2026-09-19

Under 7.5 · published price -123

Relevant score: AZ 5, NYY 3

Original analysis

ESPN confirms Cam Schlittler for New York against Brandon Pfaadt for Arizona. Schlittler carries an elite 1.95 ERA, 0.90 WHIP and 228 strikeouts over 184.2 innings, while Pfaadt is at 3.77 ERA and 1.17 WHIP. The Yankees also enter with major lineup losses, including Aaron Judge out plus Jazz Chisholm Jr., Trent Grisham and Giancarlo Stanton on the injury list. The main Under 7.5 at -123 survives the low-total gate because the strongest scoring-suppression path is Schlittler, the second path is a competent Pfaadt profile against a depleted New York lineup, and the side markets require more assumptions at Yankees -168 or Arizona +138. Model fair Under probability is 58.5% versus 55.2% implied, with estimated risk-adjusted EV about +6.1%. The Over was rejected because it lacks two independent pre-bullpen scoring paths after the New York lineup downgrade; Yankees ML was rejected because the expensive price adds side-specific bullpen and run-support exposure that the total avoids.

Risk: A 7.5 total has limited late-run tolerance. Arizona can still create damage at home, and bullpen volatility or extra innings can defeat an otherwise correct starter-suppression thesis.

Result note: The game finished with eight total runs, so the frozen Under 7.5 lost by one-half run. Arizona generated enough offense against the expected starter-suppression path to break the low-total threshold, while New York still contributed three runs despite the lineup downgrade. Primary cause was insufficient low-total blowup tolerance; secondary cause was overconfidence that the two starting-pitcher paths would suppress scoring long enough. The thesis was fundamentally wrong for this game because the 7.5 margin was too thin for the realized scoring path.

Boston Red Sox · published price +130

Relevant score: TB 2, BOS 1

Original analysis

ESPN confirms Payton Tolle for Boston against Freddy Peralta for Tampa Bay. Tolle enters at 3.02 ERA and 1.07 WHIP with 166 strikeouts in 143 innings, while Peralta is at 4.67 ERA and 1.34 WHIP. ESPN Analytics has the matchup essentially even, while offers Boston +130, so the plus price materially compensates for Tampa Bay's stronger full-season team record and home field. The model assigns Boston a 50.5% fair win probability versus 43.5% implied, producing estimated EV about +16.2% before residual variance penalties. The main 7.5 total was audited equally: Over lacks a sufficiently reliable second scoring source against Tolle, while Under is vulnerable to Peralta's run-prevention volatility. Boston +1.5 was rejected because the heavy protection price gives away too much of the outright starter-value edge.

Risk: Tampa Bay is the stronger overall team at home and can erase Boston's starter advantage once the game reaches the bullpens. Any Tolle command regression makes the underdog price look much less attractive.

Result note: Boston lost 2-1 on a ninth-inning walk-off. The frozen morning starter thesis used Payton Tolle, but Boston actually started Brayan Bello, who delivered eight innings with one earned run and 11 strikeouts. The starter-specific thesis is therefore non-comparable because the frozen starter identity changed. Primary cause of the loss was offensive non-conversion and the late walk-off; the remaining broader underdog thesis was sound but variance-defeated rather than evidence that the frozen Tolle-based starter comparison was wrong.

Atlanta Braves · published price +113

Relevant score: ATL 6, HOU 3

Original analysis

ESPN confirms Grant Holmes for Atlanta against Hayden Wesneski for Houston. Holmes owns a 3.49 ERA over 141.2 innings; Wesneski is at 3.70 ERA but with only 48.2 innings, so the smaller sample receives regression rather than being treated as a stable edge. Atlanta also enters 90-64 compared with Houston 77-77, yet prices the Braves at +113. The model assigns Atlanta about 53.0% fair win probability versus 46.9% implied, for estimated EV about +12.9%. The total market was checked on equal footing: Over 8.5 -122 is too expensive after accounting for both starters' competent run-prevention profiles, while Under 8.5 +101 lacks enough late-run tolerance given both lineups' power. Atlanta ML is the cleaner risk-adjusted expression.

Risk: Houston has home field, Yordan Alvarez-level middle-order power and a capable starter in Wesneski. If Wesneski's smaller sample is genuinely representative, Atlanta's team-record advantage may not translate cleanly into this single game.

Result note: Atlanta won 6-3. Grant Holmes allowed two earned runs across six innings, while Hayden Wesneski also pitched effectively, so the game did not validate a large starter mismatch. Atlanta's broader team-quality and plus-price case converted through stronger full-game offense and late separation. Win source was the team-level edge rather than dominant starter separation. PARTIALLY VALIDATED.

