The number that stopped me this morning was 2.54. That is the run column the model hung on the Philadelphia Phillies for this afternoon in Detroit, the lowest single-team projection anywhere on the fifteen-game board, and it belongs to a first-place-caliber offense. The projection is not an insult to Philadelphia. It is a statement about Tarik Skubal, who has issued ten walks in 70.2 innings, and about what happens to any .236 team batting average when the sequencing engine cannot find free baserunners to chain hits around. Across the hall of the same simulation, Detroit projects at 3.53 against Zack Wheeler. Add the columns and you get 6.06 combined runs in a game the market priced at 7. That 0.94-run spread is the largest full-game disagreement the model produced today, and it seeds three of the nine outputs on the Sunday card.
The Inputs: Confirmed Lineups, Live Weather, Rolling Run Columns
A quick word on what actually went into the 9:24 run, because the outputs only mean something if the inputs are honest. Lineups were confirmed through the live MLB StatsAPI feed in every park except Dodger Stadium and Petco, where projections filled in. Weather is measured, not assumed: 83 degrees with an 8 mile per hour wind out to east-northeast in Detroit, 79 and breezy in Chicago, 75 with a 10 mile per hour crosswind at Oracle Park. Each offense carries rolling seven, fifteen, and thirty-day run-scoring aggregates rather than a single season average, each starter carries his strikeout, walk, and contact-quality profile, and each bullpen carries its last-three-day workload, which matters today in San Diego, where the Padres relief corps has covered 16 innings over the past three days. No umpire assignments were available at run time, so no umpire adjustment was applied. The model does not fake inputs it does not have.
Detroit At 6.06: One Game, Three Outputs
| Projection | Market | Output |
|---|---|---|
| Phillies column: 2.54 | Team total 3.5 | Under 7 (-120, 3u), F5 under 3.5 (-115, 2u), NRFI (-155, 1u) |
| Tigers column: 3.53 | Team total 3.5 | |
| Game: 6.06 | Total 7.0 |
Wheeler at 9-1 with a 2.28 ERA, a 0.91 WHIP, and a .190 opponent average against Skubal at 3.06 with a 0.95 WHIP is the best simultaneous pitching input the model has processed this month, and the projection treats it accordingly. What makes the game bettable rather than merely admirable is the derivative structure. The first five innings under 3.5 prices the portion of the game the two starters control most directly, and the model's early-game distribution is even more lopsided than the full-game one because both bullpens carry midrange grades that widen the late innings. The first-inning derivative rests on frequency data: Philadelphia has scored a first-inning run in 3 of its last 19 games, Detroit in 5 of its last 18, and Wheeler owns a 1.29 first-inning ERA with a .143 opponent average in the frame. The counterweight is specific and worth stating exactly: Skubal's first-inning ERA is 4.50, triple his season rate, with six of his earned runs concentrated in the opening frame. That one split is why the NRFI carries the minimum stake while the wider windows carry the weight. The residual risk on the full-game under is Detroit's thirty-day form, 47 runs in nine July games, which the rolling aggregates see and the market sees too; the model simply weighs 95 games of .236 hitting more heavily than nine hot ones.
The Athletics Column: 3.64 Runs Under A 4.5 Line
The model projects the Athletics to score 3.64 runs at Rate Field today, against a team total posted at 4.5. Mechanically, the column is being dragged down by the seven-day aggregate, and the seven-day aggregate is brutal: five runs scored in the last five games, sequence 2, 1, 1, 1, 0, capped by last night's 1-0 shutout loss. This lineup has produced one run-scoring first inning in its last 17 games and has cleared four and a half runs once in its last seven. The team total under 4.5 at -135 takes 2 units. The companion output is the White Sox moneyline at -145 for 3 units, backed by a 4.50 Chicago run column against that same 3.64, a 30-17 Chicago home record, and the 15-1 combined score of the first two games of this series. Now the objection the model cannot wave away: the pitching assignments point the other direction. J.T. Ginn, at 3.10 over 98.2 innings with a .214 opponent average, is the best starter in this game, and Noah Schultz, at 6.00 with 29 walks in 48 innings, is the worst. The projection carries both facts and still lands where it lands, because an offense producing one run per night has been failing against every quality of arm it sees. If Schultz's walks meet a sudden Athletics revival, the moneyline side loses. Ninety-five games of data say the revival is not scheduled for today.
Petco Park: The Widest Gap On The Board
Toronto projects at 5.82 runs. San Diego projects at 3.60. A 2.22-run gap between two projected columns in the same game is the widest the model has produced this week, and it is the entire case for the Blue Jays moneyline at -125. The San Diego column is anchored by season-long scarcity, 374 runs scored, fewest in the major leagues, a .226 team average, 3.94 runs per game, now facing Kevin Gausman and his 108 strikeouts in 106.1 innings. The Toronto column is inflated by the opposing assignment: German Marquez carries a 5.02 ERA and a 1.43 WHIP and has averaged just over five innings across seven starts, which routes extra plate appearances to a Padres bullpen that has already absorbed 16 innings in three days. The play sits at 1 unit despite the gap, and the sizing discount is deliberate. San Diego scored eight runs last night in an 8-7 win over this exact opponent, the Padres lineup was still projected rather than confirmed at run time, and Gausman's 4-8 record reflects the same thin run support, Toronto at 4.08 runs per game, that keeps this from being a heavy position. Wide gap, soft inputs on one side, minimum stake. That is the discipline.
