rank honestly by Kelly and let you choose — not because I think it is the safest play on the list. If you want the number I'd actually defend, it is #2.

Source. ScoresAndOdds MLB player props board (bet365 column, projections, 14 Sep 2026); MLB StatsAPI schedule for confirmed probables.

#2 — Gavin Williams (CLE) Over 7.5 strikeouts @ 1.714

The math. bet365 −140 → decimal 1.714 → implied 58.33%; de-vigged fair 56.09%. Poisson on a projection of 8.72 gives P(K ≥ 8) = 64.23%. Blended estimate 59.75%. Edge = 59.75 − 58.33 = +1.42%. EV = 0.5975 × 0.714 − 0.4025 = +2.4%. Kelly 3.41%.

The evidence. MLB framework, starting-pitcher and opponent lens. The projection sits 1.22 strikeouts (14.0%) above the posted line, the second-largest positive gap on the slate, and unlike the Anderson spot the direction is confirmed by the price — bet365 is at −140 on the over, the shortest over price of any book, meaning bet365 agrees the over is live and has moved toward it. Williams faces a Chicago White Sox lineup at Progressive Field; Chicago sit 76-73 but the total on this game is the lowest on the board (7, with bet365 the only book still holding 7 rather than 6.5), which reads as a market expecting a low-contact, pitcher-dominant game. That is the same signal from a different market.

Risk factors. −140 is a lot to lay on a prop that needs eight strikeouts; a single early hook on pitch count kills it, and I have no bullpen-usage data for Cleveland's prior 2-3 days to judge how long the leash is. The edge is thin (1.4%) and sits inside the margin of my own model error.

Source. ScoresAndOdds MLB player props and MLB odds boards (bet365 column, 14 Sep 2026).

#3 — Quinn Mathews (STL) Under 4.5 strikeouts @ 2.25

The math. bet365 +125 → decimal 2.25 → implied 44.44%; de-vigged fair 42.74%. Poisson on a projection of 4.65 gives P(K ≤ 4) = 50.39%. Blend 46.18%. Edge = +1.73%. EV = 0.4618 × 1.25 − 0.5382 = +3.9%. Kelly 3.12%.

The evidence. MLB framework, opponent-quality lens. bet365 holds the best under price in the market at +125 (BetMGM's best over is −135, so the two-way market is tight and bet365 is the outlier on the side I want). Mathews draws a San Francisco club that is 62-88, the worst record on tonight's slate, and Busch Stadium is a pitcher-friendly park — but the relevant factor for an under on strikeouts is that bad, out-of-contention lineups in September often shorten at-bats rather than lengthen them, and the projection of 4.65 is close enough to the line that the plus-money price carries the play rather than the projection.

Risk factors. The projection actually leans over (4.65 > 4.5), so this is a price bet, not a model bet — the edge exists only because +125 is generous for a near-coin-flip. Thin, and it flips negative if the price shortens past +115.

Source. ScoresAndOdds MLB player props board (bet365 column, 14 Sep 2026).

#4 — Sean Newcomb (CWS) Over 1.5 strikeouts @ 2.55

The math. bet365 +155 → decimal 2.55 → implied 39.22%; de-vigged fair 37.71%. Poisson on a projection of 1.49 gives P(K ≥ 2) = 43.88%. Blend 40.49%. Edge = +1.27%. EV = +3.2%. Kelly 2.09%.

The evidence. MLB framework, role lens. A 1.5-strikeout line is the market telling you Newcomb is working as an opener or on a very short leash — the line itself is the information. This is the one prop today where I have both bet365 prices (o1.5 +155 / u1.5 −210), so the de-vig is clean rather than approximated: bet365's own two-sided market prices the over at 37.7% fair, and the projection of 1.49 strikeouts implies a genuine 43.9% chance of reaching two. Two strikeouts from an opener facing a full lineup turn is a low bar.

Risk factors. Opener usage is the least predictable role in baseball — one inning and out at 12 pitches makes this a loser regardless of stuff. The lowest hit rate of the five (40.5%), so expect it to lose more often than not even when it is correctly priced.

Source. ScoresAndOdds MLB player props board (bet365 column, both sides, 14 Sep 2026).

#5 — Nick Lodolo (CIN) Over 4.5 strikeouts @ 2.25

The math. bet365 +125 → decimal 2.25 → implied 44.44%; de-vigged fair 42.74%. Poisson on a projection of exactly 4.50 gives P(K ≥ 5) = 46.79%. Blend 44.56%. Edge = +0.12%. EV = +0.3%. Kelly 0.21%.

The evidence. MLB framework, starting-pitcher divergence lens. Lodolo's season ERA is 5.12 (StatMuse) against Skubal's 2.84 in the other dugout — a 2.28-run divergence, well past the ~1.5 flag threshold, which is exactly why Cincinnati are +185 dogs. But a poor ERA and a healthy strikeout rate coexist routinely, and the projection lands precisely on the line at 4.50, meaning the market has this as a true coin flip while bet365 pays +125 on it. Great American Ball Park's park factor punishes Lodolo's contact, not his strikeouts.

Risk factors. An edge of 0.12% is indistinguishable from zero — this is on the list only because the brief says to include any positive edge with no minimum threshold. Against a Dodgers lineup that will be hunting a struggling starter, an early exit is a live outcome. Treat as the marginal option, not a play to reach for.