Chicago White Sox · published price -136

Relevant score: CWS 3, DET 1

Original analysis

ESPN confirms Jackson Jobe for Detroit against Sean Burke for Chicago. Burke has the more established 161-inning workload with a 3.47 ERA, 1.20 WHIP and 173 strikeouts, while Jobe is at 3.89 ERA and 1.27 WHIP across only 34.2 innings and therefore receives added sample-size uncertainty. Chicago is also at home, and ESPN Analytics gives the White Sox 61.2% before price. -136 implies 57.6%; the model settles at 60.5% fair, leaving estimated EV about +5.0%. The Under 8 was rejected because bullpen and weather-related variance reduce late-run tolerance, while the Over lacks two clean pre-bullpen scoring paths against two competent starters. Chicago -1.5 was rejected because the starter edge is not large enough to require margin.

Risk: Jobe's recent skill growth could be real despite the smaller sample, and Chicago's bullpen has not been stable enough to turn a modest starter edge into a low-variance favorite position.

Result note: Chicago won 3-1. Sean Burke allowed one run over six innings and the White Sox bullpen protected the lead, while Chicago created enough offense to separate without requiring a run-line margin thesis. The expected combination of established starter workload, home field and full-game run prevention translated directly. Win source was starter performance plus lead retention. THESIS VALIDATED.

Over 7.5 · published price -116

Relevant score: MIL 1, BAL 0

Original analysis

ESPN confirms Robert Gasser for Milwaukee against Trevor Rogers for Baltimore. Their season prevention profiles are similar and ordinary rather than dominant: Gasser is at 4.30 ERA and 1.27 WHIP, Rogers at 4.22 ERA and 1.22 WHIP. Both offenses have credible pre-bullpen scoring paths, including Milwaukee's deep contact lineup and Baltimore's middle-order power, making the low main total of 7.5 more attractive than forcing a side in a game where the starters are closely matched. Over 7.5 -116 implies 53.7%; the model assigns 57.0%, for estimated EV about +6.1%. Milwaukee ML was rejected because ESPN's broader matchup context does not support treating the 96-58 record alone as sufficient against a near-even starter matchup; the Under was rejected because 7.5 leaves too little tolerance for two average starters and normal late scoring.

Risk: Both starters are capable of working efficiently into the middle innings, and Baltimore's park can play less offense-friendly than its reputation in certain conditions. A low-contact game can strand the Over despite the two-path setup.

Result note: The game finished with one total run, so Over 7.5 lost decisively. Robert Gasser threw seven scoreless innings and Trevor Rogers allowed only one run over eight innings with 10 strikeouts. The morning two-scoring-path assumption failed on both sides, and the model underweighted the possibility that two competent starters would suppress contact deep into the game. Primary cause was failure of both projected pre-bullpen scoring paths; secondary cause was insufficient respect for recent starter run-prevention form. The thesis was fundamentally wrong.

Pittsburgh Pirates · published price -118

Relevant score: PIT 6, KC 5

Original analysis

ESPN confirms Noah Cameron for Kansas City against Bubba Chandler for Pittsburgh. Cameron owns the cleaner starter line at 3.84 ERA and 1.19 WHIP versus Chandler's 4.28 ERA and 1.35 WHIP, so this is not a starter-edge favorite. Pittsburgh instead qualifies through the stronger broader team and venue structure: 77-77 overall and 40-36 at home versus Kansas City 67-87 and 28-51 away, with ESPN Analytics giving Pittsburgh 62.3%. -118 implies only 54.1%; after explicitly penalizing Pittsburgh for the starter disadvantage, the model lands at 58.5% fair and estimated EV about +8.1%. Under 8 -118 was rejected because Chandler's walk traffic and Pittsburgh's offensive path reduce late-run tolerance; Pirates -1.5 was rejected because Cameron's run prevention creates too much one-run resistance.

Risk: Cameron is the best starting pitcher in the matchup and can erase Pittsburgh's team and home-field advantages for six or seven innings. If that happens, the wager becomes overly dependent on bullpen and late-game execution.

Result note: Pittsburgh won 6-5. Noah Cameron remained a meaningful opposing-starter obstacle, while Bubba Chandler worked only four innings, so the morning starter-disadvantage warning was real. Pittsburgh nevertheless converted through the broader home and team structure and enough offense to survive a one-run game. Win source was full-game team context rather than starter dominance. PARTIALLY VALIDATED.

Chicago Cubs · published price -143

Relevant score: CHC 5, CIN 2

Original analysis

ESPN confirms Matthew Boyd for Chicago against Nick Lodolo for Cincinnati. Boyd enters at 4.16 ERA and 1.24 WHIP, while Lodolo is at 5.12 ERA and 1.49 WHIP. Chicago also owns the stronger season record, and ESPN Analytics gives the Cubs 60.6%. -143 implies 58.8%; the model assigns 61.0% fair, for estimated EV about +3.7%. Over 9 -115 was fully audited because Great American Ball Park and Lodolo create a scoring path, but nine runs still requires a reliable second path against Boyd and creates more assumptions than the side. Cubs -1.5 +113 was rejected because Cincinnati's home offense and bullpen variance make margin retention less reliable than outright win probability.