Three Quieter Gaps: St. Louis, San Francisco, Seattle
| Game | Model columns | Output |
|---|---|---|
| Braves at Cardinals | STL 4.90, ATL 4.35 | Cardinals ML -129 (1.5u) |
| Rockies at Giants | SF 5.38, COL 4.47 | Giants ML -135 (1.5u) |
| Mariners at Rays | SEA 5.01, TB 4.23 | Mariners +1.5 -170 (2u) |
St. Louis at 4.90 versus Atlanta at 4.35 looks narrow until you see the structural input driving it: Atlanta's listed starter, Danny Young, has 3.1 innings pitched this season and no starts, an opener profile that hands most of the game to relief innings, while Dustin May gives St. Louis ordinary but real length, 89 innings across 17 starts. The Braves' thirty-day offensive aggregate, 4.85 runs per game on the season, is the reason the stake stays at 1.5 units. San Francisco at 5.38 is the model's second-highest column of the day, and it is almost entirely a statement about Michael Lorenzen: a 6.46 ERA, a 1.78 WHIP, and a .327 opponent average feed the simulation more baserunners than any starter on the board. The honest tension is that Colorado's own offensive inputs are strong, a .748 team OPS and 4.79 runs per game, best of any card team, against a Giants starter with a 5.46 ERA, so this is a one-sided fade priced modestly. Seattle at 5.01 versus Tampa Bay's 4.23 is the model's quietest surprise, a projection that actually has the road team outscoring a Rays club that is 35-14 at home. The market still makes Tampa Bay the favorite, and rather than argue the full point, the card takes the run line at +1.5 -170 behind Emerson Hancock, whose 1.01 WHIP and .212 opponent average over 97.2 innings give the projection its floor. Seattle's five-game losing streak and 28-runs-in-nine-July-games form is the counterevidence, and it is why the position buys a run and a half instead of taking the price straight.
The Full Card, Ranked By Projection Confidence
| Output | Price | Stake | Core model input |
|---|---|---|---|
| Phillies/Tigers under 7 | -120 | 3u | Game projects 6.06 vs market 7 |
| White Sox moneyline | -145 | 3u | Columns 4.50 vs 3.64, home 30-17 |
| Athletics TT under 4.5 | -135 | 2u | Column 3.64, five runs in five games |
| Phillies/Tigers F5 under 3.5 | -115 | 2u | Starter-controlled window, elite WHIPs |
| Mariners +1.5 | -170 | 2u | SEA column 5.01 vs TB 4.23 |
| Cardinals moneyline | -129 | 1.5u | Opener structure vs 17-start length |
| Giants moneyline | -135 | 1.5u | Lorenzen .327 opp AVG feeds SF column 5.38 |
| Blue Jays moneyline | -125 | 1u | Widest gap: 5.82 vs 3.60 |
| Phillies/Tigers NRFI | -155 | 1u | PHI 3 of 19, DET 5 of 18 first-inning rates |
Notice that stake does not simply track gap size. The Petco game owns the widest projection gap and the smallest stake, because one lineup was unconfirmed and last night's result cut against the column. The Detroit under owns a smaller numeric gap and the largest stake, because every input beneath it, both WHIPs, both team averages, both first-inning frequencies, was confirmed and pointed the same way. Gap width measures opportunity. Input agreement measures confidence. The card is sized on the second.
What The Model Cannot See
Projected run columns are means, not destinies, and the failure modes today are concrete. Skubal's 4.50 first-inning ERA can end the NRFI inside ten minutes. Detroit's nine-game July surge, 47 runs, is exactly the kind of short-window signal the rolling aggregates dampen, and if the dampening is wrong the full-game under and the first-five under fail together, which is correlated risk the card accepts knowingly. Ginn outpitching Schultz is the single most likely path to a losing day, since it attacks 5 units across two outputs. The Braves' bats can make the opener structure irrelevant, Colorado's .748 OPS can make the Lorenzen fade feel foolish by the fourth inning, an 8-run Padres Saturday can bleed into Sunday, and laying -170 on a run line with a slumping offense means Seattle losing by exactly two is a real and painful outcome. Two lineups were projections at run time. The model publishes anyway, with sizes that reflect the doubt.
Final Verdict
The July 12 card is built on six games where the projected run columns and the posted numbers refuse to agree. The Phillies-Tigers under 7 at -120 for 3 units backs a 6.06 projection in the Wheeler-Skubal game, with the first five under 3.5 for 2 units and the NRFI at -155 for 1 unit pricing the starter-controlled windows. The White Sox moneyline at -145 for 3 units and Athletics team total under 4.5 at -135 for 2 units ride a 3.64 column produced by five runs in five days. The Cardinals at -129 bet length against an opener, the Giants at -135 fade a 1.78 WHIP, the Mariners +1.5 at -170 back the board's quietest projection inversion, and the Blue Jays at -125 take the widest gap of the week at the smallest stake. For the methodology behind the run columns, see the model methodology page, yesterday's July 11 team total model card, the latest model outputs, and the full prediction archive.