Source. StatMuse (Lodolo 5.12 ERA; Skubal 2.84 ERA, both 2026 season); ScoresAndOdds MLB player props board (bet365 column).

Why the game lines all failed

Documented here because four of the five slots above are the same market, and you should see what was screened out rather than assume it wasn't.

• Dodgers ML −225 (1.444). Below the 1.50 odds floor. Automatic no-play regardless of edge.
• Guardians ML −150 (1.667). bet365 is the best Cleveland price in the market (field: −154 to −175), which is unusual and worth noting. It still isn't enough: de-vigged consensus across the other seven books puts Cleveland at 58.56% and bet365's price implies 60.00%. Edge −1.44%.
• Orioles ML +110 (2.10). bet365 is best-of-market again (field +100 to +105), fair value lands near 47.1% against 47.62% implied — about half a point short. Separately, the scraped bet365 board shows Baltimore as the run-line favourite (−1.5 +170) while simultaneously pricing them as the moneyline underdog (+110), which cannot both be true. Data-quality no-play on top of the negative edge.
• LAD @ CIN Under 7.5 (1.909). All eight books sit on 7.5 at −110/−110 — a flat 50/50 market with no dispersion to exploit. The under also fights the park (GABP) and Lodolo's 5.12 ERA. Edge −2.38%.
• Yankees −1.5 (+130). Two independent reasons to pass: the standing Yankees tilt-fade rule, and the run line being a 1.5-run handicap that one-run games distort badly.
• SD @ COL. Coors Field. Standing no-play.
• DEN @ KC −2.5 (1.909). The NFL framework puts injury and QB status first, and I could not retrieve it from any source (see limitations). Line moved KC −155 → −130 and the total 42.5 → 43.5 since open, which is a meaningful move I cannot explain without the injury report. Weather is a non-factor: 84°F and mostly sunny at Arrowhead, no wind concern. Recommending a spread into an unexplained line move with no injury data would be guessing, so it's a no-play.

Sources

• https://statsapi.mlb.com/api/v1/schedule?sportId=1&date=2026-09-14&hydrate=probablePitcher,linescore — MLB slate, start times, confirmed probables, team records
• https://site.api.espn.com/apis/site/v2/sports/football/nfl/scoreboard?dates=20260914 — NFL slate, venue, weather, opening vs current lines
• https://sports.core.api.espn.com/v2/sports/football/leagues/nfl/events/401872931/competitions/401872931/odds — DEN @ KC line movement detail
• https://www.scoresandodds.com/mlb/odds — MLB run lines, totals and moneylines, eight-book grid including the bet365 column
• https://www.scoresandodds.com/nfl/odds — NFL spread and total, eight-book grid including the bet365 column
• https://www.scoresandodds.com/mlb/props — MLB strikeout props, best-price book attribution and projections
• https://www.scoresandodds.com/nfl — NFL Week 1 results and line movement
• https://www.statmuse.com/mlb/ask/tarik-skubal-era-last-5-starts-2026 — Skubal season ERA 2.84
• https://www.statmuse.com/mlb/ask/nick-lodolo-era-last-5-starts-2026 — Lodolo season ERA 5.12
• https://www.vegasinsider.com/mlb/odds/las-vegas/ — attempted; returned an empty board

Research limitations

This was a degraded research day and the recommendations should be weighted accordingly.

• bet365 could not be priced directly. The in-app browser refuses bet365.com outright on safety grounds, so every bet365 price in this report comes from the ScoresAndOdds eight-book comparison grid, where bet365 is the first column. Those prices are as of roughly 1:45 PM ET and may have moved. Verify on your own slip before placing.
• VegasInsider returned an empty odds board — every cell rendered N/A, and its probable-pitcher list disagreed with MLB StatsAPI in two games (it had Smith-Shawver for Atlanta and Ober for Minnesota; StatsAPI has López and Kremer). I used StatsAPI. This means the two-source starter confirmation the strategy requires was only partially satisfied — StatsAPI plus the ScoresAndOdds prop board, which agree.
• No pitcher last-5 splits. StatMuse answered the season-ERA question but ignored the last-5 filter, and the MLB StatsAPI game-log responses truncated before the game data. So the core MLB framework input — last-5 starts versus the season line — is missing for every pitcher. Season ERAs are cited where used; the hit-rate estimates lean on the ScoresAndOdds projection instead, which is why I blended toward the market rather than trusting the model.
• No bullpen usage data for any club's prior 2-3 days.
• No NFL injury report. The ESPN league-wide injuries feed truncated before reaching Denver or Kansas City, and both team-level endpoints returned empty objects. DEN @ KC was therefore not priced, and no NFL play appears in Section 3.
• Two MLB moneylines unpriced at bet365. The ScoresAndOdds page truncated partway through the moneyline block, so bet365 moneylines were available for only 4 of 10 games (LAD/CIN, CWS/CLE, DET/TOR, BAL/NYM). Run lines and totals were complete for all 10. The six missing moneylines were not screened.
• Games unscreened: 2 of 11 (SD @ COL on the Coors rule, MIA @ ARI for want of a usable price edge).
• Prop watchlist is MLB-only — no NFL props were priced, for the injury reason above.