The clearest signal on Saturday is that the market is underpricing runs — in three separate games, by a meaningful margin. LAD @ MIA is the loudest: the posted total sits at 7.5, and our model lands at 10.1, a 2.6-run gap that produces a 65.6% over probability. That's not a rounding error; that's the market and the model looking at genuinely different games. SEA @ OAK and CIN @ MIL tell a similar story, both clearing 12-plus points of model edge on their respective totals.

No moneyline side generated enough separation to qualify as a lean today. That's fine — it happens, and forcing a side onto the card just because there are 15 games would be the wrong response. The totals picture is where the model is finding disagreement, and that's what shapes the DFS build more than anything else tonight.

Nothing on the moneyline side cleared our 5-percentage-point minimum today. The totals board, however, came in with three legitimate leans — tracked edges, not official plays, and no money is implied on any of them.

• LAD @ MIA Over 7.5 — Model total: 10.1 | Over probability: 65.6% | Edge: +15.6 points. The largest gap on the board. Miami's roof status is listed as likely open, and the weather data shows a modest outbound wind. The model is seeing this as a significantly higher-scoring environment than the market does.
• SEA @ OAK Over 9.5 — Model total: 12.2 | Over probability: 63.2% | Edge: +13.2 points. Oakland is an open-air park with 13 mph blowing out to center tonight. The run and HR factors both tilt favorable.
• CIN @ MIL Over 8.5 — Model total: 10.8 | Over probability: 61.9% | Edge: +12.9 points. Milwaukee's roof is listed as likely open, with a 10 mph outbound wind. Solid edge, though the smallest of the three.

All three are leans on the board. They inform where we want DFS exposure. They are not bets.

The card is dark today. No play cleared the minimum threshold for an official stake — no moneyline side, and the totals leans above, while material in edge, are tracked-only and do not represent money committed.

A quiet card is not a problem. The three totals leans are among the cleaner model-versus-market disagreements we've logged in a while, and they shape the DFS approach meaningfully. But tracked edges are tracked edges. We note them, we build around them, and we wait for the number that earns the stake.

A significant caveat before the tables: 127 of 160 hitters on the slate are still projected, meaning lineups have not been posted for most of the board. Any hitter marked as projected below should be treated as exactly that — a model projection, not a confirmed starter. Check lineups at lock before finalizing any build.

On the pitching side, Tyler Glasnow leads the board at a 18.3-point projection with a 59.4% win probability facing MIA — and he sits inside the LAD @ MIA game where the model sees the most run-scoring activity. That's a real tension worth acknowledging: Glasnow's projection is built on strikeout volume and innings, not opponent suppression being the whole story. Gerrit Cole is the other premium arm worth noting: 17.1 projected, 62.8% win probability, and he draws the NYM in a Subway Series environment that figures to draw heavy ownership.

In the value tier, Amed Rosario and Paul Goldschmidt are the two confirmed starters generating strong value-per-dollar against Zac Thornton. Goldschmidt at $3,000 with an 8.1 projection and a 17.0 ceiling is the cleaner cash-game option of the pair. Colt Keith at $3,200 projects to 8.3 points with a 16.9 ceiling facing COL — confirmed in the lineup, which matters given how much of the board is still projected.

Three stacks stand out as the leverage targets, and the reason comes down to run environment more than projected points alone.

LAD bats against MIA carry the highest ceiling on the slate at 71.5 points for a four-bat stack at $19,100. The model sees 10.1 total runs in this game — the largest gap from the posted line anywhere today — and the park environment (HR factor 1.3, runs factor 1.1 with the roof likely open) reinforces that. If you're building a GPP lineup and want the game most likely to produce a high-scoring environment, this is the one the model points to most clearly.

The NYY stack is essentially tied in projected points (34.2) with a 70.6 ceiling at a lower price point ($17,300). Cole opposing them keeps ownership complicated, but the combination of Rosario and Goldschmidt confirmed in the lineup with a matchup against Thornton makes this a reasonable cash-game stack. DET bats vs. COL round out the top three: 34.2 projected, $16,200 — the cheapest of the group and the one the optimizer highlighted as part of its featured build alongside a BOS mini-stack.

The two parks most worth noting are on opposite ends of the temperature scale but arrive at similar HR-factor conclusions.

SF (SD @ SF) reads 18 mph blowing out to center field at 66 degrees. That's a meaningful outbound wind for a park that already plays with quirks, and the HR factor of 1.4 is the joint-highest on the slate. The cold temperature works against hitters making contact feel live, but 17 mph out to center tends to override that arithmetic on balls in the air. OAK (SEA @ OAK) also sits at an HR factor of 1.4 with 13 mph out to center at 77 degrees — the warmer and more comfortable of the two elevated-wind parks tonight, which is part of why the model's 12.2 total in that game feels grounded rather than aggressive.

Miami's roof status (likely open, 8 mph out, 83 degrees) contributes to the LAD @ MIA model total being where it is. It's not the windiest park tonight, but the combination of warm air, outbound breeze, and what the model sees as a relatively soft pitching matchup adds up.

Two angles carry the most conviction today.

First, the LAD @ MIA run environment. The model's 10.1 total against a 7.5 posted line is the single largest gap on the slate, and the weather and park data are consistent with the direction of that disagreement. It's the right game to stack around in GPP formats, and Glasnow's ceiling projection is real even with that context — he's expected to generate strikeout volume regardless of whether the offense scores six or ten. The lean is tracked on the board. The DFS stack is the executable angle.

Second, the confirmed-lineup edge in NYY vs NYM. On a slate where 127 of 160 hitters are still projected, Rosario, Goldschmidt, Judge, Cole, Soto, and Lindor are all confirmed. In cash games especially, that certainty has value when most of the board is still subject to late scratches and lineup changes. Build around the confirmed core, lean into the LAD game environment for GPP upside, and check everything at lock.