Risk: Lodolo has the strikeout ceiling to outperform his season ERA in a single start, and Cincinnati's home environment can punish any Boyd command mistake quickly. The price leaves limited room for a bullpen reversal.

Result note: Chicago won 5-2. Matthew Boyd allowed one earned run over five innings, and the Cubs produced enough offense to break a 2-2 tie and retain separation through the bullpen phase. The morning choice to use the moneyline instead of requiring -1.5 remained conservative, while the underlying team and starter edge both supported the result. Win source was starter suppression plus broader team quality. THESIS VALIDATED.

Under 7.5 · published price -123

Relevant score: MIN 5, LAA 5 (in progress)

Original analysis

ESPN confirms Joe Ryan for Minnesota against Reid Detmers for Los Angeles. Ryan is at 3.84 ERA, 1.16 WHIP and 146 strikeouts in 133.2 innings; Detmers is at 3.36 ERA, 1.05 WHIP and 199 strikeouts in 174 innings. The side market is essentially a pick'em, while ESPN Analytics slightly favors Los Angeles, so the total better captures the strongest shared information: two established strikeout starters and weak enough team context to support early suppression. Under 7.5 -123 implies 55.2%; the model assigns 57.0%, for estimated EV about +3.3%. Both MLs were rejected because the starter and team signals point in opposite directions; the Over was rejected because it lacks two reliable pre-bullpen scoring paths.

Risk: Seven and a half is a low threshold, so one short start, bullpen collapse or extra-inning inning can defeat the Under even if both starters perform acceptably.

Result note: The game was still unfinished at the grading freeze with the score tied 5-5. Although the current 10-run total has already moved beyond the frozen Under 7.5, the standing workflow requires unfinished games to remain PENDING until official completion rather than settle from an in-progress state. The item carries forward for finalization at the next morning run.

Over 7.5 · published price -114

Relevant score: CLE 12, ATH 6

Original analysis

ESPN confirms Jacob Lopez for the Athletics against Tanner Bibee for Cleveland. Lopez carries a 5.02 ERA and 1.48 WHIP, while Bibee is at 4.16 ERA. Cleveland therefore supplies the strongest scoring path against Lopez, while the Athletics retain a secondary path against a non-dominant Bibee and late relief. Over 7.5 -114 implies 53.3%; the model assigns 56.0%, for estimated EV about +5.1% before the structural penalty that keeps it below the stronger selections. Cleveland ML -188 was rejected as too expensive relative to ESPN Analytics' 61.6% home probability, and Cleveland -1.5 requires a margin assumption the offense does not clear strongly enough. Under 7.5 fails because Lopez's traffic profile leaves too little blowup tolerance.

Risk: The Athletics may fail to provide the required secondary scoring source, allowing Cleveland to win a low-scoring game such as 4-2 or 5-1. Bibee can also outperform his season ERA on any given start.

Result note: The game finished with 18 total runs, so Over 7.5 won comfortably. Cleveland immediately created the primary scoring path against Jacob Lopez, and Oakland also supplied the required secondary offense rather than leaving the Over dependent on one team. The exact two-source scoring structure required by the totals gate materialized. Win source was early Cleveland damage plus meaningful Oakland contribution. THESIS VALIDATED.

Washington Nationals · published price -107

Relevant score: WSH 8, STL 5

Original analysis

ESPN confirms Andrew Alvarez for Washington against Michael McGreevy for St. Louis. Alvarez owns the stronger prevention profile at 3.35 ERA with only five home runs allowed in 102 innings, while McGreevy is at 3.83 ERA over 160 innings. ESPN Analytics is nearly even but slightly favors Washington 50.9%, and -107 implies 51.7%; the model gives Washington 53.0% after the starter adjustment, for estimated EV about +2.5%. The main Under 8.5 -102 was audited but rejected because the hot St. Louis game environment and ordinary bullpen exposure weaken late-run tolerance, while Over 8.5 -118 is too expensive without two strong pre-bullpen scoring paths. Washington ML is the least assumption-heavy remaining market, but the edge is thin and stays last.

Risk: St. Louis has home field and a deeper full-game run-prevention structure, while McGreevy's 160-inning sample is substantially more established than Alvarez's. This is the weakest official edge and can easily flip in a close bullpen game.

Result note: Washington won 8-5 with the frozen starter identities intact. Washington created early offense against Michael McGreevy and maintained enough scoring and run prevention to convert the near-pick'em side. The starter edge was not sufficient by itself to make this a high-confidence selection, but the lower-assumption moneyline correctly captured Washington's modest edge without adding total or margin requirements. Win source was early offense plus adequate full-game execution. THESIS VALIDATED.