Daily Bet Research Report — Tuesday, September 15, 2026

Report generated at 3:10 PM ET on Tuesday, September 15, 2026. This is a re-run, replacing the 1:40 PM automated version, which was built on two defective sources. See Research Limitations.

Bankroll: $55.67 | Stake (12.5%, one play): $6.96 Record: 1-0-0 (Main bets, main leg, since 9/14 reset) | Hit rate: 100.0% (n=1 — meaningless, reported for completeness) Slate: 15 MLB games, all bookable (first pitch 6:40 PM ET) · 15 moneylines priced at bet365 · 0 props priceable

⚠ Why this report replaces the 1:40 PM version

The automated run produced five Player Prop recommendations. All five were unusable, for two separate reasons discovered this afternoon:

1. The VegasInsider strikeout prop board is serving stale data. It currently lists Skubal, Kade Anderson, Quinn Mathews, Gavin Williams, Jonah Tong, Dean Kremer and Nick Lodolo — yesterday's starters. Not one of today's fifteen probable pitchers appears on it. Cross-checked against statsapi.mlb.com for 2026-09-15.
2. Moneylines were never actually unavailable. Previous runs reported they "could not be priced due to truncation." Truncation saves the full page to a file; the complete eight-book board was always readable. It has been read in full this run.

Section 2: Shortlist Analysis

Every one of the 15 games was priced at bet365 for moneyline. None cleared. The full board and its de-vigged consensus fair value:

Best of all 28 sides shown as one per game; the opposite side was worse in every case. Every single one of the 28 bet365 moneyline sides is negative against de-vigged consensus, ranging from −0.97% (Astros) to −3.30% (Royals). BOS @ TEX had no bet365 price on the board and no confirmed Texas starter.

Games skipped without deep analysis: 0. All 15 were priced. SD @ COL is a no-play on the Coors rule regardless.

Section 3: My Recommendations

No qualifying plays today.

• Moneyline: all 15 games priced at bet365, all 28 sides negative against de-vigged consensus. Best available was Astros −155 at −0.97%. This is the expected result, not a failure — bet365 is a poor moneyline book and its prices sit consistently inside market fair value.
• Player Prop: cannot be priced. The only bet365 prop source is serving yesterday's pitchers. Recommending from it would mean pricing players who are not in today's games.
• Spread / Over-Under: no bet365 price is readable by any available tool. See Section 3c.

Allocate nothing. If you want action tonight, Section 3c gives you numbers to check in the app yourself.

Section 3b: Prop Watchlist

Empty this run. The prop board is stale; nothing on it can be honestly evaluated against today's games. Treat any prop sourced from it before this afternoon as suspect.

Section 3c: Game-Line Watch — Spread & Over/Under

bet365 spread and total prices cannot be read by any available tool, so these are screened on DraftKings and de-vigged to fair value. Required price is model-free — it is the break-even number implied by the market's own fair value, not an estimate of mine.

These are unverified at bet365. Check the app — if the price meets or beats the number, the edge is real; if not, pass.

Two line moves worth knowing when you look:

• LAD @ CIN total climbed 8 → 8.5 with the over still priced at −105. Yamamoto against Rhett Lowder in Great American Ball Park, a small park. Money is on the over; the number moving and the price staying short is the market agreeing rather than a soft spot.
• ATH @ TB total dropped 8.5 → 8. Griffin Jax against Jack Perkins at Tropicana Field. Movement toward the under.

Because no model is applied, these carry no overconfidence adjustment — but note that the required prices sit roughly 5% above DK's, which is simply the vig. bet365 would need to be materially better than DraftKings, not marginally.

Section 4: Reasoning

Why there is no recommendation.

The moneyline math is the whole story. Taking de-vigged consensus across eight books as the fair-value estimate, and bet365's own posted price as the bettable one, every side on the board prices worse than fair value. Worked example, the best of them:

Astros −155 (1.645). bet365 implied 60.8%. Consensus fair value, de-vigged from a −167 consensus against Royals +138: 59.8%. Edge = 59.8 − 60.8 = −0.97%. EV = (1.645 × 0.598) − 1 = −1.59% per dollar.

Nothing on the board is closer. The average across all 28 sides is about −2.3%, which is roughly bet365's moneyline hold on MLB. That is a structural feature of the book, not a bad slate.

Why I am not substituting props. Yesterday's card went 1-for-5, and the four losers were priced off season-rate models that omitted matchup context. Today the prop source is additionally stale. Pricing a prop for a pitcher who is not playing is not a small error, and I would rather hand you nothing than hand you that.

On the 1:40 PM version's picks. Juan Soto, Francisco Lindor and Colton Cowser are all genuinely in today's BAL @ NYM game, so those three were not stale — but they were the picks where bet365 sat 35–45 cents longer than the field across one game's RBI board. Under the rule added today, that spread is a stale-line signal rather than an edge, and three correlated legs off one suspect board is exactly the failure mode to avoid.

Sources

Fetched this run:

• https://statsapi.mlb.com/api/v1/schedule?sportId=1&date=2026-09-15&hydrate=probablePitcher — all 15 games, confirmed probable pitchers, first-pitch times. This is what exposed the stale prop board.
• https://www.vegasinsider.com/mlb/odds/las-vegas/ — eight-book moneyline board including bet365, plus opening line and consensus, all 15 games. Read in full from the saved file after truncation.
• https://www.vegasinsider.com/mlb/odds/player-props/ — fetched; found stale, not used.
• https://sports.core.api.espn.com/v2/sports/baseball/leagues/mlb/events?dates=20260915 — event IDs
• https://sports.core.api.espn.com/v2/sports/baseball/leagues/mlb/events/401816947/competitions/401816947/odds — LAD @ CIN spread/total
• https://sports.core.api.espn.com/v2/sports/baseball/leagues/mlb/events/401816943/competitions/401816943/odds — ATH @ TB spread/total

Tested and rejected as bet365 sources: scoresandodds.com/mlb, oddschecker.com/us/baseball/mlb, vegasinsider.com/mlb/odds/run-line/ (empty body), VegasInsider per-game matchup pages (moneyline only), in-app browser on VegasInsider (blocked).

Research limitations

• The bet365 player-prop source is stale and cannot currently be used. This is the most serious limitation and it needs resolving before props return to the card. Zero of today's 15 probable pitchers appear on the strikeout board.
• bet365 spread and total prices remain unreadable. Section 3c is a workaround, not a fix.
• Section 3c covers 2 of 15 games. ESPN's odds endpoint is one fetch per game; time did not allow all fifteen on a re-run. The two chosen were the earliest starts with the clearest line movement. Not a representative screen.
• No NFL, NHL, NBA, NCAAF or NCAAB games on a Tuesday in mid-September.
• Nothing was abandoned due to truncation this run.